Oracle Database SQL Language Reference Guide

User Manual:

Open the PDF directly: View PDF PDF.
Page Count: 1634

DownloadOracle Database SQL Language Reference Guide
Open PDF In BrowserView PDF
[1]
Oracle®
Database

SQL Language Reference
11g Release 2 (11.2)
E41084-04

January 2016

Oracle Database SQL Language Reference, 11g Release 2 (11.2)
E41084-04
Copyright © 1996, 2016, Oracle and/or its affiliates. All rights reserved.
Primary Authors:

Diana Lorentz, Mary Beth Roeser

Contributors: Sundeep Abraham, Angela Amor, Geeta Arora, Vikas Arora, Lance Ashdown, Hermann
Baer, Shrikanth Bellamkonda, Donna Carver, Dan Chiba, Timothy Chien, Alan Choi, Thierry Cruanes,
George Eadon, Amit Ganesh, Barb Glover, Naveen Gopal, Mike Hallas, Min-Hank Ho, Chandrasekharan
Iyer, Mark Jaeger, Vikram Kapoor, Peter Knaggs, Srinath Krishnaswamy, Andre Kruglikov, Paul Lane,
Huagang Li, Yunrui Li, Vince Liang, Bryn Llewellyn, Rich Long, Scott Lynn, Vineet Marwah, Jun
Matsuzawa, Robert McGuirk, Rahil Mir, Gopal Mulagund, Sujatha Muthulingam, Hanlin Qian, Ashish Ray,
John Russell, Laurent Schneider, Vivian Schupmann, Jia Shi, Ajeet Singh, Wayne Smith, Shanshan Song,
Vinay Srihari, Jim Stenoish, Sankar Subramanian, Seema Sundara, Mark van de Wiel, Badhri Varanasi,
William Waddington, Peter Wahl, Charles Wetherell, Sergiusz Wolicki, Daniel Wong, Tsae-feng Yu,
Mohamed Zait, Fred Zemke, Wei Zhang, Weiran Zhang
This software and related documentation are provided under a license agreement containing restrictions on
use and disclosure and are protected by intellectual property laws. Except as expressly permitted in your
license agreement or allowed by law, you may not use, copy, reproduce, translate, broadcast, modify, license,
transmit, distribute, exhibit, perform, publish, or display any part, in any form, or by any means. Reverse
engineering, disassembly, or decompilation of this software, unless required by law for interoperability, is
prohibited.
The information contained herein is subject to change without notice and is not warranted to be error-free. If
you find any errors, please report them to us in writing.
If this is software or related documentation that is delivered to the U.S. Government or anyone licensing it
on behalf of the U.S. Government, then the following notice is applicable:
U.S. GOVERNMENT END USERS: Oracle programs, including any operating system, integrated software,
any programs installed on the hardware, and/or documentation, delivered to U.S. Government end users
are "commercial computer software" pursuant to the applicable Federal Acquisition Regulation and
agency-specific supplemental regulations. As such, use, duplication, disclosure, modification, and
adaptation of the programs, including any operating system, integrated software, any programs installed on
the hardware, and/or documentation, shall be subject to license terms and license restrictions applicable to
the programs. No other rights are granted to the U.S. Government.
This software or hardware is developed for general use in a variety of information management
applications. It is not developed or intended for use in any inherently dangerous applications, including
applications that may create a risk of personal injury. If you use this software or hardware in dangerous
applications, then you shall be responsible to take all appropriate fail-safe, backup, redundancy, and other
measures to ensure its safe use. Oracle Corporation and its affiliates disclaim any liability for any damages
caused by use of this software or hardware in dangerous applications.
Oracle and Java are registered trademarks of Oracle and/or its affiliates. Other names may be trademarks of
their respective owners.
Intel and Intel Xeon are trademarks or registered trademarks of Intel Corporation. All SPARC trademarks
are used under license and are trademarks or registered trademarks of SPARC International, Inc. AMD,
Opteron, the AMD logo, and the AMD Opteron logo are trademarks or registered trademarks of Advanced
Micro Devices. UNIX is a registered trademark of The Open Group.
This software or hardware and documentation may provide access to or information about content,
products, and services from third parties. Oracle Corporation and its affiliates are not responsible for and
expressly disclaim all warranties of any kind with respect to third-party content, products, and services
unless otherwise set forth in an applicable agreement between you and Oracle. Oracle Corporation and its
affiliates will not be responsible for any loss, costs, or damages incurred due to your access to or use of
third-party content, products, or services, except as set forth in an applicable agreement between you and
Oracle.

Contents
Preface ............................................................................................................................................................... xxi
Audience..................................................................................................................................................... xxi
Documentation Accessibility ................................................................................................................... xxi
Related Documents ................................................................................................................................... xxi
Conventions .............................................................................................................................................. xxii

What's New in the SQL Language Reference? ....................................................................... xxiii
Oracle Database 11g Release 2 (11.2.0.4) New Features in the SQL Language Reference............ xxiii
Oracle Database 11g Release 2 (11.2.0.2) New Features in the SQL Language Reference............ xxiii
Oracle Database 11g Release 2 (11.2.0.1) New Features in the SQL Language Reference............ xxiii
Oracle Database 11g Release 1 New Features in the SQL Language Reference............................. xxvi

1

Introduction to Oracle SQL
History of SQL .........................................................................................................................................
SQL Standards .........................................................................................................................................
How SQL Works ...............................................................................................................................
Common Language for All Relational Databases .........................................................................
Using Enterprise Manager .....................................................................................................................
Lexical Conventions.................................................................................................................................
Tools Support ...........................................................................................................................................

2

1-1
1-1
1-2
1-2
1-3
1-3
1-3

Pseudocolumns
Hierarchical Query Pseudocolumns ....................................................................................................
CONNECT_BY_ISCYCLE Pseudocolumn ....................................................................................
CONNECT_BY_ISLEAF Pseudocolumn .......................................................................................
LEVEL Pseudocolumn ......................................................................................................................
Sequence Pseudocolumns .....................................................................................................................
Where to Use Sequence Values .......................................................................................................
How to Use Sequence Values ..........................................................................................................
Version Query Pseudocolumns ............................................................................................................
COLUMN_VALUE Pseudocolumn .....................................................................................................

2-1
2-1
2-2
2-2
2-3
2-3
2-4
2-5
2-6
iii

OBJECT_ID Pseudocolumn .................................................................................................................. 2-7
OBJECT_VALUE Pseudocolumn ......................................................................................................... 2-8
ORA_ROWSCN Pseudocolumn ........................................................................................................... 2-8
ROWID Pseudocolumn .......................................................................................................................... 2-9
ROWNUM Pseudocolumn ................................................................................................................. 2-10
XMLDATA Pseudocolumn ................................................................................................................ 2-11

3

Basic Elements of Oracle SQL
Data Types ................................................................................................................................................ 3-1
Oracle Built-in Data Types................................................................................................................ 3-6
CHAR Data Type ....................................................................................................................... 3-9
NCHAR Data Type .................................................................................................................... 3-9
NVARCHAR2 Data Type ......................................................................................................... 3-9
VARCHAR2 Data Type ......................................................................................................... 3-10
VARCHAR Data Type ........................................................................................................... 3-10
NUMBER Data Type .............................................................................................................. 3-10
FLOAT Data Type.................................................................................................................... 3-12
Floating-Point Numbers ........................................................................................................ 3-12
BINARY_FLOAT .............................................................................................................. 3-13
BINARY_DOUBLE ........................................................................................................... 3-13
Numeric Precedence ............................................................................................................... 3-14
DATE Data Type ..................................................................................................................... 3-17
Using Julian Days ............................................................................................................. 3-17
TIMESTAMP Data Type ........................................................................................................ 3-18
TIMESTAMP WITH TIME ZONE Data Type .................................................................... 3-18
TIMESTAMP WITH LOCAL TIME ZONE Data Type ..................................................... 3-19
INTERVAL YEAR TO MONTH Data Type ........................................................................ 3-19
INTERVAL DAY TO SECOND Data Type ......................................................................... 3-19
Datetime/Interval Arithmetic ............................................................................................... 3-20
Support for Daylight Saving Times ...................................................................................... 3-22
Datetime and Interval Examples ........................................................................................... 3-22
RAW and LONG RAW Data Types ..................................................................................... 3-23
BFILE Data Type ..................................................................................................................... 3-25
BLOB Data Type ...................................................................................................................... 3-26
CLOB Data Type ..................................................................................................................... 3-26
NCLOB Data Type .................................................................................................................. 3-26
Rowid Data Types .......................................................................................................................... 3-27
ROWID Data Type .................................................................................................................. 3-27
UROWID Data Type ............................................................................................................... 3-28
ANSI, DB2, and SQL/DS Data Types ......................................................................................... 3-28
User-Defined Types ....................................................................................................................... 3-29
Object Types ............................................................................................................................ 3-30
REF Data Types ....................................................................................................................... 3-30
Varrays ...................................................................................................................................... 3-30
Nested Tables .......................................................................................................................... 3-31
Oracle-Supplied Types .................................................................................................................. 3-31
Any Types ....................................................................................................................................... 3-31

iv

ANYTYPE .................................................................................................................................
ANYDATA................................................................................................................................
ANYDATASET.........................................................................................................................
XML Types ......................................................................................................................................
XMLType ..................................................................................................................................
URI Data Types .......................................................................................................................
URIFactory Package ................................................................................................................
Spatial Types ...................................................................................................................................
SDO_GEOMETRY ...................................................................................................................
SDO_TOPO_GEOMETRY .....................................................................................................
SDO_GEORASTER ..................................................................................................................
Media Types ....................................................................................................................................
Expression Filter Type....................................................................................................................
Expression .................................................................................................................................
Data Type Comparison Rules ............................................................................................................
Numeric Values ..............................................................................................................................
Date Values .....................................................................................................................................
Character Values ............................................................................................................................
Object Values ..................................................................................................................................
Varrays and Nested Tables ...........................................................................................................
Data Type Precedence ....................................................................................................................
Data Conversion .............................................................................................................................
Implicit and Explicit Data Conversion ................................................................................
Implicit Data Conversion .......................................................................................................
Implicit Data Conversion Examples......................................................................................
Explicit Data Conversion .......................................................................................................
Security Considerations for Data Conversion ............................................................................
Literals ....................................................................................................................................................
Text Literals .....................................................................................................................................
Numeric Literals .............................................................................................................................
Integer Literals .........................................................................................................................
NUMBER and Floating-Point Literals ..................................................................................
Datetime Literals ............................................................................................................................
Interval Literals................................................................................................................................
INTERVAL YEAR TO MONTH ............................................................................................
INTERVAL DAY TO SECOND .............................................................................................
Format Models ......................................................................................................................................
Number Format Models ................................................................................................................
Number Format Elements .....................................................................................................
Datetime Format Models ..............................................................................................................
Datetime Format Elements ....................................................................................................
Uppercase Letters in Date Format Elements ...............................................................
Punctuation and Character Literals in Datetime Format Models .............................
Datetime Format Elements and Globalization Support ....................................................
ISO Standard Date Format Elements ...................................................................................
The RR Datetime Format Element ........................................................................................
RR Datetime Format Examples.......................................................................................

3-31
3-32
3-32
3-32
3-32
3-32
3-33
3-34
3-34
3-34
3-34
3-35
3-36
3-36
3-36
3-36
3-37
3-37
3-39
3-39
3-39
3-40
3-40
3-40
3-42
3-43
3-44
3-45
3-45
3-47
3-47
3-47
3-50
3-53
3-53
3-54
3-56
3-57
3-57
3-60
3-60
3-61
3-61
3-65
3-65
3-65
3-66

v

Datetime Format Element Suffixes .......................................................................................
Format Model Modifiers ...............................................................................................................
Format Model Examples .........................................................................................................
String-to-Date Conversion Rules .................................................................................................
XML Format Model .......................................................................................................................
Nulls ........................................................................................................................................................
Nulls in SQL Functions ..................................................................................................................
Nulls with Comparison Conditions ............................................................................................
Nulls in Conditions ........................................................................................................................
Comments ..............................................................................................................................................
Comments Within SQL Statements .............................................................................................
Comments on Schema and Nonschema Objects .......................................................................
Hints .................................................................................................................................................
Alphabetical Listing of Hints ........................................................................................................
ALL_ROWS Hint ....................................................................................................................
APPEND Hint ..........................................................................................................................
APPEND_VALUES Hint .......................................................................................................
CACHE Hint ............................................................................................................................
CHANGE_DUPKEY_ERROR_INDEX Hint .......................................................................
CLUSTER Hint ........................................................................................................................
CURSOR_SHARING_EXACT Hint .....................................................................................
DRIVING_SITE Hint ..............................................................................................................
DYNAMIC_SAMPLING Hint ...............................................................................................
FACT Hint ................................................................................................................................
FIRST_ROWS Hint ..................................................................................................................
FULL Hint ................................................................................................................................
HASH Hint ..............................................................................................................................
IGNORE_ROW_ON_DUPKEY_INDEX Hint .....................................................................
INDEX Hint .............................................................................................................................
INDEX_ASC Hint ...................................................................................................................
INDEX_COMBINE Hint ........................................................................................................
INDEX_DESC Hint .................................................................................................................
INDEX_FFS Hint .....................................................................................................................
INDEX_JOIN Hint ..................................................................................................................
INDEX_SS Hint .......................................................................................................................
INDEX_SS_ASC Hint .............................................................................................................
INDEX_SS_DESC Hint ...........................................................................................................
LEADING Hint ........................................................................................................................
MERGE Hint ............................................................................................................................
MODEL_MIN_ANALYSIS Hint ...........................................................................................
MONITOR Hint ......................................................................................................................
NATIVE_FULL_OUTER_JOIN Hint ....................................................................................
NOAPPEND Hint ...................................................................................................................
NOCACHE Hint .....................................................................................................................
NO_EXPAND Hint .................................................................................................................
NO_FACT Hint .......................................................................................................................
NO_INDEX Hint .....................................................................................................................

vi

3-66
3-67
3-68
3-69
3-70
3-71
3-71
3-72
3-72
3-72
3-73
3-74
3-74
3-79
3-79
3-80
3-80
3-81
3-81
3-82
3-82
3-82
3-83
3-83
3-83
3-84
3-84
3-85
3-85
3-86
3-86
3-87
3-87
3-87
3-88
3-88
3-89
3-89
3-89
3-90
3-90
3-90
3-91
3-91
3-91
3-91
3-92

NO_INDEX_FFS Hint ............................................................................................................
NO_INDEX_SS Hint ...............................................................................................................
NO_MERGE Hint ...................................................................................................................
NO_MONITOR Hint ..............................................................................................................
NO_NATIVE_FULL_OUTER_JOIN Hint ............................................................................
NO_PARALLEL Hint .............................................................................................................
NOPARALLEL Hint................................................................................................................
NO_PARALLEL_INDEX Hint ..............................................................................................
NOPARALLEL_INDEX Hint .................................................................................................
NO_PUSH_PRED Hint ..........................................................................................................
NO_PUSH_SUBQ Hint ..........................................................................................................
NO_PX_JOIN_FILTER Hint ..................................................................................................
NO_QUERY_TRANSFORMATION Hint ...........................................................................
NO_RESULT_CACHE Hint ..................................................................................................
NO_REWRITE Hint ................................................................................................................
NOREWRITE Hint...................................................................................................................
NO_STAR_TRANSFORMATION Hint ...............................................................................
NO_STATEMENT_QUEUING Hint.....................................................................................
NO_UNNEST Hint .................................................................................................................
NO_USE_HASH Hint ............................................................................................................
NO_USE_MERGE Hint ..........................................................................................................
NO_USE_NL Hint ..................................................................................................................
NO_XML_QUERY_REWRITE Hint ......................................................................................
NO_XMLINDEX_REWRITE Hint ........................................................................................
OPT_PARAM Hint .................................................................................................................
ORDERED Hint .......................................................................................................................
PARALLEL Hint .....................................................................................................................
PARALLEL_INDEX Hint ....................................................................................................
PQ_DISTRIBUTE Hint .........................................................................................................
PUSH_PRED Hint .................................................................................................................
PUSH_SUBQ Hint .................................................................................................................
PX_JOIN_FILTER Hint .........................................................................................................
QB_NAME Hint ....................................................................................................................
RESULT_CACHE Hint .........................................................................................................
RETRY_ON_ROW_CHANGE Hint ...................................................................................
REWRITE Hint ......................................................................................................................
STAR_TRANSFORMATION Hint .....................................................................................
STATEMENT_QUEUING Hint ...........................................................................................
UNNEST Hint ........................................................................................................................
USE_CONCAT Hint .............................................................................................................
USE_HASH Hint ...................................................................................................................
USE_MERGE Hint ................................................................................................................
USE_NL Hint .........................................................................................................................
USE_NL_WITH_INDEX Hint .............................................................................................
Database Objects .................................................................................................................................
Schema Objects .............................................................................................................................
Nonschema Objects ......................................................................................................................

3-92
3-92
3-93
3-93
3-93
3-93
3-94
3-94
3-94
3-94
3-95
3-95
3-95
3-95
3-95
3-96
3-96
3-96
3-96
3-96
3-97
3-97
3-97
3-98
3-98
3-98
3-98
3-101
3-101
3-104
3-104
3-104
3-104
3-105
3-105
3-106
3-106
3-107
3-107
3-107
3-108
3-108
3-108
3-109
3-109
3-109
3-110

vii

Database Object Names and Qualifiers .........................................................................................
Database Object Naming Rules ..................................................................................................
Schema Object Naming Examples .............................................................................................
Schema Object Naming Guidelines ...........................................................................................
Syntax for Schema Objects and Parts in SQL Statements...........................................................
How Oracle Database Resolves Schema Object References ...................................................
References to Objects in Other Schemas ...................................................................................
References to Objects in Remote Databases .............................................................................
Creating Database Links ......................................................................................................
Database Link Names ....................................................................................................
Username and Password ...............................................................................................
Database Connect String................................................................................................
References to Database Links ..............................................................................................
References to Partitioned Tables and Indexes .........................................................................
References to Object Type Attributes and Methods ................................................................

4

3-110
3-111
3-114
3-115
3-115
3-116
3-117
3-117
3-117
3-117
3-118
3-118
3-118
3-119
3-121

Operators
About SQL Operators ..............................................................................................................................
Unary and Binary Operators ...........................................................................................................
Operator Precedence .........................................................................................................................
Arithmetic Operators ..............................................................................................................................
Concatenation Operator ..........................................................................................................................
Hierarchical Query Operators................................................................................................................
PRIOR .................................................................................................................................................
CONNECT_BY_ROOT .....................................................................................................................
Set Operators ............................................................................................................................................
Multiset Operators ..................................................................................................................................
MULTISET EXCEPT .........................................................................................................................
MULTISET INTERSECT ..................................................................................................................
MULTISET UNION ..........................................................................................................................
User-Defined Operators .........................................................................................................................

4-1
4-2
4-2
4-3
4-4
4-5
4-5
4-5
4-5
4-6
4-6
4-7
4-8
4-9

5 Functions
About SQL Functions .............................................................................................................................
Single-Row Functions ............................................................................................................................
Numeric Functions ............................................................................................................................
Character Functions Returning Character Values ........................................................................
Character Functions Returning Number Values ..........................................................................
Character Set Functions ....................................................................................................................
Datetime Functions ...........................................................................................................................
General Comparison Functions ......................................................................................................
Conversion Functions .......................................................................................................................
Large Object Functions .....................................................................................................................
Collection Functions .........................................................................................................................
Hierarchical Functions .....................................................................................................................
Data Mining Functions .....................................................................................................................
XML Functions ..................................................................................................................................
viii

5-2
5-3
5-3
5-4
5-5
5-5
5-5
5-6
5-6
5-7
5-7
5-7
5-7
5-8

Encoding and Decoding Functions ................................................................................................ 5-9
NULL-Related Functions ................................................................................................................. 5-9
Environment and Identifier Functions ........................................................................................... 5-9
Aggregate Functions ............................................................................................................................ 5-10
Analytic Functions ............................................................................................................................... 5-11
Object Reference Functions ............................................................................................................... 5-17
Model Functions ................................................................................................................................... 5-17
OLAP Functions .................................................................................................................................... 5-17
Data Cartridge Functions .................................................................................................................... 5-17
Alphabetical Listing of SQL Functions ............................................................................................ 5-17
ABS .......................................................................................................................................................... 5-18
ACOS ...................................................................................................................................................... 5-19
ADD_MONTHS ................................................................................................................................... 5-20
APPENDCHILDXML .......................................................................................................................... 5-21
ASCII ...................................................................................................................................................... 5-22
ASCIISTR .............................................................................................................................................. 5-23
ASIN ....................................................................................................................................................... 5-24
ATAN ...................................................................................................................................................... 5-25
ATAN2 .................................................................................................................................................... 5-26
AVG ........................................................................................................................................................ 5-27
BFILENAME .......................................................................................................................................... 5-29
BIN_TO_NUM ...................................................................................................................................... 5-30
BITAND ................................................................................................................................................. 5-32
CARDINALITY .................................................................................................................................... 5-34
CAST ....................................................................................................................................................... 5-35
CEIL ......................................................................................................................................................... 5-38
CHARTOROWID ................................................................................................................................. 5-39
CHR ......................................................................................................................................................... 5-40
CLUSTER_ID ........................................................................................................................................ 5-42
CLUSTER_PROBABILITY ................................................................................................................. 5-44
CLUSTER_SET ..................................................................................................................................... 5-46
COALESCE ............................................................................................................................................ 5-48
COLLECT ............................................................................................................................................... 5-50
COMPOSE ............................................................................................................................................. 5-51
CONCAT ................................................................................................................................................ 5-52
CONVERT ............................................................................................................................................. 5-53
CORR ...................................................................................................................................................... 5-55
CORR_* .................................................................................................................................................. 5-57
CORR_S ........................................................................................................................................... 5-58
CORR_K .......................................................................................................................................... 5-59
COS ......................................................................................................................................................... 5-60
COSH ...................................................................................................................................................... 5-61
COUNT ................................................................................................................................................... 5-62
COVAR_POP ........................................................................................................................................ 5-64
COVAR_SAMP ..................................................................................................................................... 5-66
CUBE_TABLE ........................................................................................................................................ 5-67
CUME_DIST ......................................................................................................................................... 5-69

ix

CURRENT_DATE ................................................................................................................................
CURRENT_TIMESTAMP ..................................................................................................................
CV .............................................................................................................................................................
DATAOBJ_TO_PARTITION .............................................................................................................
DBTIMEZONE .....................................................................................................................................
DECODE ................................................................................................................................................
DECOMPOSE .......................................................................................................................................
DELETEXML .........................................................................................................................................
DENSE_RANK .....................................................................................................................................
DEPTH ....................................................................................................................................................
DEREF .....................................................................................................................................................
DUMP .....................................................................................................................................................
EMPTY_BLOB, EMPTY_CLOB .........................................................................................................
EXISTSNODE .......................................................................................................................................
EXP ..........................................................................................................................................................
EXTRACT (datetime) ...........................................................................................................................
EXTRACT (XML) ..................................................................................................................................
EXTRACTVALUE ................................................................................................................................
FEATURE_ID ........................................................................................................................................
FEATURE_SET .....................................................................................................................................
FEATURE_VALUE ...............................................................................................................................
FIRST ....................................................................................................................................................
FIRST_VALUE ....................................................................................................................................
FLOOR ..................................................................................................................................................
FROM_TZ ............................................................................................................................................
GREATEST ..........................................................................................................................................
GROUP_ID ..........................................................................................................................................
GROUPING .........................................................................................................................................
GROUPING_ID ..................................................................................................................................
HEXTORAW .......................................................................................................................................
INITCAP ..............................................................................................................................................
INSERTCHILDXML ..........................................................................................................................
INSERTCHILDXMLAFTER .............................................................................................................
INSERTCHILDXMLBEFORE ..........................................................................................................
INSERTXMLAFTER ..........................................................................................................................
INSERTXMLBEFORE .......................................................................................................................
INSTR ...................................................................................................................................................
ITERATION_NUMBER ....................................................................................................................
LAG .......................................................................................................................................................
LAST .....................................................................................................................................................
LAST_DAY ..........................................................................................................................................
LAST_VALUE .....................................................................................................................................
LEAD .....................................................................................................................................................
LEAST ...................................................................................................................................................
LENGTH ..............................................................................................................................................
LISTAGG .............................................................................................................................................
LN ..........................................................................................................................................................

x

5-71
5-72
5-73
5-75
5-76
5-77
5-79
5-80
5-82
5-84
5-85
5-86
5-88
5-89
5-90
5-91
5-94
5-95
5-96
5-97
5-99
5-101
5-103
5-105
5-106
5-107
5-109
5-110
5-111
5-112
5-113
5-114
5-116
5-117
5-118
5-119
5-120
5-122
5-124
5-126
5-127
5-128
5-131
5-133
5-135
5-136
5-138

LNNVL .................................................................................................................................................
LOCALTIMESTAMP ........................................................................................................................
LOG .......................................................................................................................................................
LOWER .................................................................................................................................................
LPAD......................................................................................................................................................
LTRIM ..................................................................................................................................................
MAKE_REF ..........................................................................................................................................
MAX ......................................................................................................................................................
MEDIAN ..............................................................................................................................................
MIN .......................................................................................................................................................
MOD .....................................................................................................................................................
MONTHS_BETWEEN .......................................................................................................................
NANVL .................................................................................................................................................
NCHR ...................................................................................................................................................
NEW_TIME .........................................................................................................................................
NEXT_DAY ..........................................................................................................................................
NLS_CHARSET_DECL_LEN ...........................................................................................................
NLS_CHARSET_ID ...........................................................................................................................
NLS_CHARSET_NAME ...................................................................................................................
NLS_INITCAP ....................................................................................................................................
NLS_LOWER .......................................................................................................................................
NLS_UPPER ........................................................................................................................................
NLSSORT ............................................................................................................................................
NTH_VALUE ......................................................................................................................................
NTILE ...................................................................................................................................................
NULLIF .................................................................................................................................................
NUMTODSINTERVAL ....................................................................................................................
NUMTOYMINTERVAL ...................................................................................................................
NVL .......................................................................................................................................................
NVL2 .....................................................................................................................................................
ORA_DST_AFFECTED .....................................................................................................................
ORA_DST_CONVERT .....................................................................................................................
ORA_DST_ERROR ............................................................................................................................
ORA_HASH ........................................................................................................................................
PATH ....................................................................................................................................................
PERCENT_RANK ..............................................................................................................................
PERCENTILE_CONT ........................................................................................................................
PERCENTILE_DISC ..........................................................................................................................
POWER .................................................................................................................................................
POWERMULTISET ...........................................................................................................................
POWERMULTISET_BY_CARDINALITY .....................................................................................
PREDICTION .....................................................................................................................................
PREDICTION_BOUNDS .................................................................................................................
PREDICTION_COST ........................................................................................................................
PREDICTION_DETAILS .................................................................................................................
PREDICTION_PROBABILITY .......................................................................................................
PREDICTION_SET ............................................................................................................................

5-139
5-140
5-141
5-142
5-143
5-144
5-145
5-146
5-148
5-150
5-152
5-153
5-154
5-155
5-156
5-157
5-158
5-159
5-160
5-161
5-162
5-163
5-164
5-167
5-169
5-170
5-171
5-172
5-173
5-174
5-175
5-176
5-177
5-178
5-179
5-180
5-182
5-185
5-187
5-188
5-189
5-191
5-193
5-195
5-197
5-199
5-201

xi

PRESENTNNV ....................................................................................................................................
PRESENTV ...........................................................................................................................................
PREVIOUS ...........................................................................................................................................
RANK ...................................................................................................................................................
RATIO_TO_REPORT ........................................................................................................................
RAWTOHEX .......................................................................................................................................
RAWTONHEX ....................................................................................................................................
REF ........................................................................................................................................................
REFTOHEX ..........................................................................................................................................
REGEXP_COUNT ..............................................................................................................................
REGEXP_INSTR .................................................................................................................................
REGEXP_REPLACE ...........................................................................................................................
REGEXP_SUBSTR .............................................................................................................................
REGR_ (Linear Regression) Functions ...........................................................................................
REMAINDER ......................................................................................................................................
REPLACE .............................................................................................................................................
ROUND (date) ....................................................................................................................................
ROUND (number) ..............................................................................................................................
ROW_NUMBER .................................................................................................................................
ROWIDTOCHAR ...............................................................................................................................
ROWIDTONCHAR ...........................................................................................................................
RPAD ....................................................................................................................................................
RTRIM ..................................................................................................................................................
SCN_TO_TIMESTAMP ....................................................................................................................
SESSIONTIMEZONE .......................................................................................................................
SET ........................................................................................................................................................
SIGN .....................................................................................................................................................
SIN .........................................................................................................................................................
SINH .....................................................................................................................................................
SOUNDEX ...........................................................................................................................................
SQRT .....................................................................................................................................................
STATS_BINOMIAL_TEST ..............................................................................................................
STATS_CROSSTAB ..........................................................................................................................
STATS_F_TEST ..................................................................................................................................
STATS_KS_TEST ...............................................................................................................................
STATS_MODE ...................................................................................................................................
STATS_MW_TEST ............................................................................................................................
STATS_ONE_WAY_ANOVA ..........................................................................................................
STATS_T_TEST_* ..............................................................................................................................
STATS_T_TEST_ONE ..................................................................................................................
STATS_T_TEST_PAIRED ...........................................................................................................
STATS_T_TEST_INDEP and STATS_T_TEST_INDEPU .......................................................
STATS_WSR_TEST ...........................................................................................................................
STDDEV ...............................................................................................................................................
STDDEV_POP ....................................................................................................................................
STDDEV_SAMP .................................................................................................................................
SUBSTR ................................................................................................................................................

xii

5-204
5-206
5-208
5-209
5-211
5-212
5-213
5-214
5-215
5-216
5-218
5-221
5-224
5-226
5-231
5-232
5-233
5-234
5-236
5-238
5-239
5-240
5-241
5-242
5-244
5-245
5-246
5-247
5-248
5-249
5-250
5-251
5-252
5-253
5-255
5-256
5-257
5-259
5-261
5-263
5-264
5-265
5-267
5-268
5-270
5-272
5-274

SUM ......................................................................................................................................................
SYS_CONNECT_BY_PATH .............................................................................................................
SYS_CONTEXT ..................................................................................................................................
SYS_DBURIGEN ................................................................................................................................
SYS_EXTRACT_UTC ........................................................................................................................
SYS_GUID ...........................................................................................................................................
SYS_TYPEID .......................................................................................................................................
SYS_XMLAGG ...................................................................................................................................
SYS_XMLGEN ....................................................................................................................................
SYSDATE .............................................................................................................................................
SYSTIMESTAMP ...............................................................................................................................
TAN .......................................................................................................................................................
TANH ...................................................................................................................................................
TIMESTAMP_TO_SCN ....................................................................................................................
TO_BINARY_DOUBLE ....................................................................................................................
TO_BINARY_FLOAT ........................................................................................................................
TO_BLOB .............................................................................................................................................
TO_CHAR (character) ........................................................................................................................
TO_CHAR (datetime) ........................................................................................................................
TO_CHAR (number) ..........................................................................................................................
TO_CLOB .............................................................................................................................................
TO_DATE ............................................................................................................................................
TO_DSINTERVAL .............................................................................................................................
TO_LOB ...............................................................................................................................................
TO_MULTI_BYTE ..............................................................................................................................
TO_NCHAR (character) ....................................................................................................................
TO_NCHAR (datetime) .....................................................................................................................
TO_NCHAR (number) ......................................................................................................................
TO_NCLOB .........................................................................................................................................
TO_NUMBER .....................................................................................................................................
TO_SINGLE_BYTE ............................................................................................................................
TO_TIMESTAMP ...............................................................................................................................
TO_TIMESTAMP_TZ .......................................................................................................................
TO_YMINTERVAL ............................................................................................................................
TRANSLATE .......................................................................................................................................
TRANSLATE ... USING ....................................................................................................................
TREAT ..................................................................................................................................................
TRIM .....................................................................................................................................................
TRUNC (date) ......................................................................................................................................
TRUNC (number) ...............................................................................................................................
TZ_OFFSET .........................................................................................................................................
UID ........................................................................................................................................................
UNISTR ................................................................................................................................................
UPDATEXML ......................................................................................................................................
UPPER ...................................................................................................................................................
USER .....................................................................................................................................................
USERENV ............................................................................................................................................

5-276
5-278
5-279
5-286
5-287
5-288
5-289
5-290
5-291
5-292
5-293
5-294
5-295
5-296
5-297
5-299
5-300
5-301
5-302
5-305
5-307
5-308
5-310
5-312
5-313
5-314
5-315
5-316
5-317
5-318
5-319
5-320
5-321
5-323
5-324
5-325
5-327
5-328
5-329
5-330
5-331
5-332
5-333
5-334
5-336
5-337
5-338

xiii

VALUE ..................................................................................................................................................
VAR_POP .............................................................................................................................................
VAR_SAMP .........................................................................................................................................
VARIANCE .........................................................................................................................................
VSIZE ....................................................................................................................................................
WIDTH_BUCKET ..............................................................................................................................
XMLAGG .............................................................................................................................................
XMLCAST.............................................................................................................................................
XMLCDATA ........................................................................................................................................
XMLCOLATTVAL .............................................................................................................................
XMLCOMMENT ................................................................................................................................
XMLCONCAT .....................................................................................................................................
XMLDIFF .............................................................................................................................................
XMLELEMENT ...................................................................................................................................
XMLEXISTS .........................................................................................................................................
XMLFOREST .......................................................................................................................................
XMLISVALID .....................................................................................................................................
XMLPARSE .........................................................................................................................................
XMLPATCH ........................................................................................................................................
XMLPI ...................................................................................................................................................
XMLQUERY ........................................................................................................................................
XMLROOT ...........................................................................................................................................
XMLSEQUENCE ................................................................................................................................
XMLSERIALIZE .................................................................................................................................
XMLTABLE .........................................................................................................................................
XMLTRANSFORM ............................................................................................................................
ROUND and TRUNC Date Functions ............................................................................................
About User-Defined Functions .......................................................................................................
Prerequisites...................................................................................................................................
Name Precedence .........................................................................................................................
Naming Conventions ...........................................................................................................

6

5-340
5-341
5-343
5-344
5-346
5-347
5-349
5-351
5-352
5-353
5-354
5-355
5-356
5-358
5-361
5-362
5-363
5-364
5-365
5-367
5-368
5-370
5-371
5-373
5-375
5-377
5-379
5-380
5-381
5-382
5-382

Expressions
About SQL Expressions ......................................................................................................................... 6-1
Simple Expressions ................................................................................................................................. 6-3
Compound Expressions ......................................................................................................................... 6-4
CASE Expressions ................................................................................................................................... 6-5
Column Expressions ............................................................................................................................... 6-6
CURSOR Expressions.............................................................................................................................. 6-7
Datetime Expressions ............................................................................................................................. 6-8
Function Expressions ........................................................................................................................... 6-10
Interval Expressions ............................................................................................................................. 6-10
Model Expressions ................................................................................................................................ 6-11
Object Access Expressions .................................................................................................................. 6-13
Placeholder Expressions ...................................................................................................................... 6-14
Scalar Subquery Expressions ............................................................................................................. 6-14
Type Constructor Expressions ........................................................................................................... 6-14

xiv

Expression Lists .................................................................................................................................... 6-16

7

Conditions
About SQL Conditions............................................................................................................................ 7-1
Condition Precedence........................................................................................................................ 7-3
Comparison Conditions ......................................................................................................................... 7-4
Simple Comparison Conditions ...................................................................................................... 7-5
Group Comparison Conditions ...................................................................................................... 7-6
Floating-Point Conditions ..................................................................................................................... 7-7
Logical Conditions ................................................................................................................................... 7-8
Model Conditions ................................................................................................................................... 7-9
IS ANY Condition ............................................................................................................................. 7-9
IS PRESENT Condition ................................................................................................................. 7-10
Multiset Conditions ............................................................................................................................. 7-11
IS A SET Condition ........................................................................................................................ 7-12
IS EMPTY Condition ...................................................................................................................... 7-12
MEMBER Condition ...................................................................................................................... 7-13
SUBMULTISET Condition ............................................................................................................ 7-13
Pattern-matching Conditions ............................................................................................................. 7-14
LIKE Condition ............................................................................................................................... 7-14
REGEXP_LIKE Condition ............................................................................................................. 7-18
Null Conditions .................................................................................................................................... 7-19
XML Conditions ................................................................................................................................... 7-20
EQUALS_PATH Condition .......................................................................................................... 7-20
UNDER_PATH Condition ............................................................................................................ 7-21
Compound Conditions ........................................................................................................................ 7-21
BETWEEN Condition .......................................................................................................................... 7-22
EXISTS Condition ................................................................................................................................ 7-22
IN Condition ......................................................................................................................................... 7-23
IS OF type Condition ........................................................................................................................... 7-25

8

Common SQL DDL Clauses
allocate_extent_clause ............................................................................................................................ 8-2
constraint ................................................................................................................................................... 8-4
deallocate_unused_clause .................................................................................................................... 8-27
file_specification ................................................................................................................................... 8-29
logging_clause ........................................................................................................................................ 8-38
parallel_clause........................................................................................................................................ 8-41
physical_attributes_clause .................................................................................................................. 8-44
size_clause .............................................................................................................................................. 8-47
storage_clause ........................................................................................................................................ 8-48

9

SQL Queries and Subqueries
About Queries and Subqueries ............................................................................................................ 9-1
Creating Simple Queries ........................................................................................................................ 9-2
Hierarchical Queries ............................................................................................................................... 9-3

xv

Hierarchical Query Examples .......................................................................................................... 9-5
The UNION [ALL], INTERSECT, MINUS Operators ...................................................................... 9-8
Sorting Query Results ......................................................................................................................... 9-10
Joins ......................................................................................................................................................... 9-11
Join Conditions ............................................................................................................................... 9-11
Equijoins .......................................................................................................................................... 9-11
Self Joins .......................................................................................................................................... 9-12
Cartesian Products ......................................................................................................................... 9-12
Inner Joins ....................................................................................................................................... 9-12
Outer Joins ....................................................................................................................................... 9-12
Antijoins .......................................................................................................................................... 9-14
Semijoins .......................................................................................................................................... 9-14
Using Subqueries ................................................................................................................................. 9-14
Unnesting of Nested Subqueries ...................................................................................................... 9-15
Selecting from the DUAL Table ........................................................................................................ 9-16
Distributed Queries ............................................................................................................................. 9-16

10

SQL Statements: ALTER CLUSTER to ALTER JAVA
Types of SQL Statements .................................................................................................................... 10-1
Data Definition Language (DDL) Statements ............................................................................ 10-2
Data Manipulation Language (DML) Statements ..................................................................... 10-2
Transaction Control Statements ................................................................................................... 10-3
Session Control Statements ........................................................................................................... 10-3
System Control Statement ............................................................................................................. 10-3
Embedded SQL Statements .......................................................................................................... 10-3
How the SQL Statement Chapters are Organized ......................................................................... 10-4
ALTER CLUSTER ................................................................................................................................ 10-5
ALTER DATABASE ............................................................................................................................ 10-8
ALTER DATABASE LINK ............................................................................................................... 10-46
ALTER DIMENSION ........................................................................................................................ 10-48
ALTER DISKGROUP ........................................................................................................................ 10-51
ALTER FLASHBACK ARCHIVE .................................................................................................... 10-74
ALTER FUNCTION ........................................................................................................................... 10-77
ALTER INDEX .................................................................................................................................... 10-78
ALTER INDEXTYPE .......................................................................................................................... 10-97
ALTER JAVA .................................................................................................................................... 10-100

11

SQL Statements: ALTER LIBRARY to ALTER SYSTEM
ALTER LIBRARY ..................................................................................................................................
ALTER MATERIALIZED VIEW .......................................................................................................
ALTER MATERIALIZED VIEW LOG ...........................................................................................
ALTER OPERATOR ..........................................................................................................................
ALTER OUTLINE ..............................................................................................................................
ALTER PACKAGE .............................................................................................................................
ALTER PROCEDURE ........................................................................................................................
ALTER PROFILE ................................................................................................................................
ALTER RESOURCE COST ...............................................................................................................

xvi

11-2
11-3
11-18
11-25
11-28
11-30
11-31
11-32
11-35

ALTER ROLE ......................................................................................................................................
ALTER ROLLBACK SEGMENT .....................................................................................................
ALTER SEQUENCE ...........................................................................................................................
ALTER SESSION ...............................................................................................................................
Initialization Parameters and ALTER SESSION.......................................................................
Session Parameters and ALTER SESSION ...............................................................................
ALTER SYSTEM .................................................................................................................................

11-38
11-40
11-43
11-45
11-50
11-51
11-58

12 SQL Statements: ALTER TABLE to ALTER TABLESPACE
ALTER TABLE ...................................................................................................................................... 12-2
ALTER TABLESPACE ....................................................................................................................... 12-90

13

SQL Statements: ALTER TRIGGER to COMMIT
ALTER TRIGGER ................................................................................................................................
ALTER TYPE .........................................................................................................................................
ALTER USER ........................................................................................................................................
ALTER VIEW ......................................................................................................................................
ANALYZE ............................................................................................................................................
ASSOCIATE STATISTICS ..............................................................................................................
AUDIT ..................................................................................................................................................
CALL .....................................................................................................................................................
COMMENT .........................................................................................................................................
COMMIT .............................................................................................................................................

14

SQL Statements: CREATE CLUSTER to CREATE JAVA
CREATE CLUSTER ..............................................................................................................................
CREATE CONTEXT ............................................................................................................................
CREATE CONTROLFILE .................................................................................................................
CREATE DATABASE ........................................................................................................................
CREATE DATABASE LINK ............................................................................................................
CREATE DIMENSION .....................................................................................................................
CREATE DIRECTORY ......................................................................................................................
CREATE DISKGROUP .....................................................................................................................
CREATE EDITION ............................................................................................................................
CREATE FLASHBACK ARCHIVE .................................................................................................
CREATE FUNCTION ........................................................................................................................
CREATE INDEX .................................................................................................................................
CREATE INDEXTYPE .......................................................................................................................
CREATE JAVA ....................................................................................................................................

15

13-2
13-4
13-6
13-14
13-17
13-25
13-29
13-42
13-46
13-49

14-2
14-9
14-12
14-19
14-31
14-36
14-41
14-43
14-51
14-55
14-58
14-60
14-87
14-91

SQL Statements: CREATE LIBRARY to CREATE SPFILE
CREATE LIBRARY ..............................................................................................................................
CREATE MATERIALIZED VIEW ....................................................................................................
CREATE MATERIALIZED VIEW LOG ........................................................................................
CREATE OPERATOR .......................................................................................................................

15-2
15-4
15-27
15-35

xvii

CREATE OUTLINE ............................................................................................................................
CREATE PACKAGE ..........................................................................................................................
CREATE PACKAGE BODY .............................................................................................................
CREATE PFILE ...................................................................................................................................
CREATE PROCEDURE .....................................................................................................................
CREATE PROFILE .............................................................................................................................
CREATE RESTORE POINT .............................................................................................................
CREATE ROLE ...................................................................................................................................
CREATE ROLLBACK SEGMENT ..................................................................................................
CREATE SCHEMA ............................................................................................................................
CREATE SEQUENCE ........................................................................................................................
CREATE SPFILE .................................................................................................................................

16

SQL Statements: CREATE SYNONYM to CREATE TRIGGER
CREATE SYNONYM ...........................................................................................................................
CREATE TABLE ...................................................................................................................................
CREATE TABLESPACE ....................................................................................................................
CREATE TRIGGER ...........................................................................................................................

17

16-2
16-6
16-83
16-98

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT
CREATE TYPE ......................................................................................................................................
CREATE TYPE BODY .........................................................................................................................
CREATE USER ......................................................................................................................................
CREATE VIEW ...................................................................................................................................
DELETE ................................................................................................................................................
DISASSOCIATE STATISTICS .......................................................................................................
DROP CLUSTER ................................................................................................................................
DROP CONTEXT ...............................................................................................................................
DROP DATABASE ............................................................................................................................
DROP DATABASE LINK .................................................................................................................
DROP DIMENSION ..........................................................................................................................
DROP DIRECTORY ..........................................................................................................................
DROP DISKGROUP .........................................................................................................................
DROP EDITION .................................................................................................................................
DROP FLASHBACK ARCHIVE .....................................................................................................
DROP FUNCTION ............................................................................................................................
DROP INDEX.......................................................................................................................................
DROP INDEXTYPE ...........................................................................................................................
DROP JAVA ........................................................................................................................................
DROP LIBRARY .................................................................................................................................
DROP MATERIALIZED VIEW........................................................................................................
DROP MATERIALIZED VIEW LOG ............................................................................................
DROP OPERATOR ............................................................................................................................
DROP OUTLINE ................................................................................................................................
DROP PACKAGE ..............................................................................................................................
DROP PROCEDURE .........................................................................................................................
DROP PROFILE ..................................................................................................................................

xviii

15-38
15-42
15-44
15-46
15-48
15-50
15-56
15-59
15-62
15-65
15-67
15-71

17-3
17-5
17-7
17-14
17-26
17-34
17-36
17-38
17-39
17-40
17-41
17-42
17-43
17-45
17-47
17-48
17-50
17-52
17-53
17-54
17-55
17-57
17-59
17-60
17-62
17-64
17-65

DROP RESTORE POINT ................................................................................................................. 17-66
DROP ROLE ........................................................................................................................................ 17-67
DROP ROLLBACK SEGMENT ...................................................................................................... 17-68

18 SQL Statements: DROP SEQUENCE to ROLLBACK
DROP SEQUENCE ..............................................................................................................................
DROP SYNONYM ...............................................................................................................................
DROP TABLE ........................................................................................................................................
DROP TABLESPACE ..........................................................................................................................
DROP TRIGGER.................................................................................................................................
DROP TYPE .........................................................................................................................................
DROP TYPE BODY ............................................................................................................................
DROP USER ........................................................................................................................................
DROP VIEW ........................................................................................................................................
EXPLAIN PLAN ..................................................................................................................................
FLASHBACK DATABASE................................................................................................................
FLASHBACK TABLE ........................................................................................................................
GRANT .................................................................................................................................................
INSERT .................................................................................................................................................
LOCK TABLE ......................................................................................................................................
MERGE .................................................................................................................................................
NOAUDIT ............................................................................................................................................
PURGE ..................................................................................................................................................
RENAME ..............................................................................................................................................
REVOKE ...............................................................................................................................................
ROLLBACK .........................................................................................................................................

18-2
18-3
18-5
18-9
18-12
18-13
18-15
18-16
18-18
18-20
18-24
18-27
18-33
18-54
18-71
18-74
18-79
18-83
18-85
18-87
18-96

19 SQL Statements: SAVEPOINT to UPDATE
SAVEPOINT .........................................................................................................................................
SELECT ...................................................................................................................................................
SET CONSTRAINT[S] ......................................................................................................................
SET ROLE ............................................................................................................................................
SET TRANSACTION ........................................................................................................................
TRUNCATE CLUSTER .....................................................................................................................
TRUNCATE TABLE ..........................................................................................................................
UPDATE ...............................................................................................................................................

A

19-2
19-4
19-59
19-61
19-64
19-67
19-69
19-73

How to Read Syntax Diagrams
Graphic Syntax Diagrams......................................................................................................................
Required Keywords and Parameters ............................................................................................
Optional Keywords and Parameters .............................................................................................
Syntax Loops......................................................................................................................................
Multipart Diagrams .........................................................................................................................
Database Objects ..............................................................................................................................

A-1
A-2
A-3
A-3
A-4
A-4

xix

B Automatic and Manual Locking Mechanisms During SQL Operations
Automatic Locks in DML Operations .................................................................................................
Automatic Locks in DDL Operations..................................................................................................
Exclusive DDL Locks........................................................................................................................
Share DDL Locks...............................................................................................................................
Breakable Parse Locks ......................................................................................................................
Manual Data Locking .............................................................................................................................

C

Oracle and Standard SQL
ANSI Standards ......................................................................................................................................
ISO Standards ..........................................................................................................................................
Oracle Compliance To Core SQL:2008 ................................................................................................
Oracle Support for Optional Features of SQL/Foundation:2008....................................................
Oracle Compliance with SQL/CLI:2008 ............................................................................................
Oracle Compliance with SQL/PSM:2008 ..........................................................................................
Oracle Compliance with SQL/MED:2008 .........................................................................................
Oracle Compliance with SQL/OLB:2008...........................................................................................
Oracle Compliance with SQL/JRT:2008 ............................................................................................
Oracle Compliance with SQL/XML:2008 ..........................................................................................
Oracle Compliance with FIPS 127-2 .................................................................................................
Oracle Extensions to Standard SQL ..................................................................................................
Oracle Compliance with Older Standards .......................................................................................
Character Set Support...........................................................................................................................

D

B-1
B-4
B-4
B-4
B-5
B-5

C-1
C-2
C-3
C-9
C-20
C-20
C-20
C-21
C-21
C-21
C-25
C-26
C-27
C-27

Oracle Regular Expression Support
Multilingual Regular Expression Syntax .......................................................................................... D-1
Regular Expression Operator Multilingual Enhancements............................................................ D-2
Perl-influenced Extensions in Oracle Regular Expressions ........................................................... D-3

E

Oracle SQL Reserved Words and Keywords
Oracle SQL Reserved Words................................................................................................................. E-1
Oracle SQL Keywords ............................................................................................................................ E-3

F

Extended Examples
Using Extensible Indexing ................................................................................................................... F-1
Using XML in SQL Statements ............................................................................................................ F-8

Index

xx

Preface
This reference contains a complete description of the Structured Query Language
(SQL) used to manage information in an Oracle Database. Oracle SQL is a superset of
the American National Standards Institute (ANSI) and the International Organization
for Standardization (ISO) SQL:1999 standard.
This Preface contains these topics:
■

Audience

■

Documentation Accessibility

■

Related Documents

■

Conventions

Audience
The Oracle Database SQL Language Reference is intended for all users of Oracle SQL.

Documentation Accessibility
For information about Oracle's commitment to accessibility, visit the Oracle
Accessibility Program website at
http://www.oracle.com/pls/topic/lookup?ctx=acc&id=docacc.
Access to Oracle Support
Oracle customers that have purchased support have access to electronic support
through My Oracle Support. For information, visit
http://www.oracle.com/pls/topic/lookup?ctx=acc&id=info or visit
http://www.oracle.com/pls/topic/lookup?ctx=acc&id=trs if you are hearing
impaired.

Related Documents
For more information, see these Oracle resources:
■

■

Oracle Database PL/SQL Language Reference for information on PL/SQL, the
procedural language extension to Oracle SQL
Pro*C/C++ Programmer's Guide, Oracle SQL*Module for Ada Programmer's Guide, and
the Pro*COBOL Programmer's Guide for detailed descriptions of Oracle embedded
SQL

Many of the examples in this book use the sample schemas, which are installed by
default when you select the Basic Installation option with an Oracle Database
xxi

installation. Refer to Oracle Database Sample Schemas for information on how these
schemas were created and how you can use them yourself.

Conventions
The following text conventions are used in this document:

xxii

Convention

Meaning

boldface

Boldface type indicates graphical user interface elements associated
with an action, or terms defined in text or the glossary.

italic

Italic type indicates book titles, emphasis, or placeholder variables for
which you supply particular values.

monospace

Monospace type indicates commands within a paragraph, URLs, code
in examples, text that appears on the screen, or text that you enter.

What's New in the SQL Language Reference?
This section describes new features of Oracle Database 11g and provides pointers to
additional information.
For information on features that were new in earlier versions of Oracle Database, refer
to the documentation for the earlier release.

Oracle Database 11g Release 2 (11.2.0.4) New Features in the SQL
Language Reference
The following are new features in this release:
■

■

■

You can now instruct the database to optimize the storage of data in history tables.
See the [NO] OPTIMIZE DATA clause of CREATE FLASHBACK ARCHIVE on page 14-57
and the [NO] OPTIMIZE DATA clause of ALTER FLASHBACK ARCHIVE on page 10-76.
The function SYS_CONTEXT on page 5-279 enables you to query a new built-in
namespace, SYS_SESSION_ROLES, which allows you to determine if a specified role
is currently enabled for the session.
The new system privilege EXEMPT REDACTION POLICY allows you to bypass any
existing Oracle Data Redaction policies. See Table 18–1 on page 18-40.

Oracle Database 11g Release 2 (11.2.0.2) New Features in the SQL
Language Reference
The following top-level SQL statements are enhanced in this release:
■

■

CREATE TABLE and ALTER TABLE now support the clause deferred_segment_creation on
page 16-32 for partitions and subpartitions. This lets you postpone creation of a
segment until the first row of data is inserted into the partition or subpartition.
TRUNCATE TABLE has a new clause DROP ALL STORAGE on page 19-72 that lets
you deallocate all segments for a table. ALTER TABLE has a new clause DROP ALL
STORAGE on page 12-67 that lets you deallocate all segments for a partition or
subpartition.

Oracle Database 11g Release 2 (11.2.0.1) New Features in the SQL
Language Reference
Structural Changes in the SQL Language Reference
A number of sections of this book that were made up primarily of PL/SQL were
moved to Oracle Database PL/SQL Language Reference in Oracle Database 11g Release 1.
xxiii

Refer to "Structural Changes in the SQL Language Reference" on page xxvi for details
on this migration of material.
New Features in the SQL Language Reference
The following top-level SQL statements are new or enhanced in this release:
■

■

■

■

A new top-level SQL statement ALTER DATABASE LINK on page 10-46 lets you
update the fixed user password in a database link when the password of a
connection or authentication user has changed.
The ALTER DISKGROUP statement has the following changes:
–

A new disk_region_clause on page 10-65 lets you determine the Intelligent Data
Placement attribute of the disk group file.

–

New diskgroup_volume_clauses on page 10-67 let you manipulate logical Oracle
ASM Dynamic Volume Manager (Oracle ADVM) volumes corresponding to
physical volume devices.

–

Several new clauses let you control access to Oracle ASM files: usergroup_
clauses on page 10-69, user_clauses on page 10-69, file_permissions_clause on
page 10-70, and file_owner_clause on page 10-70.

AUDIT and NOAUDIT contain a new clause ALL STATEMENTS on page 13-32 that
lets you enable and disable auditing of all top-level SQL statements executed. In
AUDIT only, a new clause IN SESSION CURRENT on page 13-33 allows you to
limit auditing to the current session.
CREATE DISKGROUP and ALTER DISKGROUP have the following changes:
–

■

■

■

■

■

■

Two new statements, CREATE EDITION on page 14-51 and DROP EDITION on
page 17-45, let you use editions. An edition makes it possible to have two or more
versions of the same editionable objects in the database.
CREATE INDEXTYPE on page 14-87 and ALTER INDEXTYPE on page 10-97 have
a new clause WITH SYSTEM MANAGED STORAGE TABLES. This clause makes it possible
to create domain indexes in both range- and list-partitioned tables.
ALTER SESSION contains a new clause SYNC WITH PRIMARY on page 11-48 that
lets you synchronize the physical standby database with the primary database. A
new session parameter STANDBY_MAX_DATA_DELAY on page 11-53 lets you
specify a session-specific apply lag tolerance for queries to a physical standby
database that is in real-time query mode.
CREATE MATERIALIZED VIEW LOG has a new clause COMMIT SCN on page 15-31 that
instructs the database to use commit SCN data instead of timestamps to refresh
the materialized view, which improves the speed of the refresh.
CREATE MATERIALIZED VIEW LOG and ALTER MATERIALIZED VIEW LOG have a new
clause mv_log_purge_clause on page 15-33 that lets you specify the purge time for
the materialized view log.
CREATE TABLE and ALTER TABLE are enhanced in the following ways:
–

xxiv

A new clause QUORUM | REGULAR on page 14-45 let you designate a disk
or failure group as a quorum disk or failure group, which can contain the
voting file for Cluster Synchronization Services (CSS).

A new clause deferred_segment_creation on page 16-32 lets you postpone
creation of the table segment until the first row of data is inserted into the
table. This clause is also applicable to materialized views.

■

■

–

The clause table_compression on page 16-34 has new syntax and terminology.
Use COMPRESS FOR OLTP to specify OLTP table compression. (In earlier releases,
the syntax was COMPRESS FOR ALL OPERATIONS.) Use COMPRESS BASIC to specify
basic table compression. (In earlier releases, the syntax was COMPRESS FOR
DIRECT_LOAD OPERATIONS and this type of compression was called DSS table
compression.)

–

The nested_table_col_properties on page 16-48 provides a LOCAL keyword to
equipartition a nested table with partitioned base table. This is the default
behavior in this release. The default in earlier releases was not to equipartition
the nested table with the partitioned base table. Now you must specify the
GLOBAL keyword to store an unpartitioned nested table with a partitioned base
table.

The CREATE VIEW has a new keyword EDITIONING on page 17-17 that lets you
create an editioning view.
The statement GRANT on page 18-33 has a new EXECUTE object privilege on
directory objects. The ORACLE_LOADER access driver for external tables references
this privilege when deciding whether to execute a preprocessor program.

The following built-in functions are new or enhanced in this release:
■

■

■

For a specified measure, the function LISTAGG on page 5-136 orders data within
each group specified in an ORDER BY clause and then concatenates the values of the
measure column.
A new NTH_VALUE on page 5-167 function returns the value of a measure in a
specified row of a window of data.
Three new functions are useful when you are changing the time zone data file for
your database: ORA_DST_AFFECTED on page 5-175, ORA_DST_CONVERT on
page 5-176, and ORA_DST_ERROR on page 5-177.

The following miscellaneous features are new or enhanced in this release:
■

■

■

■

■

■

Hints, which were introduced in Oracle7, are now superseded by several Oracle
tools, including the SQL Tuning Advisor, SQL plan management, and SQL
Performance Analyzer. See "Hints" on page 3-74 for more information.
Beginning with Oracle Database 11g Release 2 (11.2.0.1), the PARALLEL, PARALLEL_
INDEX, NO_PARALLEL, and NO_PARALLEL_INDEX hints are statement-level hints and
supersede the earlier object-level hints. See "Note on Parallel Hints" on page 3-98.
A new APPEND_VALUES Hint on page 3-80 lets you use direct-path INSERT with
the VALUES clause.
When specifying a redo log file, you can use the new keyword BLOCKSIZE Clause
on page 8-34 to override the operating system-dependent sector size.
The LOB_compression_clause on page 16-46 now has a new LOW setting, which
results in significantly higher decompression and compression speeds, at the cost
of a slightly lower compression ratio.
The subquery_factoring_clause on page 19-13 now supports recursive subquery
factoring (recursive WITH), which lets you query hierarchical data. This feature is
more powerful than CONNECT BY in that it provides depth-first search and
breadth-first search, and supports multiple recursive branches. A new search_clause
on page 19-14 and cycle_clause on page 19-14 let you specify an ordering for the
rows and mark cycles in the recursion.

xxv

Oracle Database 11g Release 1 New Features in the SQL Language
Reference
Structural Changes in the SQL Language Reference
A number of SQL statements are constructed almost entirely of PL/SQL elements.
Those statements continue to appear in this reference, but the bulk of their syntax and
semantics has been moved to Oracle Database PL/SQL Language Reference. The following
table contains links to both the abbreviated SQL syntax and semantics in this book and
to the full syntax and semantics in Oracle Database PL/SQL Language Reference.
Abbreviated SQL Section

Full Syntax and Semantics

CREATE FUNCTION on page 14-58

CREATE FUNCTION

CREATE PACKAGE on page 15-42

CREATE PACKAGE

CREATE PACKAGE BODY on page 15-44

CREATE PACKAGE BODY

CREATE PROCEDURE on page 15-48

CREATE PROCEDURE

CREATE TRIGGER on page 16-98

CREATE TRIGGER

CREATE TYPE on page 17-3

CREATE TYPE

CREATE TYPE BODY on page 17-5

CREATE TYPE BODY

ALTER FUNCTION on page 10-77

ALTER FUNCTION

ALTER PACKAGE on page 11-30

ALTER PACKAGE

ALTER PROCEDURE on page 11-31

ALTER PROCEDURE

ALTER TRIGGER on page 13-2

ALTER TRIGGER

ALTER TYPE on page 13-4

ALTER TYPE

New Features in the SQL Language Reference
The following top-level SQL statements are new or enhanced in this release:
■

■

xxvi

ALTER DATABASE on page 10-8 has been enhanced as follows:
–

The clause managed_standby_recovery on page 10-22 has been greatly
simplified. A number of subclauses have been deprecated as the database now
handles much of the recovery process automatically.

–

The supplemental_db_logging on page 10-32 contains new syntax that lets you
enable or disable supplemental logging of PL/SQL calls.

–

The standby_database_clauses on page 10-34 have new syntax that lets you
convert a physical standby database into a snapshot standby database or
convert a snapshot standby database into a physical standby database.

–

The clause managed_standby_recovery on page 10-22 has new KEEP IDENTITY
syntax that lets you use the rolling upgrade feature provided by a logical
standby and also revert to the original configuration of a primary database
and a physical standby.

ALTER DISKGROUP on page 10-51 has been enhanced as follows:
–

The check_diskgroup_clause on page 10-63 has simplified syntax for checking
the consistency of disk groups, disks, and files in an Oracle ASM environment.

–

The clause diskgroup_availability on page 10-70 offers new options when
mounting a disk group.

–
■

■

■

■

■

■

■

■

■

New clauses disk_offline_clause on page 10-61 and disk_online_clause on
page 10-62 let you take a disk offline for repair and then bring it back online.

ALTER INDEX on page 10-78 has been enhanced as follows:
–

A new MIGRATE parameter lets you migrate a domain index from
user-managed storage tables to system-managed storage tables.

–

A new INVISIBLE parameter lets you modify an index so that it is invisible to
the optimizer.

–

The "PARAMETERS Clause" on page 10-89 now lets you rebuild an XMLIndex
index as well as a domain index.

ALTER SYSTEM on page 11-58 has been enhanced as follows:
–

New syntax lets you kill a session on another instance in an Oracle Real
Application Clusters (Oracle RAC) environment.

–

New rolling_migration_clauses on page 11-66 let you prepare an Oracle ASM
cluster for migration and return it to normal operation after all nodes have
migrated to the same software version.

ALTER TABLE on page 12-2 has been enhanced as follows:
–

The behavior of the add_column_clause on page 12-42 when you specify a
DEFAULT value has been enhanced for improved performance.

–

The syntax for READ ONLY | READ WRITE on page 12-39 lets you put a
table into read-only mode, to prevent DDL or DML changes during table
maintenance, and then back into read/write mode.

–

The clause add_table_partition on page 12-63 has expanded syntax to let you
add a system partition.

–

The flashback_archive_clause on page 12-39 lets you enable or disable historical
tracking for the table.

–

The add_column_clause on page 12-42 now lets you add a virtual column to a
table.

–

A new clause alter_interval_partitioning on page 12-58 lets you convert a
range-partitioned table to an interval_partitioned table.

–

A new dependent_tables_clause on page 12-74 lets you instruct the database to
cascade various partition maintenance operations on a table to
reference-partitioned child tables.

ALTER TABLESPACE on page 12-90 has new syntax that lets you shrink the space
taken by a temporary tablespace or an individual temp file.
ASSOCIATE STATISTICS on page 13-25 has syntax that lets you specify that the
database should manage storage of statistics collected on a system-managed
domain index.
AUDIT on page 13-29 has new syntax that lets you audit various activities on data
mining models.
CALL on page 13-42 now permits positional, named, and mixed notation in the
argument to the routine being called, if the routine takes any arguments.
COMMENT on page 13-46 has a new MINING MODEL clause lets you provide
descriptive comments for a data mining model.
CREATE DISKGROUP on page 14-43 and ALTER DISKGROUP on page 10-51
have new syntax that lets you set various attributes of a disk group.
xxvii

■

■

■

■

■

■

■

■

xxviii

The new statements CREATE FLASHBACK ARCHIVE on page 14-55, ALTER
FLASHBACK ARCHIVE on page 10-74, and DROP FLASHBACK ARCHIVE on
page 17-47 let you create, modify, and drop flashback data archives, which in turn
let you track historical changes to tables.
CREATE INDEX on page 14-60 has been enhanced as follows:
–

A new local_domain_index_clause on page 14-78 lets you create a locally
partitioned domain index.

–

The index_attributes on page 14-71 have been modified to let you create an
index that is invisible to the optimizer.

–

A new XMLIndex_clause on page 14-78 lets you create an XMLIndex index for
XML data.

CREATE INDEXTYPE on page 14-87 and ALTER INDEXTYPE on page 10-97 let
you specify that domain indexes built on the subject indextypes can be range
partitioned, and will have their storage tables and partition maintenance
operations managed by the database.
CREATE PFILE on page 15-46 has new syntax that lets you create a parameter file
from current system-wide parameter settings.
CREATE RESTORE POINT on page 15-56 has new syntax that lets you create a
restore point for a specified datetime or SCN in the past, and to preserve a
flashback database.
CREATE SPFILE on page 15-71 has new syntax that lets you create a system
parameter file from current system-wide parameter settings.
CREATE TABLE on page 16-6 has been enhanced as follows:
–

The flashback_archive_clause on page 16-66 lets you create the table with
tracking of historical changes enabled

–

The clause system_partitioning on page 16-61 lets you partition the table BY
SYSTEM

–

A new virtual_column_definition on page 16-29 lets you create a virtual column.

–

New syntax for XML storage lets you store XML data in binary XML format.

–

A new clause reference_partitioning on page 16-60 lets you partition a table by
reference to another partitioned table.

–

The LOB_parameters on page 16-44 now include a SECUREFILE parameter,
which lets you specify a new storage for LOBs that is faster, more efficient, and
allows for new features such as LOB compression, encryption, and
deduplication.

–

A new LOB_compression_clause on page 16-46 lets you enable or disable
server-side LOB compression for LOBs using SecureFiles storage.

–

A new LOB_deduplicate_clause on page 16-45 lets you coalesce duplicate data
into a single shared repository, reducing storage consumption and simplifying
storage management for LOBs using SecureFiles storage.

–

The LOB_parameters on page 16-44 now include ENCRYPT and DECRYPT clauses
to enable and disable encryption of LOB columns for LOBs using SecureFiles
storage.

CREATE TABLESPACE on page 16-83 has new syntax which, along with a new
ENCRYPT keyword in the storage_clause on page 8-48, lets you encrypt an entire
tablespace.

■

■

■

■
■

DROP DISKGROUP on page 17-43 has a new FORCE keyword that lets you drop a
disk group that can no longer be mounted by an Oracle ASM instance.
GRANT on page 18-33 contains several new system and object privileges that
enable the grantee to work with data mining models.
LOCK TABLE on page 18-71 has new syntax that lets you specify the maximum
number of seconds the statement should wait to obtain a DML lock on the table.
MERGE on page 18-74 now supports operations on tables with domain indexes.
SELECT on page 19-4 has new PIVOT syntax that lets you rotate rows into
columns. A new UNPIVOT operation lets you query data to rotate columns into
rows.

The following SQL built-in functions have been added or enhanced:
■

■

■

■

■

■

■

■

CUBE_TABLE on page 5-67 is a new built-in function that extracts data from a
cube or dimension and returns it in the two-dimensional format of a relational
table.
INSERTXMLAFTER on page 5-118 let you add one or more nodes of any kind
immediately after a target node that is not an attribute node.
REGEXP_INSTR on page 5-218 and REGEXP_SUBSTR on page 5-224 now have an
optional subexpr parameter that lets you target a particular substring of the
regular expression being evaluated.
REGEXP_COUNT on page 5-216 is a new built-in function that counts the number
of occurrences of a specified regular expression pattern in a source string.
PREDICTION on page 5-191, PREDICTION_COST on page 5-195, and
PREDICTION_SET on page 5-201 have been enhanced. New syntax let you specify
that the stored cost matrix should be used only if it is available, or to specify a cost
matrix inline.
PREDICTION_BOUNDS on page 5-193 is a new function that returns the lower
and upper confidence bounds for a prediction.
XMLCAST on page 5-351 and XMLEXISTS on page 5-361 are two new functions
that let you cast XML data to SQL scalar data types and determine whether an
XQuery expression returns a nonempty XQuery sequence, respectively.
XMLDIFF on page 5-356 and XMLPATCH on page 5-365 are two new functions
that provide SQL interfaces to the corresponding XMLDiff and XMLPatch C APIs.
They let you compare two XMLType documents and use the diff file to patch an
XMLType document.

The following miscellaneous changes have been made:
■

■

■

In earlier releases, one form of expression in Chapter 6, "Expressions" was the
variable expression. This form has been renamed to placeholder expression for
consistency with other books in the documentation set. See "Placeholder
Expressions" on page 6-14.
In earlier releases, the TRUNCATE statement was presented as a single statement
with separate syntactic branches for TABLE and CLUSTER. That command has now
been divided into TRUNCATE CLUSTER on page 19-67 and TRUNCATE TABLE
on page 19-69 for consistency with other top-level SQL statements. No actual
syntax or semantic changes have occurred.
Two new hints, "RESULT_CACHE Hint" on page 3-105 and "NO_RESULT_
CACHE Hint" on page 3-95, let you override settings of the RESULT_CACHE_MODE
initialization parameter.
xxix

■

■

■

xxx

"Function Expressions" on page 6-10 now permit positional, named, and mixed
notation in the argument to a user-defined function being used as an expression.
The index_partition_description syntax of ALTER TABLE on page 12-2 and
ALTER INDEX on page 10-78 now lets you specify parameters for a partition of a
domain index.
A new object type object type is supported with Oracle Multimedia. See Media
Types on page 3-35

1
1

Introduction to Oracle SQL

Structured Query Language (SQL) is the set of statements with which all programs
and users access data in an Oracle Database. Application programs and Oracle tools
often allow users access to the database without using SQL directly, but these
applications in turn must use SQL when executing the user's request. This chapter
provides background information on SQL as used by most database systems.
This chapter contains these topics:
■

History of SQL

■

SQL Standards

■

Lexical Conventions

■

Tools Support

History of SQL
Dr. E. F. Codd published the paper, "A Relational Model of Data for Large Shared Data
Banks", in June 1970 in the Association of Computer Machinery (ACM) journal,
Communications of the ACM. Codd's model is now accepted as the definitive model for
relational database management systems (RDBMS). The language, Structured English
Query Language (SEQUEL) was developed by IBM Corporation, Inc., to use Codd's
model. SEQUEL later became SQL (still pronounced "sequel"). In 1979, Relational
Software, Inc. (now Oracle) introduced the first commercially available
implementation of SQL. Today, SQL is accepted as the standard RDBMS language.

SQL Standards
Oracle strives to comply with industry-accepted standards and participates actively in
SQL standards committees. Industry-accepted committees are the American National
Standards Institute (ANSI) and the International Organization for Standardization
(ISO), which is affiliated with the International Electrotechnical Commission (IEC).
Both ANSI and the ISO/IEC have accepted SQL as the standard language for
relational databases. When a new SQL standard is simultaneously published by these
organizations, the names of the standards conform to conventions used by the
organization, but the standards are technically identical.
The latest SQL standard was adopted in July 2008 and is often called SQL:2008. The
formal names of this standard are:
■

ANSI/ISO/IEC 9075:2008, "Database Language SQL", Parts 1
("SQL/Framework"), 2 ("SQL/Foundation"), 3 ("SQL/CLI"), 4 ("SQL/PSM"), 9

Introduction to Oracle SQL 1-1

SQL Standards

("SQL/MED"), 10 ("SQL/OLB"), 11("SQL/Schemata"), 13 ("SQL/JRT"), and
ANSI/ISO/IEC 9075-14:2008, "Database Language SQL", Part 14 ("SQL/XML")
■

ISO/IEC 9075:2008, "Database Language SQL", Parts 1 ("SQL/Framework"), 2
("SQL/Foundation"), 3 ("SQL/CLI"), 4 ("SQL/PSM"), 9 ("SQL/MED"), 10
("SQL/OLB"), 11("SQL/Schemata"), 13 ("SQL/JRT"), and ISO/IEC 9075-14:2008,
"Database Language SQL", Part 14 ("SQL/XML")
See Also: Appendix C, "Oracle and Standard SQL" for a detailed
description of Oracle Database conformance to the SQL:2008
standards

How SQL Works
The strengths of SQL provide benefits for all types of users, including application
programmers, database administrators, managers, and end users. Technically
speaking, SQL is a data sublanguage. The purpose of SQL is to provide an interface to
a relational database such as Oracle Database, and all SQL statements are instructions
to the database. In this SQL differs from general-purpose programming languages like
C and BASIC. Among the features of SQL are the following:
■

It processes sets of data as groups rather than as individual units.

■

It provides automatic navigation to the data.

■

It uses statements that are complex and powerful individually, and that therefore
stand alone. Flow-control statements were not part of SQL originally, but they are
found in the recently accepted optional part of SQL, ISO/IEC 9075-5: 1996.
Flow-control statements are commonly known as "persistent stored modules"
(PSM), and the PL/SQL extension to Oracle SQL is similar to PSM.

SQL lets you work with data at the logical level. You need to be concerned with the
implementation details only when you want to manipulate the data. For example, to
retrieve a set of rows from a table, you define a condition used to filter the rows. All
rows satisfying the condition are retrieved in a single step and can be passed as a unit
to the user, to another SQL statement, or to an application. You need not deal with the
rows one by one, nor do you have to worry about how they are physically stored or
retrieved. All SQL statements use the optimizer, a part of Oracle Database that
determines the most efficient means of accessing the specified data. Oracle also
provides techniques that you can use to make the optimizer perform its job better.
SQL provides statements for a variety of tasks, including:
■

Querying data

■

Inserting, updating, and deleting rows in a table

■

Creating, replacing, altering, and dropping objects

■

Controlling access to the database and its objects

■

Guaranteeing database consistency and integrity

SQL unifies all of the preceding tasks in one consistent language.

Common Language for All Relational Databases
All major relational database management systems support SQL, so you can transfer
all skills you have gained with SQL from one database to another. In addition, all
programs written in SQL are portable. They can often be moved from one database to
another with very little modification.

1-2 Oracle Database SQL Language Reference

Tools Support

Using Enterprise Manager
Many of the operations you can accomplish using SQL syntax can be done much more
easily using Enterprise Manager. For more information, see the Oracle Enterprise
Manager documentation set, Oracle Database 2 Day DBA, or any of the Oracle Database
2 Day + books.

Lexical Conventions
The following lexical conventions for issuing SQL statements apply specifically to the
Oracle Database implementation of SQL, but are generally acceptable in other SQL
implementations.
When you issue a SQL statement, you can include one or more tabs, carriage returns,
spaces, or comments anywhere a space occurs within the definition of the statement.
Thus, Oracle Database evaluates the following two statements in the same manner:
SELECT last_name,salary*12,MONTHS_BETWEEN(SYSDATE,hire_date)
FROM employees
WHERE department_id = 30
ORDER BY last_name;
SELECT last_name,
salary * 12,
MONTHS_BETWEEN( SYSDATE, hire_date )
FROM employees
WHERE department_id=30
ORDER BY last_name;

Case is insignificant in reserved words, keywords, identifiers, and parameters.
However, case is significant in text literals and quoted names. Refer to "Text Literals"
on page 3-45 for a syntax description of text literals.
SQL statements are terminated differently in different
programming environments. This documentation set uses the default
SQL*Plus character, the semicolon (;).

Note:

Tools Support
Oracle provides a number of utilities to facilitate your SQL development process:
■

■

■

Oracle SQL Developer is a graphical tool that lets you browse, create, edit, and
delete (drop) database objects, edit and debug PL/SQL code, run SQL statements
and scripts, manipulate and export data, and create and view reports. With SQL
Developer, you can connect to any target Oracle Database schema using standard
Oracle Database authentication. Once connected, you can perform operations on
objects in the database. You can also connect to schemas for selected third-party
(non-Oracle) databases, such as MySQL, Microsoft SQL Server, and Microsoft
Access, view metadata and data in these databases, and migrate these databases to
Oracle.
SQL*Plus is an interactive and batch query tool that is installed with every Oracle
Database server or client installation. It has a command-line user interface and a
Web-based user interface called iSQL*Plus.
Oracle JDeveloper is a multiple-platform integrated development environment
supporting the complete lifecycle of development for Java, Web services, and SQL.
It provides a graphical interface for executing and tuning SQL statements and a
Introduction to Oracle SQL 1-3

Tools Support

visual schema diagrammer (database modeler). It also supports editing,
compiling, and debugging PL/SQL applications.
■

Oracle Application Express is a hosted environment for developing and deploying
database-related Web applications. SQL Workshop is a component of Oracle
Application Express that lets you view and manage database objects from a Web
browser. SQL Workshop offers quick access to a SQL command processor and a
SQL script repository.
SQL*Plus User's Guide and Reference and Oracle Application
Express Application Builder User's Guide for more information on these
products

See Also:

The Oracle Call Interface and Oracle precompilers let you embed standard SQL
statements within a procedure programming language.
■
■

The Oracle Call Interface (OCI) lets you embed SQL statements in C programs.
The Oracle precompilers, Pro*C/C++ and Pro*COBOL, interpret embedded SQL
statements and translate them into statements that can be understood by C/C++
and COBOL compilers, respectively.
See Also: Oracle C++ Call Interface Programmer's Guide, Pro*COBOL
Programmer's Guide, and Oracle Call Interface Programmer's Guide for
additional information on the embedded SQL statements allowed in
each product

Most (but not all) Oracle tools also support all features of Oracle SQL. This reference
describes the complete functionality of SQL. If the Oracle tool that you are using does
not support this complete functionality, then you can find a discussion of the
restrictions in the manual describing the tool, such as SQL*Plus User's Guide and
Reference.

1-4 Oracle Database SQL Language Reference

2
2

Pseudocolumns

A pseudocolumn behaves like a table column, but is not actually stored in the table.
You can select from pseudocolumns, but you cannot insert, update, or delete their
values. A pseudocolumn is also similar to a function without arguments (refer to
Chapter 5, "Functions"). However, functions without arguments typically return the
same value for every row in the result set, whereas pseudocolumns typically return a
different value for each row.
This chapter contains the following sections:
■

Hierarchical Query Pseudocolumns

■

Sequence Pseudocolumns

■

Version Query Pseudocolumns

■

COLUMN_VALUE Pseudocolumn

■

OBJECT_ID Pseudocolumn

■

OBJECT_VALUE Pseudocolumn

■

ORA_ROWSCN Pseudocolumn

■

ROWID Pseudocolumn

■

ROWNUM Pseudocolumn

■

XMLDATA Pseudocolumn

Hierarchical Query Pseudocolumns
The hierarchical query pseudocolumns are valid only in hierarchical queries. The
hierarchical query pseudocolumns are:
■

CONNECT_BY_ISCYCLE Pseudocolumn

■

CONNECT_BY_ISLEAF Pseudocolumn

■

LEVEL Pseudocolumn

To define a hierarchical relationship in a query, you must use the CONNECT BY clause.

CONNECT_BY_ISCYCLE Pseudocolumn
The CONNECT_BY_ISCYCLE pseudocolumn returns 1 if the current row has a child which
is also its ancestor. Otherwise it returns 0.

Pseudocolumns

2-1

Hierarchical Query Pseudocolumns

You can specify CONNECT_BY_ISCYCLE only if you have specified the NOCYCLE parameter
of the CONNECT BY clause. NOCYCLE enables Oracle to return the results of a query that
would otherwise fail because of a CONNECT BY loop in the data.
See Also: "Hierarchical Queries" on page 9-3 for more information
about the NOCYCLE parameter and "Hierarchical Query Examples" on
page 9-5 for an example that uses the CONNECT_BY_ISCYCLE
pseudocolumn

CONNECT_BY_ISLEAF Pseudocolumn
The CONNECT_BY_ISLEAF pseudocolumn returns 1 if the current row is a leaf of the tree
defined by the CONNECT BY condition. Otherwise it returns 0. This information indicates
whether a given row can be further expanded to show more of the hierarchy.
CONNECT_BY_ISLEAF Example The following example shows the first three levels
of the hr.employees table, indicating for each row whether it is a leaf row (indicated
by 1 in the IsLeaf column) or whether it has child rows (indicated by 0 in the IsLeaf
column):
SELECT last_name "Employee", CONNECT_BY_ISLEAF "IsLeaf",
LEVEL, SYS_CONNECT_BY_PATH(last_name, '/') "Path"
FROM employees
WHERE LEVEL <= 3 AND department_id = 80
START WITH employee_id = 100
CONNECT BY PRIOR employee_id = manager_id AND LEVEL <= 4
ORDER BY "Employee", "IsLeaf";
Employee
IsLeaf
LEVEL Path
------------------------- ---------- ---------- ------------------------Abel
1
3 /King/Zlotkey/Abel
Ande
1
3 /King/Errazuriz/Ande
Banda
1
3 /King/Errazuriz/Banda
Bates
1
3 /King/Cambrault/Bates
Bernstein
1
3 /King/Russell/Bernstein
Bloom
1
3 /King/Cambrault/Bloom
Cambrault
0
2 /King/Cambrault
Cambrault
1
3 /King/Russell/Cambrault
Doran
1
3 /King/Partners/Doran
Errazuriz
0
2 /King/Errazuriz
Fox
1
3 /King/Cambrault/Fox
. . .

See Also: "Hierarchical Queries" on page 9-3 and SYS_CONNECT_
BY_PATH on page 5-278

LEVEL Pseudocolumn
For each row returned by a hierarchical query, the LEVEL pseudocolumn returns 1 for a
root row, 2 for a child of a root, and so on. A root row is the highest row within an
inverted tree. A child row is any nonroot row. A parent row is any row that has
children. A leaf row is any row without children. Figure 2–1 shows the nodes of an
inverted tree with their LEVEL values.

2-2 Oracle Database SQL Language Reference

Sequence Pseudocolumns

Figure 2–1 Hierarchical Tree
root/
parent

Level 1

parent/
child

Level 2

Level 3

parent/
child

parent/
child

child/
leaf

child/
leaf

Level 4

child/
leaf

parent/
child

child/
leaf

child/
leaf

See Also: "Hierarchical Queries" on page 9-3 for information on
hierarchical queries in general and "IN Condition" on page 7-23 for
restrictions on using the LEVEL pseudocolumn

Sequence Pseudocolumns
A sequence is a schema object that can generate unique sequential values. These
values are often used for primary and unique keys. You can refer to sequence values in
SQL statements with these pseudocolumns:
■

CURRVAL: Returns the current value of a sequence

■

NEXTVAL: Increments the sequence and returns the next value

You must qualify CURRVAL and NEXTVAL with the name of the sequence:
sequence.CURRVAL
sequence.NEXTVAL

To refer to the current or next value of a sequence in the schema of another user, you
must have been granted either SELECT object privilege on the sequence or SELECT ANY
SEQUENCE system privilege, and you must qualify the sequence with the schema
containing it:
schema.sequence.CURRVAL
schema.sequence.NEXTVAL

To refer to the value of a sequence on a remote database, you must qualify the
sequence with a complete or partial name of a database link:
schema.sequence.CURRVAL@dblink
schema.sequence.NEXTVAL@dblink

A sequence can be accessed by many users concurrently with no waiting or locking.
See Also: "References to Objects in Remote Databases" on page 3-117
for more information on referring to database links

Where to Use Sequence Values
You can use CURRVAL and NEXTVAL in the following locations:
■

The select list of a SELECT statement that is not contained in a subquery,
materialized view, or view

Pseudocolumns

2-3

Sequence Pseudocolumns

■

The select list of a subquery in an INSERT statement

■

The VALUES clause of an INSERT statement

■

The SET clause of an UPDATE statement

Restrictions on Sequence Values You cannot use CURRVAL and NEXTVAL in the
following constructs:
■

A subquery in a DELETE, SELECT, or UPDATE statement

■

A query of a view or of a materialized view

■

A SELECT statement with the DISTINCT operator

■

A SELECT statement with a GROUP BY clause or ORDER BY clause

■

A SELECT statement that is combined with another SELECT statement with the
UNION, INTERSECT, or MINUS set operator

■

The WHERE clause of a SELECT statement

■

The DEFAULT value of a column in a CREATE TABLE or ALTER TABLE statement

■

The condition of a CHECK constraint

Within a single SQL statement that uses CURRVAL or NEXTVAL, all referenced LONG
columns, updated tables, and locked tables must be located on the same database.

How to Use Sequence Values
When you create a sequence, you can define its initial value and the increment
between its values. The first reference to NEXTVAL returns the initial value of the
sequence. Subsequent references to NEXTVAL increment the sequence value by the
defined increment and return the new value. Any reference to CURRVAL always returns
the current value of the sequence, which is the value returned by the last reference to
NEXTVAL.
Before you use CURRVAL for a sequence in your session, you must first initialize the
sequence with NEXTVAL. Refer to CREATE SEQUENCE on page 15-67 for information
on sequences.
Within a single SQL statement containing a reference to NEXTVAL, Oracle increments
the sequence once:
■

For each row returned by the outer query block of a SELECT statement. Such a
query block can appear in the following places:
–

A top-level SELECT statement

–

An INSERT ... SELECT statement (either single-table or multitable). For a
multitable insert, the reference to NEXTVAL must appear in the VALUES clause,
and the sequence is updated once for each row returned by the subquery, even
though NEXTVAL may be referenced in multiple branches of the multitable
insert.

–

A CREATE TABLE ... AS SELECT statement

–

A CREATE MATERIALIZED VIEW ... AS SELECT statement

■

For each row updated in an UPDATE statement

■

For each INSERT statement containing a VALUES clause

■

For each INSERT ... [ALL | FIRST] statement (multitable insert). A multitable insert
is considered a single SQL statement. Therefore, a reference to the NEXTVAL of a

2-4 Oracle Database SQL Language Reference

Version Query Pseudocolumns

sequence will increase the sequence only once for each input record coming from
the SELECT portion of the statement. If NEXTVAL is specified more than once in any
part of the INSERT ... [ALL | FIRST ] statement, then the value will be the same for
all insert branches, regardless of how often a given record might be inserted.
■

■

For each row merged by a MERGE statement. The reference to NEXTVAL can appear
in the merge_insert_clause or the merge_update_clause or both. The NEXTVALUE
value is incremented for each row updated and for each row inserted, even if the
sequence number is not actually used in the update or insert operation. If NEXTVAL
is specified more than once in any of these locations, then the sequence is
incremented once for each row and returns the same value for all occurrences of
NEXTVAL for that row.
For each input row in a multitable INSERT ALL statement. NEXTVAL is incremented
once for each row returned by the subquery, regardless of how many occurrences
of the insert_into_clause map to each row.

If any of these locations contains more than one reference to NEXTVAL, then Oracle
increments the sequence once and returns the same value for all occurrences of
NEXTVAL.
If any of these locations contains references to both CURRVAL and NEXTVAL, then Oracle
increments the sequence and returns the same value for both CURRVAL and NEXTVAL.
This example selects the next value
of the employee sequence in the sample schema hr:

Finding the next value of a sequence: Example
SELECT employees_seq.nextval
FROM DUAL;

Inserting sequence values into a table: Example This example increments the

employee sequence and uses its value for a new employee inserted into the sample
table hr.employees:
INSERT INTO employees
VALUES (employees_seq.nextval, 'John', 'Doe', 'jdoe', '555-1212',
TO_DATE(SYSDATE), 'PU_CLERK', 2500, null, null, 30);

This example adds a new order
with the next order number to the master order table. It then adds suborders with this
number to the detail order table:

Reusing the current value of a sequence: Example

INSERT INTO orders (order_id, order_date, customer_id)
VALUES (orders_seq.nextval, TO_DATE(SYSDATE), 106);
INSERT INTO order_items (order_id, line_item_id, product_id)
VALUES (orders_seq.currval, 1, 2359);
INSERT INTO order_items (order_id, line_item_id, product_id)
VALUES (orders_seq.currval, 2, 3290);
INSERT INTO order_items (order_id, line_item_id, product_id)
VALUES (orders_seq.currval, 3, 2381);

Version Query Pseudocolumns
The version query pseudocolumns are valid only in Oracle Flashback Version Query,
which is a form of Oracle Flashback Query. The version query pseudocolumns are:

Pseudocolumns

2-5

COLUMN_VALUE Pseudocolumn

■

■

■

■

VERSIONS_STARTSCN and VERSIONS_STARTTIME: Starting System Change Number
(SCN) or TIMESTAMP when the row version was created. This pseudocolumn
identifies the time when the data first had the values reflected in the row version.
Use this pseudocolumn to identify the past target time for Oracle Flashback Table
or Oracle Flashback Query. If this pseudocolumn is NULL, then the row version was
created before start.
VERSIONS_ENDSCN and VERSIONS_ENDTIME: SCN or TIMESTAMP when the row
version expired. If the pseudocolumn is NULL, then either the row version was
current at the time of the query or the row corresponds to a DELETE operation.
VERSIONS_XID: Identifier (a RAW number) of the transaction that created the row
version.
VERSIONS_OPERATION: Operation performed by the transaction: I for insertion, D
for deletion, or U for update. The version is that of the row that was inserted,
deleted, or updated; that is, the row after an INSERT operation, the row before a
DELETE operation, or the row affected by an UPDATE operation.
For user updates of an index key, Oracle Flashback Version Query might treat an
UPDATE operation as two operations, DELETE plus INSERT, represented as two
version rows with a D followed by an I VERSIONS_OPERATION.
See Also:
■

■

flashback_query_clause on page 19-17 for more information on
version queries
Oracle Database Advanced Application Developer's Guide for more
information on using Oracle Flashback Version Query

COLUMN_VALUE Pseudocolumn
When you refer to an XMLTable construct without the COLUMNS clause, or when you use
the TABLE collection expression to refer to a scalar nested table type, the database
returns a virtual table with a single column. This name of this pseudocolumn is
COLUMN_VALUE.
In the context of XMLTable, the value returned is of data type XMLType. For example,
the following two statements are equivalent, and the output for both shows COLUMN_
VALUE as the name of the column being returned:
SELECT *
FROM XMLTABLE('123');
COLUMN_VALUE
--------------------------------------123
SELECT COLUMN_VALUE
FROM (XMLTable('123'));
COLUMN_VALUE
---------------------------------------123

In the context of a TABLE collection expression, the value returned is the data type of
the collection element. The following statements create the two levels of nested tables
illustrated in "Creating a Table: Multilevel Collection Example" on page 16-73 to show
the uses of COLUMN_VALUE in this context:

2-6 Oracle Database SQL Language Reference

OBJECT_ID Pseudocolumn

CREATE TYPE phone AS TABLE OF NUMBER;
/
CREATE TYPE phone_list AS TABLE OF phone;
/

The next statement uses COLUMN_VALUE to select from the phone type:
SELECT t.COLUMN_VALUE
FROM TABLE(phone(1,2,3)) t;
COLUMN_VALUE
-----------1
2
3

In a nested type, you can use the COLUMN_VALUE pseudocolumn in both the select list
and the TABLE collection expression:
SELECT t.COLUMN_VALUE
FROM TABLE(phone_list(phone(1,2,3))) p, TABLE(p.COLUMN_VALUE) t;
COLUMN_VALUE
-----------1
2
3

The keyword COLUMN_VALUE is also the name that Oracle Database generates for the
scalar value of an inner nested table without a column or attribute name, as shown in
the example that follows. In this context, COLUMN_VALUE is not a pseudocolumn, but an
actual column name.
CREATE TABLE my_customers (
cust_id
NUMBER,
name
VARCHAR2(25),
phone_numbers phone_list,
credit_limit NUMBER)
NESTED TABLE phone_numbers STORE AS outer_ntab
(NESTED TABLE COLUMN_VALUE STORE AS inner_ntab);

See Also:
■
■

■

XMLTABLE on page 5-375 for information on that function
table_collection_expression::= on page 18-57 for information on the
TABLE collection expression
ALTER TABLE examples in "Nested Tables: Examples" on page 12-88

OBJECT_ID Pseudocolumn
The OBJECT_ID pseudocolumn returns the object identifier of a column of an object
table or view. Oracle uses this pseudocolumn as the primary key of an object table.
OBJECT_ID is useful in INSTEAD OF triggers on views and for identifying the ID of a
substitutable row in an object table.
Note: In earlier releases, this pseudocolumn was called SYS_NC_OID$.
That name is still supported for backward compatibility. However,
Oracle recommends that you use the more intuitive name OBJECT_ID.

Pseudocolumns

2-7

OBJECT_VALUE Pseudocolumn

See Also: Oracle Database Object-Relational Developer's Guide for
examples of the use of this pseudocolumn

OBJECT_VALUE Pseudocolumn
The OBJECT_VALUE pseudocolumn returns system-generated names for the columns of
an object table, XMLType table, object view, or XMLType view. This pseudocolumn is
useful for identifying the value of a substitutable row in an object table and for
creating object views with the WITH OBJECT IDENTIFIER clause.
In earlier releases, this pseudocolumn was called SYS_NC_
ROWINFO$. That name is still supported for backward compatibility.
However, Oracle recommends that you use the more intuitive name
OBJECT_VALUE.
Note:

See Also:
■

■

object_table on page 16-68 and object_view_clause on page 17-18 for
more information on the use of this pseudocolumn
Oracle Database Object-Relational Developer's Guide for examples of
the use of this pseudocolumn

ORA_ROWSCN Pseudocolumn
ORA_ROWSCN reflects the system change-number (SCN) of the most recent change to a
row. This change can be at the level of a block (coarse) or at the level of a row
(fine-grained). The latter is provided by row-level dependency tracking. Refer to
CREATE TABLE ... NOROWDEPENDENCIES | ROWDEPENDENCIES on page 16-64 for
more information on row-level dependency tracking. In the absence of row-level
dependencies, ORA_ROWSCN reflects block-level dependencies.
Whether at the block level or at the row level, the ORA_ROWSCN should not be
considered to be an exact SCN. For example, if a transaction changed row R in a block
and committed at SCN 10, it is not always true that the ORA_ROWSCN for the row would
return 10. While a value less than 10 would never be returned, any value greater than
or equal to 10 could be returned. That is, the ORA_ROWSCN of a row is not always
guaranteed to be the exact commit SCN of the transaction that last modified that row.
However, with fine-grained ORA_ROWSCN, if two transactions T1 and T2 modified the
same row R, one after another, and committed, a query on the ORA_ROWSCN of row R
after the commit of T1 will return a value lower than the value returned after the
commit of T2. If a block is queried twice, then it is possible for the value of ORA_ROWSCN
to change between the queries even though rows have not been updated in the time
between the queries. The only guarantee is that the value of ORA_ROWSCN in both
queries is greater than the commit SCN of the transaction that last modified that row.
You cannot use the ORA_ROWSCN pseudocolumn in a query to a view. However, you can
use it to refer to the underlying table when creating a view. You can also use this
pseudocolumn in the WHERE clause of an UPDATE or DELETE statement.
ORA_ROWSCN is not supported for Flashback Query. Instead, use the version query
pseudocolumns, which are provided explicitly for Flashback Query. Refer to the
SELECT ... flashback_query_clause on page 19-17 for information on Flashback Query and
"Version Query Pseudocolumns" on page 2-5 for additional information on those
pseudocolumns.

2-8 Oracle Database SQL Language Reference

ROWID Pseudocolumn

Restriction on ORA_ROWSCN: This pseudocolumn is not supported for external
tables.
Example The first statement below uses the ORA_ROWSCN pseudocolumn to get the
system change number of the last operation on the employees table. The second
statement uses the pseudocolumn with the SCN_TO_TIMESTAMP function to determine
the timestamp of the operation:
SELECT ORA_ROWSCN, last_name
FROM employees
WHERE employee_id = 188;
SELECT SCN_TO_TIMESTAMP(ORA_ROWSCN), last_name
FROM employees
WHERE employee_id = 188;

See Also:

SCN_TO_TIMESTAMP on page 5-242

ROWID Pseudocolumn
For each row in the database, the ROWID pseudocolumn returns the address of the row.
Oracle Database rowid values contain information necessary to locate a row:
■

The data object number of the object

■

The data block in the data file in which the row resides

■

The position of the row in the data block (first row is 0)

■

The data file in which the row resides (first file is 1). The file number is relative to
the tablespace.

Usually, a rowid value uniquely identifies a row in the database. However, rows in
different tables that are stored together in the same cluster can have the same rowid.
Values of the ROWID pseudocolumn have the data type ROWID or UROWID. Refer to
"Rowid Data Types" on page 3-27 and "UROWID Data Type" on page 3-28 for more
information.
Rowid values have several important uses:
■

They are the fastest way to access a single row.

■

They can show you how the rows in a table are stored.

■

They are unique identifiers for rows in a table.

You should not use ROWID as the primary key of a table. If you delete and reinsert a
row with the Import and Export utilities, for example, then its rowid may change. If
you delete a row, then Oracle may reassign its rowid to a new row inserted later.
Although you can use the ROWID pseudocolumn in the SELECT and WHERE clause of a
query, these pseudocolumn values are not actually stored in the database. You cannot
insert, update, or delete a value of the ROWID pseudocolumn.
Example This statement selects the address of all rows that contain data for
employees in department 20:
SELECT ROWID, last_name
FROM employees
WHERE department_id = 20;

Pseudocolumns

2-9

ROWNUM Pseudocolumn

ROWNUM Pseudocolumn
Note: The ROW_NUMBER built-in SQL function provides superior
support for ordering the results of a query. Refer to ROW_NUMBER
on page 5-236 for more information.

For each row returned by a query, the ROWNUM pseudocolumn returns a number
indicating the order in which Oracle selects the row from a table or set of joined rows.
The first row selected has a ROWNUM of 1, the second has 2, and so on.
You can use ROWNUM to limit the number of rows returned by a query, as in this
example:
SELECT *
FROM employees
WHERE ROWNUM < 11;

If an ORDER BY clause follows ROWNUM in the same query, then the rows will be reordered
by the ORDER BY clause. The results can vary depending on the way the rows are
accessed. For example, if the ORDER BY clause causes Oracle to use an index to access
the data, then Oracle may retrieve the rows in a different order than without the index.
Therefore, the following statement does not necessarily return the same rows as the
preceding example:
SELECT *
FROM employees
WHERE ROWNUM < 11
ORDER BY last_name;

If you embed the ORDER BY clause in a subquery and place the ROWNUM condition in the
top-level query, then you can force the ROWNUM condition to be applied after the
ordering of the rows. For example, the following query returns the employees with the
10 smallest employee numbers. This is sometimes referred to as top-N reporting:
SELECT *
FROM (SELECT * FROM employees ORDER BY employee_id)
WHERE ROWNUM < 11;

In the preceding example, the ROWNUM values are those of the top-level SELECT
statement, so they are generated after the rows have already been ordered by
employee_id in the subquery.
Conditions testing for ROWNUM values greater than a positive integer are always false.
For example, this query returns no rows:
SELECT *
FROM employees
WHERE ROWNUM > 1;

The first row fetched is assigned a ROWNUM of 1 and makes the condition false. The
second row to be fetched is now the first row and is also assigned a ROWNUM of 1 and
makes the condition false. All rows subsequently fail to satisfy the condition, so no
rows are returned.
You can also use ROWNUM to assign unique values to each row of a table, as in this
example:
UPDATE my_table
SET column1 = ROWNUM;

2-10 Oracle Database SQL Language Reference

XMLDATA Pseudocolumn

Refer to the function ROW_NUMBER on page 5-236 for an alternative method of
assigning unique numbers to rows.
Note:

Using ROWNUM in a query can affect view optimization.

XMLDATA Pseudocolumn
Oracle stores XMLType data either in LOB or object-relational columns, based on
XMLSchema information and how you specify the storage clause. The XMLDATA
pseudocolumn lets you access the underlying LOB or object relational column to
specify additional storage clause parameters, constraints, indexes, and so forth.
Example The following statements illustrate the use of this pseudocolumn. Suppose
you create a simple table of XMLType with one CLOB column:
CREATE TABLE xml_lob_tab of XMLTYPE
XMLTYPE STORE AS CLOB;

To change the storage characteristics of the underlying LOB column, you can use the
following statement:
ALTER TABLE xml_lob_tab
MODIFY LOB (XMLDATA) (STORAGE (MAXSIZE 2G) CACHE);

Now suppose you have created an XMLSchema-based table like the xwarehouses table
created in "Using XML in SQL Statements" on page F-8. You could then use the
XMLDATA column to set the properties of the underlying columns, as shown in the
following statement:
ALTER TABLE xwarehouses
ADD (UNIQUE(XMLDATA."WarehouseId"));

Pseudocolumns 2-11

XMLDATA Pseudocolumn

2-12 Oracle Database SQL Language Reference

3
3

Basic Elements of Oracle SQL

This chapter contains reference information on the basic elements of Oracle SQL.
These elements are the simplest building blocks of SQL statements. Therefore, before
using the statements described in Chapter 10 through Chapter 19, you should
familiarize yourself with the concepts covered in this chapter.
This chapter contains these sections:
■

Data Types

■

Data Type Comparison Rules

■

Literals

■

Format Models

■

Nulls

■

Comments

■

Database Objects

■

Database Object Names and Qualifiers

■

Syntax for Schema Objects and Parts in SQL Statements

Data Types
Each value manipulated by Oracle Database has a data type. The data type of a value
associates a fixed set of properties with the value. These properties cause Oracle to
treat values of one data type differently from values of another. For example, you can
add values of NUMBER data type, but not values of RAW data type.
When you create a table or cluster, you must specify a data type for each of its
columns. When you create a procedure or stored function, you must specify a data
type for each of its arguments. These data types define the domain of values that each
column can contain or each argument can have. For example, DATE columns cannot
accept the value February 29 (except for a leap year) or the values 2 or 'SHOE'. Each
value subsequently placed in a column assumes the data type of the column. For
example, if you insert '01-JAN-98' into a DATE column, then Oracle treats the
'01-JAN-98' character string as a DATE value after verifying that it translates to a valid
date.
Oracle Database provides a number of built-in data types as well as several categories
for user-defined types that can be used as data types. The syntax of Oracle data types
appears in the diagrams that follow. The text of this section is divided into the
following sections:
■

Oracle Built-in Data Types
Basic Elements of Oracle SQL 3-1

Data Types

■

ANSI, DB2, and SQL/DS Data Types

■

User-Defined Types

■

Oracle-Supplied Types

■

Data Type Comparison Rules

■

Data Conversion

A data type is either scalar or nonscalar. A scalar type contains an atomic value,
whereas a nonscalar (sometimes called a "collection") contains a set of values. A large
object (LOB) is a special form of scalar data type representing a large scalar value of
binary or character data. LOBs are subject to some restrictions that do not affect other
scalar types because of their size. Those restrictions are documented in the context of
the relevant SQL syntax.
See Also:

"Restrictions on LOB Columns" on page 3-25

The Oracle precompilers recognize other data types in embedded SQL programs.
These data types are called external data types and are associated with host variables.
Do not confuse built-in data types and user-defined types with external data types. For
information on external data types, including how Oracle converts between them and
built-in data types or user-defined types, see Pro*COBOL Programmer's Guide, and
Pro*C/C++ Programmer's Guide.
datatypes::=
Oracle_built_in_datatypes
ANSI_supported_datatypes
user_defined_types
Oracle_supplied_types

The Oracle built-in data types appear in the figures that follows. For descriptions, refer
to "Oracle Built-in Data Types" on page 3-6.
Oracle_built_in_datatypes::=
character_datatypes
number_datatypes
long_and_raw_datatypes
datetime_datatypes
large_object_datatypes
rowid_datatypes

3-2 Oracle Database SQL Language Reference

Data Types

character_datatypes::=
BYTE
CHAR
(

size

)

CHAR
BYTE
CHAR
VARCHAR2

(

size

(

size

)
)

NCHAR
NVARCHAR2

(

size

)

number_datatypes::=
,
(

scale

precision

)

NUMBER
(

precision

)

FLOAT
BINARY_FLOAT
BINARY_DOUBLE

long_and_raw_datatypes::=
LONG
LONG
RAW

RAW
(

size

)

datetime_datatypes::=
DATE
LOCAL
(

fractional_seconds_precision

)

WITH

TIME

ZONE

TIMESTAMP
(
INTERVAL

YEAR

INTERVAL

DAY

year_precision

)
TO

(

day_precision

MONTH

)

(
TO

fractional_seconds_precision

)

SECOND

Basic Elements of Oracle SQL 3-3

Data Types

large_object_datatypes::=
BLOB
CLOB
NCLOB
BFILE

rowid_datatypes::=
ROWID
(

size

)

UROWID

The ANSI-supported data types appear in the figure that follows. "ANSI, DB2, and
SQL/DS Data Types" on page 3-28 discusses the mapping of ANSI-supported data
types to Oracle built-in data types.
ANSI_supported_datatypes::=
VARYING
CHARACTER

(

size

)

CHAR
VARYING

(

size

)

NCHAR
VARCHAR

(

size

)

CHARACTER

VARYING

NATIONAL

(

size

)

CHAR
,

NUMERIC

(

scale

precision

)

DECIMAL
DEC
INTEGER
INT
SMALLINT
(

size

)

FLOAT
DOUBLE

PRECISION

REAL

For descriptions of user-defined types, refer to "User-Defined Types" on page 3-29.
The Oracle-supplied data types appear in the figures that follows. For descriptions,
refer to "Oracle-Supplied Types" on page 3-31.

3-4 Oracle Database SQL Language Reference

Data Types

Oracle_supplied_types::=
any_types
XML_types
spatial_types
media_types
expression_filter_type

For a description of the expression_filter_type, refer to "Expression Filter Type" on
page 3-36. Other Oracle-supplied types follow:
any_types::=
SYS.AnyData
SYS.AnyType
SYS.AnyDataSet

For descriptions of the Any types, refer to "Any Types" on page 3-31.
XML_types::=
XMLType
URIType

For descriptions of the XML types, refer to "XML Types" on page 3-32.
spatial_types::=
SDO_Geometry
SDO_Topo_Geometry
SDO_GeoRaster

For descriptions of the spatial types, refer to "Spatial Types" on page 3-34.
media_types::=
ORDAudio
ORDImage
ORDVideo
ORDDoc
ORDDicom
still_image_object_types

Basic Elements of Oracle SQL 3-5

Data Types

still_image_object_types::=
SI_StillImage
SI_AverageColor
SI_PositionalColor
SI_ColorHistogram
SI_Texture
SI_FeatureList
SI_Color

For descriptions of the media types, refer to "Media Types" on page 3-35.

Oracle Built-in Data Types
The table that follows summarizes Oracle built-in data types. Refer to the syntax in the
preceding sections for the syntactic elements. The codes listed for the data types are
used internally by Oracle Database. The data type code of a column or object attribute
is returned by the DUMP function.
Table 3–1

Built-in Data Type Summary

Code

Data Type

Description

1

VARCHAR2(size [BYTE | CHAR])

Variable-length character string having maximum length size
bytes or characters. Maximum size is 4000 bytes or characters,
and minimum is 1 byte or 1 character. You must specify size for
VARCHAR2.
BYTE indicates that the column will have byte length semantics.
CHAR indicates that the column will have character semantics.

1

NVARCHAR2(size)

Variable-length Unicode character string having maximum
length size characters. The number of bytes can be up to two
times size for AL16UTF16 encoding and three times size for UTF8
encoding. Maximum size is determined by the national
character set definition, with an upper limit of 4000 bytes. You
must specify size for NVARCHAR2.

2

NUMBER [ (p [, s]) ]

Number having precision p and scale s. The precision p can
range from 1 to 38. The scale s can range from -84 to 127. Both
precision and scale are in decimal digits. A NUMBER value
requires from 1 to 22 bytes.

2

FLOAT [(p)]

A subtype of the NUMBER data type having precision p. A FLOAT
value is represented internally as NUMBER. The precision p can
range from 1 to 126 binary digits. A FLOAT value requires from 1
to 22 bytes.

8

LONG

Character data of variable length up to 2 gigabytes, or 231 -1
bytes. Provided for backward compatibility.

12

DATE

Valid date range from January 1, 4712 BC, to December 31, 9999
AD. The default format is determined explicitly by the NLS_
DATE_FORMAT parameter or implicitly by the NLS_TERRITORY
parameter. The size is fixed at 7 bytes. This data type contains
the datetime fields YEAR, MONTH, DAY, HOUR, MINUTE, and SECOND. It
does not have fractional seconds or a time zone.

100

BINARY_FLOAT

32-bit floating point number. This data type requires 4 bytes.

3-6 Oracle Database SQL Language Reference

Data Types

Table 3–1 (Cont.) Built-in Data Type Summary
Code

Data Type

Description

101

BINARY_DOUBLE

64-bit floating point number. This data type requires 8 bytes.

180

TIMESTAMP [(fractional_seconds_
precision)]

Year, month, and day values of date, as well as hour, minute,
and second values of time, where fractional_seconds_
precision is the number of digits in the fractional part of the
SECOND datetime field. Accepted values of fractional_seconds_
precision are 0 to 9. The default is 6. The default format is
determined explicitly by the NLS_TIMESTAMP_FORMAT parameter
or implicitly by the NLS_TERRITORY parameter. The size is 7 or 11
bytes, depending on the precision. This data type contains the
datetime fields YEAR, MONTH, DAY, HOUR, MINUTE, and SECOND. It
contains fractional seconds but does not have a time zone.

181

TIMESTAMP [(fractional_seconds_
precision)] WITH TIME ZONE

All values of TIMESTAMP as well as time zone displacement value,
where fractional_seconds_precision is the number of digits
in the fractional part of the SECOND datetime field. Accepted
values are 0 to 9. The default is 6. The default format is
determined explicitly by the NLS_TIMESTAMP_FORMAT parameter
or implicitly by the NLS_TERRITORY parameter. The size is fixed
at 13 bytes. This data type contains the datetime fields YEAR,
MONTH, DAY, HOUR, MINUTE, SECOND, TIMEZONE_HOUR, and TIMEZONE_
MINUTE. It has fractional seconds and an explicit time zone.

231

TIMESTAMP [(fractional_seconds_
precision)] WITH LOCAL TIME ZONE

All values of TIMESTAMP WITH TIME ZONE, with the following
exceptions:
■

■

Data is normalized to the database time zone when it is
stored in the database.
When the data is retrieved, users see the data in the session
time zone.

The default format is determined explicitly by the NLS_
TIMESTAMP_FORMAT parameter or implicitly by the NLS_
TERRITORY parameter. The size is 7 or 11 bytes, depending on the
precision.
182

INTERVAL YEAR [(year_precision)]
TO MONTH

Stores a period of time in years and months, where year_
precision is the number of digits in the YEAR datetime field.
Accepted values are 0 to 9. The default is 2. The size is fixed at 5
bytes.

183

INTERVAL DAY [(day_precision)] TO
SECOND [(fractional_seconds_
precision)]

Stores a period of time in days, hours, minutes, and seconds,
where
■

■

day_precision is the maximum number of digits in the DAY
datetime field. Accepted values are 0 to 9. The default is 2.
fractional_seconds_precision is the number of digits in
the fractional part of the SECOND field. Accepted values are 0
to 9. The default is 6.

The size is fixed at 11 bytes.
23

RAW(size)

Raw binary data of length size bytes. Maximum size is 2000
bytes. You must specify size for a RAW value.

24

LONG RAW

Raw binary data of variable length up to 2 gigabytes.

69

ROWID

Base 64 string representing the unique address of a row in its
table. This data type is primarily for values returned by the
ROWID pseudocolumn.

208

UROWID [(size)]

Base 64 string representing the logical address of a row of an
index-organized table. The optional size is the size of a column
of type UROWID. The maximum size and default is 4000 bytes.

Basic Elements of Oracle SQL 3-7

Data Types

Table 3–1 (Cont.) Built-in Data Type Summary
Code

Data Type

Description

96

CHAR [(size [BYTE | CHAR])]

Fixed-length character data of length size bytes or characters.
Maximum size is 2000 bytes or characters. Default and
minimum size is 1 byte.
BYTE and CHAR have the same semantics as for VARCHAR2.

96

NCHAR[(size)]

Fixed-length character data of length size characters. The
number of bytes can be up to two times size for AL16UTF16
encoding and three times size for UTF8 encoding. Maximum
size is determined by the national character set definition, with
an upper limit of 2000 bytes. Default and minimum size is 1
character.

112

CLOB

A character large object containing single-byte or multibyte
characters. Both fixed-width and variable-width character sets
are supported, both using the database character set. Maximum
size is (4 gigabytes - 1) * (database block size).

112

NCLOB

A character large object containing Unicode characters. Both
fixed-width and variable-width character sets are supported,
both using the database national character set. Maximum size is
(4 gigabytes - 1) * (database block size). Stores national character
set data.

113

BLOB

A binary large object. Maximum size is (4 gigabytes - 1) *
(database block size).

114

BFILE

Contains a locator to a large binary file stored outside the
database. Enables byte stream I/O access to external LOBs
residing on the database server. Maximum size is 4 gigabytes.

The sections that follow describe the Oracle data types as they are stored in Oracle
Database. For information on specifying these data types as literals, refer to "Literals"
on page 3-45.
Character Data Types
Character data types store character (alphanumeric) data, which are words and
free-form text, in the database character set or national character set. They are less
restrictive than other data types and consequently have fewer properties. For example,
character columns can store all alphanumeric values, but NUMBER columns can store
only numeric values.
Character data is stored in strings with byte values corresponding to one of the
character sets, such as 7-bit ASCII or EBCDIC, specified when the database was
created. Oracle Database supports both single-byte and multibyte character sets.
These data types are used for character data:
■

CHAR Data Type

■

NCHAR Data Type

■

NVARCHAR2 Data Type

■

VARCHAR2 Data Type

For information on specifying character data types as literals, refer to "Text Literals" on
page 3-45.

3-8 Oracle Database SQL Language Reference

Data Types

CHAR Data Type
The CHAR data type specifies a fixed-length character string. Oracle ensures that all
values stored in a CHAR column have the length specified by size. If you insert a value
that is shorter than the column length, then Oracle blank-pads the value to column
length. If you try to insert a value that is too long for the column, then Oracle returns
an error.
The default length for a CHAR column is 1 byte and the maximum allowed is 2000
bytes. A 1-byte string can be inserted into a CHAR(10) column, but the string is
blank-padded to 10 bytes before it is stored.
When you create a table with a CHAR column, by default you supply the column length
in bytes. The BYTE qualifier is the same as the default. If you use the CHAR qualifier, for
example CHAR(10 CHAR), then you supply the column length in characters. A character
is technically a code point of the database character set. Its size can range from 1 byte
to 4 bytes, depending on the database character set. The BYTE and CHAR qualifiers
override the semantics specified by the NLS_LENGTH_SEMANTICS parameter, which has a
default of byte semantics. For performance reasons, Oracle recommends that you use
the NLS_LENGTH_SEMANTICS parameter to set length semantics and that you use the
BYTE and CHAR qualifiers only when necessary to override the parameter.
To ensure proper data conversion between databases with different character sets, you
must ensure that CHAR data consists of well-formed strings.
See Also: Oracle Database Globalization Support Guide for more
information on character set support and "Data Type Comparison
Rules" on page 3-36 for information on comparison semantics

NCHAR Data Type
The NCHAR data type is a Unicode-only data type. When you create a table with an
NCHAR column, you define the column length in characters. You define the national
character set when you create your database.
The maximum length of a column is determined by the national character set
definition. Width specifications of character data type NCHAR refer to the number of
characters. The maximum column size allowed is 2000 bytes.
If you insert a value that is shorter than the column length, then Oracle blank-pads the
value to column length. You cannot insert a CHAR value into an NCHAR column, nor can
you insert an NCHAR value into a CHAR column.
The following example compares the translated_description column of the
pm.product_descriptions table with a national character set string:
SELECT translated_description
FROM product_descriptions
WHERE translated_name = N'LCD Monitor 11/PM';

See Also: Oracle Database Globalization Support Guide for information
on Unicode data type support

NVARCHAR2 Data Type
The NVARCHAR2 data type is a Unicode-only data type. When you create a table with an
NVARCHAR2 column, you supply the maximum number of characters it can hold. Oracle
subsequently stores each value in the column exactly as you specify it, provided the
value does not exceed the maximum length of the column.

Basic Elements of Oracle SQL 3-9

Data Types

The maximum length of the column is determined by the national character set
definition. Width specifications of character data type NVARCHAR2 refer to the number
of characters. The maximum column size allowed is 4000 bytes.
See Also: Oracle Database Globalization Support Guide for information
on Unicode data type support.

VARCHAR2 Data Type
The VARCHAR2 data type specifies a variable-length character string. When you create a
VARCHAR2 column, you supply the maximum number of bytes or characters of data that
it can hold. Oracle subsequently stores each value in the column exactly as you specify
it, provided the value does not exceed the maximum length of the column. If you try
to insert a value that exceeds the specified length, then Oracle returns an error.
You must specify a maximum length for a VARCHAR2 column. This maximum must be
at least 1 byte, although the actual string stored is permitted to be a zero-length string
(''). You can use the CHAR qualifier, for example VARCHAR2(10 CHAR), to give the
maximum length in characters instead of bytes. A character is technically a code point
of the database character set. You can use the BYTE qualifier, for example VARCHAR2(10
BYTE), to explicitly give the maximum length in bytes. If no explicit qualifier is
included in a column or attribute definition when a database object with this column
or attribute is created, then the length semantics are determined by the value of the
NLS_LENGTH_SEMANTICS parameter of the session creating the object. Independently of
the maximum length in characters, the length of VARCHAR2 data cannot exceed 4000
bytes. Oracle compares VARCHAR2 values using nonpadded comparison semantics.
To ensure proper data conversion between databases with different character sets, you
must ensure that VARCHAR2 data consists of well-formed strings. See Oracle Database
Globalization Support Guide for more information on character set support.
"Data Type Comparison Rules" on page 3-36 for
information on comparison semantics

See Also:

VARCHAR Data Type
Do not use the VARCHAR data type. Use the VARCHAR2 data type instead. Although the
VARCHAR data type is currently synonymous with VARCHAR2, the VARCHAR data type is
scheduled to be redefined as a separate data type used for variable-length character
strings compared with different comparison semantics.
Numeric Data Types
The Oracle Database numeric data types store positive and negative fixed and
floating-point numbers, zero, infinity, and values that are the undefined result of an
operation—"not a number" or NAN. For information on specifying numeric data types
as literals, refer to "Numeric Literals" on page 3-47.

NUMBER Data Type
The NUMBER data type stores zero as well as positive and negative fixed numbers with
absolute values from 1.0 x 10-130 to but not including 1.0 x 10126. If you specify an
arithmetic expression whose value has an absolute value greater than or equal to 1.0 x
10126, then Oracle returns an error. Each NUMBER value requires from 1 to 22 bytes.
Specify a fixed-point number using the following form:
NUMBER(p,s)

where:
3-10 Oracle Database SQL Language Reference

Data Types

■

■

p is the precision, or the maximum number of significant decimal digits, where the
most significant digit is the left-most nonzero digit, and the least significant digit is
the right-most known digit. Oracle guarantees the portability of numbers with
precision of up to 20 base-100 digits, which is equivalent to 39 or 40 decimal digits
depending on the position of the decimal point.
s is the scale, or the number of digits from the decimal point to the least significant
digit. The scale can range from -84 to 127.
–

Positive scale is the number of significant digits to the right of the decimal
point to and including the least significant digit.

–

Negative scale is the number of significant digits to the left of the decimal
point, to but not including the least significant digit. For negative scale the
least significant digit is on the left side of the decimal point, because the actual
data is rounded to the specified number of places to the left of the decimal
point. For example, a specification of (10,-2) means to round to hundreds.

Scale can be greater than precision, most commonly when e notation is used. When
scale is greater than precision, the precision specifies the maximum number of
significant digits to the right of the decimal point. For example, a column defined as
NUMBER(4,5) requires a zero for the first digit after the decimal point and rounds all
values past the fifth digit after the decimal point.
It is good practice to specify the scale and precision of a fixed-point number column
for extra integrity checking on input. Specifying scale and precision does not force all
values to a fixed length. If a value exceeds the precision, then Oracle returns an error. If
a value exceeds the scale, then Oracle rounds it.
Specify an integer using the following form:
NUMBER(p)

This represents a fixed-point number with precision p and scale 0 and is equivalent to
NUMBER(p,0).
Specify a floating-point number using the following form:
NUMBER

The absence of precision and scale designators specifies the maximum range and
precision for an Oracle number.
See Also:

"Floating-Point Numbers" on page 3-12

Table 3–2 show how Oracle stores data using different precisions and scales.
Table 3–2

Storage of Scale and Precision

Actual Data

Specified As

Stored As

123.89

NUMBER

123.89

123.89

NUMBER(3)

124

123.89

NUMBER(3,2)

exceeds precision

123.89

NUMBER(4,2)

exceeds precision

123.89

NUMBER(5,2)

123.89

123.89

NUMBER(6,1)

123.9

123.89

NUMBER(6,-2)

100

Basic Elements of Oracle SQL 3-11

Data Types

Table 3–2 (Cont.) Storage of Scale and Precision
Actual Data

Specified As

Stored As

.01234

NUMBER(4,5)

.01234

.00012

NUMBER(4,5)

.00012

.000127

NUMBER(4,5)

.00013

.0000012

NUMBER(2,7)

.0000012

.00000123

NUMBER(2,7)

.0000012

1.2e-4

NUMBER(2,5)

0.00012

1.2e-5

NUMBER(2,5)

0.00001

FLOAT Data Type
The FLOAT data type is a subtype of NUMBER. It can be specified with or without
precision, which has the same definition it has for NUMBER and can range from 1 to 126.
Scale cannot be specified, but is interpreted from the data. Each FLOAT value requires
from 1 to 22 bytes.
To convert from binary to decimal precision, multiply n by 0.30103. To convert from
decimal to binary precision, multiply the decimal precision by 3.32193. The maximum
of 126 digits of binary precision is roughly equivalent to 38 digits of decimal precision.
The difference between NUMBER and FLOAT is best illustrated by example. In the
following example the same values are inserted into NUMBER and FLOAT columns:
CREATE TABLE test (col1 NUMBER(5,2), col2 FLOAT(5));
INSERT
INSERT
INSERT
INSERT

INTO
INTO
INTO
INTO

test
test
test
test

VALUES
VALUES
VALUES
VALUES

(1.23, 1.23);
(7.89, 7.89);
(12.79, 12.79);
(123.45, 123.45);

SELECT * FROM test;
COL1
COL2
---------- ---------1.23
1.2
7.89
7.9
12.79
13
123.45
120

In this example, the FLOAT value returned cannot exceed 5 binary digits. The largest
decimal number that can be represented by 5 binary digits is 31. The last row contains
decimal values that exceed 31. Therefore, the FLOAT value must be truncated so that its
significant digits do not require more than 5 binary digits. Thus 123.45 is rounded to
120, which has only two significant decimal digits, requiring only 4 binary digits.
Oracle Database uses the Oracle FLOAT data type internally when converting ANSI
FLOAT data. Oracle FLOAT is available for you to use, but Oracle recommends that you
use the BINARY_FLOAT and BINARY_DOUBLE data types instead, as they are more robust.
Refer to "Floating-Point Numbers" on page 3-12 for more information.

Floating-Point Numbers
Floating-point numbers can have a decimal point anywhere from the first to the last
digit or can have no decimal point at all. An exponent may optionally be used
following the number to increase the range, for example, 1.777 e-20. A scale value is not
3-12 Oracle Database SQL Language Reference

Data Types

applicable to floating-point numbers, because the number of digits that can appear
after the decimal point is not restricted.
Binary floating-point numbers differ from NUMBER in the way the values are stored
internally by Oracle Database. Values are stored using decimal precision for NUMBER.
All literals that are within the range and precision supported by NUMBER are stored
exactly as NUMBER. Literals are stored exactly because literals are expressed using
decimal precision (the digits 0 through 9). Binary floating-point numbers are stored
using binary precision (the digits 0 and 1). Such a storage scheme cannot represent all
values using decimal precision exactly. Frequently, the error that occurs when
converting a value from decimal to binary precision is undone when the value is
converted back from binary to decimal precision. The literal 0.1 is such an example.
Oracle Database provides two numeric data types exclusively for floating-point
numbers:
BINARY_FLOAT BINARY_FLOAT is a 32-bit, single-precision floating-point number data
type. Each BINARY_FLOAT value requires 4 bytes.
BINARY_DOUBLE BINARY_DOUBLE is a 64-bit, double-precision floating-point number
data type. Each BINARY_DOUBLE value requires 8 bytes.
In a NUMBER column, floating point numbers have decimal precision. In a BINARY_FLOAT
or BINARY_DOUBLE column, floating-point numbers have binary precision. The binary
floating-point numbers support the special values infinity and NaN (not a number).
You can specify floating-point numbers within the limits listed in Table 3–3 on
page 3-13. The format for specifying floating-point numbers is defined in "Numeric
Literals" on page 3-47.
Table 3–3

Floating Point Number Limits

Value

BINARY_FLOAT

BINARY_DOUBLE

Maximum positive finite value

3.40282E+38F

1.79769313486231E+308

Minimum positive finite value

1.17549E-38F

2.22507485850720E-308

IEEE754 Conformance The Oracle implementation of floating-point data types
conforms substantially with the Institute of Electrical and Electronics Engineers (IEEE)
Standard for Binary Floating-Point Arithmetic, IEEE Standard 754-1985 (IEEE754). The
floating-point data types conform to IEEE754 in the following areas:
■
■

■
■

The SQL function SQRT implements square root. See SQRT on page 5-250.
The SQL function REMAINDER implements remainder. See REMAINDER on
page 5-231.
Arithmetic operators conform. See "Arithmetic Operators" on page 4-3.
Comparison operators conform, except for comparisons with NaN. Oracle orders
NaN greatest with respect to all other values, and evaluates NaN equal to NaN. See
"Floating-Point Conditions" on page 7-7.

■

Conversion operators conform. See "Conversion Functions" on page 5-6.

■

The default rounding mode is supported.

■

The default exception handling mode is supported.

■

The special values INF, -INF, and NaN are supported. See "Floating-Point
Conditions" on page 7-7.

Basic Elements of Oracle SQL 3-13

Data Types

■

■

Rounding of BINARY_FLOAT and BINARY_DOUBLE values to integer-valued BINARY_
FLOAT and BINARY_DOUBLE values is provided by the SQL functions ROUND, TRUNC,
CEIL, and FLOOR.
Rounding of BINARY_FLOAT/BINARY_DOUBLE to decimal and decimal to BINARY_
FLOAT/BINARY_DOUBLE is provided by the SQL functions TO_CHAR, TO_NUMBER, TO_
NCHAR, TO_BINARY_FLOAT, TO_BINARY_DOUBLE, and CAST.

The floating-point data types do not conform to IEEE754 in the following areas:
■

-0 is coerced to +0.

■

Comparison with NaN is not supported.

■

All NaN values are coerced to either BINARY_FLOAT_NAN or BINARY_DOUBLE_NAN.

■

Non-default rounding modes are not supported.

■

Non-default exception handling mode are not supported.

Numeric Precedence
Numeric precedence determines, for operations that support numeric data types, the
data type Oracle uses if the arguments to the operation have different data types.
BINARY_DOUBLE has the highest numeric precedence, followed by BINARY_FLOAT, and
finally by NUMBER. Therefore, in any operation on multiple numeric values:
■

■

■

If any of the operands is BINARY_DOUBLE, then Oracle attempts to convert all the
operands implicitly to BINARY_DOUBLE before performing the operation.
If none of the operands is BINARY_DOUBLE but any of the operands is BINARY_FLOAT,
then Oracle attempts to convert all the operands implicitly to BINARY_FLOAT before
performing the operation.
Otherwise, Oracle attempts to convert all the operands to NUMBER before
performing the operation.

If any implicit conversion is needed and fails, then the operation fails. Refer to
Table 3–10, " Implicit Type Conversion Matrix" on page 3-40 for more information on
implicit conversion.
In the context of other data types, numeric data types have lower precedence than the
datetime/interval data types and higher precedence than character and all other data
types.
LONG Data Type
Do not create tables with LONG columns. Use LOB columns (CLOB, NCLOB, BLOB) instead.
LONG columns are supported only for backward compatibility.
LONG columns store variable-length character strings containing up to 2 gigabytes -1, or
231-1 bytes. LONG columns have many of the characteristics of VARCHAR2 columns. You
can use LONG columns to store long text strings. The length of LONG values may be
limited by the memory available on your computer. LONG literals are formed as
described for "Text Literals" on page 3-45.
Oracle also recommends that you convert existing LONG columns to LOB columns. LOB
columns are subject to far fewer restrictions than LONG columns. Further, LOB
functionality is enhanced in every release, whereas LONG functionality has been static
for several releases. See the modify_col_properties clause of ALTER TABLE on
page 12-2 and TO_LOB on page 5-312 for more information on converting LONG
columns to LOB.
You can reference LONG columns in SQL statements in these places:
3-14 Oracle Database SQL Language Reference

Data Types

■

SELECT lists

■

SET clauses of UPDATE statements

■

VALUES clauses of INSERT statements

The use of LONG values is subject to these restrictions:
■

A table can contain only one LONG column.

■

You cannot create an object type with a LONG attribute.

■

LONG columns cannot appear in WHERE clauses or in integrity constraints (except
that they can appear in NULL and NOT NULL constraints).

■

LONG columns cannot be indexed.

■

LONG data cannot be specified in regular expressions.

■

A stored function cannot return a LONG value.

■

■

■

■

You can declare a variable or argument of a PL/SQL program unit using the LONG
data type. However, you cannot then call the program unit from SQL.
Within a single SQL statement, all LONG columns, updated tables, and locked tables
must be located on the same database.
LONG and LONG RAW columns cannot be used in distributed SQL statements and
cannot be replicated.
If a table has both LONG and LOB columns, then you cannot bind more than 4000
bytes of data to both the LONG and LOB columns in the same SQL statement.
However, you can bind more than 4000 bytes of data to either the LONG or the LOB
column.

In addition, LONG columns cannot appear in these parts of SQL statements:
■

GROUP BY clauses, ORDER BY clauses, or CONNECT BY clauses or with the DISTINCT
operator in SELECT statements

■

The UNIQUE operator of a SELECT statement

■

The column list of a CREATE CLUSTER statement

■

The CLUSTER clause of a CREATE MATERIALIZED VIEW statement

■

SQL built-in functions, expressions, or conditions

■

SELECT lists of queries containing GROUP BY clauses

■

SELECT lists of subqueries or queries combined by the UNION, INTERSECT, or MINUS
set operators

■

SELECT lists of CREATE TABLE ... AS SELECT statements

■

ALTER TABLE ... MOVE statements

■

SELECT lists in subqueries in INSERT statements

Triggers can use the LONG data type in the following manner:
■
■

A SQL statement within a trigger can insert data into a LONG column.
If data from a LONG column can be converted to a constrained data type (such as
CHAR and VARCHAR2), then a LONG column can be referenced in a SQL statement
within a trigger.

■

Variables in triggers cannot be declared using the LONG data type.

■

:NEW and :OLD cannot be used with LONG columns.
Basic Elements of Oracle SQL 3-15

Data Types

You can use Oracle Call Interface functions to retrieve a portion of a LONG value from
the database.
See Also:

Oracle Call Interface Programmer's Guide

Datetime and Interval Data Types
The datetime data types are DATE, TIMESTAMP, TIMESTAMP WITH TIME ZONE, and
TIMESTAMP WITH LOCAL TIME ZONE. Values of datetime data types are sometimes called
datetimes. The interval data types are INTERVAL YEAR TO MONTH and INTERVAL DAY TO
SECOND. Values of interval data types are sometimes called intervals. For information
on expressing datetime and interval values as literals, refer to "Datetime Literals" on
page 3-50 and "Interval Literals" on page 3-53.
Both datetimes and intervals are made up of fields. The values of these fields
determine the value of the data type. Table 3–4 lists the datetime fields and their
possible values for datetimes and intervals.
To avoid unexpected results in your DML operations on datetime data, you can verify
the database and session time zones by querying the built-in SQL functions
DBTIMEZONE and SESSIONTIMEZONE. If the time zones have not been set manually, then
Oracle Database uses the operating system time zone by default. If the operating
system time zone is not a valid Oracle time zone, then Oracle uses UTC as the default
value.
Table 3–4

Datetime Fields and Values

Datetime Field

Valid Values for Datetime

Valid Values for INTERVAL

YEAR

-4712 to 9999 (excluding year 0)

Any positive or negative
integer

MONTH

01 to 12

0 to 11

DAY

01 to 31 (limited by the values of MONTH and YEAR,
according to the rules of the current NLS calendar
parameter)

Any positive or negative
integer

HOUR

00 to 23

0 to 23

MINUTE

00 to 59

0 to 59

SECOND

0 to 59.9(n), where 9(n) is the
00 to 59.9(n), where 9(n) is the precision of time
fractional seconds. The 9(n) portion is not applicable precision of interval
for DATE.
fractional seconds

TIMEZONE_HOUR

-12 to 14 (This range accommodates daylight saving Not applicable
time changes.) Not applicable for DATE or TIMESTAMP.

TIMEZONE_MINUTE

00 to 59. Not applicable for DATE or TIMESTAMP.

Not applicable

TIMEZONE_REGION

Query the TZNAME column of the V$TIMEZONE_NAMES
data dictionary view. Not applicable for DATE or
TIMESTAMP. For a complete listing of all time zone
region names, refer to Oracle Database Globalization
Support Guide.

Not applicable

TIMEZONE_ABBR

Query the TZABBREV column of the V$TIMEZONE_
NAMES data dictionary view. Not applicable for DATE
or TIMESTAMP.

Not applicable

(See note at end of table)

3-16 Oracle Database SQL Language Reference

Data Types

TIMEZONE_HOUR and TIMEZONE_MINUTE are specified together
and interpreted as an entity in the format +|- hh:mi, with values
ranging from -12:59 to +14:00. Refer to Oracle Data Provider for .NET
Developer's Guide for information on specifying time zone values for
that API.
Note:

DATE Data Type
The DATE data type stores date and time information. Although date and time
information can be represented in both character and number data types, the DATE data
type has special associated properties. For each DATE value, Oracle stores the following
information: year, month, day, hour, minute, and second.
You can specify a DATE value as a literal, or you can convert a character or numeric
value to a date value with the TO_DATE function. For examples of expressing DATE
values in both these ways, refer to "Datetime Literals" on page 3-50.
Using Julian Days A Julian day number is the number of days since January 1, 4712 BC.
Julian days allow continuous dating from a common reference. You can use the date
format model "J" with date functions TO_DATE and TO_CHAR to convert between Oracle
DATE values and their Julian equivalents.

Oracle Database uses the astronomical system of calculating
Julian days, in which the year 4713 BC is specified as -4712. The
historical system of calculating Julian days, in contrast, specifies 4713
BC as -4713. If you are comparing Oracle Julian days with values
calculated using the historical system, then take care to allow for the
365-day difference in BC dates. For more information, see
http://aa.usno.navy.mil/faq/docs/millennium.php.
Note:

The default date values are determined as follows:
■

The year is the current year, as returned by SYSDATE.

■

The month is the current month, as returned by SYSDATE.

■

The day is 01 (the first day of the month).

■

The hour, minute, and second are all 0.

These default values are used in a query that requests date values where the date itself
is not specified, as in the following example, which is issued in the month of May:
SELECT TO_DATE('2009', 'YYYY')
FROM DUAL;
TO_DATE('
--------01-MAY-09

Example

This statement returns the Julian equivalent of January 1, 2009:

SELECT TO_CHAR(TO_DATE('01-01-2009', 'MM-DD-YYYY'),'J')
FROM DUAL;
TO_CHAR
-------

Basic Elements of Oracle SQL 3-17

Data Types

2454833

See Also:

"Selecting from the DUAL Table" for a description of the

DUAL table

TIMESTAMP Data Type
The TIMESTAMP data type is an extension of the DATE data type. It stores the year,
month, and day of the DATE data type, plus hour, minute, and second values. This data
type is useful for storing precise time values and for collecting and evaluating date
information across geographic regions. Specify the TIMESTAMP data type as follows:
TIMESTAMP [(fractional_seconds_precision)]

where fractional_seconds_precision optionally specifies the number of digits
Oracle stores in the fractional part of the SECOND datetime field. When you create a
column of this data type, the value can be a number in the range 0 to 9. The default is
6.
See Also: TO_TIMESTAMP on page 5-320 for information on
converting character data to TIMESTAMP data

TIMESTAMP WITH TIME ZONE Data Type
TIMESTAMP WITH TIME ZONE is a variant of TIMESTAMP that includes a time zone region
name or a time zone offset in its value. The time zone offset is the difference (in hours
and minutes) between local time and UTC (Coordinated Universal Time—formerly
Greenwich Mean Time). This data type is useful for preserving local time zone
information.
Specify the TIMESTAMP WITH TIME ZONE data type as follows:
TIMESTAMP [(fractional_seconds_precision)] WITH TIME ZONE

where fractional_seconds_precision optionally specifies the number of digits
Oracle stores in the fractional part of the SECOND datetime field. When you create a
column of this data type, the value can be a number in the range 0 to 9. The default is
6.
Oracle time zone data is derived from the public domain information available at
http://www.iana.org/time-zones/. Oracle time zone data may not reflect the most
recent data available at this site.
See Also:
■

■

■

■

Oracle Database Globalization Support Guide for more information
on Oracle time zone data
"Support for Daylight Saving Times" on page 3-22 and Table 3–17,
" Matching Character Data and Format Models with the FX
Format Model Modifier" on page 3-68 for information on daylight
saving support
TO_TIMESTAMP_TZ on page 5-321 for information on converting
character data to TIMESTAMP WITH TIME ZONE data
ALTER SESSION on page 11-45 for information on the ERROR_ON_
OVERLAP_TIME session parameter

3-18 Oracle Database SQL Language Reference

Data Types

TIMESTAMP WITH LOCAL TIME ZONE Data Type
TIMESTAMP WITH LOCAL TIME ZONE is another variant of TIMESTAMP that is sensitive to
time zone information. It differs from TIMESTAMP WITH TIME ZONE in that data stored in
the database is normalized to the database time zone, and the time zone information is
not stored as part of the column data. When a user retrieves the data, Oracle returns it
in the user's local session time zone. This data type is useful for date information that
is always to be displayed in the time zone of the client system in a two-tier application.
Specify the TIMESTAMP WITH LOCAL TIME ZONE data type as follows:
TIMESTAMP [(fractional_seconds_precision)] WITH LOCAL TIME ZONE

where fractional_seconds_precision optionally specifies the number of digits
Oracle stores in the fractional part of the SECOND datetime field. When you create a
column of this data type, the value can be a number in the range 0 to 9. The default is
6.
Oracle time zone data is derived from the public domain information available at
http://www.iana.org/time-zones/. Oracle time zone data may not reflect the most
recent data available at this site.
See Also:
■

■

Oracle Database Globalization Support Guide for more information
on Oracle time zone data
Oracle Database Advanced Application Developer's Guide for
examples of using this data type and CAST on page 5-35 for
information on converting character data to TIMESTAMP WITH LOCAL
TIME ZONE

INTERVAL YEAR TO MONTH Data Type
INTERVAL YEAR TO MONTH stores a period of time using the YEAR and MONTH datetime
fields. This data type is useful for representing the difference between two datetime
values when only the year and month values are significant.
Specify INTERVAL YEAR TO MONTH as follows:
INTERVAL YEAR [(year_precision)] TO MONTH

where year_precision is the number of digits in the YEAR datetime field. The default
value of year_precision is 2.
You have a great deal of flexibility when specifying interval values as literals. Refer to
"Interval Literals" on page 3-53 for detailed information on specifying interval values
as literals. Also see "Datetime and Interval Examples" on page 3-22 for an example
using intervals.

INTERVAL DAY TO SECOND Data Type
INTERVAL DAY TO SECOND stores a period of time in terms of days, hours, minutes, and
seconds. This data type is useful for representing the precise difference between two
datetime values.
Specify this data type as follows:
INTERVAL DAY [(day_precision)]
TO SECOND [(fractional_seconds_precision)]

where

Basic Elements of Oracle SQL 3-19

Data Types

■

■

day_precision is the number of digits in the DAY datetime field. Accepted values
are 0 to 9. The default is 2.
fractional_seconds_precision is the number of digits in the fractional part of
the SECOND datetime field. Accepted values are 0 to 9. The default is 6.

You have a great deal of flexibility when specifying interval values as literals. Refer to
"Interval Literals" on page 3-53 for detailed information on specify interval values as
literals. Also see "Datetime and Interval Examples" on page 3-22 for an example using
intervals.

Datetime/Interval Arithmetic
You can perform a number of arithmetic operations on date (DATE), timestamp
(TIMESTAMP, TIMESTAMP WITH TIME ZONE, and TIMESTAMP WITH LOCAL TIME ZONE) and
interval (INTERVAL DAY TO SECOND and INTERVAL YEAR TO MONTH) data. Oracle calculates
the results based on the following rules:
■

■
■

■

■

■

You can use NUMBER constants in arithmetic operations on date and timestamp
values, but not interval values. Oracle internally converts timestamp values to
date values and interprets NUMBER constants in arithmetic datetime and interval
expressions as numbers of days. For example, SYSDATE + 1 is tomorrow. SYSDATE 7 is one week ago. SYSDATE + (10/1440) is ten minutes from now. Subtracting the
hire_date column of the sample table employees from SYSDATE returns the
number of days since each employee was hired. You cannot multiply or divide
date or timestamp values.
Oracle implicitly converts BINARY_FLOAT and BINARY_DOUBLE operands to NUMBER.
Each DATE value contains a time component, and the result of many date
operations include a fraction. This fraction means a portion of one day. For
example, 1.5 days is 36 hours. These fractions are also returned by Oracle built-in
functions for common operations on DATE data. For example, the MONTHS_BETWEEN
function returns the number of months between two dates. The fractional portion
of the result represents that portion of a 31-day month.
If one operand is a DATE value or a numeric value, neither of which contains time
zone or fractional seconds components, then:
–

Oracle implicitly converts the other operand to DATE data. The exception is
multiplication of a numeric value times an interval, which returns an interval.

–

If the other operand has a time zone value, then Oracle uses the session time
zone in the returned value.

–

If the other operand has a fractional seconds value, then the fractional seconds
value is lost.

When you pass a timestamp, interval, or numeric value to a built-in function that
was designed only for the DATE data type, Oracle implicitly converts the non-DATE
value to a DATE value. Refer to "Datetime Functions" on page 5-5 for information
on which functions cause implicit conversion to DATE.
When interval calculations return a datetime value, the result must be an actual
datetime value or the database returns an error. For example, the next two
statements return errors:
SELECT TO_DATE('31-AUG-2004','DD-MON-YYYY') + TO_YMINTERVAL('0-1')
FROM DUAL;
SELECT TO_DATE('29-FEB-2004','DD-MON-YYYY') + TO_YMINTERVAL('1-0')
FROM DUAL;

3-20 Oracle Database SQL Language Reference

Data Types

The first fails because adding one month to a 31-day month would result in
September 31, which is not a valid date. The second fails because adding one year
to a date that exists only every four years is not valid. However, the next statement
succeeds, because adding four years to a February 29 date is valid:
SELECT TO_DATE('29-FEB-2004', 'DD-MON-YYYY') + TO_YMINTERVAL('4-0')
FROM DUAL;
TO_DATE('
--------29-FEB-08
■

Oracle performs all timestamp arithmetic in UTC time. For TIMESTAMP WITH LOCAL
TIME ZONE, Oracle converts the datetime value from the database time zone to UTC
and converts back to the database time zone after performing the arithmetic. For
TIMESTAMP WITH TIME ZONE, the datetime value is always in UTC, so no conversion
is necessary.

Table 3–5 is a matrix of datetime arithmetic operations. Dashes represent operations
that are not supported.
Table 3–5

Matrix of Datetime Arithmetic

Operand & Operator

DATE

TIMESTAMP

INTERVAL

Numeric

+

—

—

DATE

DATE

-

NUMBER

INTERVAL

DATE

DATE

*

—

—

—

—

/

—

—

—

—

+

—

—

TIMESTAMP

DATE

-

INTERVAL

INTERVAL

TIMESTAMP

DATE

*

—

—

—

—

/

—

—

—

—

+

DATE

TIMESTAMP

INTERVAL

—

-

—

—

INTERVAL

—

*

—

—

—

INTERVAL

/

—

—

—

INTERVAL

+

DATE

DATE

—

NA

-

—

—

—

NA

*

—

—

INTERVAL

NA

/

—

—

—

NA

DATE

TIMESTAMP

INTERVAL

Numeric

You can add an interval value expression to a start time. Consider the
sample table oe.orders with a column order_date. The following statement adds 30
days to the value of the order_date column:

Examples

Basic Elements of Oracle SQL 3-21

Data Types

SELECT order_id, order_date + INTERVAL '30' DAY AS "Due Date"
FROM orders
ORDER BY order_id, "Due Date";

Support for Daylight Saving Times
Oracle Database automatically determines, for any given time zone region, whether
daylight saving is in effect and returns local time values accordingly. The datetime
value is sufficient for Oracle to determine whether daylight saving time is in effect for
a given region in all cases except boundary cases. A boundary case occurs during the
period when daylight saving goes into or comes out of effect. For example, in the
US-Pacific region, when daylight saving goes into effect, the time changes from 2:00
a.m. to 3:00 a.m. The one hour interval between 2 and 3 a.m. does not exist. When
daylight saving goes out of effect, the time changes from 2:00 a.m. back to 1:00 a.m.,
and the one-hour interval between 1 and 2 a.m. is repeated.
To resolve these boundary cases, Oracle uses the TZR and TZD format elements, as
described in Table 3–17. TZR represents the time zone region name in datetime input
strings. Examples are 'Australia/North', 'UTC', and 'Singapore'. TZD represents an
abbreviated form of the time zone region name with daylight saving information.
Examples are 'PST' for US/Pacific standard time and 'PDT' for US/Pacific daylight time.
To see a listing of valid values for the TZR and TZD format elements, query the TZNAME
and TZABBREV columns of the V$TIMEZONE_NAMES dynamic performance view.

Time zone region names are needed by the daylight saving
feature. These names are stored in two types of time zone files: one
large and one small. One of these files is the default file, depending
on your environment and the release of Oracle Database you are
using. For more information regarding time zone files and names,
see Oracle Database Globalization Support Guide.

Note:

For a complete listing of the time zone region names in both files, refer to Oracle
Database Globalization Support Guide.
Oracle time zone data is derived from the public domain information available at
http://www.iana.org/time-zones/. Oracle time zone data may not reflect the most
recent data available at this site.
See Also:
■

■

■

"Datetime Format Models" on page 3-60 for information on the
format elements and the session parameter ERROR_ON_
OVERLAP_TIME on page 11-51.
Oracle Database Globalization Support Guide for more information
on Oracle time zone data
Oracle Database Reference for information on the dynamic
performance views

Datetime and Interval Examples
The following example shows how to specify some datetime and interval data types.
CREATE TABLE time_table
(start_time
TIMESTAMP,
duration_1
INTERVAL DAY (6) TO SECOND (5),
duration_2
INTERVAL YEAR TO MONTH);

3-22 Oracle Database SQL Language Reference

Data Types

The start_time column is of type TIMESTAMP. The implicit fractional seconds precision
of TIMESTAMP is 6.
The duration_1 column is of type INTERVAL DAY TO SECOND. The maximum number of
digits in field DAY is 6 and the maximum number of digits in the fractional second is 5.
The maximum number of digits in all other datetime fields is 2.
The duration_2 column is of type INTERVAL YEAR TO MONTH. The maximum number of
digits of the value in each field (YEAR and MONTH) is 2.
Interval data types do not have format models. Therefore, to adjust their presentation,
you must combine character functions such as EXTRACT and concatenate the
components. For example, the following examples query the hr.employees and
oe.orders tables, respectively, and change interval output from the form "yy-mm" to
"yy years mm months" and from "dd-hh" to "dddd days hh hours":
SELECT last_name, EXTRACT(YEAR FROM (SYSDATE - hire_date) YEAR TO MONTH)
|| ' years '
|| EXTRACT(MONTH FROM (SYSDATE - hire_date) YEAR TO MONTH)
|| ' months' "Interval"
FROM employees;
LAST_NAME
------------------------OConnell
Grant
Whalen
Hartstein
Fay
Mavris
Baer
Higgins
Gietz
. . .

Interval
-------------------2 years 3 months
1 years 9 months
6 years 1 months
5 years 8 months
4 years 2 months
7 years 4 months
7 years 4 months
7 years 4 months
7 years 4 months

SELECT order_id, EXTRACT(DAY FROM (SYSDATE - order_date) DAY TO SECOND)
|| ' days '
|| EXTRACT(HOUR FROM (SYSDATE - order_date) DAY TO SECOND)
|| ' hours' "Interval"
FROM orders;
ORDER_ID
---------2458
2397
2454
2354
2358
2381
2440
2357
2394
2435
. . .

Interval
-------------------780 days 23 hours
685 days 22 hours
733 days 21 hours
447 days 20 hours
635 days 20 hours
508 days 18 hours
765 days 17 hours
1365 days 16 hours
602 days 15 hours
763 days 15 hours

RAW and LONG RAW Data Types
The RAW and LONG RAW data types store data that is not to be explicitly converted by
Oracle Database when moving data between different systems. These data types are
intended for binary data or byte strings. For example, you can use LONG RAW to store
Basic Elements of Oracle SQL 3-23

Data Types

graphics, sound, documents, or arrays of binary data, for which the interpretation is
dependent on the use.
Oracle strongly recommends that you convert LONG RAW columns to binary LOB (BLOB)
columns. LOB columns are subject to far fewer restrictions than LONG columns. See
TO_LOB on page 5-312 for more information.
RAW is a variable-length data type like VARCHAR2, except that Oracle Net (which
connects client software to a database or one database to another) and the Oracle
import and export utilities do not perform character conversion when transmitting RAW
or LONG RAW data. In contrast, Oracle Net and the Oracle import and export utilities
automatically convert CHAR, VARCHAR2, and LONG data between different database
character sets, if data is transported between databases, or between the database
character set and the client character set, if data is transported between a database and
a client. The client character set is determined by the type of the client interface, such
as OCI or JDBC, and the client configuration (for example, the NLS_LANG environment
variable).
When Oracle implicitly converts RAW or LONG RAW data to character data, the resulting
character value contains a hexadecimal representation of the binary input, where each
character is a hexadecimal digit (0-9, A-F) representing four consecutive bits of RAW
data. For example, one byte of RAW data with bits 11001011 becomes the value CB.
When Oracle implicitly converts character data to RAW or LONG RAW, it interprets each
consecutive input character as a hexadecimal representation of four consecutive bits of
binary data and builds the resulting RAW or LONG RAW value by concatenating those bits.
If any of the input characters is not a hexadecimal digit (0-9, A-F, a-f), then an error is
reported. If the number of characters is odd, then the result is undefined.
The SQL functions RAWTOHEX and HEXTORAW perform explicit conversions that are
equivalent to the above implicit conversions. Other types of conversions between RAW
and character data are possible with functions in the Oracle-supplied PL/SQL
packages UTL_RAW and UTL_I18N.
Large Object (LOB) Data Types
The built-in LOB data types BLOB, CLOB, and NCLOB (stored internally) and BFILE
(stored externally) can store large and unstructured data such as text, image, video,
and spatial data. The size of BLOB, CLOB, and NCLOB data can be up to (232-1 bytes) * (the
value of the CHUNK parameter of LOB storage). If the tablespaces in your database are
of standard block size, and if you have used the default value of the CHUNK parameter
of LOB storage when creating a LOB column, then this is equivalent to (232-1 bytes) *
(database block size). BFILE data can be up to 264-1 bytes, although your operating
system may impose restrictions on this maximum.
When creating a table, you can optionally specify different tablespace and storage
characteristics for LOB columns or LOB object attributes from those specified for the
table.
CLOB, NCLOB, and BLOB values up to approximately 4000 bytes are stored inline if you
enable storage in row at the time the LOB column is created. LOBs greater than 4000
bytes are always stored externally. Refer to ENABLE STORAGE IN ROW on
page 16-44 for more information.
LOB columns contain LOB locators that can refer to internal (in the database) or
external (outside the database) LOB values. Selecting a LOB from a table actually
returns the LOB locator and not the entire LOB value. The DBMS_LOB package and
Oracle Call Interface (OCI) operations on LOBs are performed through these locators.
LOBs are similar to LONG and LONG RAW types, but differ in the following ways:

3-24 Oracle Database SQL Language Reference

Data Types

■
■

■
■

LOBs can be attributes of an object type (user-defined data type).
The LOB locator is stored in the table column, either with or without the actual
LOB value. BLOB, NCLOB, and CLOB values can be stored in separate tablespaces.
BFILE data is stored in an external file on the server.
When you access a LOB column, the locator is returned.
A LOB can be up to (232-1 bytes)*(database block size) in size. BFILE data can be up
to 264-1 bytes, although your operating system may impose restrictions on this
maximum.

■

LOBs permit efficient, random, piece-wise access to and manipulation of data.

■

You can define more than one LOB column in a table.

■

With the exception of NCLOB, you can define one or more LOB attributes in an
object.

■

You can declare LOB bind variables.

■

You can select LOB columns and LOB attributes.

■

■

■

You can insert a new row or update an existing row that contains one or more LOB
columns or an object with one or more LOB attributes. In update operations, you
can set the internal LOB value to NULL, empty, or replace the entire LOB with data.
You can set the BFILE to NULL or make it point to a different file.
You can update a LOB row-column intersection or a LOB attribute with another
LOB row-column intersection or LOB attribute.
You can delete a row containing a LOB column or LOB attribute and thereby also
delete the LOB value. For BFILEs, the actual operating system file is not deleted.

You can access and populate rows of an inline LOB column (a LOB column stored in
the database) or a LOB attribute (an attribute of an object type column stored in the
database) simply by issuing an INSERT or UPDATE statement.
Restrictions on LOB Columns LOB columns are subject to a number of rules and
restrictions. See Oracle Database SecureFiles and Large Objects Developer's Guide for a
complete listing.
See Also:
■

■

Oracle Database PL/SQL Packages and Types Reference and Oracle Call
Interface Programmer's Guide for more information about these
interfaces and LOBs
the modify_col_properties clause of ALTER TABLE on page 12-2
and TO_LOB on page 5-312 for more information on converting
LONG columns to LOB columns

BFILE Data Type
The BFILE data type enables access to binary file LOBs that are stored in file systems
outside Oracle Database. A BFILE column or attribute stores a BFILE locator, which
serves as a pointer to a binary file on the server file system. The locator maintains the
directory name and the filename.
You can change the filename and path of a BFILE without affecting the base table by
using the BFILENAME function. Refer to BFILENAME on page 5-29 for more
information on this built-in SQL function.

Basic Elements of Oracle SQL 3-25

Data Types

Binary file LOBs do not participate in transactions and are not recoverable. Rather, the
underlying operating system provides file integrity and durability. BFILE data can be
up to 264-1 bytes, although your operating system may impose restrictions on this
maximum.
The database administrator must ensure that the external file exists and that Oracle
processes have operating system read permissions on the file.
The BFILE data type enables read-only support of large binary files. You cannot
modify or replicate such a file. Oracle provides APIs to access file data. The primary
interfaces that you use to access file data are the DBMS_LOB package and Oracle Call
Interface (OCI).
See Also: Oracle Database SecureFiles and Large Objects Developer's
Guide and Oracle Call Interface Programmer's Guide for more
information about LOBs and CREATE DIRECTORY on page 14-41

BLOB Data Type
The BLOB data type stores unstructured binary large objects. BLOB objects can be
thought of as bitstreams with no character set semantics. BLOB objects can store binary
data up to (4 gigabytes -1) * (the value of the CHUNK parameter of LOB storage). If the
tablespaces in your database are of standard block size, and if you have used the
default value of the CHUNK parameter of LOB storage when creating a LOB column,
then this is equivalent to (4 gigabytes - 1) * (database block size).
BLOB objects have full transactional support. Changes made through SQL, the DBMS_
LOB package, or Oracle Call Interface (OCI) participate fully in the transaction. BLOB
value manipulations can be committed and rolled back. However, you cannot save a
BLOB locator in a PL/SQL or OCI variable in one transaction and then use it in another
transaction or session.

CLOB Data Type
The CLOB data type stores single-byte and multibyte character data. Both fixed-width
and variable-width character sets are supported, and both use the database character
set. CLOB objects can store up to (4 gigabytes -1) * (the value of the CHUNK parameter of
LOB storage) of character data. If the tablespaces in your database are of standard
block size, and if you have used the default value of the CHUNK parameter of LOB
storage when creating a LOB column, then this is equivalent to (4 gigabytes - 1) *
(database block size).
CLOB objects have full transactional support. Changes made through SQL, the DBMS_
LOB package, or Oracle Call Interface (OCI) participate fully in the transaction. CLOB
value manipulations can be committed and rolled back. However, you cannot save a
CLOB locator in a PL/SQL or OCI variable in one transaction and then use it in another
transaction or session.

NCLOB Data Type
The NCLOB data type stores Unicode data. Both fixed-width and variable-width
character sets are supported, and both use the national character set. NCLOB objects can
store up to (4 gigabytes -1) * (the value of the CHUNK parameter of LOB storage) of
character text data. If the tablespaces in your database are of standard block size, and if
you have used the default value of the CHUNK parameter of LOB storage when creating
a LOB column, then this is equivalent to (4 gigabytes - 1) * (database block size).
NCLOB objects have full transactional support. Changes made through SQL, the DBMS_
LOB package, or OCI participate fully in the transaction. NCLOB value manipulations

3-26 Oracle Database SQL Language Reference

Data Types

can be committed and rolled back. However, you cannot save an NCLOB locator in a
PL/SQL or OCI variable in one transaction and then use it in another transaction or
session.
See Also: Oracle Database Globalization Support Guide for information
on Unicode data type support

Rowid Data Types
Each row in the database has an address. The sections that follow describe the two
forms of row address in an Oracle Database.

ROWID Data Type
The rows in heap-organized tables that are native to Oracle Database have row
addresses called rowids. You can examine a rowid row address by querying the
pseudocolumn ROWID. Values of this pseudocolumn are strings representing the
address of each row. These strings have the data type ROWID. You can also create tables
and clusters that contain actual columns having the ROWID data type. Oracle Database
does not guarantee that the values of such columns are valid rowids. Refer to
Chapter 2, "Pseudocolumns" for more information on the ROWID pseudocolumn.
Beginning with Oracle8, Oracle SQL incorporated an extended
format for rowids to efficiently support partitioned tables and indexes
and tablespace-relative data block addresses without ambiguity. If you
are running Version 7 of the database and you intend to upgrade, use
the DBMS_ROWID package to migrate rowids in your data to the
extended format. Refer to Oracle Database PL/SQL Packages and Types
Reference for information on DBMS_ROWID and to Oracle Database
Upgrade Guide for information on upgrading from Oracle7.
Note:

Rowids contain the following information:
■

■
■

■

The data block of the data file containing the row. The length of this string
depends on your operating system.
The row in the data block.
The database file containing the row. The first data file has the number 1. The
length of this string depends on your operating system.
The data object number, which is an identification number assigned to every
database segment. You can retrieve the data object number from the data
dictionary views USER_OBJECTS, DBA_OBJECTS, and ALL_OBJECTS. Objects that share
the same segment (clustered tables in the same cluster, for example) have the same
object number.

Rowids are stored as base 64 values that can contain the characters A-Z, a-z, 0-9, and
the plus sign (+) and forward slash (/). Rowids are not available directly. You can use
the supplied package DBMS_ROWID to interpret rowid contents. The package functions
extract and provide information on the four rowid elements listed above.
See Also: Oracle Database PL/SQL Packages and Types Reference for
information on the functions available with the DBMS_ROWID package
and how to use them

Basic Elements of Oracle SQL 3-27

Data Types

UROWID Data Type
The rows of some tables have addresses that are not physical or permanent or were not
generated by Oracle Database. For example, the row addresses of index-organized
tables are stored in index leaves, which can move. Rowids of foreign tables (such as
DB2 tables accessed through a gateway) are not standard Oracle rowids.
Oracle uses universal rowids (urowids) to store the addresses of index-organized and
foreign tables. Index-organized tables have logical urowids and foreign tables have
foreign urowids. Both types of urowid are stored in the ROWID pseudocolumn (as are
the physical rowids of heap-organized tables).
Oracle creates logical rowids based on the primary key of the table. The logical rowids
do not change as long as the primary key does not change. The ROWID pseudocolumn
of an index-organized table has a data type of UROWID. You can access this
pseudocolumn as you would the ROWID pseudocolumn of a heap-organized table
(using a SELECT ... ROWID statement). If you want to store the rowids of an
index-organized table, then you can define a column of type UROWID for the table and
retrieve the value of the ROWID pseudocolumn into that column.

ANSI, DB2, and SQL/DS Data Types
SQL statements that create tables and clusters can also use ANSI data types and data
types from the IBM products SQL/DS and DB2. Oracle recognizes the ANSI or IBM
data type name that differs from the Oracle Database data type name. It converts the
data type to the equivalent Oracle data type, records the Oracle data type as the name
of the column data type, and stores the column data in the Oracle data type based on
the conversions shown in the tables that follow.
Table 3–6

ANSI Data Types Converted to Oracle Data Types

ANSI SQL Data Type

Oracle Data Type

CHARACTER(n)

CHAR(n)

CHAR(n)
CHARACTER VARYING(n)

VARCHAR2(n)

CHAR VARYING(n)
NATIONAL CHARACTER(n)

NCHAR(n)

NATIONAL CHAR(n)
NCHAR(n)
NATIONAL CHARACTER VARYING(n)

NVARCHAR2(n)

NATIONAL CHAR VARYING(n)
NCHAR VARYING(n)
NUMERIC[(p,s)]

NUMBER(p,s)

DECIMAL[(p,s)] (Note 1)
INTEGER

NUMBER(p,0)

INT
SMALLINT
FLOAT (Note 2)

FLOAT(126)

DOUBLE PRECISION (Note 3)

FLOAT(126)

REAL (Note 4)

FLOAT(63)

3-28 Oracle Database SQL Language Reference

Data Types

Notes:
1. The NUMERIC and DECIMAL data types can specify only fixed-point numbers. For
those data types, the scale (s) defaults to 0.
2.

The FLOAT data type is a floating-point number with a binary precision b. The
default precision for this data type is 126 binary, or 38 decimal.

3.

The DOUBLE PRECISION data type is a floating-point number with binary precision
126.

4.

The REAL data type is a floating-point number with a binary precision of 63, or 18
decimal.

Do not define columns with the following SQL/DS and DB2 data types, because they
have no corresponding Oracle data type:
■

GRAPHIC

■

LONG VARGRAPHIC

■

VARGRAPHIC

■

TIME

Note that data of type TIME can also be expressed as Oracle datetime data.
See Also:

Table 3–7

"Datetime and Interval Data Types" on page 3-16

SQL/DS and DB2 Data Types Converted to Oracle Data Types

SQL/DS or DB2 Data Type

Oracle Data Type

CHARACTER(n)

CHAR(n)

VARCHAR(n)

VARCHAR(n)

LONG VARCHAR

LONG

DECIMAL(p,s) (Note 1)

NUMBER(p,s)

INTEGER

NUMBER(p,0)

SMALLINT
FLOAT (Note 2)

NUMBER

Notes:
1. The DECIMAL data type can specify only fixed-point numbers. For this data type, s
defaults to 0.
2.

The FLOAT data type is a floating-point number with a binary precision b. The
default precision for this data type is 126 binary or 38 decimal.

User-Defined Types
User-defined data types use Oracle built-in data types and other user-defined data
types as the building blocks of object types that model the structure and behavior of
data in applications. The sections that follow describe the various categories of
user-defined types.

Basic Elements of Oracle SQL 3-29

Data Types

See Also:
■

■

■

Oracle Database Concepts for information about Oracle built-in data
types
CREATE TYPE on page 17-3 and the CREATE TYPE BODY on
page 17-5 for information about creating user-defined types
Oracle Database Object-Relational Developer's Guide for information
about using user-defined types

Object Types
Object types are abstractions of the real-world entities, such as purchase orders, that
application programs deal with. An object type is a schema object with three kinds of
components:
■
■

■

A name, which identifies the object type uniquely within that schema.
Attributes, which are built-in types or other user-defined types. Attributes model
the structure of the real-world entity.
Methods, which are functions or procedures written in PL/SQL and stored in the
database, or written in a language like C or Java and stored externally. Methods
implement operations the application can perform on the real-world entity.

REF Data Types
An object identifier (represented by the keyword OID) uniquely identifies an object
and enables you to reference the object from other objects or from relational tables. A
data type category called REF represents such references. A REF data type is a container
for an object identifier. REF values are pointers to objects.
When a REF value points to a nonexistent object, the REF is said to be "dangling". A
dangling REF is different from a null REF. To determine whether a REF is dangling or
not, use the condition IS [NOT] DANGLING. For example, given object view oc_orders in
the sample schema oe, the column customer_ref is of type REF to type customer_typ,
which has an attribute cust_email:
SELECT o.customer_ref.cust_email
FROM oc_orders o
WHERE o.customer_ref IS NOT DANGLING;

Varrays
An array is an ordered set of data elements. All elements of a given array are of the
same data type. Each element has an index, which is a number corresponding to the
position of the element in the array.
The number of elements in an array is the size of the array. Oracle arrays are of
variable size, which is why they are called varrays. You must specify a maximum size
when you declare the varray.
When you declare a varray, it does not allocate space. It defines a type, which you can
use as:
■

The data type of a column of a relational table

■

An object type attribute

■

A PL/SQL variable, parameter, or function return type

Oracle normally stores an array object either in line (as part of the row data) or out of
line (in a LOB), depending on its size. However, if you specify separate storage
3-30 Oracle Database SQL Language Reference

Data Types

characteristics for a varray, then Oracle stores it out of line, regardless of its size. Refer
to the varray_col_properties of CREATE TABLE on page 16-47 for more information
about varray storage.

Nested Tables
A nested table type models an unordered set of elements. The elements may be
built-in types or user-defined types. You can view a nested table as a single-column
table or, if the nested table is an object type, as a multicolumn table, with a column for
each attribute of the object type.
A nested table definition does not allocate space. It defines a type, which you can use
to declare:
■

The data type of a column of a relational table

■

An object type attribute

■

A PL/SQL variable, parameter, or function return type

When a nested table appears as the type of a column in a relational table or as an
attribute of the underlying object type of an object table, Oracle stores all of the nested
table data in a single table, which it associates with the enclosing relational or object
table.

Oracle-Supplied Types
Oracle provides SQL-based interfaces for defining new types when the built-in or
ANSI-supported types are not sufficient. The behavior for these types can be
implemented in C/C++, Java, or PL/ SQL. Oracle Database automatically provides the
low-level infrastructure services needed for input-output, heterogeneous client-side
access for new data types, and optimizations for data transfers between the application
and the database.
These interfaces can be used to build user-defined (or object) types and are also used
by Oracle to create some commonly useful data types. Several such data types are
supplied with the server, and they serve both broad horizontal application areas (for
example, the Any types) and specific vertical ones (for example, the spatial types).
The Oracle-supplied types, along with cross-references to the documentation of their
implementation and use, are described in the following sections:
■

Any Types

■

XML Types

■

Spatial Types

■

Media Types

Any Types
The Any types provide highly flexible modeling of procedure parameters and table
columns where the actual type is not known. These data types let you dynamically
encapsulate and access type descriptions, data instances, and sets of data instances of
any other SQL type. These types have OCI and PL/SQL interfaces for construction
and access.

ANYTYPE
This type can contain a type description of any named SQL type or unnamed transient
type.
Basic Elements of Oracle SQL 3-31

Data Types

ANYDATA
This type contains an instance of a given type, with data, plus a description of the
type. ANYDATA can be used as a table column data type and lets you store
heterogeneous values in a single column. The values can be of SQL built-in types as
well as user-defined types.

ANYDATASET
This type contains a description of a given type plus a set of data instances of that
type. ANYDATASET can be used as a procedure parameter data type where such
flexibility is needed. The values of the data instances can be of SQL built-in types as
well as user-defined types.
See Also: Oracle Database PL/SQL Packages and Types Reference for
information on the ANYTYPE, ANYDATA, and ANYDATASET types

XML Types
Extensible Markup Language (XML) is a standard format developed by the World
Wide Web Consortium (W3C) for representing structured and unstructured data on
the World Wide Web. Universal resource identifiers (URIs) identify resources such as
Web pages anywhere on the Web. Oracle provides types to handle XML and URI data,
as well as a class of URIs called DBURIRef types to access data stored within the
database itself. It also provides a set of types to store and access both external and
internal URIs from within the database.

XMLType
This Oracle-supplied type can be used to store and query XML data in the database.
XMLType has member functions you can use to access, extract, and query the XML data
using XPath expressions. XPath is another standard developed by the W3C committee
to traverse XML documents. Oracle XMLType functions support many W3C XPath
expressions. Oracle also provides a set of SQL functions and PL/SQL packages to
create XMLType values from existing relational or object-relational data.
XMLType is a system-defined type, so you can use it as an argument of a function or as
the data type of a table or view column. You can also create tables and views of
XMLType. When you create an XMLType column in a table, you can choose to store the
XML data in a CLOB column, as binary XML (stored internally as a CLOB), or object
relationally.
You can also register the schema (using the DBMS_XMLSCHEMA package) and create a
table or column conforming to the registered schema. In this case Oracle stores the
XML data in underlying object-relational columns by default, but you can specify
storage in a CLOB or binary XML column even for schema-based data.
Queries and DML on XMLType columns operate the same regardless of the storage
mechanism.
See Also: Oracle XML DB Developer's Guide for information about
using XMLType columns

URI Data Types
Oracle supplies a family of URI types—URIType, DBURIType, XDBURIType, and
HTTPURIType—which are related by an inheritance hierarchy. URIType is an object type
and the others are subtypes of URIType. Since URIType is the supertype, you can create

3-32 Oracle Database SQL Language Reference

Data Types

columns of this type and store DBURIType or HTTPURIType type instances in this
column.
HTTPURIType You can use HTTPURIType to store URLs to external Web pages or to
files. Oracle accesses these files using HTTP (Hypertext Transfer Protocol).
XDBURIType You can use XDBURIType to expose documents in the XML database
hierarchy as URIs that can be embedded in any URIType column in a table. The
XDBURIType consists of a URL, which comprises the hierarchical name of the XML
document to which it refers and an optional fragment representing the XPath syntax.
The fragment is separated from the URL part by a pound sign (#). The following lines
are examples of XDBURIType:
/home/oe/doc1.xml
/home/oe/doc1.xml#/orders/order_item

DBURIType DBURIType can be used to store DBURIRef values, which reference data

inside the database. Storing DBURIRef values lets you reference data stored inside or
outside the database and access the data consistently.
DBURIRef values use an XPath-like representation to reference data inside the database.
If you imagine the database as an XML tree, then you would see the tables, rows, and
columns as elements in the XML document. For example, the sample human resources
user hr would see the following XML tree:




205
Higgins
12008
.. 

... 





The DBURIRef is an XPath expression over this virtual XML document. So to reference
the SALARY value in the EMPLOYEES table for the employee with employee number 205,
you can write a DBURIRef as,
/HR/EMPLOYEES/ROW[EMPLOYEE_ID=205]/SALARY

Using this model, you can reference data stored in CLOB columns or other columns and
expose them as URLs to the external world.

URIFactory Package
Oracle also provides the URIFactory package, which can create and return instances of
the various subtypes of the URITypes. The package analyzes the URL string, identifies
the type of URL (HTTP, DBURI, and so on), and creates an instance of the subtype. To
create a DBURI instance, the URL must start with the prefix /oradb. For example,
URIFactory.getURI('/oradb/HR/EMPLOYEES') would create a DBURIType instance and
URIFactory.getUri('/sys/schema') would create an XDBURIType instance.

Basic Elements of Oracle SQL 3-33

Data Types

See Also:
■

■

■

Oracle Database Object-Relational Developer's Guide for general
information on object types and type inheritance
Oracle XML DB Developer's Guide for more information about
these supplied types and their implementation
Oracle Streams Advanced Queuing User's Guide for information
about using XMLType with Oracle Advanced Queuing

Spatial Types
Oracle Spatial is designed to make spatial data management easier and more natural
to users of location-enabled applications, geographic information system (GIS)
applications, and geoimaging applications. After the spatial data is stored in an Oracle
Database, you can easily manipulate, retrieve, and relate it to all the other data stored
in the database. The following data types are available only if you have installed
Oracle Spatial.

SDO_GEOMETRY
The geometric description of a spatial object is stored in a single row, in a single
column of object type SDO_GEOMETRY in a user-defined table. Any table that has a
column of type SDO_GEOMETRY must have another column, or set of columns, that
defines a unique primary key for that table. Tables of this sort are sometimes called
geometry tables.
The SDO_GEOMETRY object type has the following definition:
CREATE TYPE SDO_GEOMETRY AS OBJECT
(sgo_gtype
NUMBER,
sdo_srid
NUMBER,
sdo_point
SDO_POINT_TYPE,
sdo_elem_info
SDO_ELEM_INFO_ARRAY,
sdo_ordinates
SDO_ORDINATE_ARRAY);
/

SDO_TOPO_GEOMETRY
This type describes a topology geometry, which is stored in a single row, in a single
column of object type SDO_TOPO_GEOMETRY in a user-defined table.
The SDO_TOPO_GEOMETRY object type has the following definition:
CREATE TYPE SDO_TOPO_GEOMETRY AS OBJECT
(tg_type
NUMBER,
tg_id
NUMBER,
tg_layer_id
NUMBER,
topology_id
NUMBER);
/

SDO_GEORASTER
In the GeoRaster object-relational model, a raster grid or image object is stored in a
single row, in a single column of object type SDO_GEORASTER in a user-defined table.
Tables of this sort are called GeoRaster tables.
The SDO_GEORASTER object type has the following definition:
CREATE TYPE SDO_GEORASTER AS OBJECT
(rasterType
NUMBER,

3-34 Oracle Database SQL Language Reference

Data Types

spatialExtent
rasterDataTable
rasterID
metadata

SDO_GEOMETRY,
VARCHAR2(32),
NUMBER,
XMLType);

/

See Also: Oracle Spatial Developer's Guide, Oracle Spatial Topology and
Network Data Models Developer's Guide, and Oracle Spatial GeoRaster
Developer's Guide for information on the full implementation of the
spatial data types and guidelines for using them

Media Types
Oracle Multimedia uses object types, similar to Java or C++ classes, to describe
multimedia data. An instance of these object types consists of attributes, including
metadata and the media data, and methods. The Multimedia data types are created in
the ORDSYS schema. Public synonyms exist for all the data types, so you can access
them without specifying the schema name.
Oracle Multimedia provides the following object types:
■

ORDAudio
Supports the storage and management of audio data.

■

ORDDicom
Supports the storage and management of Digital Imaging and Communications in
Medicine (DICOM), the format universally recognized as the standard for medical
imaging.

■

ORDDoc
Supports storage and management of any type of media data, including audio,
image and video data. Use this type when you want all media to be stored in a
single column.

■

ORDImage
Supports the storage and management of image data.

■

ORDVideo
Supports the storage and management of video data.

■

ORDImageSignature
The ORDImageSignature object type has been deprecated and should no longer be
introduced into your code. Existing occurrences of this object type will continue to
function as in the past.

The following data types provide compliance with the ISO-IEC 13249-5 Still Image
standard, commonly referred to as SQL/MM StillImage:
■

SI_AverageColor
Represents a feature that characterizes an image by its average color.

■

SI_Color
Encapsulates color values.

■

SI_ColorHistogram
Represents a feature that characterizes an image by the relative frequencies of the
colors exhibited by samples of the raw image.

Basic Elements of Oracle SQL 3-35

Data Type Comparison Rules

■

SI_FeatureList
A list containing up to four of the image features represented by the preceding
object types (SI_AverageColor, SI_ColorHistogram, SI_PositionalColor, and SI_
Texture), where each feature is associated with a feature weight.

■

SI_PositionalColor
Given an image divided into n by m rectangles, the SI_PositionalColor object
type represents the feature that characterizes an image by the n by m most
significant colors of the rectangles.

■

SI_StillImage
Represents digital images with inherent image characteristics such as height,
width, and format.

■

SI_Texture
Represents a feature that characterizes an image by the size of repeating items
(coarseness), brightness variations (contrast), and predominant direction
(directionality).
See Also:
■

■

Oracle Multimedia DICOM Developer's Guide for information on the
ORDDicom object type
Oracle Multimedia Reference for information on all other object
types listed in this section

Expression Filter Type
The Oracle Expression Filter allows application developers to manage and evaluate
conditional expressions that describe users' interests in data. The Expression Filter
includes the following data type:

Expression
Expression Filter uses a virtual data type called Expression to manage and evaluate
conditional expressions as data in database tables. The Expression Filter creates a
column of Expression data type from a VARCHAR2 column by assigning an attribute set
to the column. This assignment enables a data constraint that ensures the validity of
expressions stored in the column.
You can define conditions using the EVALUATE operator on an Expression data type to
evaluate the expressions stored in a column for some data. If you are using Enterprise
Edition, then you can also define an Expression Filter index on a column of
Expression data type to process queries using the EVALUATE operator.
See Also: Oracle Database Rules Manager and Expression Filter
Developer's Guide for more information on the Expression Filter

Data Type Comparison Rules
This section describes how Oracle Database compares values of each data type.

Numeric Values
A larger value is considered greater than a smaller one. All negative numbers are less
than zero and all positive numbers. Thus, -1 is less than 100; -100 is less than -1.

3-36 Oracle Database SQL Language Reference

Data Type Comparison Rules

The floating-point value NaN (not a number) is greater than any other numeric value
and is equal to itself.
See Also: "Numeric Precedence" on page 3-14 and "Floating-Point
Numbers" on page 3-12 for more information on comparison
semantics

Date Values
A later date is considered greater than an earlier one. For example, the date equivalent
of '29-MAR-2005' is less than that of '05-JAN-2006' and '05-JAN-2006 1:35pm' is greater
than '05-JAN-2005 10:09am'.

Character Values
Character values are compared on the basis of two measures:
■

Binary or linguistic sorting

■

Blank-padded or nonpadded comparison semantics

The following subsections describe the two measures.
Binary and Linguistic Comparisons
In binary comparison, which is the default, Oracle compares character strings
according to the concatenated value of the numeric codes of the characters in the
database character set. One character is greater than another if it has a greater numeric
value than the other in the character set. Oracle considers blanks to be less than any
character, which is true in most character sets.
These are some common character sets:
■

7-bit ASCII (American Standard Code for Information Interchange)

■

EBCDIC Code (Extended Binary Coded Decimal Interchange Code)

■

ISO 8859/1 (International Organization for Standardization)

■

JEUC Japan Extended UNIX

Linguistic comparison is useful if the binary sequence of numeric codes does not
match the linguistic sequence of the characters you are comparing. Linguistic
comparison is used if the NLS_SORT parameter has a setting other than BINARY and the
NLS_COMP parameter is set to LINGUISTIC. In linguistic sorting, all SQL sorting and
comparison are based on the linguistic rule specified by NLS_SORT.
See Also: Oracle Database Globalization Support Guide for more
information about linguistic sorting

Blank-Padded and Nonpadded Comparison Semantics
With blank-padded semantics, if the two values have different lengths, then Oracle
first adds blanks to the end of the shorter one so their lengths are equal. Oracle then
compares the values character by character up to the first character that differs. The
value with the greater character in the first differing position is considered greater. If
two values have no differing characters, then they are considered equal. This rule
means that two values are equal if they differ only in the number of trailing blanks.
Oracle uses blank-padded comparison semantics only when both values in the
comparison are either expressions of data type CHAR, NCHAR, text literals, or values
returned by the USER function.

Basic Elements of Oracle SQL 3-37

Data Type Comparison Rules

With nonpadded semantics, Oracle compares two values character by character up to
the first character that differs. The value with the greater character in that position is
considered greater. If two values of different length are identical up to the end of the
shorter one, then the longer value is considered greater. If two values of equal length
have no differing characters, then the values are considered equal. Oracle uses
nonpadded comparison semantics whenever one or both values in the comparison
have the data type VARCHAR2 or NVARCHAR2.
The results of comparing two character values using different comparison semantics
may vary. The table that follows shows the results of comparing five pairs of character
values using each comparison semantic. Usually, the results of blank-padded and
nonpadded comparisons are the same. The last comparison in the table illustrates the
differences between the blank-padded and nonpadded comparison semantics.
Blank-Padded

Nonpadded

'ac' > 'ab'

'ac' > 'ab'

'ab' > 'a '

'ab' > 'a

'ab' > 'a'

'ab' > 'a'

'ab' = 'ab'

'ab' = 'ab'

'a

'a

' = 'a'

'

' > 'a'

Portions of the ASCII and EBCDIC character sets appear in Table 3–8 and Table 3–9.
Uppercase and lowercase letters are not equivalent. The numeric values for the
characters of a character set may not match the linguistic sequence for a particular
language.
Table 3–8

ASCII Character Set

Symbol

Decimal value

Symbol

Decimal value

blank

32

;

59

!

33

<

60

"

34

=

61

#

35

>

62

$

36

?

63

%

37

@

64

&

38

A-Z

65-90

'

39

[

91

(

40

\

92

)

41

]

93

*

42

^

94

+

43

_

95

,

44

'

96

-

45

a-z

97-122

.

46

{

123

/

47

|

124

0-9

48-57

}

125

3-38 Oracle Database SQL Language Reference

Data Type Comparison Rules

Table 3–8 (Cont.) ASCII Character Set
Symbol
:

Decimal value
58

Table 3–9

Symbol
~

Decimal value
126

EBCDIC Character Set

Symbol

Decimal value

Symbol

Decimal value

blank

64

%

108

¢

74

_

109

.

75

>

110

<

76

?

111

(

77

:

122

+

78

#

123

|

79

@

124

&

80

'

125

!

90

=

126

$

91

"

127

*

92

a-i

129-137

)

93

j-r

145-153

;

94

s-z

162-169

ÿ

95

A-I

193-201

-

96

J-R

209-217

/

97

S-Z

226-233

Object Values
Object values are compared using one of two comparison functions: MAP and ORDER.
Both functions compare object type instances, but they are quite different from one
another. These functions must be specified as part of any object type that will be
compared with other object types.
CREATE TYPE on page 17-3 for a description of MAP and
ORDER methods and the values they return
See Also:

Varrays and Nested Tables
Comparison of nested tables is described in "Comparison Conditions" on page 7-4.

Data Type Precedence
Oracle uses data type precedence to determine implicit data type conversion, which is
discussed in the section that follows. Oracle data types take the following precedence:
■

Datetime and interval data types

■

BINARY_DOUBLE

■

BINARY_FLOAT

■

NUMBER

Basic Elements of Oracle SQL 3-39

Data Type Comparison Rules

■

Character data types

■

All other built-in data types

Data Conversion
Generally an expression cannot contain values of different data types. For example, an
expression cannot multiply 5 by 10 and then add 'JAMES'. However, Oracle supports
both implicit and explicit conversion of values from one data type to another.

Implicit and Explicit Data Conversion
Oracle recommends that you specify explicit conversions, rather than rely on implicit
or automatic conversions, for these reasons:
■

■

■

■

■

SQL statements are easier to understand when you use explicit data type
conversion functions.
Implicit data type conversion can have a negative impact on performance,
especially if the data type of a column value is converted to that of a constant
rather than the other way around.
Implicit conversion depends on the context in which it occurs and may not work
the same way in every case. For example, implicit conversion from a datetime
value to a VARCHAR2 value may return an unexpected year depending on the value
of the NLS_DATE_FORMAT parameter.
Algorithms for implicit conversion are subject to change across software releases
and among Oracle products. Behavior of explicit conversions is more predictable.
If implicit data type conversion occurs in an index expression, then Oracle
Database might not use the index because it is defined for the pre-conversion data
type. This can have a negative impact on performance.

Implicit Data Conversion
Oracle Database automatically converts a value from one data type to another when
such a conversion makes sense.
Table 3–10 is a matrix of Oracle implicit conversions. The table shows all possible
conversions, without regard to the direction of the conversion or the context in which
it is made. The rules governing these details follow the table.

NCLOB

BLOB

CLOB

ROWID

RAW

LONG

BINARY_DOUBLE

BINARY_FLOAT

NUMBER

INTERVAL

DATETIME/

DATE

NVARCHAR2

NCHAR

VARCHAR2

Implicit Type Conversion Matrix

CHAR

Table 3–10

CHAR

--

X

X

X

X

X

X

X

X

X

X

--

X

X

X

VARCHAR2

X

--

X

X

X

X

X

X

X

X

X

X

X

--

X

NCHAR

X

X

--

X

X

X

X

X

X

X

X

X

X

--

X

NVARCHAR2

X

X

X

--

X

X

X

X

X

X

X

X

X

--

X

DATE

X

X

X

X

--

--

--

--

--

--

--

--

--

--

--

DATETIME/
INTERVAL

X

X

X

X

--

--

--

--

--

X

--

--

--

--

--

NUMBER

X

X

X

X

--

--

--

X

X

--

--

--

--

--

--

3-40 Oracle Database SQL Language Reference

Data Type Comparison Rules

NCLOB

BLOB

CLOB

ROWID

RAW

LONG

BINARY_DOUBLE

BINARY_FLOAT

NUMBER

INTERVAL

DATETIME/

DATE

NVARCHAR2

NCHAR

CHAR

VARCHAR2

Table 3–10 (Cont.) Implicit Type Conversion Matrix

BINARY_
FLOAT

X

X

X

X

--

--

X

--

X

--

--

--

--

--

--

BINARY_
DOUBLE

X

X

X

X

--

--

X

X

--

--

--

--

--

--

--

LONG

X

X

X

X

--

X1

--

--

--

--

X

--

X

--

X

RAW

X

X

X

X

--

--

--

--

--

X

--

--

--

X

--

ROWID

--

X

X

X

--

--

--

--

--

--

--

--

--

--

--

CLOB

X

X

X

X

--

--

--

--

--

X

--

--

--

--

X

BLOB

--

--

--

--

--

--

--

--

--

--

X

--

--

--

--

NCLOB

X

X

X

X

--

--

--

--

--

X

--

--

X

--

--

1

You cannot convert LONG to INTERVAL directly, but you can convert LONG to VARCHAR2 using TO_CHAR(interval), and then convert
the resulting VARCHAR2 value to INTERVAL.

The following rules govern implicit data type conversions:
■

■

■

■

■

■

■

■
■

■

During INSERT and UPDATE operations, Oracle converts the value to the data type
of the affected column.
During SELECT FROM operations, Oracle converts the data from the column to the
type of the target variable.
When manipulating numeric values, Oracle usually adjusts precision and scale to
allow for maximum capacity. In such cases, the numeric data type resulting from
such operations can differ from the numeric data type found in the underlying
tables.
When comparing a character value with a numeric value, Oracle converts the
character data to a numeric value.
Conversions between character values or NUMBER values and floating-point
number values can be inexact, because the character types and NUMBER use decimal
precision to represent the numeric value, and the floating-point numbers use
binary precision.
When converting a CLOB value into a character data type such as VARCHAR2, or
converting BLOB to RAW data, if the data to be converted is larger than the target
data type, then the database returns an error.
During conversion from a timestamp value to a DATE value, the fractional seconds
portion of the timestamp value is truncated. This behavior differs from earlier
releases of Oracle Database, when the fractional seconds portion of the timestamp
value was rounded.
Conversions from BINARY_FLOAT to BINARY_DOUBLE are exact.
Conversions from BINARY_DOUBLE to BINARY_FLOAT are inexact if the BINARY_
DOUBLE value uses more bits of precision that supported by the BINARY_FLOAT.
When comparing a character value with a DATE value, Oracle converts the
character data to DATE.

Basic Elements of Oracle SQL 3-41

Data Type Comparison Rules

■

■

■

■

■

■

■

When you use a SQL function or operator with an argument of a data type other
than the one it accepts, Oracle converts the argument to the accepted data type.
When making assignments, Oracle converts the value on the right side of the
equal sign (=) to the data type of the target of the assignment on the left side.
During concatenation operations, Oracle converts from noncharacter data types to
CHAR or NCHAR.
During arithmetic operations on and comparisons between character and
noncharacter data types, Oracle converts from any character data type to a
numeric, date, or rowid, as appropriate. In arithmetic operations between
CHAR/VARCHAR2 and NCHAR/NVARCHAR2, Oracle converts to a NUMBER.
Most SQL character functions are enabled to accept CLOBs as parameters, and
Oracle performs implicit conversions between CLOB and character types. Therefore,
functions that are not yet enabled for CLOBs can accept CLOBs through implicit
conversion. In such cases, Oracle converts the CLOBs to CHAR or VARCHAR2 before the
function is invoked. If the CLOB is larger than 4000 bytes, then Oracle converts only
the first 4000 bytes to CHAR.
When converting RAW or LONG RAW data to and from character data, the binary data
is represented in hexadecimal form, with one hexadecimal character representing
every four bits of RAW data. Refer to "RAW and LONG RAW Data Types" on
page 3-23 for more information.
Comparisons between CHAR and VARCHAR2 and between NCHAR and NVARCHAR2
types may entail different character sets. The default direction of conversion in
such cases is from the database character set to the national character set.
Table 3–11 shows the direction of implicit conversions between different character
types.

Table 3–11

Conversion Direction of Different Character Types
to CHAR

to VARCHAR2

to NCHAR

to NVARCHAR2

from CHAR

--

VARCHAR2

NCHAR

NVARCHAR2

from VARCHAR2

VARCHAR2

--

NVARCHAR2

NVARCHAR2

from NCHAR

NCHAR

NCHAR

--

NVARCHAR2

from NVARCHAR2

NVARCHAR2

NVARCHAR2

NVARCHAR2

--

User-defined types such as collections cannot be implicitly converted, but must be
explicitly converted using CAST ... MULTISET.

Implicit Data Conversion Examples
The text literal '10' has data type CHAR. Oracle implicitly
converts it to the NUMBER data type if it appears in a numeric expression as in the
following statement:
Text Literal Example

SELECT salary + '10'
FROM employees;

When a condition compares a character
value and a NUMBER value, Oracle implicitly converts the character value to a NUMBER
value, rather than converting the NUMBER value to a character value. In the following
statement, Oracle implicitly converts '200' to 200:
Character and Number Values Example

SELECT last_name

3-42 Oracle Database SQL Language Reference

Data Type Comparison Rules

FROM employees
WHERE employee_id = '200';

In the following statement, Oracle implicitly converts '24-JUN-06' to a
DATE value using the default date format 'DD-MON-YY':
Date Example

SELECT last_name
FROM employees
WHERE hire_date = '24-JUN-06';

Explicit Data Conversion
You can explicitly specify data type conversions using SQL conversion functions.
Table 3–12 shows SQL functions that explicitly convert a value from one data type to
another.

TO_
NUMBER

TO_NCHAR
(char.)

TO_DATE

HEXTORAW

CHARTO=R -OWID

TO_CLOB

--

--

--

TO_TIMESTAMP

to BINARY_DOUBLE

to BINARY_FLOAT

BLOB

to CLOB, NCLOB,

LONG RAW

to LONG,

to ROWID

to RAW

to Datetime/

to NUMBER

NVARCHAR2

TO_CHAR
(char.)

NCHAR,

from CHAR,
VARCHAR2,
NCHAR,
NVARCHAR2

VARCHAR2,

Explicit Type Conversions

to CHAR,

Table 3–12

Interval

You cannot specify LONG and LONG RAW values in cases in which Oracle can perform
implicit data type conversion. For example, LONG and LONG RAW values cannot appear in
expressions with functions or operators. Refer to "LONG Data Type" on page 3-14 for
information on the limitations on LONG and LONG RAW data types.

TO_
BINARY_
FLOAT

TO_
BINARY_
DOUBLE

--

TO_
BINARY_
FLOAT

TO_
BINARY_
DOUBLE

TO_NCLOB

TO_TIMESTAMP_
TZ
TO_YMINTERVAL
TO_DSINTERVAL

from NUMBER

TO_CHAR
(number)

--

NUMTOYMINTERVAL

TO_NCHAR
(number)

from Datetime/
Interval

TO_CHAR
(date)

TO_DATE

NUMTODSINTERVAL
--

--

--

--

--

--

--

--

--

--

--

--

--

TO_BLOB

--

--

TO_NCHAR
(datetime)
from RAW

RAWTOHEX
RAWTONHEX

from ROWID

ROWIDTOCHAR

--

--

--

--

--

--

--

--

from LONG /
LONG RAW

--

--

--

--

--

--

TO_LOB

--

--

from CLOB,
NCLOB, BLOB

TO_CHAR

--

--

--

--

--

TO_CLOB

--

--

TO_NCHAR

TO_NCLOB

Basic Elements of Oracle SQL 3-43

Data Type Comparison Rules

from CLOB,
NCLOB, BLOB

TO_CHAR

from BINARY_
FLOAT

TO_CHAR
(char.)

--

--

--

--

--

TO_NCHAR

TO_CLOB

TO_CHAR
(char.)

to BINARY_DOUBLE

to BINARY_FLOAT

BLOB
--

--

TO_NCLOB
TO_
NUMBER

--

--

--

--

--

TO_
BINARY_
FLOAT

TO_
BINARY_
DOUBLE

TO_
NUMBER

--

--

--

--

--

TO_
BINARY_
FLOAT

TO_
BINARY_
DOUBLE

TO_NCHAR
(char.)
from BINARY_
DOUBLE

to CLOB, NCLOB,

LONG RAW

to LONG,

to ROWID

to RAW

Interval

to Datetime/

to NUMBER

NVARCHAR2

NCHAR,

VARCHAR2,

to CHAR,

Table 3–12 (Cont.) Explicit Type Conversions

TO_NCHAR
(char.)

See Also: "Conversion Functions" on page 5-6 for details on all of
the explicit conversion functions

Security Considerations for Data Conversion
When a datetime value is converted to text, either by implicit conversion or by explicit
conversion that does not specify a format model, the format model is defined by one of
the globalization session parameters. Depending on the source data type, the
parameter name is NLS_DATE_FORMAT, NLS_TIMESTAMP_FORMAT, or NLS_TIMESTAMP_TZ_
FORMAT. The values of these parameters can be specified in the client environment or in
an ALTER SESSION statement.
The dependency of format models on session parameters can have a negative impact
on database security when conversion without an explicit format model is applied to a
datetime value that is being concatenated to text of a dynamic SQL statement.
Dynamic SQL statements are those statements whose text is concatenated from
fragments before being passed to a database for execution. Dynamic SQL is frequently
associated with the built-in PL/SQL package DBMS_SQL or with the PL/SQL statement
EXECUTE IMMEDIATE, but these are not the only places where dynamically constructed
SQL text may be passed as argument. For example:
EXECUTE IMMEDIATE
'SELECT last_name FROM employees WHERE hire_date > ''' || start_date || '''';

where start_date has the data type DATE.
In the above example, the value of start_date is converted to text using a format
model specified in the session parameter NLS_DATE_FORMAT. The result is concatenated
into SQL text. A datetime format model can consist simply of literal text enclosed in
double quotation marks. Therefore, any user who can explicitly set globalization
parameters for a session can decide what text is produced by the above conversion. If
the SQL statement is executed by a PL/SQL procedure, the procedure becomes
vulnerable to SQL injection through the session parameter. If the procedure runs with
definer's rights, with higher privileges than the session itself, the user can gain
unauthorized access to sensitive data.

3-44 Oracle Database SQL Language Reference

Literals

Oracle Database PL/SQL Language Reference for further
examples and for recommendations on avoiding this security risk

See Also:

This security risk also applies to middle-tier applications that
construct SQL text from datetime values converted to text by the
database or by OCI datetime functions. Those applications are
vulnerable if session globalization parameters are obtained from a
user preference.

Note:

Implicit and explicit conversion for numeric values may also suffer from the analogous
problem, as the conversion result may depend on the session parameter NLS_NUMERIC_
CHARACTERS. This parameter defines the decimal and group separator characters. If the
decimal separator is defined to be the quotation mark or the double quotation mark,
some potential for SQL injection emerges.
See Also:
■

■

Oracle Database Globalization Support Guide for detailed
descriptions of the session globalization parameters
"Format Models" on page 3-56 for information on the format
models

Literals
The terms literal and constant value are synonymous and refer to a fixed data value.
For example, 'JACK', 'BLUE ISLAND', and '101' are all character literals; 5001 is a
numeric literal. Character literals are enclosed in single quotation marks so that Oracle
can distinguish them from schema object names.
This section contains these topics:
■

Text Literals

■

Numeric Literals

■

Datetime Literals

■

Interval Literals

Many SQL statements and functions require you to specify character and numeric
literal values. You can also specify literals as part of expressions and conditions. You
can specify character literals with the 'text' notation, national character literals with
the N'text' notation, and numeric literals with the integer, or number notation,
depending on the context of the literal. The syntactic forms of these notations appear
in the sections that follow.
To specify a datetime or interval data type as a literal, you must take into account any
optional precisions included in the data types. Examples of specifying datetime and
interval data types as literals are provided in the relevant sections of "Data Types" on
page 3-1.

Text Literals
Use the text literal notation to specify values whenever 'string' appears in the syntax
of expressions, conditions, SQL functions, and SQL statements in other parts of this
reference. This reference uses the terms text literal, character literal, and string
interchangeably. Text, character, and string literals are always surrounded by single
Basic Elements of Oracle SQL 3-45

Literals

quotation marks. If the syntax uses the term char, then you can specify either a text
literal or another expression that resolves to character data — for example, the last_
name column of the hr.employees table. When char appears in the syntax, the single
quotation marks are not used.
The syntax of text literals or strings follows:
string::=
N
n

c
’

’
Q
’

quote_delimiter

c

quote_delimiter

’

q

where N or n specifies the literal using the national character set (NCHAR or NVARCHAR2
data). By default, text entered using this notation is translated into the national
character set by way of the database character set when used by the server. To avoid
potential loss of data during the text literal conversion to the database character set, set
the environment variable ORA_NCHAR_LITERAL_REPLACE to TRUE. Doing so transparently
replaces the n' internally and preserves the text literal for SQL processing.
See Also: Oracle Database Globalization Support Guide for more
information about N-quoted literals

In the top branch of the syntax:
■

■

c is any member of the user's character set. A single quotation mark (') within the
literal must be preceded by an escape character. To represent one single quotation
mark within a literal, enter two single quotation marks.
' ' are two single quotation marks that begin and end text literals.

In the bottom branch of the syntax:
■

■

■

■

Q or q indicates that the alternative quoting mechanism will be used. This
mechanism allows a wide range of delimiters for the text string.
The outermost ' ' are two single quotation marks that precede and follow,
respectively, the opening and closing quote_delimiter.
c is any member of the user's character set. You can include quotation marks (") in
the text literal made up of c characters. You can also include the quote_delimiter,
as long as it is not immediately followed by a single quotation mark.
quote_delimiter is any single- or multibyte character except space, tab, and
return. The quote_delimiter can be a single quotation mark. However, if the
quote_delimiter appears in the text literal itself, ensure that it is not immediately
followed by a single quotation mark.
If the opening quote_delimiter is one of [, {, <, or (, then the closing quote_
delimiter must be the corresponding ], }, >, or ). In all other cases, the opening
and closing quote_delimiter must be the same character.

Text literals have properties of both the CHAR and VARCHAR2 data types:
■

■

Within expressions and conditions, Oracle treats text literals as though they have
the data type CHAR by comparing them using blank-padded comparison semantics.
A text literal can have a maximum length of 4000 bytes.

3-46 Oracle Database SQL Language Reference

Literals

Here are some valid text literals:
'Hello'
'ORACLE.dbs'
'Jackie''s raincoat'
'09-MAR-98'
N'nchar literal'

Here are some valid text literals using the alternative quoting mechanism:
q'!name LIKE '%DBMS_%%'!'
q'<'So,' she said, 'It's finished.'>'
q'{SELECT * FROM employees WHERE last_name = 'Smith';}'
nq'ï Ÿ1234 ï'
q'"name like '['"'

"Blank-Padded and Nonpadded Comparison Semantics"
on page 3-37

See Also:

Numeric Literals
Use numeric literal notation to specify fixed and floating-point numbers.

Integer Literals
You must use the integer notation to specify an integer whenever integer appears in
expressions, conditions, SQL functions, and SQL statements described in other parts of
this reference.
The syntax of integer follows:
integer::=
+
–
digit

where digit is one of 0, 1, 2, 3, 4, 5, 6, 7, 8, 9.
An integer can store a maximum of 38 digits of precision.
Here are some valid integers:
7
+255

NUMBER and Floating-Point Literals
You must use the number or floating-point notation to specify values whenever
number or n appears in expressions, conditions, SQL functions, and SQL statements in
other parts of this reference.
The syntax of number follows:

Basic Elements of Oracle SQL 3-47

Literals

number::=
+

.

–

digit

digit
.

digit

f
e

+

E

–

F
d
digit

D

where
■

■
■

■

■

+ or - indicates a positive or negative value. If you omit the sign, then a positive
value is the default.
digit is one of 0, 1, 2, 3, 4, 5, 6, 7, 8 or 9.
e or E indicates that the number is specified in scientific notation. The digits after
the E specify the exponent. The exponent can range from -130 to 125.
f or F indicates that the number is a 32-bit binary floating point number of type
BINARY_FLOAT.
d or D indicates that the number is a 64-bit binary floating point number of type
BINARY_DOUBLE.
If you omit f or F and d or D, then the number is of type NUMBER.
The suffixes f (F) and d (D) are supported only in floating-point number literals,
not in character strings that are to be converted to NUMBER. For example, if Oracle is
expecting a NUMBER and it encounters the string '9', then it converts the string to
the number 9. However, if Oracle encounters the string '9f', then conversion fails
and an error is returned.

A number of type NUMBER can store a maximum of 38 digits of precision. If the literal
requires more precision than provided by NUMBER, BINARY_FLOAT, or BINARY_DOUBLE,
then Oracle truncates the value. If the range of the literal exceeds the range supported
by NUMBER, BINARY_FLOAT, or BINARY_DOUBLE, then Oracle raises an error.
Numeric literals are SQL syntax elements, which are not sensitive to NLS settings. The
decimal separator character in numeric literals is always the period (.). However, if a
text literal is specified where a numeric value is expected, then the text literal is
implicitly converted to a number in an NLS-sensitive way. The decimal separator
contained in the text literal must be the one established with the initialization
parameter NLS_NUMERIC_CHARACTERS. Oracle recommends that you use numeric literals
in SQL scripts to make them work independently of the NLS environment.
The following examples illustrate the behavior of decimal separators in numeric
literals and text literals. These examples assume that you have established the comma
(,) as the NLS decimal separator for the current session with the following statement:
ALTER SESSION SET NLS_NUMERIC_CHARACTERS=',.';

The previous statement also establishes the period (.) as the NLS group separator, but
that is irrelevant for these examples.

3-48 Oracle Database SQL Language Reference

Literals

This example uses the required decimal separator (.) in the numeric literal 1.23 and
the established NLS decimal separator (,) in the text literal '2,34'. The text literal is
converted to the numeric value 2.34, and the output is displayed using commas for the
decimal separators.
SELECT 2 * 1.23, 3 * '2,34' FROM DUAL;
2*1.23
3*'2,34'
---------- ---------2,46
7,02

The next example shows that a comma is not treated as part of a numeric literal.
Rather, the comma is treated as the delimiter in a list of two numeric expressions: 2*1
and 23.
SELECT 2 * 1,23 FROM DUAL;
2*1
23
---------- ---------2
23

The next example shows that the decimal separator in a text literal must match the
NLS decimal separator in order for implicit text-to-number conversion to succeed. The
following statement fails because the decimal separator (.) does not match the
established NLS decimal separator (,):
SELECT 3 * '2.34' FROM DUAL;
*
ERROR at line 1:
ORA-01722: invalid number

See Also: ALTER SESSION on page 11-45 and Oracle Database
Reference

Here are some valid NUMBER literals:
25
+6.34
0.5
25e-03
-1

Here are some valid floating-point number literals:
25f
+6.34F
0.5d
-1D

You can also use the following supplied floating-point literals in situations where a
value cannot be expressed as a numeric literal:

Literal

Meaning

Example

binary_float_nan

A value of type
SELECT COUNT(*)
BINARY_FLOAT for
FROM employees
which the condition
WHERE TO_BINARY_FLOAT(commission_pct)
IS NAN is true
!= BINARY_FLOAT_NAN;

Basic Elements of Oracle SQL 3-49

Literals

Literal

Meaning

Example

binary_float_
infinity

Single-precision
positive infinity

SELECT COUNT(*)
FROM employees
WHERE salary < BINARY_FLOAT_INFINITY;

binary_double_nan

A value of type
SELECT COUNT(*)
BINARY_DOUBLE for
FROM employees
which the condition
WHERE TO_BINARY_FLOAT(commission_pct)
IS NAN is true
!= BINARY_FLOAT_NAN;

binary_double_
infinity

Double-precision
positive infinity

SELECT COUNT(*)
FROM employees
WHERE salary < BINARY_DOUBLE_INFINITY;

Datetime Literals
Oracle Database supports four datetime data types: DATE, TIMESTAMP, TIMESTAMP WITH
TIME ZONE, and TIMESTAMP WITH LOCAL TIME ZONE.
You can specify a DATE value as a string literal, or you can convert a
character or numeric value to a date value with the TO_DATE function. DATE literals are
the only case in which Oracle Database accepts a TO_DATE expression in place of a
string literal.

Date Literals

To specify a DATE value as a literal, you must use the Gregorian calendar. You can
specify an ANSI literal, as shown in this example:
DATE '1998-12-25'

The ANSI date literal contains no time portion, and must be specified in the format
'YYYY-MM-DD'. Alternatively you can specify an Oracle date value, as in the following
example:
TO_DATE('98-DEC-25 17:30','YY-MON-DD HH24:MI')

The default date format for an Oracle DATE value is specified by the initialization
parameter NLS_DATE_FORMAT. This example date format includes a two-digit number
for the day of the month, an abbreviation of the month name, the last two digits of the
year, and a 24-hour time designation.
Oracle automatically converts character values that are in the default date format into
date values when they are used in date expressions.
If you specify a date value without a time component, then the default time is
midnight (00:00:00 or 12:00:00 for 24-hour and 12-hour clock time, respectively). If you
specify a date value without a date, then the default date is the first day of the current
month.
Oracle DATE columns always contain both the date and time fields. Therefore, if you
query a DATE column, then you must either specify the time field in your query or
ensure that the time fields in the DATE column are set to midnight. Otherwise, Oracle
may not return the query results you expect. You can use the TRUNC date function to set
the time field to midnight, or you can include a greater-than or less-than condition in
the query instead of an equality or inequality condition.
Here are some examples that assume a table my_table with a number column row_num
and a DATE column datecol:
INSERT INTO my_table VALUES (1, SYSDATE);
INSERT INTO my_table VALUES (2, TRUNC(SYSDATE));

3-50 Oracle Database SQL Language Reference

Literals

SELECT *
FROM my_table;
ROW_NUM
---------1
2

DATECOL
--------03-OCT-02
03-OCT-02

SELECT *
FROM my_table
WHERE datecol > TO_DATE('02-OCT-02', 'DD-MON-YY');
ROW_NUM
---------1
2

DATECOL
--------03-OCT-02
03-OCT-02

SELECT *
FROM my_table
WHERE datecol = TO_DATE('03-OCT-02','DD-MON-YY');
ROW_NUM DATECOL
---------- --------2 03-OCT-02

If you know that the time fields of your DATE column are set to midnight, then you can
query your DATE column as shown in the immediately preceding example, or by using
the DATE literal:
SELECT *
FROM my_table
WHERE datecol = DATE '2002-10-03';

ROW_NUM DATECOL
---------- --------2 03-OCT-02

However, if the DATE column contains values other than midnight, then you must filter
out the time fields in the query to get the correct result. For example:
SELECT *
FROM my_table
WHERE TRUNC(datecol) = DATE '2002-10-03';

ROW_NUM
---------1
2

DATECOL
--------03-OCT-02
03-OCT-02

Oracle applies the TRUNC function to each row in the query, so performance is better if
you ensure the midnight value of the time fields in your data. To ensure that the time
fields are set to midnight, use one of the following methods during inserts and
updates:
■

Use the TO_DATE function to mask out the time fields:
INSERT INTO my_table
VALUES (3, TO_DATE('3-OCT-2002','DD-MON-YYYY'));

Basic Elements of Oracle SQL 3-51

Literals

■

Use the DATE literal:
INSERT INTO my_table
VALUES (4, '03-OCT-02');

■

Use the TRUNC function:
INSERT INTO my_table
VALUES (5, TRUNC(SYSDATE));

The date function SYSDATE returns the current system date and time. The function
CURRENT_DATE returns the current session date. For information on SYSDATE, the TO_*
datetime functions, and the default date format, see "Datetime Functions" on page 5-5.
TIMESTAMP Literals The TIMESTAMP data type stores year, month, day, hour, minute,
and second, and fractional second values. When you specify TIMESTAMP as a literal, the
fractional_seconds_precision value can be any number of digits up to 9, as follows:
TIMESTAMP '1997-01-31 09:26:50.124'

TIMESTAMP WITH TIME ZONE Literals The TIMESTAMP WITH TIME ZONE data type is a
variant of TIMESTAMP that includes a time zone region name or time zone offset. When
you specify TIMESTAMP WITH TIME ZONE as a literal, the fractional_seconds_precision
value can be any number of digits up to 9. For example:
TIMESTAMP '1997-01-31 09:26:56.66 +02:00'

Two TIMESTAMP WITH TIME ZONE values are considered identical if they represent the
same instant in UTC, regardless of the TIME ZONE offsets stored in the data. For
example,
TIMESTAMP '1999-04-15 8:00:00 -8:00'

is the same as
TIMESTAMP '1999-04-15 11:00:00 -5:00'

8:00 a.m. Pacific Standard Time is the same as 11:00 a.m. Eastern Standard Time.
You can replace the UTC offset with the TZR (time zone region name) format element.
For example, the following example has the same value as the preceding example:
TIMESTAMP '1999-04-15 8:00:00 US/Pacific'

To eliminate the ambiguity of boundary cases when the daylight saving time switches,
use both the TZR and a corresponding TZD format element. The following example
ensures that the preceding example will return a daylight saving time value:
TIMESTAMP '1999-10-29 01:30:00 US/Pacific PDT'

You can also express the time zone offset using a datetime expression:
SELECT TIMESTAMP '2009-10-29 01:30:00' AT TIME ZONE 'US/Pacific'
FROM DUAL;

See Also:

"Datetime Expressions" on page 6-8 for more information

If you do not add the TZD format element, and the datetime value is ambiguous, then
Oracle returns an error if you have the ERROR_ON_OVERLAP_TIME session parameter set
to TRUE. If that parameter is set to FALSE, then Oracle interprets the ambiguous
datetime as standard time in the specified region.
3-52 Oracle Database SQL Language Reference

Literals

TIMESTAMP WITH LOCAL TIME ZONE Literals The TIMESTAMP WITH LOCAL TIME ZONE
data type differs from TIMESTAMP WITH TIME ZONE in that data stored in the database is
normalized to the database time zone. The time zone offset is not stored as part of the
column data. There is no literal for TIMESTAMP WITH LOCAL TIME ZONE. Rather, you
represent values of this data type using any of the other valid datetime literals. The
table that follows shows some of the formats you can use to insert a value into a
TIMESTAMP WITH LOCAL TIME ZONE column, along with the corresponding value returned
by a query.
Value Specified in INSERT Statement

Value Returned by Query

'19-FEB-2004'

19-FEB-2004.00.00.000000 AM

SYSTIMESTAMP

19-FEB-04 02.54.36.497659 PM

TO_TIMESTAMP('19-FEB-2004', 'DD-MON-YYYY')

19-FEB-04 12.00.00.000000 AM

SYSDATE

19-FEB-04 02.55.29.000000 PM

TO_DATE('19-FEB-2004', 'DD-MON-YYYY')

19-FEB-04 12.00.00.000000 AM

TIMESTAMP'2004-02-19 8:00:00 US/Pacific'

19-FEB-04 08.00.00.000000 AM

Notice that if the value specified does not include a time component (either explicitly
or implicitly), then the value returned defaults to midnight.

Interval Literals
An interval literal specifies a period of time. You can specify these differences in terms
of years and months, or in terms of days, hours, minutes, and seconds. Oracle
Database supports two types of interval literals, YEAR TO MONTH and DAY TO SECOND.
Each type contains a leading field and may contain a trailing field. The leading field
defines the basic unit of date or time being measured. The trailing field defines the
smallest increment of the basic unit being considered. For example, a YEAR TO MONTH
interval considers an interval of years to the nearest month. A DAY TO MINUTE interval
considers an interval of days to the nearest minute.
If you have date data in numeric form, then you can use the NUMTOYMINTERVAL or
NUMTODSINTERVAL conversion function to convert the numeric data into interval values.
Interval literals are used primarily with analytic functions.
"Analytic Functions" on page 5-11, NUMTODSINTERVAL
on page 5-171, and NUMTOYMINTERVAL on page 5-172

See Also:

INTERVAL YEAR TO MONTH
Specify YEAR TO MONTH interval literals using the following syntax:
interval_year_to_month::=
–
INTERVAL

’

integer

integer

’

YEAR
TO
YEAR

(

precision

)

MONTH

MONTH

Basic Elements of Oracle SQL 3-53

Literals

where
■

■

'integer [-integer]' specifies integer values for the leading and optional
trailing field of the literal. If the leading field is YEAR and the trailing field is MONTH,
then the range of integer values for the month field is 0 to 11.
precision is the maximum number of digits in the leading field. The valid range
of the leading field precision is 0 to 9 and its default value is 2.

Restriction on the Leading Field If you specify a trailing field, then it must be less
significant than the leading field. For example, INTERVAL '0-1' MONTH TO YEAR is not
valid.

The following INTERVAL YEAR TO MONTH literal indicates an interval of 123 years, 2
months:
INTERVAL '123-2' YEAR(3) TO MONTH

Examples of the other forms of the literal follow, including some abbreviated versions:
Form of Interval Literal

Interpretation

INTERVAL '123-2' YEAR(3) TO MONTH

An interval of 123 years, 2 months. You must
specify the leading field precision if it is
greater than the default of 2 digits.

INTERVAL '123' YEAR(3)

An interval of 123 years 0 months.

INTERVAL '300' MONTH(3)

An interval of 300 months.

INTERVAL '4' YEAR

Maps to INTERVAL '4-0' YEAR TO MONTH and
indicates 4 years.

INTERVAL '50' MONTH

Maps to INTERVAL '4-2' YEAR TO MONTH and
indicates 50 months or 4 years 2 months.

INTERVAL '123' YEAR

Returns an error, because the default precision
is 2, and '123' has 3 digits.

You can add or subtract one INTERVAL YEAR TO MONTH literal to or from another to yield
another INTERVAL YEAR TO MONTH literal. For example:
INTERVAL '5-3' YEAR TO MONTH + INTERVAL'20' MONTH =
INTERVAL '6-11' YEAR TO MONTH

INTERVAL DAY TO SECOND
Specify DAY TO SECOND interval literals using the following syntax:

3-54 Oracle Database SQL Language Reference

Literals

interval_day_to_second::=
integer
INTERVAL

’

integer

time_expr

’

time_expr

DAY

(

leading_precision

)

HOUR
MINUTE
,
(

fractional_seconds_precision

leading_precision

)

SECOND

DAY
HOUR
TO

MINUTE
(

fractional_seconds_precision

)

SECOND

where
■

■

■

■

integer specifies the number of days. If this value contains more digits than the
number specified by the leading precision, then Oracle returns an error.
time_expr specifies a time in the format HH[:MI[:SS[.n]]] or MI[:SS[.n]] or SS[.n],
where n specifies the fractional part of a second. If n contains more digits than the
number specified by fractional_seconds_precision, then n is rounded to the
number of digits specified by the fractional_seconds_precision value. You can
specify time_expr following an integer and a space only if the leading field is DAY.
leading_precision is the number of digits in the leading field. Accepted values
are 0 to 9. The default is 2.
fractional_seconds_precision is the number of digits in the fractional part of
the SECOND datetime field. Accepted values are 1 to 9. The default is 6.

If you specify a trailing field, then it must be less
significant than the leading field. For example, INTERVAL MINUTE TO DAY is not valid. As
a result of this restriction, if SECOND is the leading field, the interval literal cannot have
any trailing field.
Restriction on the Leading Field:

The valid range of values for the trailing field are as follows:
■

HOUR: 0 to 23

■

MINUTE: 0 to 59

■

SECOND: 0 to 59.999999999

Examples of the various forms of INTERVAL DAY TO SECOND literals follow, including
some abbreviated versions:

Basic Elements of Oracle SQL 3-55

Format Models

Form of Interval Literal

Interpretation

INTERVAL '4 5:12:10.222' DAY TO
SECOND(3)

4 days, 5 hours, 12 minutes, 10 seconds, and
222 thousandths of a second.

INTERVAL '4 5:12' DAY TO MINUTE

4 days, 5 hours and 12 minutes.

INTERVAL '400 5' DAY(3) TO HOUR

400 days 5 hours.

INTERVAL '400' DAY(3)

400 days.

INTERVAL '11:12:10.2222222' HOUR TO
SECOND(7)

11 hours, 12 minutes, and 10.2222222 seconds.

INTERVAL '11:20' HOUR TO MINUTE

11 hours and 20 minutes.

INTERVAL '10' HOUR

10 hours.

INTERVAL '10:22' MINUTE TO SECOND

10 minutes 22 seconds.

INTERVAL '10' MINUTE

10 minutes.

INTERVAL '4' DAY

4 days.

INTERVAL '25' HOUR

25 hours.

INTERVAL '40' MINUTE

40 minutes.

INTERVAL '120' HOUR(3)

120 hours.

INTERVAL '30.12345' SECOND(2,4)

30.1235 seconds. The fractional second '12345'
is rounded to '1235' because the precision is 4.

You can add or subtract one DAY TO SECOND interval literal from another DAY TO SECOND
literal. For example.
INTERVAL'20' DAY - INTERVAL'240' HOUR = INTERVAL'10-0' DAY TO SECOND

Format Models
A format model is a character literal that describes the format of datetime or numeric
data stored in a character string. A format model does not change the internal
representation of the value in the database. When you convert a character string into a
date or number, a format model determines how Oracle Database interprets the string.
In SQL statements, you can use a format model as an argument of the TO_CHAR and TO_
DATE functions to specify:
■

The format for Oracle to use to return a value from the database

■

The format for a value you have specified for Oracle to store in the database

For example:
■

The datetime format model for the string '17:45:29' is 'HH24:MI:SS'.

■

The datetime format model for the string '11-Nov-1999' is 'DD-Mon-YYYY'.

■

The number format model for the string '$2,304.25' is '$9,999.99'.

For lists of number and datetime format model elements, see Table 3–13, " Number
Format Elements" on page 3-58 and Table 3–15, " Datetime Format Elements" on
page 3-61.
The values of some formats are determined by the value of initialization parameters.
For such formats, you can specify the characters returned by these format elements
implicitly using the initialization parameter NLS_TERRITORY. You can change the
default date format for your session with the ALTER SESSION statement.

3-56 Oracle Database SQL Language Reference

Format Models

See Also:
■

■

■

ALTER SESSION on page 11-45 for information on changing the
values of these parameters and Format Model Examples on
page 3-68 for examples of using format models
TO_CHAR (datetime) on page 5-302, TO_CHAR (number) on
page 5-305, and TO_DATE on page 5-308
Oracle Database Reference and Oracle Database Globalization Support
Guide for information on these parameters

This remainder of this section describes how to use the following format models:
■

Number Format Models

■

Datetime Format Models

■

Format Model Modifiers

Number Format Models
You can use number format models in the following functions:
■

■

■

In the TO_CHAR function to translate a value of NUMBER, BINARY_FLOAT, or BINARY_
DOUBLE data type to VARCHAR2 data type
In the TO_NUMBER function to translate a value of CHAR or VARCHAR2 data type to
NUMBER data type
In the TO_BINARY_FLOAT and TO_BINARY_DOUBLE functions to translate CHAR and
VARCHAR2 expressions to BINARY_FLOAT or BINARY_DOUBLE values

All number format models cause the number to be rounded to the specified number of
significant digits. If a value has more significant digits to the left of the decimal place
than are specified in the format, then pound signs (#) replace the value. This event
typically occurs when you are using TO_CHAR with a restrictive number format string,
causing a rounding operation.
■

■

If a positive NUMBER value is extremely large and cannot be represented in the
specified format, then the infinity sign (~) replaces the value. Likewise, if a
negative NUMBER value is extremely small and cannot be represented by the
specified format, then the negative infinity sign replaces the value (-~).
If a BINARY_FLOAT or BINARY_DOUBLE value is converted to CHAR or NCHAR, and the
input is either infinity or NaN (not a number), then Oracle always returns the
pound signs to replace the value. However, if you omit the format model, then
Oracle returns either Inf or Nan as a string.

Number Format Elements
A number format model is composed of one or more number format elements. The
tables that follow list the elements of a number format model and provide some
examples.
Negative return values automatically contain a leading negative sign and positive
values automatically contain a leading space unless the format model contains the MI,
S, or PR format element.

Basic Elements of Oracle SQL 3-57

Format Models

Table 3–13

Number Format Elements

Element

Example

Description

, (comma)

9,999

Returns a comma in the specified position. You can specify multiple commas in a
number format model.
Restrictions:
■
■

. (period)

99.99

A comma element cannot begin a number format model.
A comma cannot appear to the right of a decimal character or period in a
number format model.

Returns a decimal point, which is a period (.) in the specified position.
Restriction: You can specify only one period in a number format model.

$

$9999

Returns value with a leading dollar sign.

0

0999

Returns leading zeros.

9990

Returns trailing zeros.

9

9999

Returns value with the specified number of digits with a leading space if positive
or with a leading minus if negative. Leading zeros are blank, except for a zero
value, which returns a zero for the integer part of the fixed-point number.

B

B9999

Returns blanks for the integer part of a fixed-point number when the integer part
is zero (regardless of zeros in the format model).

C

C999

Returns in the specified position the ISO currency symbol (the current value of the
NLS_ISO_CURRENCY parameter).

D

99D99

Returns in the specified position the decimal character, which is the current value
of the NLS_NUMERIC_CHARACTER parameter. The default is a period (.).
Restriction: You can specify only one decimal character in a number format model.

EEEE

9.9EEEE

Returns a value using in scientific notation.

G

9G999

Returns in the specified position the group separator (the current value of the NLS_
NUMERIC_CHARACTER parameter). You can specify multiple group separators in a
number format model.
Restriction: A group separator cannot appear to the right of a decimal character or
period in a number format model.

L

L999

Returns in the specified position the local currency symbol (the current value of
the NLS_CURRENCY parameter).

MI

9999MI

Returns negative value with a trailing minus sign (-).
Returns positive value with a trailing blank.
Restriction: The MI format element can appear only in the last position of a
number format model.

PR

9999PR

Returns negative value in .
Returns positive value with a leading and trailing blank.
Restriction: The PR format element can appear only in the last position of a
number format model.

RN

RN

Returns a value as Roman numerals in uppercase.

rn

rn

Returns a value as Roman numerals in lowercase.
Value can be an integer between 1 and 3999.

3-58 Oracle Database SQL Language Reference

Format Models

Table 3–13 (Cont.) Number Format Elements
Element

Example

Description

S

S9999

Returns negative value with a leading minus sign (-).
Returns positive value with a leading plus sign (+).

9999S

Returns negative value with a trailing minus sign (-).
Returns positive value with a trailing plus sign (+).
Restriction: The S format element can appear only in the first or last position of a
number format model.

TM

The text minimum number format model returns (in decimal output) the smallest
number of characters possible. This element is case insensitive.

TM

The default is TM9, which returns the number in fixed notation unless the output
exceeds 64 characters. If the output exceeds 64 characters, then Oracle Database
automatically returns the number in scientific notation.
Restrictions:
You cannot precede this element with any other element.

■

You can follow this element only with one 9 or one E (or e), but not with any
combination of these. The following statement returns an error:

■

SELECT TO_CHAR(1234, 'TM9e') FROM DUAL;
U

U9999

Returns in the specified position the Euro (or other) dual currency symbol,
determined by the current value of the NLS_DUAL_CURRENCY parameter.

V

999V99

Returns a value multiplied by 10n (and if necessary, round it up), where n is the
number of 9's after the V.

X

XXXX

Returns the hexadecimal value of the specified number of digits. If the specified
number is not an integer, then Oracle Database rounds it to an integer.

xxxx

Restrictions:
This element accepts only positive values or 0. Negative values return an
error.

■

You can precede this element only with 0 (which returns leading zeroes) or
FM. Any other elements return an error. If you specify neither 0 nor FM with
X, then the return always has one leading blank. Refer to the format model
modifier FM on page 3-67 for more information.

■

Table 3–14 shows the results of the following query for different values of number and
'fmt':
SELECT TO_CHAR(number, 'fmt')
FROM DUAL;
Table 3–14

Results of Number Conversions

number

'fmt'

Result

-1234567890

9999999999S

'1234567890-'

0

99.99

'

.00'

+0.1

99.99

'

.10'

-0.2

99.99

'

-.20'

0

90.99

'

0.00'

+0.1

90.99

'

0.10'

-0.2

90.99

' -0.20'

0

9999

'

0'

Basic Elements of Oracle SQL 3-59

Format Models

Table 3–14 (Cont.) Results of Number Conversions
number

'fmt'

Result

1

9999

'

1'

0

B9999

'

'

1

B9999

'

1'

0

B90.99

'

'

+123.456

999.999

' 123.456'

-123.456

999.999

'-123.456'

+123.456

FM999.009

'123.456'

+123.456

9.9EEEE

'

9.9EEEE

' 1.0E+123'

+123.456

FM9.9EEEE

'1.2E+02'

+123.45

FM999.009

'123.45'

+123.0

FM999.009

'123.00'

+123.45

L999.99

'

+123.45

FML999.99

'$123.45'

9999999999S

'1234567890+'

+1E+123

+1234567890

1.2E+02'

$123.45'

Datetime Format Models
You can use datetime format models in the following functions:
■

■

In the TO_* datetime functions to translate a character value that is in a format
other than the default format into a datetime value. (The TO_* datetime functions
are TO_DATE, TO_TIMESTAMP, and TO_TIMESTAMP_TZ.)
In the TO_CHAR function to translate a datetime value into a character value that is
in a format other than the default format (for example, to print the date from an
application)

The total length of a datetime format model cannot exceed 22 characters.
The default datetime formats are specified either explicitly with the NLS session
parameters NLS_DATE_FORMAT, NLS_TIMESTAMP_FORMAT, and NLS_TIMESTAMP_TZ_FORMAT,
or implicitly with the NLS session parameter NLS_TERRITORY. You can change the
default datetime formats for your session with the ALTER SESSION statement.
ALTER SESSION on page 11-45 and Oracle Database
Globalization Support Guide for information on the NLS parameters

See Also:

Datetime Format Elements
A datetime format model is composed of one or more datetime format elements as
listed in Table 3–15, " Datetime Format Elements" on page 3-61.
■

■

For input format models, format items cannot appear twice, and format items that
represent similar information cannot be combined. For example, you cannot use
'SYYYY' and 'BC' in the same format string.
The second column indicates whether the format element can be used in the TO_*
datetime functions. All format elements can be used in the TO_CHAR function.

3-60 Oracle Database SQL Language Reference

Format Models

■

■

The following datetime format elements can be used in timestamp and interval
format models, but not in the original DATE format model: FF, TZD, TZH, TZM, and
TZR.
Many datetime format elements are padded with blanks or leading zeroes to a
specific length. Refer to the format model modifier FM on page 3-67 for more
information.

Oracle recommends that you use the 4-digit year element
(YYYY) instead of the shorter year elements for these reasons:

Note:
■
■

The 4-digit year element eliminates ambiguity.
The shorter year elements may affect query optimization because
the year is not known at query compile time and can only be
determined at run time.

Uppercase Letters in Date Format Elements Capitalization in a spelled-out word,
abbreviation, or Roman numeral follows capitalization in the corresponding format
element. For example, the date format model 'DAY' produces capitalized words like
'MONDAY'; 'Day' produces 'Monday'; and 'day' produces 'monday'.
Punctuation and Character Literals in Datetime Format Models You can include these
characters in a date format model:
■

Punctuation such as hyphens, slashes, commas, periods, and colons

■

Character literals, enclosed in double quotation marks

These characters appear in the return value in the same location as they appear in the
format model.
Table 3–15

Datetime Format Elements
TO_*
datetime
functions?

Description

/
,
.
;
:
"text"

Yes

Punctuation and quoted text is reproduced in the result.

AD
A.D.

Yes

AD indicator with or without periods.

AM
A.M.

Yes

Meridian indicator with or without periods.

BC
B.C.

Yes

BC indicator with or without periods.

Element

Basic Elements of Oracle SQL 3-61

Format Models

Table 3–15 (Cont.) Datetime Format Elements

Element

TO_*
datetime
functions?

Description
Century.

CC
SCC

■

■

If the last 2 digits of a 4-digit year are between 01 and 99 (inclusive), then the
century is one greater than the first 2 digits of that year.
If the last 2 digits of a 4-digit year are 00, then the century is the same as the first
2 digits of that year.

For example, 2002 returns 21; 2000 returns 20.
D

Yes

Day of week (1-7). This element depends on the NLS territory of the session.

DAY

Yes

Name of day.

DD

Yes

Day of month (1-31).

DDD

Yes

Day of year (1-366).

DL

Yes

Returns a value in the long date format, which is an extension of the Oracle
Database DATE format, determined by the current value of the NLS_DATE_FORMAT
parameter. Makes the appearance of the date components (day name, month
number, and so forth) depend on the NLS_TERRITORY and NLS_LANGUAGE parameters.
For example, in the AMERICAN_AMERICA locale, this is equivalent to specifying the
format 'fmDay, Month dd, yyyy'. In the GERMAN_GERMANY locale, it is equivalent to
specifying the format 'fmDay, dd. Month yyyy'.
Restriction: You can specify this format only with the TS element, separated by
white space.

DS

Yes

Returns a value in the short date format. Makes the appearance of the date
components (day name, month number, and so forth) depend on the NLS_TERRITORY
and NLS_LANGUAGE parameters. For example, in the AMERICAN_AMERICA locale, this is
equivalent to specifying the format 'MM/DD/RRRR'. In the ENGLISH_UNITED_KINGDOM
locale, it is equivalent to specifying the format 'DD/MM/RRRR'.
Restriction: You can specify this format only with the TS element, separated by
white space.

DY

Yes

Abbreviated name of day.

E

Yes

Abbreviated era name (Japanese Imperial, ROC Official, and Thai Buddha
calendars).

EE

Yes

Full era name (Japanese Imperial, ROC Official, and Thai Buddha calendars).

FF [1..9]

Yes

Fractional seconds; no radix character is printed. Use the X format element to add
the radix character. Use the numbers 1 to 9 after FF to specify the number of digits in
the fractional second portion of the datetime value returned. If you do not specify a
digit, then Oracle Database uses the precision specified for the datetime data type or
the data type's default precision. Valid in timestamp and interval formats, but not in
DATE formats.
Examples: 'HH:MI:SS.FF'
SELECT TO_CHAR(SYSTIMESTAMP, 'SS.FF3') from DUAL;

FM

Yes

Returns a value with no leading or trailing blanks.
See Also: Additional discussion on this format model modifier in the Oracle
Database SQL Language Reference

FX

Yes

Requires exact matching between the character data and the format model.
See Also: Additional discussion on this format model modifier in the Oracle
Database SQL Language Reference

HH
HH12

Yes

Hour of day (1-12).

3-62 Oracle Database SQL Language Reference

Format Models

Table 3–15 (Cont.) Datetime Format Elements

Element

TO_*
datetime
functions?

Description

HH24

Yes

Hour of day (0-23).
Calendar week of year (1-52 or 1-53), as defined by the ISO 8601 standard.

IW

■

A calendar week starts on Monday.

■

The first calendar week of the year includes January 4.

■

The first calendar week of the year may include December 29, 30, and 31.

■

The last calendar week of the year may include January 1, 2, and 3.

IYYY

4-digit year of the year containing the calendar week, as defined by the ISO 8601
standard.

IYY
IY
I

Last 3, 2, or 1 digit(s) of the year containing the calendar week, as defined by the ISO
8601 standard.

J

Yes

Julian day; the number of days since January 1, 4712 BC. Number specified with J
must be integers.

MI

Yes

Minute (0-59).

MM

Yes

Month (01-12; January = 01).

MON

Yes

Abbreviated name of month.

MONTH

Yes

Name of month.

PM
P.M.

Yes

Meridian indicator with or without periods.
Quarter of year (1, 2, 3, 4; January - March = 1).

Q
RM

Yes

Roman numeral month (I-XII; January = I).

RR

Yes

Lets you store 20th century dates in the 21st century using only two digits.
See Also: "The RR Datetime Format Element" on page 3-65

RRRR

Yes

Round year. Accepts either 4-digit or 2-digit input. If 2-digit, provides the same
return as RR. If you do not want this functionality, then enter the 4-digit year.

SS

Yes

Second (0-59).

SSSSS

Yes

Seconds past midnight (0-86399).

TS

Yes

Returns a value in the short time format. Makes the appearance of the time
components (hour, minutes, and so forth) depend on the NLS_TERRITORY and NLS_
LANGUAGE initialization parameters.
Restriction: You can specify this format only with the DL or DS element, separated by
white space.

TZD

Yes

Daylight saving information. The TZD value is an abbreviated time zone string with
daylight saving information. It must correspond with the region specified in TZR.
Valid in timestamp and interval formats, but not in DATE formats.
Example: PST (for US/Pacific standard time); PDT (for US/Pacific daylight time).

TZH

Yes

Time zone hour. (See TZM format element.) Valid in timestamp and interval formats,
but not in DATE formats.
Example: 'HH:MI:SS.FFTZH:TZM'.

TZM

Yes

Time zone minute. (See TZH format element.) Valid in timestamp and interval
formats, but not in DATE formats.
Example: 'HH:MI:SS.FFTZH:TZM'.

Basic Elements of Oracle SQL 3-63

Format Models

Table 3–15 (Cont.) Datetime Format Elements

Element

TO_*
datetime
functions?

TZR

Yes

Description
Time zone region information. The value must be one of the time zone region names
supported in the database. Valid in timestamp and interval formats, but not in DATE
formats.
Example: US/Pacific

WW

Week of year (1-53) where week 1 starts on the first day of the year and continues to
the seventh day of the year.

W

Week of month (1-5) where week 1 starts on the first day of the month and ends on
the seventh.

X

Yes

Local radix character.
Example: 'HH:MI:SSXFF'.

Y,YYY

Yes

Year with comma in this position.
Year, spelled out; S prefixes BC dates with a minus sign (-).

YEAR
SYEAR
YYYY
SYYYY

Yes

4-digit year; S prefixes BC dates with a minus sign.

YYY
YY
Y

Yes

Last 3, 2, or 1 digit(s) of year.

Oracle Database converts strings to dates with some flexibility. For example, when the
TO_DATE function is used, a format model containing punctuation characters matches
an input string lacking some or all of these characters, provided each numerical
element in the input string contains the maximum allowed number of digits—for
example, two digits '05' for 'MM' or four digits '2007' for 'YYYY'. The following
statement does not return an error:
SELECT TO_CHAR(TO_DATE('0207','MM/YY'), 'MM/YY') FROM DUAL;
TO_CH
----02/07

However, the following format string does return an error, because the FX (format
exact) format modifier requires an exact match of the expression and the format string:
SELECT TO_CHAR(TO_DATE('0207', 'fxmm/yy'), 'mm/yy') FROM DUAL;
SELECT TO_CHAR(TO_DATE('0207', 'fxmm/yy'), 'mm/yy') FROM DUAL;
*
ERROR at line 1:
ORA-01861: literal does not match format string

Any non-alphanumeric character is allowed to match the punctuation characters in the
format model. For example, the following statement does not return an error:
SELECT TO_CHAR (TO_DATE('02#07','MM/YY'), 'MM/YY') FROM DUAL;
TO_CH
----02/07

3-64 Oracle Database SQL Language Reference

Format Models

See Also: "Format Model Modifiers" on page 3-67 and
"String-to-Date Conversion Rules" on page 3-69 for more information

Datetime Format Elements and Globalization Support
The functionality of some datetime format elements depends on the country and
language in which you are using Oracle Database. For example, these datetime format
elements return spelled values:
■

MONTH

■

MON

■

DAY

■

DY

■

BC or AD or B.C. or A.D.

■

AM or PM or A.M or P.M.

The language in which these values are returned is specified either explicitly with the
initialization parameter NLS_DATE_LANGUAGE or implicitly with the initialization
parameter NLS_LANGUAGE. The values returned by the YEAR and SYEAR datetime format
elements are always in English.
The datetime format element D returns the number of the day of the week (1-7). The
day of the week that is numbered 1 is specified implicitly by the initialization
parameter NLS_TERRITORY.
Oracle Database Reference and Oracle Database Globalization
Support Guide for information on globalization support initialization
parameters

See Also:

ISO Standard Date Format Elements
Oracle calculates the values returned by the datetime format elements IYYY, IYY, IY, I,
and IW according to the ISO standard. For information on the differences between
these values and those returned by the datetime format elements YYYY, YYY, YY, Y,
and WW, see the discussion of globalization support in Oracle Database Globalization
Support Guide.

The RR Datetime Format Element
The RR datetime format element is similar to the YY datetime format element, but it
provides additional flexibility for storing date values in other centuries. The RR
datetime format element lets you store 20th century dates in the 21st century by
specifying only the last two digits of the year.
If you use the TO_DATE function with the YY datetime format element, then the year
returned always has the same first 2 digits as the current year. If you use the RR
datetime format element instead, then the century of the return value varies according
to the specified two-digit year and the last two digits of the current year.
That is:
■

If the specified two-digit year is 00 to 49, then
–

If the last two digits of the current year are 00 to 49, then the returned year has
the same first two digits as the current year.

–

If the last two digits of the current year are 50 to 99, then the first 2 digits of
the returned year are 1 greater than the first 2 digits of the current year.

Basic Elements of Oracle SQL 3-65

Format Models

■

If the specified two-digit year is 50 to 99, then
–

If the last two digits of the current year are 00 to 49, then the first 2 digits of
the returned year are 1 less than the first 2 digits of the current year.

–

If the last two digits of the current year are 50 to 99, then the returned year has
the same first two digits as the current year.

The following examples demonstrate the behavior of the RR datetime format element.
RR Datetime Format Examples
Assume these queries are issued between 1950 and 1999:
SELECT TO_CHAR(TO_DATE('27-OCT-98', 'DD-MON-RR'), 'YYYY') "Year" FROM DUAL;
Year
---1998
SELECT TO_CHAR(TO_DATE('27-OCT-17', 'DD-MON-RR'), 'YYYY') "Year" FROM DUAL;
Year
---2017

Now assume these queries are issued between 2000 and 2049:
SELECT TO_CHAR(TO_DATE('27-OCT-98', 'DD-MON-RR'), 'YYYY') "Year" FROM DUAL;
Year
---1998
SELECT TO_CHAR(TO_DATE('27-OCT-17', 'DD-MON-RR'), 'YYYY') "Year" FROM DUAL;
Year
---2017

Note that the queries return the same values regardless of whether they are issued
before or after the year 2000. The RR datetime format element lets you write SQL
statements that will return the same values from years whose first two digits are
different.

Datetime Format Element Suffixes
Table 3–16 lists suffixes that can be added to datetime format elements:
Table 3–16

Date Format Element Suffixes

Suffix

Meaning

Example Element

Example Value

TH

Ordinal Number

DDTH

4TH

SP

Spelled Number

DDSP

FOUR

SPTH or THSP

Spelled, ordinal number

DDSPTH

FOURTH

Notes on date format element suffixes:
■
When you add one of these suffixes to a datetime format element, the return value
is always in English.

3-66 Oracle Database SQL Language Reference

Format Models

■

Datetime suffixes are valid only to format output. You cannot use them to insert a
date into the database.

Format Model Modifiers
The FM and FX modifiers, used in format models in the TO_CHAR function, control blank
padding and exact format checking.
A modifier can appear in a format model more than once. In such a case, each
subsequent occurrence toggles the effects of the modifier. Its effects are enabled for the
portion of the model following its first occurrence, and then disabled for the portion
following its second, and then reenabled for the portion following its third, and so on.
Fill mode. Oracle uses trailing blank characters and leading zeroes to fill format
elements to a constant width. The width is equal to the display width of the largest
element for the relevant format model:

FM

■

■

■

■

Numeric elements are padded with leading zeros to the width of the maximum
value allowed for the element. For example, the YYYY element is padded to four
digits (the length of '9999'), HH24 to two digits (the length of '23'), and DDD to three
digits (the length of '366').
The character elements MONTH, MON, DAY, and DY are padded with trailing blanks to
the width of the longest full month name, the longest abbreviated month name,
the longest full date name, or the longest abbreviated day name, respectively,
among valid names determined by the values of NLS_DATE_LANGUAGE and NLS_
CALENDAR parameters. For example, when NLS_DATE_LANGUAGE is AMERICAN and
NLS_CALENDAR is GREGORIAN (the default), the largest element for MONTH is
SEPTEMBER, so all values of the MONTH format element are padded to nine display
characters. The values of the NLS_DATE_LANGUAGE and NLS_CALENDAR parameters
are specified in the third argument to TO_CHAR and TO_* datetime functions or they
are retrieved from the NLS environment of the current session.
The character element RM is padded with trailing blanks to the length of 4, which
is the length of 'viii'.
Other character elements and spelled-out numbers (SP, SPTH, and THSP suffixes)
are not padded.

The FM modifier suppresses the above padding in the return value of the TO_CHAR
function.
FX Format exact. This modifier specifies exact matching for the character argument
and datetime format model of a TO_DATE function:
■

■

■

Punctuation and quoted text in the character argument must exactly match (except
for case) the corresponding parts of the format model.
The character argument cannot have extra blanks. Without FX, Oracle ignores extra
blanks.
Numeric data in the character argument must have the same number of digits as
the corresponding element in the format model. Without FX, numbers in the
character argument can omit leading zeros.
When FX is enabled, you can disable this check for leading zeros by using the FM
modifier as well.

If any portion of the character argument violates any of these conditions, then Oracle
returns an error message.

Basic Elements of Oracle SQL 3-67

Format Models

Format Model Examples
The following statement uses a date format model to return a character expression:
SELECT TO_CHAR(SYSDATE, 'fmDDTH') || ' of ' ||
TO_CHAR(SYSDATE, 'fmMonth') || ', ' ||
TO_CHAR(SYSDATE, 'YYYY') "Ides"
FROM DUAL;
Ides
-----------------3RD of April, 2008

The preceding statement also uses the FM modifier. If FM is omitted, then the month is
blank-padded to nine characters:
SELECT TO_CHAR(SYSDATE, 'DDTH') || ' of ' ||
TO_CHAR(SYSDATE, 'Month') || ', ' ||
TO_CHAR(SYSDATE, 'YYYY') "Ides"
FROM DUAL;
Ides
----------------------03RD of April
, 2008

The following statement places a single quotation mark in the return value by using a
date format model that includes two consecutive single quotation marks:
SELECT TO_CHAR(SYSDATE, 'fmDay') || '''s Special' "Menu"
FROM DUAL;
Menu
----------------Tuesday's Special

Two consecutive single quotation marks can be used for the same purpose within a
character literal in a format model.
Table 3–17 shows whether the following statement meets the matching conditions for
different values of char and 'fmt' using FX (the table named table has a column date_
column of data type DATE):
UPDATE table
SET date_column = TO_DATE(char, 'fmt');
Table 3–17
Modifier

Matching Character Data and Format Models with the FX Format Model
char

'fmt'

Match or Error?

'15/ JAN /1998'

'DD-MON-YYYY'

Match

' 15! JAN % /1998'

'DD-MON-YYYY'

Error

'15/JAN/1998'

'FXDD-MON-YYYY'

Error

'15-JAN-1998'

'FXDD-MON-YYYY'

Match

'1-JAN-1998'

'FXDD-MON-YYYY'

Error

'01-JAN-1998'

'FXDD-MON-YYYY'

Match

'1-JAN-1998'

'FXFMDD-MON-YYYY'

Match

3-68 Oracle Database SQL Language Reference

Format Models

Format of Return Values: Examples You can use a format model to specify the
format for Oracle to use to return values from the database to you.

The following statement selects the salaries of the employees in Department 80 and
uses the TO_CHAR function to convert these salaries into character values with the
format specified by the number format model '$99,990.99':
SELECT last_name employee, TO_CHAR(salary, '$99,990.99')
FROM employees
WHERE department_id = 80;

Because of this format model, Oracle returns salaries with leading dollar signs,
commas every three digits, and two decimal places.
The following statement selects the date on which each employee from Department 20
was hired and uses the TO_CHAR function to convert these dates to character strings
with the format specified by the date format model 'fmMonth DD, YYYY':
SELECT last_name employee, TO_CHAR(hire_date,'fmMonth DD, YYYY') hiredate
FROM employees
WHERE department_id = 20;

With this format model, Oracle returns the hire dates without blank padding (as
specified by fm), two digits for the day, and the century included in the year.
See Also: "Format Model Modifiers" on page 3-67 for a description
of the fm format element

When you insert or update a
column value, the data type of the value that you specify must correspond to the
column data type of the column. You can use format models to specify the format of a
value that you are converting from one data type to another data type required for a
column.

Supplying the Correct Format Model: Examples

For example, a value that you insert into a DATE column must be a value of the DATE
data type or a character string in the default date format (Oracle implicitly converts
character strings in the default date format to the DATE data type). If the value is in
another format, then you must use the TO_DATE function to convert the value to the
DATE data type. You must also use a format model to specify the format of the
character string.
The following statement updates Hunold's hire date using the TO_DATE function with
the format mask 'YYYY MM DD' to convert the character string '2008 05 20' to a DATE
value:
UPDATE employees
SET hire_date = TO_DATE('2008 05 20','YYYY MM DD')
WHERE last_name = 'Hunold';

String-to-Date Conversion Rules
The following additional formatting rules apply when converting string values to date
values (unless you have used the FX or FXFM modifiers in the format model to control
exact format checking):
■

■

You can omit punctuation included in the format string from the date string if all
the digits of the numerical format elements, including leading zeros, are specified.
For example, specify 02 and not 2 for two-digit format elements such as MM, DD,
and YY.
You can omit time fields found at the end of a format string from the date string.
Basic Elements of Oracle SQL 3-69

Format Models

■

■

You can use any non-alphanumeric character in the date string to match the
punctuation symbol in the format string.
If a match fails between a datetime format element and the corresponding
characters in the date string, then Oracle attempts alternative format elements, as
shown in Table 3–18.

Table 3–18

Oracle Format Matching

Original Format Element

Additional Format Elements to Try in Place of the Original

'MM'

'MON' and 'MONTH'

'MON

'MONTH'

'MONTH'

'MON'

'YY'

'YYYY'

'RR'

'RRRR'

XML Format Model
The SYS_XMLGEN function returns an instance of type XMLType containing an XML
document. Oracle provides the XMLFormat object, which lets you format the output of
the SYS_XMLGEN function.
Table 3–19 lists and describes the attributes of the XMLFormat object. The function that
implements this type follows the table.
See Also:
■

■

Table 3–19

SYS_XMLGEN on page 5-291 for information on the SYS_XMLGEN
function
Oracle XML Developer's Kit Programmer's Guide for more
information on the implementation of the XMLFormat object and its
use

Attributes of the XMLFormat Object

Attribute

Data Type

Purpose

enclTag

VARCHAR2(4000)

The name of the enclosing tag for the result of the SYS_XMLGEN
function. If the input to the function is a column name, then the
default is the column name. Otherwise the default is ROW. When
schemaType is set to USE_GIVEN_SCHEMA, this attribute also gives the
name of the XMLSchema element.

schemaType

VARCHAR2(100)

The type of schema generation for the output document. Valid values
are 'NO_SCHEMA' and 'USE_GIVEN_SCHEMA'. The default is 'NO_SCHEMA'.

schemaName

VARCHAR2(4000)

The name of the target schema Oracle uses if the value of the
schemaType is 'USE_GIVEN_SCHEMA'. If you specify schemaName, then
Oracle uses the enclosing tag as the element name.

targetNameSpace

VARCHAR2(4000)

The target namespace if the schema is specified (that is, schemaType is
GEN_SCHEMA_*, or USE_GIVEN_SCHEMA)

dburlPrefix

VARCHAR2(4000)

The URL to the database to use if WITH_SCHEMA is specified. If this
attribute is not specified, then Oracle declares the URL to the types as
a relative URL reference.

processingIns

VARCHAR2(4000)

User-provided processing instructions, which are appended to the top
of the function output before the element.

3-70 Oracle Database SQL Language Reference

Nulls

The function that implements the XMLFormat object follows:
STATIC FUNCTION createFormat(
enclTag IN varchar2 := 'ROWSET',
schemaType IN varchar2 := 'NO_SCHEMA',
schemaName IN varchar2 := null,
targetNameSpace IN varchar2 := null,
dburlPrefix IN varchar2 := null,
processingIns IN varchar2 := null) RETURN XMLGenFormatType
deterministic parallel_enable,
MEMBER PROCEDURE genSchema (spec IN varchar2),
MEMBER PROCEDURE setSchemaName(schemaName IN varchar2),
MEMBER PROCEDURE setTargetNameSpace(targetNameSpace IN varchar2),
MEMBER PROCEDURE setEnclosingElementName(enclTag IN varchar2),
MEMBER PROCEDURE setDbUrlPrefix(prefix IN varchar2),
MEMBER PROCEDURE setProcessingIns(pi IN varchar2),
CONSTRUCTOR FUNCTION XMLGenFormatType (
enclTag IN varchar2 := 'ROWSET',
schemaType IN varchar2 := 'NO_SCHEMA',
schemaName IN varchar2 := null,
targetNameSpace IN varchar2 := null,
dbUrlPrefix IN varchar2 := null,
processingIns IN varchar2 := null) RETURN SELF AS RESULT
deterministic parallel_enable,
STATIC function createFormat2(
enclTag in varchar2 := 'ROWSET',
flags in raw) return sys.xmlgenformattype
deterministic parallel_enable
);

Nulls
If a column in a row has no value, then the column is said to be null, or to contain null.
Nulls can appear in columns of any data type that are not restricted by NOT NULL or
PRIMARY KEY integrity constraints. Use a null when the actual value is not known or
when a value would not be meaningful.
Oracle Database treats a character value with a length of zero as null. However, do not
use null to represent a numeric value of zero, because they are not equivalent.
Oracle Database currently treats a character value with a
length of zero as null. However, this may not continue to be true in
future releases, and Oracle recommends that you do not treat empty
strings the same as nulls.

Note:

Any arithmetic expression containing a null always evaluates to null. For example,
null added to 10 is null. In fact, all operators (except concatenation) return null when
given a null operand.

Nulls in SQL Functions
For information on null handling in SQL functions, see "Nulls in SQL Functions" on
page 5-2.

Basic Elements of Oracle SQL 3-71

Comments

Nulls with Comparison Conditions
To test for nulls, use only the comparison conditions IS NULL and IS NOT NULL. If you
use any other condition with nulls and the result depends on the value of the null,
then the result is UNKNOWN. Because null represents a lack of data, a null cannot be equal
or unequal to any value or to another null. However, Oracle considers two nulls to be
equal when evaluating a DECODE function. Refer to DECODE on page 5-77 for syntax
and additional information.
Oracle also considers two nulls to be equal if they appear in compound keys. That is,
Oracle considers identical two compound keys containing nulls if all the non-null
components of the keys are equal.

Nulls in Conditions
A condition that evaluates to UNKNOWN acts almost like FALSE. For example, a SELECT
statement with a condition in the WHERE clause that evaluates to UNKNOWN returns no
rows. However, a condition evaluating to UNKNOWN differs from FALSE in that further
operations on an UNKNOWN condition evaluation will evaluate to UNKNOWN. Thus, NOT
FALSE evaluates to TRUE, but NOT UNKNOWN evaluates to UNKNOWN.
Table 3–20 shows examples of various evaluations involving nulls in conditions. If the
conditions evaluating to UNKNOWN were used in a WHERE clause of a SELECT statement,
then no rows would be returned for that query.
Table 3–20

Conditions Containing Nulls

Condition

Value of A

Evaluation

a IS NULL

10

FALSE

a IS NOT NULL

10

TRUE

a IS NULL

NULL

TRUE

a IS NOT NULL

NULL

FALSE

a = NULL

10

UNKNOWN

a != NULL

10

UNKNOWN

a = NULL

NULL

UNKNOWN

a != NULL

NULL

UNKNOWN

a = 10

NULL

UNKNOWN

a != 10

NULL

UNKNOWN

For the truth tables showing the results of logical conditions containing nulls, see
Table 7–5 on page 7-9, Table 7–6 on page 7-9, and Table 7–7 on page 7-9.

Comments
You can create two types of comments:
■

■

Comments within SQL statements are stored as part of the application code that
executes the SQL statements.
Comments associated with individual schema or nonschema objects are stored in
the data dictionary along with metadata on the objects themselves.

3-72 Oracle Database SQL Language Reference

Comments

Comments Within SQL Statements
Comments can make your application easier for you to read and maintain. For
example, you can include a comment in a statement that describes the purpose of the
statement within your application. With the exception of hints, comments within SQL
statements do not affect the statement execution. Refer to "Hints" on page 3-74 on
using this particular form of comment.
A comment can appear between any keywords, parameters, or punctuation marks in a
statement. You can include a comment in a statement in two ways:
■

■

Begin the comment with a slash and an asterisk (/*). Proceed with the text of the
comment. This text can span multiple lines. End the comment with an asterisk and
a slash (*/). The opening and terminating characters need not be separated from
the text by a space or a line break.
Begin the comment with -- (two hyphens). Proceed with the text of the comment.
This text cannot extend to a new line. End the comment with a line break.

Some of the tools used to enter SQL have additional restrictions. For example, if you
are using SQL*Plus, by default you cannot have a blank line inside a multiline
comment. For more information, refer to the documentation for the tool you use as an
interface to the database.
A SQL statement can contain multiple comments of both styles. The text of a comment
can contain any printable characters in your database character set.
Example

These statements contain many comments:

SELECT last_name, employee_id, salary + NVL(commission_pct, 0),
job_id, e.department_id
/* Select all employees whose compensation is
greater than that of Pataballa.*/
FROM employees e, departments d
/*The DEPARTMENTS table is used to get the department name.*/
WHERE e.department_id = d.department_id
AND salary + NVL(commission_pct,0) >
/* Subquery:
*/
(SELECT salary + NVL(commission_pct,0)
/* total compensation is salary + commission_pct */
FROM employees
WHERE last_name = 'Pataballa')
ORDER BY last_name, employee_id;
SELECT last_name,
employee_id
salary + NVL(commission_pct, 0),
job_id,
e.department_id
FROM employees e,
departments d
WHERE e.department_id = d.department_id
AND salary + NVL(commission_pct, 0) >
(SELECT salary + NVL(commission_pct,0)
FROM employees
WHERE last_name = 'Pataballa')
ORDER BY last_name
employee_id

-------

select the name
employee id
total compensation
job
and department
of all employees

-- whose compensation
-- is greater than
-- the compensation
-- of Pataballa
-- and order by last name
-- and employee id.

;

Basic Elements of Oracle SQL 3-73

Comments

Comments on Schema and Nonschema Objects
You can use the COMMENT command to associate a comment with a schema object (table,
view, materialized view, operator, indextype, mining model) or a nonschema object
(edition) using the COMMENT command. You can also create a comment on a column,
which is part of a table schema object. Comments associated with schema and
nonschema objects are stored in the data dictionary. Refer to COMMENT on
page 13-46 for a description of this form of comment.

Hints
Hints are comments in a SQL statement that pass instructions to the Oracle Database
optimizer. The optimizer uses these hints to choose an execution plan for the
statement, unless some condition exists that prevents the optimizer from doing so.
Hints were introduced in Oracle7, when users had little recourse if the optimizer
generated suboptimal plans. Now Oracle provides a number of tools, including the
SQL Tuning Advisor, SQL plan management, and SQL Performance Analyzer, to help
you address performance problems that are not solved by the optimizer. Oracle
strongly recommends that you use those tools rather than hints. The tools are far
superior to hints, because when used on an ongoing basis, they provide fresh solutions
as your data and database environment change.
Hints should be used sparingly, and only after you have collected statistics on the
relevant tables and evaluated the optimizer plan without hints using the EXPLAIN PLAN
statement. Changing database conditions as well as query performance enhancements
in subsequent releases can have significant impact on how hints in your code affect
performance.
The remainder of this section provides information on some commonly used hints. If
you decide to use hints rather than the more advanced tuning tools, be aware that any
short-term benefit resulting from the use of hints may not continue to result in
improved performance over the long term.
Using Hints
A statement block can have only one comment containing hints, and that comment
must follow the SELECT, UPDATE, INSERT, MERGE, or DELETE keyword.
The following syntax diagram shows hints contained in both styles of comments that
Oracle supports within a statement block. The hint syntax must follow immediately
after an INSERT, UPDATE, DELETE, SELECT, or MERGE keyword that begins the statement
block.
hint::=
string
/*+

hint

*/
string

––+

hint

where:
■

The plus sign (+) causes Oracle to interpret the comment as a list of hints. The plus
sign must follow immediately after the comment delimiter. No space is permitted.

3-74 Oracle Database SQL Language Reference

Comments

■

■

hint is one of the hints discussed in this section. The space between the plus sign
and the hint is optional. If the comment contains multiple hints, then separate the
hints by at least one space.
string is other commenting text that can be interspersed with the hints.

The --+ syntax requires that the entire comment be on a single line.
Oracle Database ignores hints and does not return an error under the following
circumstances:
■

■

■

■

■

The hint contains misspellings or syntax errors. However, the database does
consider other correctly specified hints in the same comment.
The comment containing the hint does not follow a DELETE, INSERT, MERGE, SELECT,
or UPDATE keyword.
A combination of hints conflict with each other. However, the database does
consider other hints in the same comment.
The database environment uses PL/SQL version 1, such as Forms version 3
triggers, Oracle Forms 4.5, and Oracle Reports 2.5.
A global hint refers to multiple query blocks. Refer to "Specifying Multiple Query
Blocks in a Global Hint" on page 3-76 for more information.

Specifying a Query Block in a Hint
You can specify an optional query block name in many hints to specify the query block
to which the hint applies. This syntax lets you specify in the outer query a hint that
applies to an inline view.
The syntax of the query block argument is of the form @queryblock, where queryblock
is an identifier that specifies a query block in the query. The queryblock identifier can
either be system-generated or user-specified. When you specify a hint in the query
block itself to which the hint applies, you omit the @queryblock syntax.
■

■

The system-generated identifier can be obtained by using EXPLAIN PLAN for the
query. Pretransformation query block names can be determined by running
EXPLAIN PLAN for the query using the NO_QUERY_TRANSFORMATION hint. See "NO_
QUERY_TRANSFORMATION Hint" on page 3-95.
The user-specified name can be set with the QB_NAME hint. See "QB_NAME Hint"
on page 3-104.

Specifying Global Hints
Many hints can apply both to specific tables or indexes and more globally to tables
within a view or to columns that are part of indexes. The syntactic elements tablespec
and indexspec define these global hints.
tablespec::=
view

.
table

You must specify the table to be accessed exactly as it appears in the statement. If the
statement uses an alias for the table, then use the alias rather than the table name in
the hint. However, do not include the schema name with the table name within the
hint, even if the schema name appears in the statement.

Basic Elements of Oracle SQL 3-75

Comments

Note: Specifying a global hint using the tablespec clause does not
work for queries that use ANSI joins, because the optimizer generates
additional views during parsing. Instead, specify @queryblock to
indicate the query block to which the hint applies.

indexspec::=
index
table

.

(

column

)

When tablespec is followed by indexspec in the specification of a hint, a comma
separating the table name and index name is permitted but not required. Commas are
also permitted, but not required, to separate multiple occurrences of indexspec.
Specifying Multiple Query Blocks in a Global Hint Oracle Database ignores global
hints that refer to multiple query blocks. To avoid this issue, Oracle recommends that
you specify the object alias in the hint instead of using tablespec and indexspec.

For example, consider the following view v and table t:
CREATE VIEW v AS
SELECT e.last_name, e.department_id, d.location_id
FROM employees e, departments d
WHERE e.department_id = d.department_id;
CREATE TABLE t AS
SELECT * from employees
WHERE employee_id < 200;

Note: The following examples use the EXPLAIN PLAN statement,
which enables you to display the execution plan and determine if a
hint is honored or ignored. Refer to EXPLAIN PLAN on page 18-20
for more information.

The LEADING hint is ignored in the following query because it refers to multiple query
blocks, that is, the main query block containing table t and the view query block v:
EXPLAIN PLAN
SET STATEMENT_ID = 'Test 1'
INTO plan_table FOR
(SELECT /*+ LEADING(v.e v.d t) */ *
FROM t, v
WHERE t.department_id = v.department_id);

The following SELECT statement returns the execution plan, which shows that the
LEADING hint was ignored:
SELECT id, LPAD(' ',2*(LEVEL-1))||operation operation, options, object_name,
object_alias
FROM plan_table
START WITH id = 0 AND statement_id = 'Test 1'
CONNECT BY PRIOR id = parent_id AND statement_id = 'Test 1'
ORDER BY id;

3-76 Oracle Database SQL Language Reference

Comments

ID OPERATION
--- -------------------0 SELECT STATEMENT
1
HASH JOIN
2
HASH JOIN
3
TABLE ACCESS
4
TABLE ACCESS
5
TABLE ACCESS

OPTIONS
OBJECT_NAME
OBJECT_ALIAS
---------- ------------- --------------------

FULL
FULL
FULL

DEPARTMENTS
EMPLOYEES
T

D@SEL$2
E@SEL$2
T@SEL$1

The LEADING hint is honored in the following query because it refers to object aliases,
which can be found in the execution plan that was returned by the previous query:
EXPLAIN PLAN
SET STATEMENT_ID = 'Test 2'
INTO plan_table FOR
(SELECT /*+ LEADING(E@SEL$2 D@SEL$2 T@SEL$1) */ *
FROM t, v
WHERE t.department_id = v.department_id);

The following SELECT statement returns the execution plan, which shows that the
LEADING hint was honored:
SELECT id, LPAD(' ',2*(LEVEL-1))||operation operation, options,
object_name, object_alias
FROM plan_table
START WITH id = 0 AND statement_id = 'Test 2'
CONNECT BY PRIOR id = parent_id AND statement_id = 'Test 2'
ORDER BY id;
ID OPERATION
--- -------------------0 SELECT STATEMENT
1
HASH JOIN
2
HASH JOIN
3
TABLE ACCESS
4
TABLE ACCESS
5
TABLE ACCESS

OPTIONS
OBJECT_NAME
OBJECT_ALIAS
---------- ------------- --------------------

FULL
FULL
FULL

EMPLOYEES
DEPARTMENTS
T

E@SEL$2
D@SEL$2
T@SEL$1

See Also: Oracle Database Performance Tuning Guide for information
on the following topics:
■
■

■

When to use global hints and how Oracle interprets them
Using EXPLAIN PLAN to learn how the optimizer is executing a
query
References in hints to tables within views

Hints by Functional Category
Table 3–21 lists the hints by functional category and contains cross-references to the
syntax and semantics for each hint. An alphabetical reference of the hints follows the
table.
Table 3–21

Hints by Functional Category

Hint

Link to Syntax and Semantics

Optimization Goals and
Approaches

ALL_ROWS Hint on page 3-79

Access Path Hints

CLUSTER Hint on page 3-82

FIRST_ROWS Hint on page 3-83

Basic Elements of Oracle SQL 3-77

Comments

Table 3–21 (Cont.) Hints by Functional Category
Hint

Link to Syntax and Semantics

--

FULL Hint on page 3-84

--

HASH Hint on page 3-84

--

INDEX Hint on page 3-85
NO_INDEX Hint on page 3-92

--

INDEX_ASC Hint on page 3-86
INDEX_DESC Hint on page 3-87

--

INDEX_COMBINE Hint on page 3-86

--

INDEX_JOIN Hint on page 3-87

--

INDEX_FFS Hint on page 3-87

--

INDEX_SS Hint on page 3-88

--

INDEX_SS_ASC Hint on page 3-88

--

INDEX_SS_DESC Hint on page 3-89

--

NATIVE_FULL_OUTER_JOIN Hint on page 3-90
NO_NATIVE_FULL_OUTER_JOIN Hint on page 3-93

--

NO_INDEX_FFS Hint on page 3-92

--

NO_INDEX_SS Hint on page 3-92

Join Order Hints

ORDERED Hint on page 3-98

--

LEADING Hint on page 3-89

Join Operation Hints

USE_HASH Hint on page 3-108
NO_USE_HASH Hint on page 3-96

--

USE_MERGE Hint on page 3-108
NO_USE_MERGE Hint on page 3-97

--

USE_NL Hint on page 3-108
USE_NL_WITH_INDEX Hint on page 3-109
NO_USE_NL Hint on page 3-97

Parallel Execution Hints

PARALLEL Hint on page 3-98
NO_PARALLEL Hint on page 3-93

--

PARALLEL_INDEX Hint on page 3-101
NO_PARALLEL_INDEX Hint on page 3-94

--

PQ_DISTRIBUTE Hint on page 3-101

Online Application
Upgrade Hints

CHANGE_DUPKEY_ERROR_INDEX Hint on page 3-81

--

IGNORE_ROW_ON_DUPKEY_INDEX Hint on page 3-85

--

RETRY_ON_ROW_CHANGE Hint on page 3-105

Query Transformation
Hints

FACT Hint on page 3-83

--

MERGE Hint on page 3-89

NO_FACT Hint on page 3-91

NO_MERGE Hint on page 3-93

3-78 Oracle Database SQL Language Reference

Comments

Table 3–21 (Cont.) Hints by Functional Category
Hint

Link to Syntax and Semantics

--

NO_EXPAND Hint on page 3-91
USE_CONCAT Hint on page 3-107

--

REWRITE Hint on page 3-106
NO_REWRITE Hint on page 3-95

--

UNNEST Hint on page 3-107
NO_UNNEST Hint on page 3-96

--

STAR_TRANSFORMATION Hint on page 3-106
NO_STAR_TRANSFORMATION Hint on page 3-96

--

NO_QUERY_TRANSFORMATION Hint on page 3-95

XML Hints

NO_XMLINDEX_REWRITE Hint on page 3-98

--

NO_XML_QUERY_REWRITE Hint on page 3-97

Other Hints

APPEND Hint on page 3-80
APPEND_VALUES Hint on page 3-80
NOAPPEND Hint on page 3-91

--

CACHE Hint on page 3-81
NOCACHE Hint on page 3-91

--

CURSOR_SHARING_EXACT Hint on page 3-82

--

DRIVING_SITE Hint on page 3-82

--

DYNAMIC_SAMPLING Hint on page 3-83

--

MODEL_MIN_ANALYSIS Hint on page 3-90

--

MONITOR Hint on page 3-90

--

NO_MONITOR Hint on page 3-93

--

OPT_PARAM Hint on page 3-98

--

PUSH_PRED Hint on page 3-104
NO_PUSH_PRED Hint on page 3-94

--

PUSH_SUBQ Hint on page 3-104
NO_PUSH_SUBQ Hint on page 3-95

--

PX_JOIN_FILTER Hint on page 3-104
NO_PX_JOIN_FILTER Hint on page 3-95

--

QB_NAME Hint on page 3-104

--

RESULT_CACHE Hint on page 3-105
NO_RESULT_CACHE Hint on page 3-95

Alphabetical Listing of Hints
This section provides syntax and semantics for all hints in alphabetical order.

ALL_ROWS Hint
/*+

ALL_ROWS

*/

Basic Elements of Oracle SQL 3-79

Comments

The ALL_ROWS hint instructs the optimizer to optimize a statement block with a goal of
best throughput, which is minimum total resource consumption. For example, the
optimizer uses the query optimization approach to optimize this statement for best
throughput:
SELECT /*+ ALL_ROWS */ employee_id, last_name, salary, job_id
FROM employees
WHERE employee_id = 107;

If you specify either the ALL_ROWS or the FIRST_ROWS hint in a SQL statement, and if the
data dictionary does not have statistics about tables accessed by the statement, then
the optimizer uses default statistical values, such as allocated storage for such tables,
to estimate the missing statistics and to subsequently choose an execution plan. These
estimates might not be as accurate as those gathered by the DBMS_STATS package, so
you should use the DBMS_STATS package to gather statistics.
If you specify hints for access paths or join operations along with either the ALL_ROWS
or FIRST_ROWS hint, then the optimizer gives precedence to the access paths and join
operations specified by the hints.

APPEND Hint
/*+

APPEND

*/

The APPEND hint instructs the optimizer to use direct-path INSERT with the subquery
syntax of the INSERT statement.
■

■

Conventional INSERT is the default in serial mode. In serial mode, direct path can
be used only if you include the APPEND hint.
Direct-path INSERT is the default in parallel mode. In parallel mode, conventional
insert can be used only if you specify the NOAPPEND hint.

The decision whether the INSERT will go parallel or not is independent of the APPEND
hint.
In direct-path INSERT, data is appended to the end of the table, rather than using
existing space currently allocated to the table. As a result, direct-path INSERT can be
considerably faster than conventional INSERT.
The APPEND hint is only supported with the subquery syntax of the INSERT statement,
not the VALUES clause. If you specify the APPEND hint with the VALUES clause, it is
ignored and conventional insert will be used. To use direct-path INSERT with the
VALUES clause, refer to "APPEND_VALUES Hint" on page 3-80.
See Also: "NOAPPEND Hint" on page 3-91 for information on
that hint and Oracle Database Administrator's Guide for information
on direct-path inserts

APPEND_VALUES Hint
/*+

APPEND_VALUES

*/

The APPEND_VALUES hint instructs the optimizer to use direct-path INSERT with the
VALUES clause. If you do not specify this hint, then conventional INSERT is used.
In direct-path INSERT, data is appended to the end of the table, rather than using
existing space currently allocated to the table. As a result, direct-path INSERT can be
considerably faster than conventional INSERT.

3-80 Oracle Database SQL Language Reference

Comments

The APPEND_VALUES hint can be used to greatly enhance performance. Some examples
of its uses are:
■

■

In an Oracle Call Interface (OCI) program, when using large array binds or array
binds with row callbacks
In PL/SQL, when loading a large number of rows with a FORALL loop that has an
INSERT statement with a VALUES clause

The APPEND_VALUES hint is only supported with the VALUES clause of the INSERT
statement. If you specify the APPEND_VALUES hint with the subquery syntax of the
INSERT statement, it is ignored and conventional insert will be used. To use direct-path
INSERT with a subquery, refer to "APPEND Hint" on page 3-80.
See Also: Oracle Database Administrator's Guide for information on
direct-path inserts

CACHE Hint
@
/*+

CACHE

queryblock

(

tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
The CACHE hint instructs the optimizer to place the blocks retrieved for the table at the
most recently used end of the LRU list in the buffer cache when a full table scan is
performed. This hint is useful for small lookup tables.
In the following example, the CACHE hint overrides the default caching specification of
the table:
SELECT /*+ FULL (hr_emp) CACHE(hr_emp) */ last_name
FROM employees hr_emp;

The CACHE and NOCACHE hints affect system statistics table scans (long tables) and
table scans (short tables), as shown in the V$SYSSTAT data dictionary view.

CHANGE_DUPKEY_ERROR_INDEX Hint
table
/*+

CHANGE_DUPKEY_ERROR_INDEX

,

(

index
,

table

(

column

)

*/

)

The CHANGE_DUPKEY_ERROR_INDEX, IGNORE_ROW_ON_DUPKEY_
INDEX, and RETRY_ON_ROW_CHANGE hints are unlike other hints in that
they have a semantic effect. The general philosophy explained in
"Hints" on page 3-74 does not apply for these three hints.
Note:

The CHANGE_DUPKEY_ERROR_INDEX hint provides a mechanism to unambiguously
identify a unique key violation for a specified set of columns or for a specified index.
When a unique key violation occurs for the specified index, an ORA-38911 error is
reported instead of an ORA-001.
This hint applies to INSERT, UPDATE operations. If you specify an index, then the index
must exist and be unique. If you specify a column list instead of an index, then a

Basic Elements of Oracle SQL 3-81

Comments

unique index whose columns match the specified columns in number and order must
exist.
This use of this hint results in error messages if specific rules are violated. Refer to
IGNORE_ROW_ON_DUPKEY_INDEX Hint on page 3-85 for details.
Note:

This hint disables both APPEND mode and parallel DML.

See Also: IGNORE_ROW_ON_DUPKEY_INDEX Hint on page 3-85
for information on that hint and Oracle Database Performance Tuning
Guide for more information on using the online application upgrade
related hints

CLUSTER Hint
@
/*+

CLUSTER

queryblock

(

tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
The CLUSTER hint instructs the optimizer to use a cluster scan to access the specified
table. This hint applies only to tables in an indexed cluster.

CURSOR_SHARING_EXACT Hint
/*+

CURSOR_SHARING_EXACT

*/

Oracle can replace literals in SQL statements with bind variables, when it is safe to do
so. This replacement is controlled with the CURSOR_SHARING initialization parameter.
The CURSOR_SHARING_EXACT hint instructs the optimizer to switch this behavior off.
When you specify this hint, Oracle executes the SQL statement without any attempt to
replace literals with bind variables.

DRIVING_SITE Hint
@
/*+

DRIVING_SITE

(

queryblock
tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
The DRIVING_SITE hint instructs the optimizer to execute the query at a different site
than that selected by the database. This hint is useful if you are using distributed
query optimization.
For example:
SELECT /*+ DRIVING_SITE(departments) */ *
FROM employees, departments@rsite
WHERE employees.department_id = departments.department_id;

If this query is executed without the hint, then rows from departments are sent to the
local site, and the join is executed there. With the hint, the rows from employees are
sent to the remote site, and the query is executed there and the result set is returned to
the local site.

3-82 Oracle Database SQL Language Reference

Comments

DYNAMIC_SAMPLING Hint
@
/*+

DYNAMIC_SAMPLING

queryblock

tablespec

(

integer

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
The DYNAMIC_SAMPLING hint instructs the optimizer how to control dynamic sampling
to improve server performance by determining more accurate predicate selectivity and
statistics for tables and indexes.
You can set the value of DYNAMIC_SAMPLING to a value from 0 to 10. The higher the
level, the more effort the compiler puts into dynamic sampling and the more broadly it
is applied. Sampling defaults to cursor level unless you specify tablespec.
The integer value is 0 to 10, indicating the degree of sampling.
If a cardinality statistic already exists for the table, then the optimizer uses it.
Otherwise, the optimizer enables dynamic sampling to estimate the cardinality
statistic.
If you specify tablespec and the cardinality statistic already exists, then:
■

If there is no single-table predicate (a WHERE clause that evaluates only one table),
then the optimizer trusts the existing statistics and ignores this hint. For example,
the following query will not result in any dynamic sampling if employees is
analyzed:
SELECT /*+ DYNAMIC_SAMPLING(e 1) */ count(*)
FROM employees e;

■

If there is a single-table predicate, then the optimizer uses the existing cardinality
statistic and estimates the selectivity of the predicate using the existing statistics.

To apply dynamic sampling to a specific table, use the following form of the hint:
SELECT /*+ DYNAMIC_SAMPLING(employees 1) */ *
FROM employees
WHERE ...

See Also: Oracle Database Performance Tuning Guide for
information about dynamic sampling and the sampling levels that
you can set

FACT Hint
@
/*+

FACT

queryblock

(

tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
The FACT hint is used in the context of the star transformation. It instructs the
optimizer that the table specified in tablespec should be considered as a fact table.

FIRST_ROWS Hint
/*+

FIRST_ROWS

(

integer

)

*/

Basic Elements of Oracle SQL 3-83

Comments

The FIRST_ROWS hint instructs Oracle to optimize an individual SQL statement for fast
response, choosing the plan that returns the first n rows most efficiently. For integer,
specify the number of rows to return.
For example, the optimizer uses the query optimization approach to optimize the
following statement for best response time:
SELECT /*+ FIRST_ROWS(10) */ employee_id, last_name, salary, job_id
FROM employees
WHERE department_id = 20;

In this example each department contains many employees. The user wants the first 10
employees of department 20 to be displayed as quickly as possible.
The optimizer ignores this hint in DELETE and UPDATE statement blocks and in SELECT
statement blocks that include any blocking operations, such as sorts or groupings.
Such statements cannot be optimized for best response time, because Oracle Database
must retrieve all rows accessed by the statement before returning the first row. If you
specify this hint in any such statement, then the database optimizes for best
throughput.
See Also: "ALL_ROWS Hint" on page 3-79 for additional
information on the FIRST_ROWS hint and statistics

FULL Hint
@
/*+

FULL

queryblock

(

tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
The FULL hint instructs the optimizer to perform a full table scan for the specified table.
For example:
SELECT /*+ FULL(e) */ employee_id, last_name
FROM hr.employees e
WHERE last_name LIKE :b1;

Oracle Database performs a full table scan on the employees table to execute this
statement, even if there is an index on the last_name column that is made available by
the condition in the WHERE clause.
The employees table has alias e in the FROM clause, so the hint must refer to the table by
its alias rather than by its name. Do not specify schema names in the hint even if they
are specified in the FROM clause.

HASH Hint
@
/*+

HASH

(

queryblock
tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
The HASH hint instructs the optimizer to use a hash scan to access the specified table.
This hint applies only to tables in a hash cluster.

3-84 Oracle Database SQL Language Reference

Comments

IGNORE_ROW_ON_DUPKEY_INDEX Hint
table
/*+

IGNORE_ROW_ON_DUPKEY_INDEX

,

(

index
,

table

(

column

)

*/

)

The CHANGE_DUPKEY_ERROR_INDEX, IGNORE_ROW_ON_DUPKEY_
INDEX, and RETRY_ON_ROW_CHANGE hints are unlike other hints in that
they have a semantic effect. The general philosophy explained in
"Hints" on page 3-74 does not apply for these three hints.
Note:

The IGNORE_ROW_ON_DUPKEY_INDEX hint applies only to single-table INSERT operations.
It is not supported for UPDATE, DELETE, MERGE, or multitable insert operations. IGNORE_
ROW_ON_DUPKEY_INDEX causes the statement to ignore a unique key violation for a
specified set of columns or for a specified index. When a unique key violation is
encountered, a row-level rollback occurs and execution resumes with the next input
row. If you specify this hint when inserting data with DML error logging enabled, then
the unique key violation is not logged and does not cause statement termination.
The semantic effect of this hint results in error messages if specific rules are violated:
■

■

■

If you specify index, then the index must exist and be unique. Otherwise, the
statement causes ORA-38913.
You must specify exactly one index. If you specify no index, then the statement
causes ORA-38912. If you specify more than one index, then the statement causes
ORA-38915.
You can specify either a CHANGE_DUPKEY_ERROR_INDEX or IGNORE_ROW_ON_DUPKEY_
INDEX hint in an INSERT statement, but not both. If you specify both, then the
statement causes ORA-38915.

As with all hints, a syntax error in the hint causes it to be silently ignored. The result
will be that ORA-00001 will be caused, just as if no hint were used.
Note:

This hint disables both APPEND mode and parallel DML.

See Also: CHANGE_DUPKEY_ERROR_INDEX Hint on page 3-81
for information on that hint and Oracle Database Performance Tuning
Guide for more information on using the online application upgrade
related hints

INDEX Hint
@
/*+

INDEX

(

queryblock

indexspec
tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75,
indexspec::= on page 3-76)
The INDEX hint instructs the optimizer to use an index scan for the specified table. You
can use the INDEX hint for function-based, domain, B-tree, bitmap, and bitmap join
indexes.

Basic Elements of Oracle SQL 3-85

Comments

The behavior of the hint depends on the indexspec specification:
■

■

■

If the INDEX hint specifies a single available index, then the database performs a
scan on this index. The optimizer does not consider a full table scan or a scan of
another index on the table.
For a hint on a combination of multiple indexes, Oracle recommends using INDEX_
COMBINE rather than INDEX, because it is a more versatile hint. If the INDEX hint
specifies a list of available indexes, then the optimizer considers the cost of a scan
on each index in the list and then performs the index scan with the lowest cost.
The database can also choose to scan multiple indexes from this list and merge the
results, if such an access path has the lowest cost. The database does not consider a
full table scan or a scan on an index not listed in the hint.
If the INDEX hint specifies no indexes, then the optimizer considers the cost of a
scan on each available index on the table and then performs the index scan with
the lowest cost. The database can also choose to scan multiple indexes and merge
the results, if such an access path has the lowest cost. The optimizer does not
consider a full table scan.

For example:
SELECT /*+ INDEX (employees emp_department_ix)*/ employee_id, department_id
FROM employees
WHERE department_id > 50;

INDEX_ASC Hint
@
/*+

INDEX_ASC

queryblock

(

indexspec
tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75,
indexspec::= on page 3-76)
The INDEX_ASC hint instructs the optimizer to use an index scan for the specified table.
If the statement uses an index range scan, then Oracle Database scans the index entries
in ascending order of their indexed values. Each parameter serves the same purpose as
in "INDEX Hint" on page 3-85.
The default behavior for a range scan is to scan index entries in ascending order of
their indexed values, or in descending order for a descending index. This hint does not
change the default order of the index, and therefore does not specify anything more
than the INDEX hint. However, you can use the INDEX_ASC hint to specify ascending
range scans explicitly should the default behavior change.

INDEX_COMBINE Hint
@
/*+

INDEX_COMBINE

(

queryblock

indexspec
tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75,
indexspec::= on page 3-76)
The INDEX_COMBINE hint instructs the optimizer to use a bitmap access path for the
table. If indexspec is omitted from the INDEX_COMBINE hint, then the optimizer uses
whatever Boolean combination of indexes has the best cost estimate for the table. If
you specify indexspec, then the optimizer tries to use some Boolean combination of
the specified indexes. Each parameter serves the same purpose as in "INDEX Hint" on
3-86 Oracle Database SQL Language Reference

Comments

page 3-85. For example:
SELECT /*+ INDEX_COMBINE(e emp_manager_ix emp_department_ix) */ *
FROM employees e
WHERE manager_id = 108
OR department_id = 110;

INDEX_DESC Hint
@
/*+

INDEX_DESC

queryblock

(

indexspec
tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75,
indexspec::= on page 3-76)
The INDEX_DESC hint instructs the optimizer to use a descending index scan for the
specified table. If the statement uses an index range scan and the index is ascending,
then Oracle scans the index entries in descending order of their indexed values. In a
partitioned index, the results are in descending order within each partition. For a
descending index, this hint effectively cancels out the descending order, resulting in a
scan of the index entries in ascending order. Each parameter serves the same purpose
as in "INDEX Hint" on page 3-85. For example:
SELECT /*+ INDEX_DESC(e emp_name_ix) */ *
FROM employees e;

See Also: Oracle Database Performance Tuning Guide for information
on full scans

INDEX_FFS Hint
@
/*+

INDEX_FFS

queryblock

(

indexspec
tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75,
indexspec::= on page 3-76)
The INDEX_FFS hint instructs the optimizer to perform a fast full index scan rather than
a full table scan.
Each parameter serves the same purpose as in "INDEX Hint" on page 3-85. For
example:
SELECT /*+ INDEX_FFS(e emp_name_ix) */ first_name
FROM employees e;

INDEX_JOIN Hint
@
/*+

INDEX_JOIN

(

queryblock

indexspec
tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75,
indexspec::= on page 3-76)
The INDEX_JOIN hint instructs the optimizer to use an index join as an access path. For
the hint to have a positive effect, a sufficiently small number of indexes must exist that
contain all the columns required to resolve the query.
Basic Elements of Oracle SQL 3-87

Comments

Each parameter serves the same purpose as in "INDEX Hint" on page 3-85. For
example, the following query uses an index join to access the manager_id and
department_id columns, both of which are indexed in the employees table.
SELECT /*+ INDEX_JOIN(e emp_manager_ix emp_department_ix) */ department_id
FROM employees e
WHERE manager_id < 110
AND department_id < 50;

INDEX_SS Hint
@
/*+

INDEX_SS

queryblock

(

indexspec
tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75,
indexspec::= on page 3-76)
The INDEX_SS hint instructs the optimizer to perform an index skip scan for the
specified table. If the statement uses an index range scan, then Oracle scans the index
entries in ascending order of their indexed values. In a partitioned index, the results
are in ascending order within each partition.
Each parameter serves the same purpose as in "INDEX Hint" on page 3-85. For
example:
SELECT /*+ INDEX_SS(e emp_name_ix) */ last_name
FROM employees e
WHERE first_name = 'Steven';

See Also: Oracle Database Performance Tuning Guide for information
on index skip scans

INDEX_SS_ASC Hint
@
/*+

INDEX_SS_ASC

(

queryblock

indexspec
tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75,
indexspec::= on page 3-76)
The INDEX_SS_ASC hint instructs the optimizer to perform an index skip scan for the
specified table. If the statement uses an index range scan, then Oracle Database scans
the index entries in ascending order of their indexed values. In a partitioned index, the
results are in ascending order within each partition. Each parameter serves the same
purpose as in "INDEX Hint" on page 3-85.
The default behavior for a range scan is to scan index entries in ascending order of
their indexed values, or in descending order for a descending index. This hint does not
change the default order of the index, and therefore does not specify anything more
than the INDEX_SS hint. However, you can use the INDEX_SS_ASC hint to specify
ascending range scans explicitly should the default behavior change.
See Also: Oracle Database Performance Tuning Guide for information
on index skip scans

3-88 Oracle Database SQL Language Reference

Comments

INDEX_SS_DESC Hint
@
/*+

INDEX_SS_DESC

queryblock

indexspec

(

tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75,
indexspec::= on page 3-76)
The INDEX_SS_DESC hint instructs the optimizer to perform an index skip scan for the
specified table. If the statement uses an index range scan and the index is ascending,
then Oracle scans the index entries in descending order of their indexed values. In a
partitioned index, the results are in descending order within each partition. For a
descending index, this hint effectively cancels out the descending order, resulting in a
scan of the index entries in ascending order.
Each parameter serves the same purpose as in the "INDEX Hint" on page 3-85. For
example:
SELECT /*+ INDEX_SS_DESC(e emp_name_ix) */ last_name
FROM employees e
WHERE first_name = 'Steven';

See Also: Oracle Database Performance Tuning Guide for information
on index skip scans

LEADING Hint
@
/*+

LEADING

queryblock

(

tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
The LEADING hint instructs the optimizer to use the specified set of tables as the prefix
in the execution plan. This hint is more versatile than the ORDERED hint. For example:
SELECT /*+ LEADING(e j) */ *
FROM employees e, departments d, job_history j
WHERE e.department_id = d.department_id
AND e.hire_date = j.start_date;

The LEADING hint is ignored if the tables specified cannot be joined first in the order
specified because of dependencies in the join graph. If you specify two or more
conflicting LEADING hints, then all of them are ignored. If you specify the ORDERED hint,
it overrides all LEADING hints.

MERGE Hint
@
(

queryblock
@

)

queryblock
tablespec

/*+

MERGE

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
The MERGE hint lets you merge views in a query.
If a view's query block contains a GROUP BY clause or DISTINCT operator in the SELECT
list, then the optimizer can merge the view into the accessing statement only if
Basic Elements of Oracle SQL 3-89

Comments

complex view merging is enabled. Complex merging can also be used to merge an IN
subquery into the accessing statement if the subquery is uncorrelated.
For example:
SELECT /*+ MERGE(v) */ e1.last_name, e1.salary, v.avg_salary
FROM employees e1,
(SELECT department_id, avg(salary) avg_salary
FROM employees e2
GROUP BY department_id) v
WHERE e1.department_id = v.department_id
AND e1.salary > v.avg_salary
ORDER BY e1.last_name;

When the MERGE hint is used without an argument, it should be placed in the view
query block. When MERGE is used with the view name as an argument, it should be
placed in the surrounding query.

MODEL_MIN_ANALYSIS Hint
/*+

MODEL_MIN_ANALYSIS

*/

The MODEL_MIN_ANALYSIS hint instructs the optimizer to omit some compile-time
optimizations of spreadsheet rules—primarily detailed dependency graph analysis.
Other spreadsheet optimizations, such as creating filters to selectively populate
spreadsheet access structures and limited rule pruning, are still used by the optimizer.
This hint reduces compilation time because spreadsheet analysis can be lengthy if the
number of spreadsheet rules is more than several hundreds.

MONITOR Hint
/*+

MONITOR

*/

The MONITOR hint forces real-time SQL monitoring for the query, even if the statement
is not long running. This hint is valid only when the parameter CONTROL_MANAGEMENT_
PACK_ACCESS is set to DIAGNOSTIC+TUNING.
See Also: Oracle Database Performance Tuning Guide for more
information about real-time SQL monitoring

NATIVE_FULL_OUTER_JOIN Hint
/*+

NATIVE_FULL_OUTER_JOIN

*/

The NATIVE_FULL_OUTER_JOIN hint instructs the optimizer to use native full outer join,
which is a native execution method based on a hash join.
See Also:
■
■

NO_NATIVE_FULL_OUTER_JOIN Hint on page 3-93
Oracle Database Performance Tuning Guide for more information
about native full outer joins

3-90 Oracle Database SQL Language Reference

Comments

NOAPPEND Hint
/*+

NOAPPEND

*/

The NOAPPEND hint instructs the optimizer to use conventional INSERT by disabling
parallel mode for the duration of the INSERT statement. Conventional INSERT is the
default in serial mode, and direct-path INSERT is the default in parallel mode.

NOCACHE Hint
@
/*+

NOCACHE

queryblock

(

tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
The NOCACHE hint instructs the optimizer to place the blocks retrieved for the table at
the least recently used end of the LRU list in the buffer cache when a full table scan is
performed. This is the normal behavior of blocks in the buffer cache. For example:
SELECT /*+ FULL(hr_emp) NOCACHE(hr_emp) */ last_name
FROM employees hr_emp;

The CACHE and NOCACHE hints affect system statistics table scans(long tables) and
table scans(short tables), as shown in the V$SYSSTAT view.
See Also: Oracle Database Performance Tuning Guide for information
on automatic caching of tables, depending on their size

NO_EXPAND Hint
(
/*+

@

queryblock

NO_EXPAND

)
*/

(See "Specifying a Query Block in a Hint" on page 3-75)
The NO_EXPAND hint instructs the optimizer not to consider OR-expansion for queries
having OR conditions or IN-lists in the WHERE clause. Usually, the optimizer considers
using OR expansion and uses this method if it decides that the cost is lower than not
using it. For example:
SELECT /*+ NO_EXPAND */ *
FROM employees e, departments d
WHERE e.manager_id = 108
OR d.department_id = 110;

See Also: The "USE_CONCAT Hint" on page 3-107, which is the
opposite of this hint

NO_FACT Hint
@
/*+

NO_FACT

(

queryblock
tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
The NO_FACT hint is used in the context of the star transformation. It instruct the
optimizer that the queried table should not be considered as a fact table.

Basic Elements of Oracle SQL 3-91

Comments

NO_INDEX Hint
@
/*+

NO_INDEX

queryblock

(

indexspec
tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75,
indexspec::= on page 3-76)
The NO_INDEX hint instructs the optimizer not to use one or more indexes for the
specified table. For example:
SELECT /*+ NO_INDEX(employees emp_empid) */ employee_id
FROM employees
WHERE employee_id > 200;

Each parameter serves the same purpose as in "INDEX Hint" on page 3-85 with the
following modifications:
■

■

■

If this hint specifies a single available index, then the optimizer does not consider a
scan on this index. Other indexes not specified are still considered.
If this hint specifies a list of available indexes, then the optimizer does not consider
a scan on any of the specified indexes. Other indexes not specified in the list are
still considered.
If this hint specifies no indexes, then the optimizer does not consider a scan on any
index on the table. This behavior is the same as a NO_INDEX hint that specifies a list
of all available indexes for the table.

The NO_INDEX hint applies to function-based, B-tree, bitmap, cluster, or domain
indexes. If a NO_INDEX hint and an index hint (INDEX, INDEX_ASC, INDEX_DESC, INDEX_
COMBINE, or INDEX_FFS) both specify the same indexes, then the database ignores both
the NO_INDEX hint and the index hint for the specified indexes and considers those
indexes for use during execution of the statement.

NO_INDEX_FFS Hint
@
/*+

NO_INDEX_FFS

queryblock

(

indexspec
tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75,
indexspec::= on page 3-76)
The NO_INDEX_FFS hint instructs the optimizer to exclude a fast full index scan of the
specified indexes on the specified table. Each parameter serves the same purpose as in
the "NO_INDEX Hint" on page 3-92. For example:
SELECT /*+ NO_INDEX_FFS(items item_order_ix) */ order_id
FROM order_items items;

NO_INDEX_SS Hint
@
/*+

NO_INDEX_SS

(

queryblock

indexspec
tablespec

)

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75,
indexspec::= on page 3-76)

3-92 Oracle Database SQL Language Reference

*/

Comments

The NO_INDEX_SS hint instructs the optimizer to exclude a skip scan of the specified
indexes on the specified table. Each parameter serves the same purpose as in the "NO_
INDEX Hint" on page 3-92.
See Also: Oracle Database Performance Tuning Guide for information
on index skip scans

NO_MERGE Hint
@
(

queryblock
@

)

queryblock
tablespec

/*+

NO_MERGE

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
The NO_MERGE hint instructs the optimizer not to combine the outer query and any
inline view queries into a single query.
This hint lets you have more influence over the way in which the view is accessed. For
example, the following statement causes view seattle_dept not to be merged:
SELECT /*+ NO_MERGE(seattle_dept) */ e1.last_name, seattle_dept.department_name
FROM employees e1,
(SELECT location_id, department_id, department_name
FROM departments
WHERE location_id = 1700) seattle_dept
WHERE e1.department_id = seattle_dept.department_id;

When you use the NO_MERGE hint in the view query block, specify it without an
argument. When you specify NO_MERGE in the surrounding query, specify it with the
view name as an argument.

NO_MONITOR Hint
/*+

NO_MONITOR

*/

The NO_MONITOR hint disables real-time SQL monitoring for the query, even if the query
is long running.

NO_NATIVE_FULL_OUTER_JOIN Hint
/*+

NO_NATIVE_FULL_OUTER_JOIN

*/

The NO_NATIVE_FULL_OUTER_JOIN hint instructs the optimizer to exclude the native
execution method when joining each specified table. Instead, the full outer join is
executed as a union of left outer join and anti-join.
See Also:

NATIVE_FULL_OUTER_JOIN Hint on page 3-90

NO_PARALLEL Hint
@
/*+

NO_PARALLEL

(

queryblock
tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
Basic Elements of Oracle SQL 3-93

Comments

The NO_PARALLEL hint instructs the optimizer to run the statement serially. This hint
overrides the value of the PARALLEL_DEGREE_POLICY initialization parameter. It also
overrides a PARALLEL parameter in the DDL that created or altered the table. For
example, the following SELECT statement will run serially:
ALTER TABLE employees PARALLEL 8;
SELECT /*+ NO_PARALLEL(hr_emp) */ last_name
FROM employees hr_emp;

See Also:
■

■

"Note on Parallel Hints" on page 3-98 for more information on the
parallel hints
Oracle Database Reference for more information on the PARALLEL_
DEGREE_POLICY initialization parameter

NOPARALLEL Hint
The NOPARALLEL hint has been deprecated. Use the NO_PARALLEL hint instead.

NO_PARALLEL_INDEX Hint
@
/*+

NO_PARALLEL_INDEX

queryblock

(

indexspec
tablespec

)

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75,
indexspec::= on page 3-76)
The NO_PARALLEL_INDEX hint overrides a PARALLEL parameter in the DDL that created
or altered the index, thus avoiding a parallel index scan operation.
See Also: "Note on Parallel Hints" on page 3-98 for more
information on the parallel hints

NOPARALLEL_INDEX Hint
The NOPARALLEL_INDEX hint has been deprecated. Use the NO_PARALLEL_INDEX hint
instead.

NO_PUSH_PRED Hint
@
(

queryblock
@

)

queryblock
tablespec

/*+

NO_PUSH_PRED

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
The NO_PUSH_PRED hint instructs the optimizer not to push a join predicate into the
view. For example:
SELECT /*+ NO_MERGE(v) NO_PUSH_PRED(v) */ *
FROM employees e,
(SELECT manager_id
FROM employees) v
WHERE e.manager_id = v.manager_id(+)

3-94 Oracle Database SQL Language Reference

*/

Comments

AND e.employee_id = 100;

NO_PUSH_SUBQ Hint
(
/*+

@

queryblock

)

NO_PUSH_SUBQ

*/

(See "Specifying a Query Block in a Hint" on page 3-75)
The NO_PUSH_SUBQ hint instructs the optimizer to evaluate nonmerged subqueries as
the last step in the execution plan. Doing so can improve performance if the subquery
is relatively expensive or does not reduce the number of rows significantly.

NO_PX_JOIN_FILTER Hint
/*+

NO_PX_JOIN_FILTER

(

tablespec

)

*/

This hint prevents the optimizer from using parallel join bitmap filtering.

NO_QUERY_TRANSFORMATION Hint
/*+

NO_QUERY_TRANSFORMATION

*/

The NO_QUERY_TRANSFORMATION hint instructs the optimizer to skip all query
transformations, including but not limited to OR-expansion, view merging, subquery
unnesting, star transformation, and materialized view rewrite. For example:
SELECT /*+ NO_QUERY_TRANSFORMATION */ employee_id, last_name
FROM (SELECT * FROM employees e) v
WHERE v.last_name = 'Smith';

NO_RESULT_CACHE Hint
/*+

NO_RESULT_CACHE

*/

The optimizer caches query results in the result cache if the RESULT_CACHE_MODE
initialization parameter is set to FORCE. In this case, the NO_RESULT_CACHE hint disables
such caching for the current query.
If the query is executed from OCI client and OCI client result cache is enabled, then the
NO_RESULT_CACHE hint disables caching for the current query.

NO_REWRITE Hint
(
/*+

NO_REWRITE

@

queryblock

)
*/

(See "Specifying a Query Block in a Hint" on page 3-75)
The NO_REWRITE hint instructs the optimizer to disable query rewrite for the query
block, overriding the setting of the parameter QUERY_REWRITE_ENABLED. For example:
SELECT /*+ NO_REWRITE */ sum(s.amount_sold) AS dollars
FROM sales s, times t
WHERE s.time_id = t.time_id
GROUP BY t.calendar_month_desc;

Basic Elements of Oracle SQL 3-95

Comments

NOREWRITE Hint
The NOREWRITE hint has been deprecated. Use the NO_REWRITE hint instead.

NO_STAR_TRANSFORMATION Hint
(
/*+

@

queryblock

)

NO_STAR_TRANSFORMATION

*/

(See "Specifying a Query Block in a Hint" on page 3-75)
The NO_STAR_TRANSFORMATION hint instructs the optimizer not to perform star query
transformation.

NO_STATEMENT_QUEUING Hint
/*+

NO_STATEMENT_QUEUING

*/

The NO_STATEMENT_QUEUING hint influences whether or not a statement is queued with
parallel statement queuing.
When PARALLEL_DEGREE_POLICY is set to AUTO, this hint enables a statement to bypass
the parallel statement queue. However, a statement that bypasses the statement queue
can potentially cause the system to exceed the maximum number of parallel execution
servers defined by the value of the PARALLEL_SERVERS_TARGET initialization parameter,
which determines the limit at which parallel statement queuing is initiated.
There is no guarantee that the statement that bypasses the parallel statement queue
receives the number of parallel execution servers requested because only the number
of parallel execution servers available on the system, up to the value of the PARALLEL_
MAX_SERVERS initialization parameter, can be allocated.
For example:
SELECT /*+ NO_STATEMENT_QUEUING */ emp.last_name, dpt.department_name
FROM employees emp, departments dpt
WHERE emp.department_id = dpt.department_id;

See Also:

"STATEMENT_QUEUING Hint" on page 3-107

NO_UNNEST Hint
(
/*+

@

queryblock

)

NO_UNNEST

*/

(See "Specifying a Query Block in a Hint" on page 3-75)
Use of the NO_UNNEST hint turns off unnesting .

NO_USE_HASH Hint
@
/*+

NO_USE_HASH

(

queryblock
tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)

3-96 Oracle Database SQL Language Reference

Comments

The NO_USE_HASH hint instructs the optimizer to exclude hash joins when joining each
specified table to another row source using the specified table as the inner table. For
example:
SELECT /*+ NO_USE_HASH(e d) */ *
FROM employees e, departments d
WHERE e.department_id = d.department_id;

NO_USE_MERGE Hint
@
/*+

NO_USE_MERGE

queryblock

(

tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
The NO_USE_MERGE hint instructs the optimizer to exclude sort-merge joins when
joining each specified table to another row source using the specified table as the inner
table. For example:
SELECT /*+ NO_USE_MERGE(e d) */ *
FROM employees e, departments d
WHERE e.department_id = d.department_id
ORDER BY d.department_id;

NO_USE_NL Hint
@
/*+

NO_USE_NL

queryblock

(

tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
The NO_USE_NL hint instructs the optimizer to exclude nested loops joins when joining
each specified table to another row source using the specified table as the inner table.
For example:
SELECT /*+ NO_USE_NL(l h) */ *
FROM orders h, order_items l
WHERE l.order_id = h.order_id
AND l.order_id > 2400;

When this hint is specified, only hash join and sort-merge joins are considered for the
specified tables. However, in some cases tables can be joined only by using nested
loops. In such cases, the optimizer ignores the hint for those tables.

NO_XML_QUERY_REWRITE Hint
/*+

NO_XML_QUERY_REWRITE

*/

The NO_XML_QUERY_REWRITE hint instructs the optimizer to prohibit the rewriting of
XPath expressions in SQL statements. By prohibiting the rewriting of XPath
expressions, this hint also prohibits the use of any XMLIndexes for the current query.
For example:
SELECT /*+NO_XML_QUERY_REWRITE*/ XMLQUERY('' RETURNING CONTENT)
FROM DUAL;

See Also:

"NO_XMLINDEX_REWRITE Hint" on page 3-98

Basic Elements of Oracle SQL 3-97

Comments

NO_XMLINDEX_REWRITE Hint
/*+

NO_XMLINDEX_REWRITE

*/

The NO_XMLINDEX_REWRITE hint instructs the optimizer not to use any XMLIndex
indexes for the current query. For example:
SELECT /*+NO_XMLINDEX_REWRITE*/ count(*)
FROM warehouses
WHERE existsNode(warehouse_spec, '/Warehouse/Building') = 1;

See Also: "NO_XML_QUERY_REWRITE Hint" on page 3-97 for
another way to disable the use of XMLIndexes

OPT_PARAM Hint
,
/*+

OPT_PARAM

(

parameter_name

parameter_value

)

*/

The OPT_PARAM hint lets you set an initialization parameter for the duration of the
current query only. This hint is valid only for the following parameters: OPTIMIZER_
DYNAMIC_SAMPLING, OPTIMIZER_INDEX_CACHING, OPTIMIZER_INDEX_COST_ADJ,
OPTIMIZER_SECURE_VIEW_MERGING, and STAR_TRANSFORMATION_ENABLED. For example,
the following hint sets the parameter STAR_TRANSFORMATION_ENABLED to TRUE for the
statement to which it is added:
SELECT /*+ OPT_PARAM('star_transformation_enabled' 'true') */ *
FROM ... ;

Parameter values that are strings are enclosed in single quotation marks. Numeric
parameter values are specified without quotation marks.

ORDERED Hint
/*+

ORDERED

*/

The ORDERED hint instructs Oracle to join tables in the order in which they appear in the
FROM clause. Oracle recommends that you use the LEADING hint, which is more versatile
than the ORDERED hint.
When you omit the ORDERED hint from a SQL statement requiring a join, the optimizer
chooses the order in which to join the tables. You might want to use the ORDERED hint
to specify a join order if you know something that the optimizer does not know about
the number of rows selected from each table. Such information lets you choose an
inner and outer table better than the optimizer could.
The following query is an example of the use of the ORDERED hint:
SELECT /*+ ORDERED */ o.order_id, c.customer_id, l.unit_price * l.quantity
FROM customers c, order_items l, orders o
WHERE c.cust_last_name = 'Taylor'
AND o.customer_id = c.customer_id
AND o.order_id = l.order_id;

PARALLEL Hint
Note on Parallel Hints Beginning with Oracle Database 11g Release 2 (11.2.0.1), the

PARALLEL and NO_PARALLEL hints are statement-level hints and supersede the earlier
3-98 Oracle Database SQL Language Reference

Comments

object-level hints: PARALLEL_INDEX, NO_PARALLEL_INDEX, and previously specified
PARALLEL and NO_PARALLEL hints. For PARALLEL, if you specify integer, then that
degree of parallelism will be used for the statement. If you omit integer, then the
database computes the degree of parallelism. All the access paths that can use
parallelism will use the specified or computed degree of parallelism.
In the syntax diagrams below, parallel_hint_statement shows the syntax for
statement-level hints, and parallel_hint_object shows the syntax for object-level
hints. Object-level hints are supported for backward compatibility, and are superseded
by statement-level hints.
parallel_hint_statement::=
DEFAULT
AUTO
(

)
MANUAL
integer

/*+

PARALLEL

*/

parallel_hint_object::=
integer
@
/*+

PARALLEL

queryblock

(

DEFAULT
tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
The PARALLEL hint instructs the optimizer to use the specified number of concurrent
servers for a parallel operation. This hint overrides the value of the PARALLEL_DEGREE_
POLICY initialization parameter. The hint applies to the SELECT, INSERT, MERGE, UPDATE,
and DELETE portions of a statement, as well as to the table scan portion. If any parallel
restrictions are violated, then the hint is ignored.
The number of servers that can be used is twice the value in
the PARALLEL hint, if sorting or grouping operations also take place.

Note:

For a statement-level PARALLEL hint:
■

■

■

■

■

PARALLEL: The statement always is run parallel, and the database computes the
degree of parallelism, which can be 2 or greater.
PARALLEL (DEFAULT): The optimizer calculates a degree of parallelism equal to the
number of CPUs available on all participating instances times the value of the
PARALLEL_THREADS_PER_CPU initialization parameter.
PARALLEL (AUTO): The database computes the degree of parallelism, which can be 1
or greater. If the computed degree of parallelism is 1, then the statement runs
serially.
PARALLEL (MANUAL): The optimizer is forced to use the parallel settings of the
objects in the statement.
PARALLEL (integer): The optimizer uses the degree of parallelism specified by
integer.

Basic Elements of Oracle SQL 3-99

Comments

In the following example, the optimizer calculates the degree of parallelism. The
statement always runs in parallel.
SELECT /*+ PARALLEL */ last_name
FROM employees;

In the following example, the optimizer calculates the degree of parallelism, but that
degree may be 1, in which case the statement will run serially.
SELECT /*+ PARALLEL (AUTO) */ last_name
FROM employees;

In the following example, the PARALLEL hint advises the optimizer to use the degree of
parallelism currently in effect for the table itself, which is 5:
CREATE TABLE parallel_table (col1 number, col2 VARCHAR2(10)) PARALLEL 5;
SELECT /*+ PARALLEL (MANUAL) */ col2
FROM parallel_table;

For an object-level PARALLEL hint:
■

■

■

PARALLEL: The query coordinator should examine the settings of the initialization
parameters to determine the default degree of parallelism.
PARALLEL (integer): The optimizer uses the degree of parallelism specified by
integer.
PARALLEL (DEFAULT): The optimizer calculates a degree of parallelism equal to the
number of CPUs available on all participating instances times the value of the
PARALLEL_THREADS_PER_CPU initialization parameter.

In the following example, the PARALLEL hint overrides the degree of parallelism
specified in the employees table definition:
SELECT /*+ FULL(hr_emp) PARALLEL(hr_emp, 5) */ last_name
FROM employees hr_emp;

In the next example, the PARALLEL hint overrides the degree of parallelism specified in
the employees table definition and instructs the optimizer to calculate a degree of
parallelism equal to the number of CPUs available on all participating instances times
the value of the PARALLEL_THREADS_PER_CPU initialization parameter.
SELECT /*+ FULL(hr_emp) PARALLEL(hr_emp, DEFAULT) */ last_name
FROM employees hr_emp;

Refer to CREATE TABLE on page 16-6 and Oracle Database Concepts for more
information on parallel execution.

3-100 Oracle Database SQL Language Reference

Comments

See Also:
■

■

■

■

CREATE TABLE on page 16-6 and Oracle Database Concepts for
more information on parallel execution.
Oracle Database PL/SQL Packages and Types Reference for
information on the DBMS_PARALLEL_EXECUTE package, which
provides methods to apply table changes in chunks of rows.
Changes to each chunk are independently committed when there
are no errors.
Oracle Database Reference for more information on the PARALLEL_
DEGREE_POLICY initialization parameter
NO_PARALLEL Hint on page 3-93

PARALLEL_INDEX Hint
@
/*+

PARALLEL_INDEX

queryblock

(

indexspec
tablespec

integer
DEFAULT
)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75,
indexspec::= on page 3-76)
The PARALLEL_INDEX hint instructs the optimizer to use the specified number of
concurrent servers to parallelize index range scans, full scans, and fast full scans for
partitioned indexes.
The integer value indicates the degree of parallelism for the specified index.
Specifying DEFAULT or no value signifies that the query coordinator should examine
the settings of the initialization parameters to determine the default degree of
parallelism. For example, the following hint indicates three parallel execution
processes are to be used:
SELECT /*+ PARALLEL_INDEX(table1, index1, 3) */

See Also: "Note on Parallel Hints" on page 3-98 for more
information on the parallel hints

PQ_DISTRIBUTE Hint
@
/*+

PQ_DISTRIBUTE

(

queryblock

distribution
tablespec

)
outer_distribution

*/

inner_distribution

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
The PQ_DISTRIBUTE hint instructs the optimizer how to distribute rows among
producer and consumer query servers. You can control the distribution of rows for
either joins or for load.

Basic Elements of Oracle SQL 3-101

Comments

You can control the distribution of rows for parallel
INSERT ... SELECT and parallel CREATE TABLE ... AS SELECT statements to direct how rows
should be distributed between the producer (query) and the consumer (load) servers.
Use the upper branch of the syntax by specifying a single distribution method. The
values of the distribution methods and their semantics are described in Table 3–22 on
page 3-102.
Control of Distribution for Load

Table 3–22

Distribution Values for Load

Distribution

Description

NONE

No distribution. That is the query and load operation are
combined into each query server. All servers will load all
partitions. This lack of distribution is useful to avoid the
overhead of distributing rows where there is no skew. Skew can
occur due to empty segments or to a predicate in the statement
that filters out all rows evaluated by the query. If skew occurs
due to using this method, then use either RANDOM or RANDOM_
LOCAL distribution instead.
Note: Use this distribution with care. Each partition loaded
requires a minimum of 512 KB per process of PGA memory. If
you also use compression, then approximately 1.5 MB of PGA
memory is consumer per server.

PARTITION

This method uses the partitioning information of tablespec to
distribute the rows from the query servers to the load servers.
Use this distribution method when it is not possible or desirable
to combine the query and load operations, when the number of
partitions being loaded is greater than or equal to the number of
load servers, and the input data will be evenly distributed across
the partitions being loaded—that is, there is no skew.

RANDOM

This method distributes the rows from the producers in a
round-robin fashion to the consumers. Use this distribution
method when the input data is highly skewed.

RANDOM_LOCAL

This method distributes the rows from the producers to a set of
servers that are responsible for maintaining a given set of
partitions. Two or more servers can be loading the same
partition, but no servers are loading all partitions. Use this
distribution method when the input data is skewed and
combining query and load operations is not possible due to
memory constraints.

For example, in the following direct-load insert operation, the query and load portions
of the operation are combined into each query server:
INSERT /*+ APPEND PARALLEL(target_table, 16) PQ_DISTRIBUTE(target_table, NONE) */
INTO target_table
SELECT * FROM source_table;

In the following table creation example, the optimizer uses the partitioning of target_
table to distribute the rows:
CREATE /*+ PQ_DISTRIBUTE(target_table, PARTITION) */ TABLE target_table
NOLOGGING PARALLEL 16
PARTITION BY HASH (l_orderkey) PARTITIONS 512
AS SELECT * FROM source_table;

Control of Distribution for Joins You control the distribution method for joins by
specifying two distribution methods, as shown in the lower branch of the syntax
diagram, one distribution for the outer table and one distribution for the inner table.
■

outer_distribution is the distribution for the outer table.

3-102 Oracle Database SQL Language Reference

Comments

■

inner_distribution is the distribution for the inner table.

The values of the distributions are HASH, BROADCAST, PARTITION, and NONE. Only six
combinations table distributions are valid, as described in Table 3–23:
Table 3–23

Distribution Values for Joins

Distribution

Description

HASH, HASH

The rows of each table are mapped to consumer query servers,
using a hash function on the join keys. When mapping is
complete, each query server performs the join between a pair of
resulting partitions. This distribution is recommended when the
tables are comparable in size and the join operation is
implemented by hash-join or sort merge join.

BROADCAST, NONE

All rows of the outer table are broadcast to each query server.
The inner table rows are randomly partitioned. This distribution
is recommended when the outer table is very small compared
with the inner table. As a general rule, use this distribution
when the inner table size multiplied by the number of query
servers is greater than the outer table size.

NONE, BROADCAST

All rows of the inner table are broadcast to each consumer query
server. The outer table rows are randomly partitioned. This
distribution is recommended when the inner table is very small
compared with the outer table. As a general rule, use this
distribution when the inner table size multiplied by the number
of query servers is less than the outer table size.

PARTITION, NONE

The rows of the outer table are mapped using the partitioning of
the inner table. The inner table must be partitioned on the join
keys. This distribution is recommended when the number of
partitions of the outer table is equal to or nearly equal to a
multiple of the number of query servers; for example, 14
partitions and 15 query servers.
Note: The optimizer ignores this hint if the inner table is not
partitioned or not equijoined on the partitioning key.

NONE, PARTITION

The rows of the inner table are mapped using the partitioning of
the outer table. The outer table must be partitioned on the join
keys. This distribution is recommended when the number of
partitions of the outer table is equal to or nearly equal to a
multiple of the number of query servers; for example, 14
partitions and 15 query servers.
Note: The optimizer ignores this hint if the outer table is not
partitioned or not equijoined on the partitioning key.

NONE, NONE

Each query server performs the join operation between a pair of
matching partitions, one from each table. Both tables must be
equipartitioned on the join keys.

For example, given two tables r and s that are joined using a hash join, the following
query contains a hint to use hash distribution:
SELECT /*+ORDERED PQ_DISTRIBUTE(s HASH, HASH) USE_HASH (s)*/ column_list
FROM r,s
WHERE r.c=s.c;

To broadcast the outer table r, the query is:
SELECT /*+ORDERED PQ_DISTRIBUTE(s BROADCAST, NONE) USE_HASH (s) */ column_list
FROM r,s
WHERE r.c=s.c;

Basic Elements of Oracle SQL 3-103

Comments

See Also: Oracle Database Performance Tuning Guide for more
information on how Oracle parallelizes join operations

PUSH_PRED Hint
@
(

queryblock
@

)

queryblock
tablespec

/*+

PUSH_PRED

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
The PUSH_PRED hint instructs the optimizer to push a join predicate into the view. For
example:
SELECT /*+ NO_MERGE(v) PUSH_PRED(v) */ *
FROM employees e,
(SELECT manager_id
FROM employees) v
WHERE e.manager_id = v.manager_id(+)
AND e.employee_id = 100;

PUSH_SUBQ Hint
(
/*+

@

queryblock

)

PUSH_SUBQ

*/

(See "Specifying a Query Block in a Hint" on page 3-75)
The PUSH_SUBQ hint instructs the optimizer to evaluate nonmerged subqueries at the
earliest possible step in the execution plan. Generally, subqueries that are not merged
are executed as the last step in the execution plan. If the subquery is relatively
inexpensive and reduces the number of rows significantly, then evaluating the
subquery earlier can improve performance.
This hint has no effect if the subquery is applied to a remote table or one that is joined
using a merge join.

PX_JOIN_FILTER Hint
/*+

PX_JOIN_FILTER

(

tablespec

)

*/

This hint forces the optimizer to use parallel join bitmap filtering.

QB_NAME Hint
/*+

QB_NAME

(

queryblock

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75)
Use the QB_NAME hint to define a name for a query block. This name can then be used
in a hint in the outer query or even in a hint in an inline view to affect query execution
on the tables appearing in the named query block.
If two or more query blocks have the same name, or if the same query block is hinted
twice with different names, then the optimizer ignores all the names and the hints

3-104 Oracle Database SQL Language Reference

Comments

referencing that query block. Query blocks that are not named using this hint have
unique system-generated names. These names can be displayed in the plan table and
can also be used in hints within the query block, or in query block hints. For example:
SELECT /*+ QB_NAME(qb) FULL(@qb e) */ employee_id, last_name
FROM employees e
WHERE last_name = 'Smith';

RESULT_CACHE Hint
/*+

RESULT_CACHE

*/

The RESULT_CACHE hint instructs the database to cache the results of the current query
or query fragment in memory and then to use the cached results in future executions
of the query or query fragment. The hint is recognized in the top-level query, the
subquery_factoring_clause, or FROM clause inline view. The cached results reside in
the result cache memory portion of the shared pool.
A cached result is automatically invalidated whenever a database object used in its
creation is successfully modified. This hint takes precedence over settings of the
RESULT_CACHE_MODE initialization parameter.
The query is eligible for result caching only if all functions entailed in the query—for
example, built-in or user-defined functions or virtual columns—are deterministic.
If the query is executed from OCI client and OCI client result cache is enabled, then
RESULT_CACHE hint enables client caching for the current query.
See Also: Oracle Database Performance Tuning Guide for information
about using this hint, Oracle Database Reference for information about
the RESULT_CACHE_MODE initialization parameter, and Oracle Call
Interface Programmer's Guide for more information about the OCI result
cache and usage guidelines

RETRY_ON_ROW_CHANGE Hint
/*+

RETRY_ON_ROW_CHANGE

*/

The CHANGE_DUPKEY_ERROR_INDEX, IGNORE_ROW_ON_DUPKEY_
INDEX, and RETRY_ON_ROW_CHANGE hints are unlike other hints in that
they have a semantic effect. The general philosophy explained in
"Hints" on page 3-74 does not apply for these three hints.
Note:

This hint is valid only for UPDATE and DELETE operations. It is not supported for INSERT
or MERGE operations. When you specify this hint, the operation is retried when the ORA_
ROWSCN for one or more rows in the set has changed from the time the set of rows to be
modified is determined to the time the block is actually modified.
See Also: IGNORE_ROW_ON_DUPKEY_INDEX Hint on page 3-85
and CHANGE_DUPKEY_ERROR_INDEX Hint on page 3-81 for
information on those hints and Oracle Database Performance Tuning
Guide for more information on using the online application upgrade
related hints

Basic Elements of Oracle SQL 3-105

Comments

REWRITE Hint
@

queryblock

(
/*+

view

)

REWRITE

*/

(See "Specifying a Query Block in a Hint" on page 3-75)
The REWRITE hint instructs the optimizer to rewrite a query in terms of materialized
views, when possible, without cost consideration. Use the REWRITE hint with or
without a view list. If you use REWRITE with a view list and the list contains an eligible
materialized view, then Oracle uses that view regardless of its cost.
Oracle does not consider views outside of the list. If you do not specify a view list,
then Oracle searches for an eligible materialized view and always uses it regardless of
the cost of the final plan.
See Also:
■

■

Oracle Database Concepts and Oracle Database Advanced
Replication for more information on materialized views
Oracle Database Data Warehousing Guide for more information on
using REWRITE with materialized views

STAR_TRANSFORMATION Hint
(
/*+

@

queryblock

)

STAR_TRANSFORMATION

*/

(See "Specifying a Query Block in a Hint" on page 3-75)
The STAR_TRANSFORMATION hint instructs the optimizer to use the best plan in which
the transformation has been used. Without the hint, the optimizer could make a query
optimization decision to use the best plan generated without the transformation,
instead of the best plan for the transformed query. For example:
SELECT /*+ STAR_TRANSFORMATION */ s.time_id, s.prod_id, s.channel_id
FROM sales s, times t, products p, channels c
WHERE s.time_id = t.time_id
AND s.prod_id = p.prod_id
AND s.channel_id = c.channel_id
AND c.channel_desc = 'Tele Sales';

Even if the hint is specified, there is no guarantee that the transformation will take
place. The optimizer generates the subqueries only if it seems reasonable to do so. If no
subqueries are generated, then there is no transformed query, and the best plan for the
untransformed query is used, regardless of the hint.
See Also:
■

■

Oracle Database Data Warehousing Guide for a full discussion of
star transformation.
Oracle Database Reference for more information on the STAR_
TRANSFORMATION_ENABLED initialization parameter.

3-106 Oracle Database SQL Language Reference

Comments

STATEMENT_QUEUING Hint
/*+

STATEMENT_QUEUING

*/

The NO_STATEMENT_QUEUING hint influences whether or not a statement is queued with
parallel statement queuing.
When PARALLEL_DEGREE_POLICY is not set to AUTO, this hint enables a statement to be
considered for parallel statement queuing, but to run only when enough parallel
processes are available to run at the requested DOP. The number of available parallel
execution servers, before queuing is enabled, is equal to the difference between the
number of parallel execution servers in use and the maximum number allowed in the
system, which is defined by the PARALLEL_SERVERS_TARGET initialization parameter.
For example:
SELECT /*+ STATEMENT_QUEUING */ emp.last_name, dpt.department_name
FROM employees emp, departments dpt
WHERE emp.department_id = dpt.department_id;

See Also:

"NO_STATEMENT_QUEUING Hint" on page 3-96

UNNEST Hint
(
/*+

@

queryblock

)

UNNEST

*/

(See "Specifying a Query Block in a Hint" on page 3-75)
The UNNEST hint instructs the optimizer to unnest and merge the body of the subquery
into the body of the query block that contains it, allowing the optimizer to consider
them together when evaluating access paths and joins.
Before a subquery is unnested, the optimizer first verifies whether the statement is
valid. The statement must then pass heuristic and query optimization tests. The
UNNEST hint instructs the optimizer to check the subquery block for validity only. If the
subquery block is valid, then subquery unnesting is enabled without checking the
heuristics or costs.
See Also:
■

■

"Collection Unnesting: Examples" on page 19-54 for more
information on unnesting nested subqueries and the conditions
that make a subquery block valid
Oracle Database Performance Tuning Guide for additional
information on subquery unnesting

USE_CONCAT Hint
(
/*+

USE_CONCAT

@

queryblock

)
*/

(See "Specifying a Query Block in a Hint" on page 3-75)
The USE_CONCAT hint instructs the optimizer to transform combined OR-conditions in
the WHERE clause of a query into a compound query using the UNION ALL set operator.
Without this hint, this transformation occurs only if the cost of the query using the

Basic Elements of Oracle SQL 3-107

Comments

concatenations is cheaper than the cost without them. The USE_CONCAT hint overrides
the cost consideration. For example:
SELECT /*+ USE_CONCAT */ *
FROM employees e
WHERE manager_id = 108
OR department_id = 110;

See Also: The "NO_EXPAND Hint" on page 3-91, which is the
opposite of this hint

USE_HASH Hint
@
/*+

USE_HASH

queryblock

(

tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
The USE_HASH hint instructs the optimizer to join each specified table with another row
source using a hash join. For example:
SELECT /*+ USE_HASH(l h) */ *
FROM orders h, order_items l
WHERE l.order_id = h.order_id
AND l.order_id > 2400;

USE_MERGE Hint
@
/*+

USE_MERGE

queryblock

(

tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
The USE_MERGE hint instructs the optimizer to join each specified table with another
row source using a sort-merge join. For example:
SELECT /*+ USE_MERGE(employees departments) */ *
FROM employees, departments
WHERE employees.department_id = departments.department_id;

Use of the USE_NL and USE_MERGE hints is recommended with the LEADING and ORDERED
hints. The optimizer uses those hints when the referenced table is forced to be the
inner table of a join. The hints are ignored if the referenced table is the outer table.

USE_NL Hint
The USE_NL hint instructs the optimizer to join each specified table to another row
source with a nested loops join, using the specified table as the inner table.
@
/*+

USE_NL

(

queryblock
tablespec

)

*/

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75)
The USE_NL hint instructs the optimizer to join each specified table to another row
source with a nested loops join, using the specified table as the inner table.
Use of the USE_NL and USE_MERGE hints is recommended with the LEADING and ORDERED
hints. The optimizer uses those hints when the referenced table is forced to be the
inner table of a join. The hints are ignored if the referenced table is the outer table.
3-108 Oracle Database SQL Language Reference

Database Objects

In the following example, where a nested loop is forced through a hint, orders is
accessed through a full table scan and the filter condition l.order_id = h.order_id is
applied to every row. For every row that meets the filter condition, order_items is
accessed through the index order_id.
SELECT /*+ USE_NL(l h) */ h.customer_id, l.unit_price * l.quantity
FROM orders h, order_items l
WHERE l.order_id = h.order_id;

Adding an INDEX hint to the query could avoid the full table scan on orders, resulting
in an execution plan similar to one used on larger systems, even though it might not be
particularly efficient here.

USE_NL_WITH_INDEX Hint
@
/*+

USE_NL_WITH_INDEX

(

queryblock

indexspec
tablespec

)

(See "Specifying a Query Block in a Hint" on page 3-75, tablespec::= on page 3-75,
indexspec::= on page 3-76)
The USE_NL_WITH_INDEX hint instructs the optimizer to join the specified table to
another row source with a nested loops join using the specified table as the inner table.
For example:
SELECT /*+ USE_NL_WITH_INDEX(l item_product_ix) */ *
FROM orders h, order_items l
WHERE l.order_id = h.order_id
AND l.order_id > 2400;

The following conditions apply:
■

■

If no index is specified, then the optimizer must be able to use some index with at
least one join predicate as the index key.
If an index is specified, then the optimizer must be able to use that index with at
least one join predicate as the index key.

Database Objects
Oracle Database recognizes objects that are associated with a particular schema and
objects that are not associated with any particular schema, as described in the sections
that follow.

Schema Objects
A schema is a collection of logical structures of data, or schema objects. A schema is
owned by a database user and has the same name as that user. Each user owns a single
schema. Schema objects can be created and manipulated with SQL and include the
following types of objects:
Clusters
Constraints
Database links
Database triggers
Dimensions
External procedure libraries
Index-organized tables
Basic Elements of Oracle SQL 3-109

*/

Database Object Names and Qualifiers

Indexes
Indextypes
Java classes, Java resources, Java sources
Materialized views
Materialized view logs
Mining models
Object tables
Object types
Object views
Operators
Packages
Sequences
Stored functions, stored procedures
Synonyms
Tables
Views

Nonschema Objects
Other types of objects are also stored in the database and can be created and
manipulated with SQL but are not contained in a schema:
Contexts
Directories
Editions
Restore points
Roles
Rollback segments
Tablespaces
Users
In this reference, each type of object is described in Chapter 10 through Chapter 19, in
the section devoted to the statement that creates the database object. These statements
begin with the keyword CREATE. For example, for the definition of a cluster, see
CREATE CLUSTER on page 14-2.
See Also:

Oracle Database Concepts for an overview of database

objects
You must provide names for most types of database objects when you create them.
These names must follow the rules listed in the sections that follow.

Database Object Names and Qualifiers
Some database objects are made up of parts that you can or must name, such as the
columns in a table or view, index and table partitions and subpartitions, integrity
constraints on a table, and objects that are stored within a package, including
procedures and stored functions. This section provides:
■

Rules for naming database objects and database object location qualifiers

■

Guidelines for naming database objects and qualifiers

3-110 Oracle Database SQL Language Reference

Database Object Names and Qualifiers

Note: Oracle uses system-generated names beginning with "SYS_"
for implicitly generated database objects and subobjects, and names
beginning with "ORA_" for some Oracle-supplied objects. Oracle
discourages you from using these prefixes in the names you explicitly
provide to your database objects and subobjects to avoid possible
conflict in name resolution.

Database Object Naming Rules
Every database object has a name. In a SQL statement, you represent the name of an
object with a quoted identifier or a nonquoted identifier.
■

■

A quoted identifier begins and ends with double quotation marks ("). If you name
a schema object using a quoted identifier, then you must use the double quotation
marks whenever you refer to that object.
A nonquoted identifier is not surrounded by any punctuation.

You can use either quoted or nonquoted identifiers to name any database object.
However, database names, global database names, and database link names are
always case insensitive and are stored as uppercase. If you specify such names as
quoted identifiers, then the quotation marks are silently ignored. Refer to CREATE
USER on page 17-7 for additional rules for naming users and passwords.

Oracle does not recommend using quoted identifiers for
database object names. These quoted identifiers are accepted by
SQL*Plus, but they may not be valid when using other tools that
manage database objects.

Note:

The following list of rules applies to both quoted and nonquoted identifiers unless
otherwise indicated:
1.

Names must be from 1 to 30 bytes long with these exceptions:
■

Names of databases are limited to 8 bytes.

■

Names of database links can be as long as 128 bytes.

If an identifier includes multiple parts separated by periods, then each attribute
can be up to 30 bytes long. Each period separator, as well as any surrounding
double quotation marks, counts as one byte. For example, suppose you identify a
column like this:
"schema"."table"."column"

The schema name can be 30 bytes, the table name can be 30 bytes, and the column
name can be 30 bytes. Each of the quotation marks and periods is a single-byte
character, so the total length of the identifier in this example can be up to 98 bytes.
2.

Nonquoted identifiers cannot be Oracle SQL reserved words. Quoted identifiers
can be reserved words, although this is not recommended.
Depending on the Oracle product you plan to use to access a database object,
names might be further restricted by other product-specific reserved words.

Basic Elements of Oracle SQL 3-111

Database Object Names and Qualifiers

The reserved word ROWID is an exception to this rule. You
cannot use the uppercase word ROWID, either quoted or nonquoted, as
a column name. However, you can use the uppercase word as a
quoted identifier that is not a column name, and you can use the word
with one or more lowercase letters (for example, "Rowid" or "rowid") as
any quoted identifier, including a column name.
Note:

See Also:
■

■

3.

"Oracle SQL Reserved Words" on page E-1 for a listing of all
Oracle SQL reserved words
The manual for a specific product, such as Oracle Database PL/SQL
Language Reference, for a list of the reserved words of that product

The Oracle SQL language contains other words that have special meanings. These
words include data types, schema names, function names, the dummy system
table DUAL, and keywords (the uppercase words in SQL statements, such as
DIMENSION, SEGMENT, ALLOCATE, DISABLE, and so forth). These words are not
reserved. However, Oracle uses them internally in specific ways. Therefore, if you
use these words as names for objects and object parts, then your SQL statements
may be more difficult to read and may lead to unpredictable results.
In particular, do not use words beginning with SYS_ or ORA_ as schema object
names, and do not use the names of SQL built-in functions for the names of
schema objects or user-defined functions.
See Also:
■

■

"Oracle SQL Keywords" on page E-3 for information how to
obtain a list of keywords
"Data Types" on page 3-1, About SQL Functions on page 5-2, and
"Selecting from the DUAL Table" on page 9-16

4.

You should use ASCII characters in database names, global database names, and
database link names, because ASCII characters provide optimal compatibility
across different platforms and operating systems.

5.

You can include multibyte characters in passwords.

6.

Nonquoted identifiers must begin with an alphabetic character from your
database character set. Quoted identifiers can begin with any character.

7.

Nonquoted identifiers can contain only alphanumeric characters from your
database character set and the underscore (_), dollar sign ($), and pound sign (#).
Database links can also contain periods (.) and "at" signs (@). Oracle strongly
discourages you from using $ and # in nonquoted identifiers.
Quoted identifiers can contain any characters and punctuations marks as well as
spaces. However, neither quoted nor nonquoted identifiers can contain double
quotation marks or the null character (\0).

8.

Within a namespace, no two objects can have the same name.
The following schema objects share one namespace:
■

Packages

■

Private synonyms

3-112 Oracle Database SQL Language Reference

Database Object Names and Qualifiers

■

Sequences

■

Stand-alone procedures

■

Stand-alone stored functions

■

User-defined operators

■

User-defined types

■

Tables

■

Views

Each of the following schema objects has its own namespace:
■

Clusters

■

Constraints

■

Database triggers

■

Dimensions

■

Indexes

■

■

Materialized views (When you create a materialized view, the database creates
an internal table of the same name. This table has the same namespace as the
other tables in the schema. Therefore, a schema cannot contain a table and a
materialized view of the same name.)
Private database links

Because tables and sequences are in the same namespace, a table and a sequence in
the same schema cannot have the same name. However, tables and indexes are in
different namespaces. Therefore, a table and an index in the same schema can have
the same name.
Each schema in the database has its own namespaces for the objects it contains.
This means, for example, that two tables in different schemas are in different
namespaces and can have the same name.
Each of the following nonschema objects also has its own namespace:
■

Editions

■

Parameter files (PFILEs) and server parameter files (SPFILEs)

■

Profiles

■

Public database links

■

Public synonyms

■

Tablespaces

■

User roles

Because the objects in these namespaces are not contained in schemas, these
namespaces span the entire database.
9.

Nonquoted identifiers are not case sensitive. Oracle interprets them as uppercase.
Quoted identifiers are case sensitive.
By enclosing names in double quotation marks, you can give the following names
to different objects in the same namespace:
"employees"
"Employees"

Basic Elements of Oracle SQL 3-113

Database Object Names and Qualifiers

"EMPLOYEES"

Note that Oracle interprets the following names the same, so they cannot be used
for different objects in the same namespace:
employees
EMPLOYEES
"EMPLOYEES"
10. When Oracle stores or compares identifiers in uppercase, the uppercase form of

each character in the identifiers is determined by applying the uppercasing rules
of the database character set. Language-specific rules determined by the session
setting NLS_SORT are not considered. This behavior corresponds to applying the
SQL function UPPER to the identifier rather than the function NLS_UPPER.
The database character set uppercasing rules can yield results that are incorrect
when viewed as being in a certain natural language. For example, small letter
sharp s ("ß"), used in German, does not have an uppercase form according to the
database character set uppercasing rules. It is not modified when an identifier is
converted into uppercase, while the expected uppercase form in German is the
sequence of two characters capital letter S ("SS"). Similarly, the uppercase form of
small letter i, according to the database character set uppercasing rules, is capital
letter I. However, the expected uppercase form in Turkish and Azerbaijani is
capital letter I with dot above.
The database character set uppercasing rules ensure that identifiers are interpreted
the same in any linguistic configuration of a session. If you want an identifier to
look correctly in a certain natural language, then you can quote it to preserve the
lowercase form or you can use the linguistically correct uppercase form whenever
you use that identifier.
11. Columns in the same table or view cannot have the same name. However,

columns in different tables or views can have the same name.
12. Procedures or functions contained in the same package can have the same name, if

their arguments are not of the same number and data types. Creating multiple
procedures or functions with the same name in the same package with different
arguments is called overloading the procedure or function.

Schema Object Naming Examples
The following examples are valid schema object names:
last_name
horse
hr.hire_date
"EVEN THIS & THAT!"
a_very_long_and_valid_name

All of these examples adhere to the rules listed in "Database Object Naming Rules" on
page 3-111. The following example is not valid, because it exceeds 30 characters:
a_very_very_long_and_not_valid_name

Although column aliases, table aliases, usernames, and passwords are not objects or
parts of objects, they must also follow these naming rules unless otherwise specified in
the rules themselves.

3-114 Oracle Database SQL Language Reference

Syntax for Schema Objects and Parts in SQL Statements

Schema Object Naming Guidelines
Here are several helpful guidelines for naming objects and their parts:
■

Use full, descriptive, pronounceable names (or well-known abbreviations).

■

Use consistent naming rules.

■

Use the same name to describe the same entity or attribute across tables.

When naming objects, balance the objective of keeping names short and easy to use
with the objective of making names as descriptive as possible. When in doubt, choose
the more descriptive name, because the objects in the database may be used by many
people over a period of time. Your counterpart ten years from now may have difficulty
understanding a table column with a name like pmdd instead of payment_due_date.
Using consistent naming rules helps users understand the part that each table plays in
your application. One such rule might be to begin the names of all tables belonging to
the FINANCE application with fin_.
Use the same names to describe the same things across tables. For example, the
department number columns of the sample employees and departments tables are
both named department_id.

Syntax for Schema Objects and Parts in SQL Statements
This section tells you how to refer to schema objects and their parts in the context of a
SQL statement. This section shows you:
■

The general syntax for referring to an object

■

How Oracle resolves a reference to an object

■

How to refer to objects in schemas other than your own

■

How to refer to objects in remote databases

■

How to refer to table and index partitions and subpartitions

The following diagram shows the general syntax for referring to an object or a part:
database_object_or_part::=
schema

.

.

part

@

dblink

object

(dblink::= on page 3-118)
where:
■
■

object is the name of the object.
schema is the schema containing the object. The schema qualifier lets you refer to
an object in a schema other than your own. You must be granted privileges to refer
to objects in other schemas. If you omit schema, then Oracle assumes that you are
referring to an object in your own schema.
Only schema objects can be qualified with schema. Schema objects are shown with
list item 8 on page 3-112. Nonschema objects, also shown with list item 8, cannot
be qualified with schema because they are not schema objects. An exception is
public synonyms, which can optionally be qualified with "PUBLIC". The quotation
marks are required.

Basic Elements of Oracle SQL 3-115

Syntax for Schema Objects and Parts in SQL Statements

■

■

part is a part of the object. This identifier lets you refer to a part of a schema
object, such as a column or a partition of a table. Not all types of objects have
parts.
dblink applies only when you are using the Oracle Database distributed
functionality. This is the name of the database containing the object. The dblink
qualifier lets you refer to an object in a database other than your local database. If
you omit dblink, then Oracle assumes that you are referring to an object in your
local database. Not all SQL statements allow you to access objects on remote
databases.

You can include spaces around the periods separating the components of the reference
to the object, but it is conventional to omit them.

How Oracle Database Resolves Schema Object References
When you refer to an object in a SQL statement, Oracle considers the context of the
SQL statement and locates the object in the appropriate namespace. After locating the
object, Oracle performs the operation specified by the statement on the object. If the
named object cannot be found in the appropriate namespace, then Oracle returns an
error.
The following example illustrates how Oracle resolves references to objects within SQL
statements. Consider this statement that adds a row of data to a table identified by the
name departments:
INSERT INTO departments
VALUES (280, 'ENTERTAINMENT_CLERK', 206, 1700);

Based on the context of the statement, Oracle determines that departments can be:
■

A table in your own schema

■

A view in your own schema

■

A private synonym for a table or view

■

A public synonym

Oracle always attempts to resolve an object reference within the namespaces in your
own schema before considering namespaces outside your schema. In this example,
Oracle attempts to resolve the name departments as follows:
1.

First, Oracle attempts to locate the object in the namespace in your own schema
containing tables, views, and private synonyms. If the object is a private synonym,
then Oracle locates the object for which the synonym stands. This object could be
in your own schema, another schema, or on another database. The object could
also be another synonym, in which case Oracle locates the object for which this
synonym stands.

2.

If the object is in the namespace, then Oracle attempts to perform the statement on
the object. In this example, Oracle attempts to add the row of data to departments.
If the object is not of the correct type for the statement, then Oracle returns an
error. In this example, departments must be a table, view, or a private synonym
resolving to a table or view. If departments is a sequence, then Oracle returns an
error.

3.

If the object is not in any namespace searched in thus far, then Oracle searches the
namespace containing public synonyms. If the object is in that namespace, then
Oracle attempts to perform the statement on it. If the object is not of the correct

3-116 Oracle Database SQL Language Reference

Syntax for Schema Objects and Parts in SQL Statements

type for the statement, then Oracle returns an error. In this example, if
departments is a public synonym for a sequence, then Oracle returns an error.
If a public synonym has any dependent tables or user-defined types, then you cannot
create an object with the same name as the synonym in the same schema as the
dependent objects.
If a synonym does not have any dependent tables or user-defined types, then you can
create an object with the same name in the same schema as the dependent objects.
Oracle invalidates any dependent objects and attempts to revalidate them when they
are next accessed.
Oracle Database PL/SQL Language Reference for information
about how PL/SQL resolves identifier names

See Also:

References to Objects in Other Schemas
To refer to objects in schemas other than your own, prefix the object name with the
schema name:
schema.object

For example, this statement drops the employees table in the sample schema hr:
DROP TABLE hr.employees;

References to Objects in Remote Databases
To refer to objects in databases other than your local database, follow the object name
with the name of the database link to that database. A database link is a schema object
that causes Oracle to connect to a remote database to access an object there. This
section tells you:
■

How to create database links

■

How to use database links in your SQL statements

Creating Database Links
You create a database link with the statement CREATE DATABASE LINK on
page 14-31. The statement lets you specify this information about the database link:
■

The name of the database link

■

The database connect string to access the remote database

■

The username and password to connect to the remote database

Oracle stores this information in the data dictionary.
Database Link Names When you create a database link, you must specify its name.
Database link names are different from names of other types of objects. They can be as
long as 128 bytes and can contain periods (.) and the "at" sign (@).
The name that you give to a database link must correspond to the name of the
database to which the database link refers and the location of that database in the
hierarchy of database names. The following syntax diagram shows the form of the
name of a database link:

Basic Elements of Oracle SQL 3-117

Syntax for Schema Objects and Parts in SQL Statements

dblink::=
.

domain

@

connection_qualifier

database

where:
■

■

■

database should specify the name portion of the global name of the remote
database to which the database link connects. This global name is stored in the
data dictionary of the remote database. You can see this name in the GLOBAL_NAME
data dictionary view.
domain should specify the domain portion of the global name of the remote
database to which the database link connects. If you omit domain from the name of
a database link, then Oracle qualifies the database link name with the domain of
your local database as it currently exists in the data dictionary.
connection_qualifier lets you further qualify a database link. Using connection
qualifiers, you can create multiple database links to the same database. For
example, you can use connection qualifiers to create multiple database links to
different instances of the Oracle Real Application Clusters that access the same
database.
See Also: Oracle Database Administrator's Guide and Oracle Database
Advanced Replication for more information on connection qualifiers

The combination database.domain is sometimes called the service name.
See Also:

Oracle Database Net Services Administrator's Guide

Username and Password Oracle uses the username and password to connect to the
remote database. The username and password for a database link are optional.
Database Connect String The database connect string is the specification used by Oracle
Net to access the remote database. For information on writing database connect
strings, see the Oracle Net documentation for your specific network protocol. The
database connect string for a database link is optional.

References to Database Links
Database links are available only if you are using Oracle distributed functionality.
When you issue a SQL statement that contains a database link, you can specify the
database link name in one of these forms:
■

■

The complete database link name as stored in the data dictionary, including the
database, domain, and optional connection_qualifier components.
The partial database link name is the database and optional connection_
qualifier components, but not the domain component.

Oracle performs these tasks before connecting to the remote database:
1.

If the database link name specified in the statement is partial, then Oracle expands
the name to contain the domain of the local database as found in the global
database name stored in the data dictionary. (You can see the current global
database name in the GLOBAL_NAME data dictionary view.)

3-118 Oracle Database SQL Language Reference

Syntax for Schema Objects and Parts in SQL Statements

2.

Oracle first searches for a private database link in your own schema with the same
name as the database link in the statement. Then, if necessary, it searches for a
public database link with the same name.
■

■

Oracle always determines the username and password from the first matching
database link (either private or public). If the first matching database link has
an associated username and password, then Oracle uses it. If it does not have
an associated username and password, then Oracle uses your current
username and password.
If the first matching database link has an associated database string, then
Oracle uses it. Otherwise Oracle searches for the next matching (public)
database link. If no matching database link is found, or if no matching link has
an associated database string, then Oracle returns an error.

3.

Oracle uses the database string to access the remote database. After accessing the
remote database, if the value of the GLOBAL_NAMES parameter is true, then Oracle
verifies that the database.domain portion of the database link name matches the
complete global name of the remote database. If this condition is true, then Oracle
proceeds with the connection, using the username and password chosen in Step 2.
If not, Oracle returns an error.

4.

If the connection using the database string, username, and password is successful,
then Oracle attempts to access the specified object on the remote database using
the rules for resolving object references and referring to objects in other schemas
discussed earlier in this section.

You can disable the requirement that the database.domain portion of the database link
name must match the complete global name of the remote database by setting to FALSE
the initialization parameter GLOBAL_NAMES or the GLOBAL_NAMES parameter of the ALTER
SYSTEM or ALTER SESSION statement.
See Also: Oracle Database Administrator's Guide for more information
on remote name resolution

References to Partitioned Tables and Indexes
Tables and indexes can be partitioned. When partitioned, these schema objects consist
of a number of parts called partitions, all of which have the same logical attributes.
For example, all partitions in a table share the same column and constraint definitions,
and all partitions in an index share the same index columns.
Partition-extended and subpartition-extended names let you perform some
partition-level and subpartition-level operations, such as deleting all rows from a
partition or subpartition, on only one partition or subpartition. Without extended
names, such operations would require that you specify a predicate (WHERE clause). For
range- and list-partitioned tables, trying to phrase a partition-level operation with a
predicate can be cumbersome, especially when the range partitioning key uses more
than one column. For hash partitions and subpartitions, using a predicate is more
difficult still, because these partitions and subpartitions are based on a system-defined
hash function.
Partition-extended names let you use partitions as if they were tables. An advantage of
this method, which is most useful for range-partitioned tables, is that you can build
partition-level access control mechanisms by granting (or revoking) privileges on these
views to (or from) other users or roles. To use a partition as a table, create a view by
selecting data from a single partition, and then use the view as a table.

Basic Elements of Oracle SQL 3-119

Syntax for Schema Objects and Parts in SQL Statements

Syntax You can specify partition-extended or subpartition-extended table names in
any SQL statements in which the partition_extended_name or subpartition_
extended_name element appears in the syntax.

partition_extended_name::=
PARTITION

partition
,

PARTITION

FOR

(

partition_key_value

)

subpartition_extended_name::=
SUBPARTITION

subpartition
,

SUBPARTITION

FOR

(

subpartition_key_value

)

The DML statements INSERT, UPDATE, and DELETE and the ANALYZE statement requires
parentheses around the partition or subpartition name. This small distinction is
reflected in the partition_extension_clause:
partition_extension_clause::=
(

partition

)

PARTITION

,
FOR
(

(

partition_key_value

subpartition

SUBPARTITION

)

)
,

FOR

(

subpartition_key_value

)

In partition_extended_name, subpartition_extended_name, and partition_
extension_clause, the PARTITION FOR and SUBPARTITION FOR clauses let you refer to a
partition without using its name. They are valid with any type of partitioning and are
especially useful for interval partitions. Interval partitions are created automatically as
needed when data is inserted into a table.
For the respective partition_key_value or subpartition_key_value, specify one
value for each partitioning key column. For multicolumn partitioning keys, specify
one value for each partitioning key. For composite partitions, specify one value for
each partitioning key, followed by one value for each subpartitioning key. All
partitioning key values are comma separated. For interval partitions, you can specify
only one partition_key_value, and it must be a valid NUMBER or datetime value. Your
SQL statement will operate on the partition or subpartitions that contain the values
you specify.
See Also: The CREATE TABLE "INTERVAL Clause" on page 16-52 for
more information on interval partitions

Currently, the use of partition-extended and
subpartition-extended table names has the following restrictions:

Restrictions on Extended Names

3-120 Oracle Database SQL Language Reference

Syntax for Schema Objects and Parts in SQL Statements

■

■

■

■

■

No remote tables: A partition-extended or subpartition-extended table name
cannot contain a database link (dblink) or a synonym that translates to a table with
a dblink. To use remote partitions and subpartitions, create a view at the remote
site that uses the extended table name syntax and then refer to the remote view.
No synonyms: A partition or subpartition extension must be specified with a base
table. You cannot use synonyms, views, or any other objects.
The PARTITION FOR and SUBPARTITION FOR clauses are not valid for DDL
operations on views.
In the PARTITION FOR and SUBPARTITION FOR clauses, you cannot specify the
keywords DEFAULT or MAXVALUE or a bind variable for the partition_key_value or
subpartition_key_value.
In the PARTITION and SUBPARTITION clauses, you cannot specify a bind variable for
the partition or subpartition name.

In the following statement, sales is a partitioned table with partition
sales_q1_2000. You can create a view of the single partition sales_q1_2000, and then
use it as if it were a table. This example deletes rows from the partition.
Example

CREATE VIEW Q1_2000_sales AS
SELECT *
FROM sales PARTITION (SALES_Q1_2000);
DELETE FROM Q1_2000_sales
WHERE amount_sold < 0;

References to Object Type Attributes and Methods
To refer to object type attributes or methods in a SQL statement, you must fully qualify
the reference with a table alias. Consider the following example from the sample
schema oe, which contains a type cust_address_typ and a table customers with a
cust_address column based on the cust_address_typ:
CREATE TYPE cust_address_typ
OID '82A4AF6A4CD1656DE034080020E0EE3D'
AS OBJECT
(street_address
VARCHAR2(40),
postal_code
VARCHAR2(10),
city
VARCHAR2(30),
state_province
VARCHAR2(10),
country_id
CHAR(2));
/
CREATE TABLE customers
(customer_id
NUMBER(6),
cust_first_name
VARCHAR2(20) CONSTRAINT cust_fname_nn NOT NULL,
cust_last_name
VARCHAR2(20) CONSTRAINT cust_lname_nn NOT NULL,
cust_address
cust_address_typ,
. . .

In a SQL statement, reference to the postal_code attribute must be fully qualified
using a table alias, as illustrated in the following example:
SELECT c.cust_address.postal_code
FROM customers c;
UPDATE customers c
SET c.cust_address.postal_code = '14621-2604'
WHERE c.cust_address.city = 'Rochester'

Basic Elements of Oracle SQL 3-121

Syntax for Schema Objects and Parts in SQL Statements

AND c.cust_address.state_province = 'NY';

To reference a member method that does not accept arguments, you must provide
empty parentheses. For example, the sample schema oe contains an object table
categories_tab, based on catalog_typ, which contains the member function
getCatalogName. In order to call this method in a SQL statement, you must provide
empty parentheses as shown in this example:
SELECT TREAT(VALUE(c) AS catalog_typ).getCatalogName() "Catalog Type"
FROM categories_tab c
WHERE category_id = 90;
Catalog Type
-----------------------------------online catalog

3-122 Oracle Database SQL Language Reference

4
4

Operators

An operator manipulates data items and returns a result. Syntactically, an operator
appears before or after an operand or between two operands.
This chapter contains these sections:
■

About SQL Operators

■

Arithmetic Operators

■

Concatenation Operator

■

Hierarchical Query Operators

■

Set Operators

■

Multiset Operators

■

User-Defined Operators

This chapter discusses nonlogical (non-Boolean) operators. These operators cannot by
themselves serve as the condition of a WHERE or HAVING clause in queries or subqueries.
For information on logical operators, which serve as conditions, refer to Chapter 7,
"Conditions".

About SQL Operators
Operators manipulate individual data items called operands or arguments. Operators
are represented by special characters or by keywords. For example, the multiplication
operator is represented by an asterisk (*).
If you have installed Oracle Text, then you can use the SCORE operator, which is part of
that product, in Oracle Text queries. You can also create conditions with the built-in
Text operators, including CONTAINS, CATSEARCH, and MATCHES. For more information on
these Oracle Text elements, refer to Oracle Text Reference.
If you are using Oracle Expression Filter, then you can create conditions with the
built-in EVALUATE operator that is part of that product. For more information, refer to
Oracle Database Rules Manager and Expression Filter Developer's Guide.

Operators

4-1

About SQL Operators

The combined values of the NLS_COMP and NLS_SORT settings
determine the rules by which characters are sorted and compared. If
NLS_COMP is set to LINGUISTIC for your database, then all entities in
this chapter will be interpreted according to the rules specified by the
NLS_SORT parameter. If NLS_COMP is not set to LINGUISTIC, then the
functions are interpreted without regard to the NLS_SORT setting. NLS_
SORT can be explicitly set. If it is not set explicitly, it is derived from
NLS_LANGUAGE. Refer to Oracle Database Globalization Support Guide for
more information on these settings.
Note:

Unary and Binary Operators
The two general classes of operators are:
■

unary: A unary operator operates on only one operand. A unary operator typically
appears with its operand in this format:
operator operand

■

binary: A binary operator operates on two operands. A binary operator appears
with its operands in this format:
operand1 operator operand2

Other operators with special formats accept more than two operands. If an operator is
given a null operand, then the result is always null. The only operator that does not
follow this rule is concatenation (||).

Operator Precedence
Precedence is the order in which Oracle Database evaluates different operators in the
same expression. When evaluating an expression containing multiple operators,
Oracle evaluates operators with higher precedence before evaluating those with lower
precedence. Oracle evaluates operators with equal precedence from left to right within
an expression.
Table 4–1 lists the levels of precedence among SQL operators from high to low.
Operators listed on the same line have the same precedence.
Table 4–1

SQL Operator Precedence

Operator

Operation

+, - (as unary operators), PRIOR, CONNECT_BY_
ROOT

Identity, negation, location in hierarchy

*, /

Multiplication, division

+, - (as binary operators), ||

Addition, subtraction, concatenation

SQL conditions are evaluated after SQL
operators

See "Condition Precedence" on page 7-3

Precedence Example In the following expression, multiplication has a higher
precedence than addition, so Oracle first multiplies 2 by 3 and then adds the result to
1.
1+2*3

4-2 Oracle Database SQL Language Reference

Arithmetic Operators

You can use parentheses in an expression to override operator precedence. Oracle
evaluates expressions inside parentheses before evaluating those outside.
SQL also supports set operators (UNION, UNION ALL, INTERSECT, and MINUS), which
combine sets of rows returned by queries, rather than individual data items. All set
operators have equal precedence.
"Hierarchical Query Operators" on page 4-5 and
"Hierarchical Queries" on page 9-3 for information on the PRIOR
operator, which is used only in hierarchical queries

See Also:

Arithmetic Operators
You can use an arithmetic operator with one or two arguments to negate, add,
subtract, multiply, and divide numeric values. Some of these operators are also used in
datetime and interval arithmetic. The arguments to the operator must resolve to
numeric data types or to any data type that can be implicitly converted to a numeric
data type.
Unary arithmetic operators return the same data type as the numeric data type of the
argument. For binary arithmetic operators, Oracle determines the argument with the
highest numeric precedence, implicitly converts the remaining arguments to that data
type, and returns that data type. Table 4–2 lists arithmetic operators.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion, "Numeric
Precedence" on page 3-14 for information on numeric precedence, and
"Datetime/Interval Arithmetic" on page 3-20
Table 4–2

Arithmetic Operators

Operator

Purpose

Example

+-

When these denote a positive or
negative expression, they are unary
operators.

SELECT *
FROM order_items
WHERE quantity = -1
ORDER BY order_id,
line_item_id, product_id;
SELECT *
FROM employees
WHERE -salary < 0
ORDER BY employee_id;

+-

When they add or subtract, they are
binary operators.

SELECT hire_date
FROM employees
WHERE SYSDATE - hire_date > 365
ORDER BY hire_date;

*/

Multiply, divide. These are binary
operators.

UPDATE employees
SET salary = salary * 1.1;

Do not use two consecutive minus signs (--) in arithmetic expressions to indicate
double negation or the subtraction of a negative value. The characters -- are used to
begin comments within SQL statements. You should separate consecutive minus signs
with a space or parentheses. Refer to "Comments" on page 3-72 for more information
on comments within SQL statements.

Operators

4-3

Concatenation Operator

Concatenation Operator
The concatenation operator manipulates character strings and CLOB data. Table 4–3
describes the concatenation operator.
Table 4–3

Concatenation Operator

Operator

Purpose

Example

||

Concatenates character strings
and CLOB data.

SELECT 'Name is ' || last_name
FROM employees
ORDER BY last_name;

The result of concatenating two character strings is another character string. If both
character strings are of data type CHAR, then the result has data type CHAR and is limited
to 2000 characters. If either string is of data type VARCHAR2, the result has data type
VARCHAR2 and is limited to 4000 characters. If either argument is a CLOB, the result is a
temporary CLOB. Trailing blanks in character strings are preserved by concatenation,
regardless of the data types of the string or CLOB.
On most platforms, the concatenation operator is two solid vertical bars, as shown in
Table 4–3. However, some IBM platforms use broken vertical bars for this operator.
When moving SQL script files between systems having different character sets, such as
between ASCII and EBCDIC, vertical bars might not be translated into the vertical bar
required by the target Oracle Database environment. Oracle provides the CONCAT
character function as an alternative to the vertical bar operator for cases when it is
difficult or impossible to control translation performed by operating system or
network utilities. Use this function in applications that will be moved between
environments with differing character sets.
Although Oracle treats zero-length character strings as nulls, concatenating a
zero-length character string with another operand always results in the other operand,
so null can result only from the concatenation of two null strings. However, this may
not continue to be true in future versions of Oracle Database. To concatenate an
expression that might be null, use the NVL function to explicitly convert the expression
to a zero-length string.
See Also:
■

■
■

"Character Data Types" on page 3-8 for more information on the
differences between the CHAR and VARCHAR2 data types
The functions CONCAT on page 5-52 and NVL on page 5-173
Oracle Database SecureFiles and Large Objects Developer's Guide for
more information about CLOBs

This example creates a table with both CHAR and VARCHAR2
columns, inserts values both with and without trailing blanks, and then selects these
values and concatenates them. Note that for both CHAR and VARCHAR2 columns, the
trailing blanks are preserved.
Concatenation Example

CREATE TABLE tab1 (col1 VARCHAR2(6), col2 CHAR(6),
col3 VARCHAR2(6), col4 CHAR(6));
INSERT INTO tab1 (col1, col2,
VALUES ('abc', 'def

col3,
', 'ghi

col4)
', 'jkl');

SELECT col1 || col2 || col3 || col4 "Concatenation"

4-4 Oracle Database SQL Language Reference

Set Operators

FROM tab1;
Concatenation
-----------------------abcdef
ghi
jkl

Hierarchical Query Operators
Two operators, PRIOR and CONNECT_BY_ROOT, are valid only in hierarchical queries.

PRIOR
In a hierarchical query, one expression in the CONNECT BY condition must be qualified
by the PRIOR operator. If the CONNECT BY condition is compound, then only one
condition requires the PRIOR operator, although you can have multiple PRIOR
conditions. PRIOR evaluates the immediately following expression for the parent row
of the current row in a hierarchical query.
PRIOR is most commonly used when comparing column values with the equality
operator. (The PRIOR keyword can be on either side of the operator.) PRIOR causes
Oracle to use the value of the parent row in the column. Operators other than the
equal sign (=) are theoretically possible in CONNECT BY clauses. However, the conditions
created by these other operators can result in an infinite loop through the possible
combinations. In this case Oracle detects the loop at run time and returns an error.
Refer to "Hierarchical Queries" on page 9-3 for more information on this operator,
including examples.

CONNECT_BY_ROOT
CONNECT_BY_ROOT is a unary operator that is valid only in hierarchical queries. When
you qualify a column with this operator, Oracle returns the column value using data
from the root row. This operator extends the functionality of the CONNECT BY [PRIOR]
condition of hierarchical queries.
You cannot specify this operator in the START
WITH condition or the CONNECT BY condition.
Restriction on CONNECT_BY_ROOT

See Also:

"CONNECT_BY_ROOT Examples" on page 9-7

Set Operators
Set operators combine the results of two component queries into a single result.
Queries containing set operators are called compound queries. Table 4–4 lists SQL set
operators. They are fully described, including examples and restrictions on these
operators, in "The UNION [ALL], INTERSECT, MINUS Operators" on page 9-8.
Table 4–4

Set Operators

Operator

Returns

UNION

All distinct rows selected by either query

UNION ALL

All rows selected by either query, including all duplicates

INTERSECT

All distinct rows selected by both queries

MINUS

All distinct rows selected by the first query but not the second

Operators

4-5

Multiset Operators

Multiset Operators
Multiset operators combine the results of two nested tables into a single nested table.
The examples related to multiset operators require that two nested tables be created
and loaded with data as follows:
First, make a copy of the oe.customers table called customers_demo:
CREATE TABLE customers_demo AS
SELECT * FROM customers;

Next, create a table type called cust_address_tab_typ. This type will be used when
creating the nested table columns.
CREATE TYPE cust_address_tab_typ AS
TABLE OF cust_address_typ;
/

Now, create two nested table columns in the customers_demo table:
ALTER TABLE customers_demo
ADD (cust_address_ntab cust_address_tab_typ,
cust_address2_ntab cust_address_tab_typ)
NESTED TABLE cust_address_ntab STORE AS cust_address_ntab_store
NESTED TABLE cust_address2_ntab STORE AS cust_address2_ntab_store;

Finally, load data into the two new nested table columns using data from the cust_
address column of the oe.customers table:
UPDATE customers_demo cd
SET cust_address_ntab =
CAST(MULTISET(SELECT cust_address
FROM customers c
WHERE c.customer_id =
cd.customer_id) as cust_address_tab_typ);
UPDATE customers_demo cd
SET cust_address2_ntab =
CAST(MULTISET(SELECT cust_address
FROM customers c
WHERE c.customer_id =
cd.customer_id) as cust_address_tab_typ);

MULTISET EXCEPT
MULTISET EXCEPT takes as arguments two nested tables and returns a nested table
whose elements are in the first nested table but not in the second nested table. The two
input nested tables must be of the same type, and the returned nested table is of the
same type as well.
ALL
DISTINCT
nested_table1

■

MULTISET

EXCEPT

nested_table2

The ALL keyword instructs Oracle to return all elements in nested_table1 that are
not in nested_table2. For example, if a particular element occurs m times in
nested_table1 and n times in nested_table2, then the result will have (m-n)
occurrences of the element if m >n and 0 occurrences if m<=n. ALL is the default.

4-6 Oracle Database SQL Language Reference

Multiset Operators

■

■

The DISTINCT keyword instructs Oracle to eliminate any element in nested_
table1 which is also in nested_table2, regardless of the number of occurrences.
The element types of the nested tables must be comparable. Refer to "Comparison
Conditions" on page 7-4 for information on the comparability of nonscalar types.

Example
The following example compares two nested tables and returns a nested table of those
elements found in the first nested table but not in the second nested table:
SELECT customer_id, cust_address_ntab
MULTISET EXCEPT DISTINCT cust_address2_ntab multiset_except
FROM customers_demo
ORDER BY customer_id;
CUSTOMER_ID
----------101
102
103
104
105
. . .

MULTISET_EXCEPT(STREET_ADDRESS, POSTAL_CODE, CITY, STATE_PROVINCE, COUNTRY_ID)
-------------------------------------------------------------------------------CUST_ADDRESS_TAB_TYP()
CUST_ADDRESS_TAB_TYP()
CUST_ADDRESS_TAB_TYP()
CUST_ADDRESS_TAB_TYP()
CUST_ADDRESS_TAB_TYP()

The preceding example requires the table customers_demo and two nested table
columns containing data. Refer to "Multiset Operators" on page 4-6 to create this table
and nested table columns.

MULTISET INTERSECT
MULTISET INTERSECT takes as arguments two nested tables and returns a nested table
whose values are common in the two input nested tables. The two input nested tables
must be of the same type, and the returned nested table is of the same type as well.
ALL
DISTINCT
nested_table1

■

■

■

MULTISET

INTERSECT

nested_table2

The ALL keyword instructs Oracle to return all common occurrences of elements
that are in the two input nested tables, including duplicate common values and
duplicate common NULL occurrences. For example, if a particular value occurs m
times in nested_table1 and n times in nested_table2, then the result would
contain the element min(m,n) times. ALL is the default.
The DISTINCT keyword instructs Oracle to eliminate duplicates from the returned
nested table, including duplicates of NULL, if they exist.
The element types of the nested tables must be comparable. Refer to "Comparison
Conditions" on page 7-4 for information on the comparability of nonscalar types.

Example
The following example compares two nested tables and returns a nested table of those
elements found in both input nested tables:
SELECT customer_id, cust_address_ntab
MULTISET INTERSECT DISTINCT cust_address2_ntab multiset_intersect
FROM customers_demo
ORDER BY customer_id;
CUSTOMER_ID MULTISET_INTERSECT(STREET_ADDRESS, POSTAL_CODE, CITY, STATE_PROVINCE, COUNTRY_ID
----------- -----------------------------------------------------------------------------------

Operators

4-7

Multiset Operators

101
102
103
104
105

CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP('514 W Superior St', '46901', 'Kokomo', 'IN', 'US'))
CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP('2515 Bloyd Ave', '46218', 'Indianapolis', 'IN', 'US'))
CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP('8768 N State Rd 37', '47404', 'Bloomington', 'IN', 'US'))
CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP('6445 Bay Harbor Ln', '46254', 'Indianapolis', 'IN', 'US'))
CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP('4019 W 3Rd St', '47404', 'Bloomington', 'IN', 'US'))

. . .

The preceding example requires the table customers_demo and two nested table
columns containing data. Refer to "Multiset Operators" on page 4-6 to create this table
and nested table columns.

MULTISET UNION
MULTISET UNION takes as arguments two nested tables and returns a nested table whose
values are those of the two input nested tables. The two input nested tables must be of
the same type, and the returned nested table is of the same type as well.
ALL
DISTINCT
nested_table1

■

■

■

MULTISET

UNION

nested_table2

The ALL keyword instructs Oracle to return all elements that are in the two input
nested tables, including duplicate values and duplicate NULL occurrences. This is
the default.
The DISTINCT keyword instructs Oracle to eliminate duplicates from the returned
nested table, including duplicates of NULL, if they exist.
The element types of the nested tables must be comparable. Refer to "Comparison
Conditions" on page 7-4 for information on the comparability of nonscalar types.

Example
The following example compares two nested tables and returns a nested table of
elements from both input nested tables:
SELECT customer_id, cust_address_ntab
MULTISET UNION cust_address2_ntab multiset_union
FROM customers_demo
ORDER BY customer_id;
CUSTOMER_ID MULTISET_UNION(STREET_ADDRESS, POSTAL_CODE, CITY, STATE_PROVINCE, COUNTRY_ID)
----------- ------------------------------------------------------------------------------101 CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP('514 W Superior St', '46901', 'Kokomo', 'IN', 'US'),
CUST_ADDRESS_TYP('514 W Superior St', '46901', 'Kokomo', 'IN', 'US'))
102 CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP('2515 Bloyd Ave', '46218', 'Indianapolis', 'IN', 'US'),
CUST_ADDRESS_TYP('2515 Bloyd Ave', '46218', 'Indianapolis', 'IN','US'))
103 CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP('8768 N State Rd 37', '47404', 'Bloomington', 'IN', 'US'),
CUST_ADDRESS_TYP('8768 N State Rd 37', '47404', 'Bloomington', 'IN', 'US'))
104 CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP('6445 Bay Harbor Ln', '46254', 'Indianapolis', 'IN', 'US'),
CUST_ADDRESS_TYP('6445 Bay Harbor Ln', '46254', 'Indianapolis', 'IN', 'US'))
105 CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP('4019 W 3Rd St', '47404', 'Bloomington', 'IN', 'US'),
CUST_ADDRESS_TYP('4019 W 3Rd St', '47404', 'Bloomington', 'IN', 'US'))
. . .

The preceding example requires the table customers_demo and two nested table
columns containing data. Refer to "Multiset Operators" on page 4-6 to create this table
and nested table columns.

4-8 Oracle Database SQL Language Reference

User-Defined Operators

User-Defined Operators
Like built-in operators, user-defined operators take a set of operands as input and
return a result. However, you create them with the CREATE OPERATOR statement, and
they are identified by user-defined names. They reside in the same namespace as
tables, views, types, and standalone functions.
After you have defined a new operator, you can use it in SQL statements like any other
built-in operator. For example, you can use user-defined operators in the select list of a
SELECT statement, the condition of a WHERE clause, or in ORDER BY clauses and GROUP BY
clauses. However, you must have EXECUTE privilege on the operator to do so, because
it is a user-defined object.
See Also: CREATE OPERATOR on page 15-35 for an example of
creating an operator and Oracle Database Data Cartridge Developer's
Guide for more information on user-defined operators

Operators

4-9

User-Defined Operators

4-10 Oracle Database SQL Language Reference

5
Functions
5

Functions are similar to operators in that they manipulate data items and return a
result. Functions differ from operators in the format of their arguments. This format
enables them to operate on zero, one, two, or more arguments:
function(argument, argument, ...)

A function without any arguments is similar to a pseudocolumn (refer to Chapter 2,
"Pseudocolumns"). However, a pseudocolumn typically returns a different value for
each row in the result set, whereas a function without any arguments typically returns
the same value for each row.
This chapter contains these sections:
■

About SQL Functions

■

Single-Row Functions
–

Numeric Functions

–

Character Functions Returning Character Values

–

Character Functions Returning Number Values

–

Character Set Functions

–

Datetime Functions

–

General Comparison Functions

–

Conversion Functions

–

Large Object Functions

–

Collection Functions

–

Hierarchical Functions

–

Data Mining Functions

–

XML Functions

–

Encoding and Decoding Functions

–

NULL-Related Functions

–

Environment and Identifier Functions

■

Aggregate Functions

■

Analytic Functions

■

Object Reference Functions

Functions 5-1

About SQL Functions

■

Model Functions

■

OLAP Functions

■

Data Cartridge Functions

■

About User-Defined Functions

About SQL Functions
SQL functions are built into Oracle Database and are available for use in various
appropriate SQL statements. Do not confuse SQL functions with user-defined
functions written in PL/SQL.
If you call a SQL function with an argument of a data type other than the data type
expected by the SQL function, then Oracle attempts to convert the argument to the
expected data type before performing the SQL function.
"About User-Defined Functions" on page 5-380 for
information on user functions and "Data Conversion" on page 3-40 for
implicit conversion of data types
See Also:

Nulls in SQL Functions Most scalar functions return null when given a null
argument. You can use the NVL function to return a value when a null occurs. For
example, the expression NVL(commission_pct,0) returns 0 if commission_pct is null or
the value of commission_pct if it is not null.

For information on how aggregate functions handle nulls, see "Aggregate Functions"
on page 5-10.
In the syntax diagrams for SQL functions, arguments are
indicated by their data types. When the parameter function appears in SQL syntax,
replace it with one of the functions described in this section. Functions are grouped by
the data types of their arguments and their return values.
Syntax for SQL Functions

When you apply SQL functions to LOB columns, Oracle
Database creates temporary LOBs during SQL and PL/SQL
processing. You should ensure that temporary tablespace quota is
sufficient for storing these temporary LOBs for your application.

Note:

The combined values of the NLS_COMP and NLS_SORT settings
determine the rules by which characters are sorted and compared. If
NLS_COMP is set to LINGUISTIC for your database, then all entities in
this chapter will be interpreted according to the rules specified by the
NLS_SORT parameter. If NLS_COMP is not set to LINGUISTIC, then the
functions are interpreted without regard to the NLS_SORT setting. NLS_
SORT can be explicitly set. If it is not set explicitly, it is derived from
NLS_LANGUAGE. Refer to Oracle Database Globalization Support Guide for
more information on these settings.
Note:

The syntax showing the categories of functions follows:

5-2 Oracle Database SQL Language Reference

Single-Row Functions

function::=
single_row_function
aggregate_function
analytic_function
object_reference_function
model_function
OLAP_function
data_cartridge_function
user_defined_function

single_row_function::=
numeric_function
character_function
datetime_function
comparison_function
conversion_function
large_object_function
collection_function
hierarchical_function
data_mining_function
XML_function
encoding_decoding_function
NULL_related_function
environment_id_function

The sections that follow list the built-in SQL functions in each of the groups illustrated
in the preceding diagrams except user-defined functions. All of the built-in SQL
functions are then described in alphabetical order.
"About User-Defined Functions" on page 5-380 and
CREATE FUNCTION on page 14-58

See Also:

Single-Row Functions
Single-row functions return a single result row for every row of a queried table or
view. These functions can appear in select lists, WHERE clauses, START WITH and CONNECT
BY clauses, and HAVING clauses.

Numeric Functions
Numeric functions accept numeric input and return numeric values. Most numeric
functions return NUMBER values that are accurate to 38 decimal digits. The
transcendental functions COS, COSH, EXP, LN, LOG, SIN, SINH, SQRT, TAN, and TANH are
Functions 5-3

Single-Row Functions

accurate to 36 decimal digits. The transcendental functions ACOS, ASIN, ATAN, and ATAN2
are accurate to 30 decimal digits. The numeric functions are:
ABS
ACOS
ASIN
ATAN
ATAN2
BITAND
CEIL
COS
COSH
EXP
FLOOR
LN
LOG
MOD
NANVL
POWER
REMAINDER
ROUND (number)
SIGN
SIN
SINH
SQRT
TAN
TANH
TRUNC (number)
WIDTH_BUCKET

Character Functions Returning Character Values
Character functions that return character values return values of the following data
types unless otherwise documented:
■

If the input argument is CHAR or VARCHAR2, then the value returned is VARCHAR2.

■

If the input argument is NCHAR or NVARCHAR2, then the value returned is NVARCHAR2.

The length of the value returned by the function is limited by the maximum length of
the data type returned.
■

■

For functions that return CHAR or VARCHAR2, if the length of the return value
exceeds the limit, then Oracle Database truncates it and returns the result without
an error message.
For functions that return CLOB values, if the length of the return values exceeds the
limit, then Oracle raises an error and returns no data.

The character functions that return character values are:
CHR
CONCAT
INITCAP
LOWER
LPAD
LTRIM
NCHR
NLS_INITCAP

5-4 Oracle Database SQL Language Reference

Single-Row Functions

NLS_LOWER
NLS_UPPER
NLSSORT
REGEXP_REPLACE
REGEXP_SUBSTR
REPLACE
RPAD
RTRIM
SOUNDEX
SUBSTR
TRANSLATE
TRANSLATE ... USING
TRIM
UPPER

Character Functions Returning Number Values
Character functions that return number values can take as their argument any
character data type. The character functions that return number values are:
ASCII
INSTR
LENGTH
REGEXP_COUNT
REGEXP_INSTR

Character Set Functions
The character set functions return information about the character set. The character
set functions are:
NLS_CHARSET_DECL_LEN
NLS_CHARSET_ID
NLS_CHARSET_NAME

Datetime Functions
Datetime functions operate on date (DATE), timestamp (TIMESTAMP, TIMESTAMP WITH
TIME ZONE, and TIMESTAMP WITH LOCAL TIME ZONE), and interval (INTERVAL DAY TO
SECOND, INTERVAL YEAR TO MONTH) values.
Some of the datetime functions were designed for the Oracle DATE data type (ADD_
MONTHS, CURRENT_DATE, LAST_DAY, NEW_TIME, and NEXT_DAY). If you provide a
timestamp value as their argument, then Oracle Database internally converts the input
type to a DATE value and returns a DATE value. The exceptions are the MONTHS_BETWEEN
function, which returns a number, and the ROUND and TRUNC functions, which do not
accept timestamp or interval values at all.
The remaining datetime functions were designed to accept any of the three types of
data (date, timestamp, and interval) and to return a value of one of these types.
All of the datetime functions that return current system datetime information, such as
SYSDATE, SYSTIMESTAMP, CURRENT_TIMESTAMP, and so forth, are evaluated once for each
SQL statement, regardless how many times they are referenced in that statement.
The datetime functions are:
ADD_MONTHS
CURRENT_DATE
Functions 5-5

Single-Row Functions

CURRENT_TIMESTAMP
DBTIMEZONE
EXTRACT (datetime)
FROM_TZ
LAST_DAY
LOCALTIMESTAMP
MONTHS_BETWEEN
NEW_TIME
NEXT_DAY
NUMTODSINTERVAL
NUMTOYMINTERVAL
ORA_DST_AFFECTED
ORA_DST_CONVERT
ORA_DST_ERROR
ROUND (date)
SESSIONTIMEZONE
SYS_EXTRACT_UTC
SYSDATE
SYSTIMESTAMP
TO_CHAR (datetime)
TO_DSINTERVAL
TO_TIMESTAMP
TO_TIMESTAMP_TZ
TO_YMINTERVAL
TRUNC (date)
TZ_OFFSET

General Comparison Functions
The general comparison functions determine the greatest and or least value from a set
of values. The general comparison functions are:
GREATEST
LEAST

Conversion Functions
Conversion functions convert a value from one data type to another. Generally, the
form of the function names follows the convention datatype TO datatype. The first
data type is the input data type. The second data type is the output data type. The SQL
conversion functions are:
ASCIISTR
BIN_TO_NUM
CAST
CHARTOROWID
COMPOSE
CONVERT
DECOMPOSE
HEXTORAW
NUMTODSINTERVAL
NUMTOYMINTERVAL
RAWTOHEX
RAWTONHEX
ROWIDTOCHAR
ROWIDTONCHAR

5-6 Oracle Database SQL Language Reference

Single-Row Functions

SCN_TO_TIMESTAMP
TIMESTAMP_TO_SCN
TO_BINARY_DOUBLE
TO_BINARY_FLOAT
TO_BLOB
TO_CHAR (character)
TO_CHAR (datetime)
TO_CHAR (number)
TO_CLOB
TO_DATE
TO_DSINTERVAL
TO_LOB
TO_MULTI_BYTE
TO_NCHAR (character)
TO_NCHAR (datetime)
TO_NCHAR (number)
TO_NCLOB
TO_NUMBER
TO_SINGLE_BYTE
TO_TIMESTAMP
TO_TIMESTAMP_TZ
TO_YMINTERVAL
TREAT
UNISTR

Large Object Functions
The large object functions operate on LOBs. The large object functions are:
BFILENAME
EMPTY_BLOB, EMPTY_CLOB

Collection Functions
The collection functions operate on nested tables and varrays. The SQL collection
functions are:
CARDINALITY
COLLECT
POWERMULTISET
POWERMULTISET_BY_CARDINALITY
SET

Hierarchical Functions
Hierarchical functions applies hierarchical path information to a result set. The
hierarchical function is:
SYS_CONNECT_BY_PATH

Data Mining Functions
The data mining functions operate on models that have been built using the DBMS_
DATA_MINING package or the Oracle Data Mining Java API. The SQL data mining
functions are:
CLUSTER_ID

Functions 5-7

Single-Row Functions

CLUSTER_PROBABILITY
CLUSTER_SET
FEATURE_ID
FEATURE_SET
FEATURE_VALUE
PREDICTION
PREDICTION_BOUNDS
PREDICTION_COST
PREDICTION_DETAILS
PREDICTION_PROBABILITY
PREDICTION_SET
See Also: Oracle Data Mining Application Developer's Guide for
information on SQL data mining functions

XML Functions
The XML functions operate on or return XML documents or fragments. These
functions use arguments that are not defined as part of the ANSI/ISO/IEC SQL
Standard but are defined as part of the World Wide Web Consortium (W3C) standards.
The processing and operations that the functions perform are defined by the relevant
W3C standards. The table below provides a link to the appropriate section of the W3C
standard for the rules and guidelines that apply to each of these XML-related
arguments. A SQL statement that uses one of these XML functions, where any of the
arguments does not conform to the relevant W3C syntax, will result in an error. Of
special note is the fact that not every character that is allowed in the value of a
database column is considered legal in XML.
Syntax Element

W3C Standard URL

value_expr

http://www.w3.org/TR/2006/REC-xml-20060816

Xpath_string

http://www.w3.org/TR/1999/REC-xpath-19991116

XQuery_string

http://www.w3.org/TR/2007/REC-xquery-semantics-20070123/

namespace_string

http://www.w3.org/TR/2006/REC-xml-names-20060816/

identifier

http://www.w3.org/TR/2006/REC-xml-20060816/#NT-Nmtoken

For more information about selecting and querying XML data using these functions,
including information on formatting output, refer to Oracle XML DB Developer's Guide.
The SQL XML functions are:
APPENDCHILDXML
DELETEXML
DEPTH
EXISTSNODE
EXTRACT (XML)
EXTRACTVALUE
INSERTCHILDXML
INSERTCHILDXMLAFTER
INSERTCHILDXMLBEFORE
INSERTXMLAFTER
INSERTXMLBEFORE
PATH
SYS_DBURIGEN
SYS_XMLAGG
5-8 Oracle Database SQL Language Reference

Single-Row Functions

SYS_XMLGEN
UPDATEXML
XMLAGG
XMLCAST
XMLCDATA
XMLCOLATTVAL
XMLCOMMENT
XMLCONCAT
XMLDIFF
XMLELEMENT
XMLEXISTS
XMLFOREST
XMLISVALID
XMLPARSE
XMLPATCH
XMLPI
XMLQUERY
XMLROOT
XMLSEQUENCE
XMLSERIALIZE
XMLTABLE
XMLTRANSFORM

Encoding and Decoding Functions
The encoding and decoding functions let you inspect and decode data in the database.
The encoding and decoding functions are:
DECODE
DUMP
ORA_HASH
VSIZE

NULL-Related Functions
The NULL-related functions facilitate null handling. The NULL-related functions are:
COALESCE
LNNVL
NANVL
NULLIF
NVL
NVL2

Environment and Identifier Functions
The environment and identifier functions provide information about the instance and
session. The environment and identifier functions are:
SYS_CONTEXT
SYS_GUID
SYS_TYPEID
UID
USER
USERENV

Functions 5-9

Aggregate Functions

Aggregate Functions
Aggregate functions return a single result row based on groups of rows, rather than on
single rows. Aggregate functions can appear in select lists and in ORDER BY and HAVING
clauses. They are commonly used with the GROUP BY clause in a SELECT statement,
where Oracle Database divides the rows of a queried table or view into groups. In a
query containing a GROUP BY clause, the elements of the select list can be aggregate
functions, GROUP BY expressions, constants, or expressions involving one of these.
Oracle applies the aggregate functions to each group of rows and returns a single
result row for each group.
If you omit the GROUP BY clause, then Oracle applies aggregate functions in the select
list to all the rows in the queried table or view. You use aggregate functions in the
HAVING clause to eliminate groups from the output based on the results of the
aggregate functions, rather than on the values of the individual rows of the queried
table or view.
See Also: "Using the GROUP BY Clause: Examples" on page 19-42
and the "HAVING Clause" on page 19-28 for more information on the
GROUP BY clause and HAVING clauses in queries and subqueries

Many (but not all) aggregate functions that take a single argument accept these
clauses:
■

■

DISTINCT and UNIQUE, which are synonymous, cause an aggregate function to
consider only distinct values of the argument expression. The syntax diagrams for
aggregate functions in this chapter use the keyword DISTINCT for simplicity.
ALL causes an aggregate function to consider all values, including all duplicates.

For example, the DISTINCT average of 1, 1, 1, and 3 is 2. The ALL average is 1.5. If you
specify neither, then the default is ALL.
Some aggregate functions allow the windowing_clause, which is part of the syntax of
analytic functions. Refer to windowing_clause on page 5-15 for information about this
clause. In the listing of aggregate functions at the end of this section, the functions that
allow the windowing_clause are followed by an asterisk (*)
All aggregate functions except COUNT(*), GROUPING, and GROUPING_ID ignore nulls. You
can use the NVL function in the argument to an aggregate function to substitute a value
for a null. COUNT and REGR_COUNT never return null, but return either a number or zero.
For all the remaining aggregate functions, if the data set contains no rows, or contains
only rows with nulls as arguments to the aggregate function, then the function returns
null.
The aggregate functions MIN, MAX, SUM, AVG, COUNT, VARIANCE, and STDDEV, when
followed by the KEEP keyword, can be used in conjunction with the FIRST or LAST
function to operate on a set of values from a set of rows that rank as the FIRST or LAST
with respect to a given sorting specification. Refer to FIRST on page 5-101 for more
information.
You can nest aggregate functions. For example, the following example calculates the
average of the maximum salaries of all the departments in the sample schema hr:
SELECT AVG(MAX(salary))
FROM employees
GROUP BY department_id;
AVG(MAX(SALARY))
---------------10926.3333
5-10 Oracle Database SQL Language Reference

Analytic Functions

This calculation evaluates the inner aggregate (MAX(salary)) for each group defined by
the GROUP BY clause (department_id), and aggregates the results again.
In the list of aggregate functions that follows, functions followed by an asterisk (*)
allow the windowing_clause.
AVG
COLLECT
CORR
CORR_*
COUNT
COVAR_POP
COVAR_SAMP
CUME_DIST
DENSE_RANK
FIRST
GROUP_ID
GROUPING
GROUPING_ID
LAST
LISTAGG
MAX
MEDIAN
MIN
PERCENT_RANK
PERCENTILE_CONT
PERCENTILE_DISC
RANK
REGR_ (Linear Regression) Functions
STATS_BINOMIAL_TEST
STATS_CROSSTAB
STATS_F_TEST
STATS_KS_TEST
STATS_MODE
STATS_MW_TEST
STATS_ONE_WAY_ANOVA
STATS_T_TEST_*
STATS_WSR_TEST
STDDEV
STDDEV_POP
STDDEV_SAMP
SUM
SYS_XMLAGG
VAR_POP
VAR_SAMP
VARIANCE
XMLAGG

Analytic Functions
Analytic functions compute an aggregate value based on a group of rows. They differ
from aggregate functions in that they return multiple rows for each group. The group
of rows is called a window and is defined by the analytic_clause. For each row, a
sliding window of rows is defined. The window determines the range of rows used to

Functions 5-11

Analytic Functions

perform the calculations for the current row. Window sizes can be based on either a
physical number of rows or a logical interval such as time.
Analytic functions are the last set of operations performed in a query except for the
final ORDER BY clause. All joins and all WHERE, GROUP BY, and HAVING clauses are
completed before the analytic functions are processed. Therefore, analytic functions
can appear only in the select list or ORDER BY clause.
Analytic functions are commonly used to compute cumulative, moving, centered, and
reporting aggregates.
analytic_function::=
arguments
analytic_function

(

)

OVER

(

analytic_clause

)

analytic_clause::=
windowing_clause
query_partition_clause

order_by_clause

query_partition_clause::=
,
expr
PARTITION

BY

,
(

expr

)

order_by_clause::=
,

expr

SIBLINGS
ORDER

BY

ASC

NULLS

FIRST

DESC

NULLS

LAST

position
c_alias

windowing_clause::=
UNBOUNDED
BETWEEN

CURRENT

PRECEDING
ROW

UNBOUNDED
AND

CURRENT

PRECEDING
value_expr

ROWS
RANGE

ROW
PRECEDING

value_expr
FOLLOWING

UNBOUNDED

FOLLOWING

FOLLOWING

PRECEDING

CURRENT

ROW

value_expr

PRECEDING

The semantics of this syntax are discussed in the sections that follow.

5-12 Oracle Database SQL Language Reference

Analytic Functions

analytic_function
Specify the name of an analytic function (see the listing of analytic functions following
this discussion of semantics).

arguments
Analytic functions take 0 to 3 arguments. The arguments can be any numeric data type
or any nonnumeric data type that can be implicitly converted to a numeric data type.
Oracle determines the argument with the highest numeric precedence and implicitly
converts the remaining arguments to that data type. The return type is also that data
type, unless otherwise noted for an individual function.
See Also: "Numeric Precedence" on page 3-14 for information on
numeric precedence and Table 3–10, " Implicit Type Conversion
Matrix" on page 3-40 for more information on implicit conversion

analytic_clause
Use OVER analytic_clause to indicate that the function operates on a query result set.
This clause is computed after the FROM, WHERE, GROUP BY, and HAVING clauses. You can
specify analytic functions with this clause in the select list or ORDER BY clause. To filter
the results of a query based on an analytic function, nest these functions within the
parent query, and then filter the results of the nested subquery.
Notes on the analytic_clause:
■

■

The following notes apply to the analytic_clause:

You cannot nest analytic functions by specifying any analytic function in any part
of the analytic_clause. However, you can specify an analytic function in a
subquery and compute another analytic function over it.
You can specify OVER analytic_clause with user-defined analytic functions as
well as built-in analytic functions. See CREATE FUNCTION on page 14-58.

query_partition_clause
Use the PARTITION BY clause to partition the query result set into groups based on one
or more value_expr. If you omit this clause, then the function treats all rows of the
query result set as a single group.
To use the query_partition_clause in an analytic function, use the upper branch of
the syntax (without parentheses). To use this clause in a model query (in the model_
column_clauses) or a partitioned outer join (in the outer_join_clause), use the lower
branch of the syntax (with parentheses).
You can specify multiple analytic functions in the same query, each with the same or
different PARTITION BY keys.
If the objects being queried have the parallel attribute, and if you specify an analytic
function with the query_partition_clause, then the function computations are
parallelized as well.
Valid values of value_expr are constants, columns, nonanalytic functions, function
expressions, or expressions involving any of these.
order_by_clause
Use the order_by_clause to specify how data is ordered within a partition. For all
analytic functions you can order the values in a partition on multiple keys, each
defined by a value_expr and each qualified by an ordering sequence.

Functions 5-13

Analytic Functions

Within each function, you can specify multiple ordering expressions. Doing so is
especially useful when using functions that rank values, because the second
expression can resolve ties between identical values for the first expression.
Whenever the order_by_clause results in identical values for multiple rows, the
function behaves as follows:
■

■

■

CUME_DIST, DENSE_RANK, NTILE, PERCENT_RANK, and RANK return the same result for
each of the rows.
ROW_NUMBER assigns each row a distinct value even if there is a tie based on the
order_by_clause. The value is based on the order in which the row is processed,
which may be nondeterministic if the ORDER BY does not guarantee a total ordering.
For all other analytic functions, the result depends on the window specification. If
you specify a logical window with the RANGE keyword, then the function returns
the same result for each of the rows. If you specify a physical window with the
ROWS keyword, then the result is nondeterministic.

Restrictions on the ORDER BY Clause The following restrictions apply to the ORDER

BY clause:
■

■

When used in an analytic function, the order_by_clause must take an expression
(expr). The SIBLINGS keyword is not valid (it is relevant only in hierarchical
queries). Position (position) and column aliases (c_alias) are also invalid.
Otherwise this order_by_clause is the same as that used to order the overall
query or subquery.
An analytic function that uses the RANGE keyword can use multiple sort keys in its
ORDER BY clause if it specifies any of the following windows:
–

RANGE BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW. The short form of this
is RANGE UNBOUNDED PRECEDING.

–

RANGE BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING

–

RANGE BETWEEN CURRENT ROW AND CURRENT ROW

–

RANGE BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING

Window boundaries other than these four can have only one sort key in the ORDER
BY clause of the analytic function. This restriction does not apply to window
boundaries specified by the ROW keyword.
ASC | DESC

Specify the ordering sequence (ascending or descending). ASC is the

default.
NULLS FIRST | NULLS LAST Specify whether returned rows containing nulls should
appear first or last in the ordering sequence.

NULLS LAST is the default for ascending order, and NULLS FIRST is the default for
descending order.
Analytic functions always operate on rows in the order specified in the order_by_
clause of the function. However, the order_by_clause of the function does not
guarantee the order of the result. Use the order_by_clause of the query to guarantee
the final result ordering.
See Also: order_by_clause of SELECT on page 19-33 for more
information on this clause

5-14 Oracle Database SQL Language Reference

Analytic Functions

windowing_clause
Some analytic functions allow the windowing_clause. In the listing of analytic
functions at the end of this section, the functions that allow the windowing_clause are
followed by an asterisk (*).
ROWS | RANGE These keywords define for each row a window (a physical or logical
set of rows) used for calculating the function result. The function is then applied to all
the rows in the window. The window moves through the query result set or partition
from top to bottom.
■

ROWS specifies the window in physical units (rows).

■

RANGE specifies the window as a logical offset.

You cannot specify this clause unless you have specified the order_by_clause. Some
window boundaries defined by the RANGE clause let you specify only one expression in
the order_by_clause. Refer to "Restrictions on the ORDER BY Clause" on page 5-14.
The value returned by an analytic function with a logical offset is always deterministic.
However, the value returned by an analytic function with a physical offset may
produce nondeterministic results unless the ordering expression results in a unique
ordering. You may have to specify multiple columns in the order_by_clause to
achieve this unique ordering.
BETWEEN ... AND Use the BETWEEN ... AND clause to specify a start point and end
point for the window. The first expression (before AND) defines the start point and the
second expression (after AND) defines the end point.

If you omit BETWEEN and specify only one end point, then Oracle considers it the start
point, and the end point defaults to the current row.
Specify UNBOUNDED PRECEDING to indicate that the
window starts at the first row of the partition. This is the start point specification and
cannot be used as an end point specification.

UNBOUNDED PRECEDING

UNBOUNDED FOLLOWING Specify UNBOUNDED FOLLOWING to indicate that the
window ends at the last row of the partition. This is the end point specification and
cannot be used as a start point specification.

As a start point, CURRENT ROW specifies that the window begins at the
current row or value (depending on whether you have specified ROW or RANGE,
respectively). In this case the end point cannot be value_expr PRECEDING.
CURRENT ROW

As an end point, CURRENT ROW specifies that the window ends at the current row or
value (depending on whether you have specified ROW or RANGE, respectively). In this
case the start point cannot be value_expr FOLLOWING.
value_expr PRECEDING or value_expr FOLLOWING For RANGE or ROW:
■

■

If value_expr FOLLOWING is the start point, then the end point must be value_expr
FOLLOWING.
If value_expr PRECEDING is the end point, then the start point must be value_expr
PRECEDING.

If you are defining a logical window defined by an interval of time in numeric format,
then you may need to use conversion functions.

Functions 5-15

Analytic Functions

NUMTOYMINTERVAL on page 5-172 and
NUMTODSINTERVAL on page 5-171 for information on converting
numeric times into intervals

See Also:

If you specified ROWS:
■

■

value_expr is a physical offset. It must be a constant or expression and must
evaluate to a positive numeric value.
If value_expr is part of the start point, then it must evaluate to a row before the
end point.

If you specified RANGE:
■

■
■

■

value_expr is a logical offset. It must be a constant or expression that evaluates to
a positive numeric value or an interval literal. Refer to "Literals" on page 3-45 for
information on interval literals.
You can specify only one expression in the order_by_clause.
If value_expr evaluates to a numeric value, then the ORDER BY expr must be a
numeric or DATE data type.
If value_expr evaluates to an interval value, then the ORDER BY expr must be a
DATE data type.

If you omit the windowing_clause entirely, then the default is RANGE BETWEEN
UNBOUNDED PRECEDING AND CURRENT ROW.
Analytic functions are commonly used in data warehousing environments. In the list
of analytic functions that follows, functions followed by an asterisk (*) allow the full
syntax, including the windowing_clause.
AVG *
CORR *
COUNT *
COVAR_POP *
COVAR_SAMP *
CUME_DIST
DENSE_RANK
FIRST
FIRST_VALUE *
LAG
LAST
LAST_VALUE *
LEAD
LISTAGG
MAX *
MEDIAN
MIN *
NTH_VALUE *
NTILE
PERCENT_RANK
PERCENTILE_CONT
PERCENTILE_DISC
RANK
RATIO_TO_REPORT
REGR_ (Linear Regression) Functions *
ROW_NUMBER
STDDEV *
5-16 Oracle Database SQL Language Reference

Alphabetical Listing of SQL Functions

STDDEV_POP *
STDDEV_SAMP *
SUM *
VAR_POP *
VAR_SAMP *
VARIANCE *
See Also: Oracle Database Data Warehousing Guide for more
information on these functions and for scenarios illustrating their use

Object Reference Functions
Object reference functions manipulate REF values, which are references to objects of
specified object types. The object reference functions are:
DEREF
MAKE_REF
REF
REFTOHEX
VALUE
See Also: Oracle Database Object-Relational Developer's Guide for more
information about REF data types

Model Functions
Model functions can be used only in the model_clause of the SELECT statement. The
model functions are:
CV
ITERATION_NUMBER
PRESENTNNV
PRESENTV
PREVIOUS

OLAP Functions
OLAP functions returns data from a dimensional object in two-dimension relational
format. The OLAP function is:
CUBE_TABLE

Data Cartridge Functions
Data Cartridge functions are useful for Data Cartridge developers. The Data Cartridge
function is:
DATAOBJ_TO_PARTITION

Alphabetical Listing of SQL Functions
The SQL functions are described in alphabetical order.

Functions 5-17

ABS

ABS
Syntax
ABS

(

n

)

Purpose
ABS returns the absolute value of n.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. The function returns the
same data type as the numeric data type of the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

Examples
The following example returns the absolute value of -15:
SELECT ABS(-15) "Absolute"
FROM DUAL;
Absolute
---------15

5-18 Oracle Database SQL Language Reference

ACOS

ACOS
Syntax
ACOS

(

n

)

Purpose
ACOS returns the arc cosine of n. The argument n must be in the range of -1 to 1, and the
function returns a value in the range of 0 to pi, expressed in radians.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. If the argument is
BINARY_FLOAT, then the function returns BINARY_DOUBLE. Otherwise the function
returns the same numeric data type as the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

Examples
The following example returns the arc cosine of .3:
SELECT ACOS(.3)"Arc_Cosine"
FROM DUAL;
Arc_Cosine
---------1.26610367

Functions 5-19

ADD_MONTHS

ADD_MONTHS
Syntax
ADD_MONTHS

(

date

,

integer

)

Purpose
ADD_MONTHS returns the date date plus integer months. A month is defined by the
session parameter NLS_CALENDAR. The date argument can be a datetime value or any
value that can be implicitly converted to DATE. The integer argument can be an
integer or any value that can be implicitly converted to an integer. The return type is
always DATE, regardless of the data type of date. If date is the last day of the month or
if the resulting month has fewer days than the day component of date, then the result
is the last day of the resulting month. Otherwise, the result has the same day
component as date.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

Examples
The following example returns the month after the hire_date in the sample table
employees:
SELECT TO_CHAR(ADD_MONTHS(hire_date, 1), 'DD-MON-YYYY') "Next month"
FROM employees
WHERE last_name = 'Baer';
Next Month
----------07-JUL-2002

5-20 Oracle Database SQL Language Reference

APPENDCHILDXML

APPENDCHILDXML
Syntax
,
APPENDCHILDXML

(

XMLType_instance

,

XPath_string

,

namespace_string

value_expr

)

Purpose
APPENDCHILDXML appends a user-supplied value onto the target XML as the child of the
node indicated by an XPath expression.
■
■

■
■

XMLType_instance is an instance of XMLType.
XPath_string is an Xpath expression indicating one or more nodes onto which
one or more child nodes are to be appended. You can specify an absolute XPath_
string with an initial slash or a relative XPath_string by omitting the initial slash.
If you omit the initial slash, then the context of the relative path defaults to the
root node.
value_expr specifies one or more nodes of XMLType. It must resolve to a string.
The optional namespace_string provides namespace information for the XPath_
string. This parameter must be of type VARCHAR2.
See Also: Oracle XML DB Developer's Guide for more information
about this function

Examples
The following example adds an /Owner node to the /Warehouse/Building node of
warehouse_spec in the oe.warehouses table if the value of the /Building node is
"Rented":
UPDATE warehouses
SET warehouse_spec = APPENDCHILDXML(warehouse_spec, 'Warehouse/Building',
XMLType('Grandco'))
WHERE EXTRACTVALUE(warehouse_spec, '/Warehouse/Building') = 'Rented';
SELECT warehouse_id,
warehouse_name,
EXTRACTVALUE(warehouse_spec, '/Warehouse/Building/Owner') "Prop.Owner"
FROM warehouses
WHERE EXISTSNODE(warehouse_spec, '/Warehouse/Building/Owner') = 1;
WAREHOUSE_ID
-----------2
3

WAREHOUSE_NAME
--------------San Francisco
New Jersey

Prop.Owner
---------Grandco
Grandco

Functions 5-21

ASCII

ASCII
Syntax
ASCII

(

char

)

Purpose
ASCII returns the decimal representation in the database character set of the first
character of char.
char can be of data type CHAR, VARCHAR2, NCHAR, or NVARCHAR2. The value returned is of
data type NUMBER. If your database character set is 7-bit ASCII, then this function
returns an ASCII value. If your database character set is EBCDIC Code, then this
function returns an EBCDIC value. There is no corresponding EBCDIC character
function.
This function does not support CLOB data directly. However, CLOBs can be passed in as
arguments through implicit data conversion.
See Also: "Data Type Comparison Rules" on page 3-36 for more
information

Examples
The following example returns employees whose last names begin with the letter L,
whose ASCII equivalent is 76:
SELECT last_name
FROM employees
WHERE ASCII(SUBSTR(last_name, 1, 1)) = 76
ORDER BY last_name;
LAST_NAME
------------------------Ladwig
Landry
Lee
Livingston
Lorentz

5-22 Oracle Database SQL Language Reference

ASCIISTR

ASCIISTR
Syntax
ASCIISTR

(

char

)

Purpose
ASCIISTR takes as its argument a string, or an expression that resolves to a string, in
any character set and returns an ASCII version of the string in the database character
set. Non-ASCII characters are converted to the form \xxxx, where xxxx represents a
UTF-16 code unit.
See Also: Oracle Database Globalization Support Guide for information
on Unicode character sets and character semantics

Examples
The following example returns the ASCII string equivalent of the text string "ABÄCDE":
SELECT ASCIISTR('ABÄCDE')
FROM DUAL;
ASCIISTR('
---------AB\00C4CDE

Functions 5-23

ASIN

ASIN
Syntax
ASIN

(

n

)

Purpose
ASIN returns the arc sine of n. The argument n must be in the range of -1 to 1, and the
function returns a value in the range of -pi/2 to pi/2, expressed in radians.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. If the argument is
BINARY_FLOAT, then the function returns BINARY_DOUBLE. Otherwise the function
returns the same numeric data type as the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

Examples
The following example returns the arc sine of .3:
SELECT ASIN(.3) "Arc_Sine"
FROM DUAL;
Arc_Sine
---------.304692654

5-24 Oracle Database SQL Language Reference

ATAN

ATAN
Syntax
ATAN

(

n

)

Purpose
ATAN returns the arc tangent of n. The argument n can be in an unbounded range and
returns a value in the range of -pi/2 to pi/2, expressed in radians.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. If the argument is
BINARY_FLOAT, then the function returns BINARY_DOUBLE. Otherwise the function
returns the same numeric data type as the argument.
See Also: ATAN2 on page 5-26 for information about the ATAN2
function and Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

Examples
The following example returns the arc tangent of .3:
SELECT ATAN(.3) "Arc_Tangent"
FROM DUAL;
Arc_Tangent
---------.291456794

Functions 5-25

ATAN2

ATAN2
Syntax
ATAN2

(

n1

,

n2

)

Purpose
ATAN2 returns the arc tangent of n1 and n2. The argument n1 can be in an unbounded
range and returns a value in the range of -pi to pi, depending on the signs of n1 and n2,
expressed in radians.
This function takes as arguments any numeric data type or any nonnumeric data type
that can be implicitly converted to a numeric data type. If any argument is BINARY_
FLOAT or BINARY_DOUBLE, then the function returns BINARY_DOUBLE. Otherwise the
function returns NUMBER.
See Also: ATAN on page 5-25 for information on the ATAN function
and Table 3–10, " Implicit Type Conversion Matrix" on page 3-40 for
more information on implicit conversion

Examples
The following example returns the arc tangent of .3 and .2:
SELECT ATAN2(.3, .2) "Arc_Tangent2"
FROM DUAL;
Arc_Tangent2
-----------.982793723

5-26 Oracle Database SQL Language Reference

AVG

AVG
Syntax
DISTINCT
ALL
AVG

(

OVER
expr

(

analytic_clause

)

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions

Purpose
AVG returns average value of expr.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. The function returns the
same data type as the numeric data type of the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

If you specify DISTINCT, then you can specify only the query_partition_clause of the
analytic_clause. The order_by_clause and windowing_clause are not allowed.
See Also: "About SQL Expressions" on page 6-1 for information on
valid forms of expr and "Aggregate Functions" on page 5-10

Aggregate Example
The following example calculates the average salary of all employees in the
hr.employees table:
SELECT AVG(salary) "Average"
FROM employees;
Average
-------------6461.83178

Analytic Example
The following example calculates, for each employee in the employees table, the
average salary of the employees reporting to the same manager who were hired in the
range just before through just after the employee:
SELECT manager_id, last_name, hire_date, salary,
AVG(salary) OVER (PARTITION BY manager_id ORDER BY hire_date
ROWS BETWEEN 1 PRECEDING AND 1 FOLLOWING) AS c_mavg
FROM employees
ORDER BY manager_id, hire_date, salary;
MANAGER_ID
---------100
100
100
100

LAST_NAME
------------------------De Haan
Raphaely
Kaufling
Hartstein

HIRE_DATE
SALARY
C_MAVG
--------- ---------- ---------13-JAN-01
17000
14000
07-DEC-02
11000 11966.6667
01-MAY-03
7900 10633.3333
17-FEB-04
13000 9633.33333

Functions 5-27

AVG

100
100
100
100

Weiss
Russell
Partners
Errazuriz

. . .

5-28 Oracle Database SQL Language Reference

18-JUL-04
01-OCT-04
05-JAN-05
10-MAR-05

8000
14000
13500
12000

11666.6667
11833.3333
13166.6667
11233.3333

BFILENAME

BFILENAME
Syntax
BFILENAME

(

’

directory

’

,

’

filename

’

)

Purpose
BFILENAME returns a BFILE locator that is associated with a physical LOB binary file on
the server file system.
■

■

'directory' is a database object that serves as an alias for a full path name on the
server file system where the files are actually located.
'filename' is the name of the file in the server file system.

You must create the directory object and associate a BFILE value with a physical file
before you can use them as arguments to BFILENAME in a SQL or PL/SQL statement,
DBMS_LOB package, or OCI operation.
You can use this function in two ways:
■
■

In a DML statement to initialize a BFILE column
In a programmatic interface to access BFILE data by assigning a value to the BFILE
locator

The directory argument is case sensitive. You must ensure that you specify the
directory object name exactly as it exists in the data dictionary. For example, if an
"Admin" directory object was created using mixed case and a quoted identifier in the
CREATE DIRECTORY statement, then when using the BFILENAME function you must refer
to the directory object as 'Admin'. You must specify the filename argument according
to the case and punctuation conventions for your operating system.
See Also:
■

■

Oracle Database SecureFiles and Large Objects Developer's Guide and
Oracle Call Interface Programmer's Guide for more information on
LOBs and for examples of retrieving BFILE data
CREATE DIRECTORY on page 14-41

Examples
The following example inserts a row into the sample table pm.print_media. The
example uses the BFILENAME function to identify a binary file on the server file system
in the directory /demo/schema/product_media. The example shows how the directory
database object media_dir was created in the pm schema.
CREATE DIRECTORY media_dir AS '/demo/schema/product_media';
INSERT INTO print_media (product_id, ad_id, ad_graphic)
VALUES (3000, 31001, BFILENAME('MEDIA_DIR', 'modem_comp_ad.gif'));

Functions 5-29

BIN_TO_NUM

BIN_TO_NUM
Syntax
,
BIN_TO_NUM

(

expr

)

Purpose
BIN_TO_NUM converts a bit vector to its equivalent number. Each argument to this
function represents a bit in the bit vector. This function takes as arguments any
numeric data type, or any nonnumeric data type that can be implicitly converted to
NUMBER. Each expr must evaluate to 0 or 1. This function returns Oracle NUMBER.
BIN_TO_NUM is useful in data warehousing applications for selecting groups of interest
from a materialized view using grouping sets.
See Also:
■

■

■

group_by_clause on page 19-27 for information on GROUPING SETS
syntax
Table 3–10, " Implicit Type Conversion Matrix" on page 3-40 for
more information on implicit conversion
Oracle Database Data Warehousing Guide for information on data
aggregation in general

Examples
The following example converts a binary value to a number:
SELECT BIN_TO_NUM(1,0,1,0)
FROM DUAL;
BIN_TO_NUM(1,0,1,0)
------------------10

The next example converts three values into a single binary value and uses BIN_TO_NUM
to convert that binary into a number. The example uses a PL/SQL declaration to
specify the original values. These would normally be derived from actual data sources.
SELECT order_status
FROM orders
WHERE order_id = 2441;
ORDER_STATUS
-----------5
DECLARE
warehouse NUMBER := 1;
ground
NUMBER := 1;
insured
NUMBER := 1;
result
NUMBER;
BEGIN
SELECT BIN_TO_NUM(warehouse, ground, insured) INTO result FROM DUAL;
UPDATE orders SET order_status = result WHERE order_id = 2441;
END;

5-30 Oracle Database SQL Language Reference

BIN_TO_NUM

/
PL/SQL procedure successfully completed.
SELECT order_status
FROM orders
WHERE order_id = 2441;
ORDER_STATUS
-----------7

Refer to the examples for BITAND on page 5-32 for information on reversing this
process by extracting multiple values from a single column value.

Functions 5-31

BITAND

BITAND
Syntax
BITAND

(

expr1

,

expr2

)

Purpose
The BITAND function treats its inputs and its output as vectors of bits; the output is the
bitwise AND of the inputs.
The types of expr1 and expr2 are NUMBER, and the result is of type NUMBER. If either
argument to BITAND is NULL, the result is NULL.
The arguments must be in the range -(2(n-1)) .. ((2(n-1))-1). If an argument is out of this
range, the result is undefined.
The result is computed in several steps. First, each argument A is replaced with the
value SIGN(A)*FLOOR(ABS(A)). This conversion has the effect of truncating each
argument towards zero. Next, each argument A (which must now be an integer value)
is converted to an n-bit two's complement binary integer value. The two bit values are
combined using a bitwise AND operation. Finally, the resulting n-bit two's complement
value is converted back to NUMBER.
Notes on the BITAND Function
■
The current implementation of BITAND defines n = 128.
■

PL/SQL supports an overload of BITAND for which the types of the inputs and of
the result are all BINARY_INTEGER and for which n = 32.

Examples
The following example performs an AND operation on the numbers 6 (binary 1,1,0) and
3 (binary 0,1,1):
SELECT BITAND(6,3)
FROM DUAL;
BITAND(6,3)
----------2

This is the same as the following example, which shows the binary values of 6 and 3.
The BITAND function operates only on the significant digits of the binary values:
SELECT BITAND(
BIN_TO_NUM(1,1,0),
BIN_TO_NUM(0,1,1)) "Binary"
FROM DUAL;
Binary
---------2

Refer to the example for BIN_TO_NUM on page 5-30 for information on encoding
multiple values in a single column value.
The following example supposes that the order_status column of the sample table
oe.orders encodes several choices as individual bits within a single numeric value.

5-32 Oracle Database SQL Language Reference

BITAND

For example, an order still in the warehouse is represented by a binary value 001
(decimal 1). An order being sent by ground transportation is represented by a binary
value 010 (decimal 2). An insured package is represented by a binary value 100
(decimal 4). The example uses the DECODE function to provide two values for each of
the three bits in the order_status value, one value if the bit is turned on and one if it
is turned off.
SELECT order_id, customer_id, order_status,
DECODE(BITAND(order_status, 1), 1, 'Warehouse', 'PostOffice') "Location",
DECODE(BITAND(order_status, 2), 2, 'Ground', 'Air') "Method",
DECODE(BITAND(order_status, 4), 4, 'Insured', 'Certified') "Receipt"
FROM orders
WHERE sales_rep_id = 160
ORDER BY order_id;
ORDER_ID CUSTOMER_ID ORDER_STATUS Location
Method Receipt
---------- ----------- ------------ ---------- ------ --------2416
104
6 PostOffice Ground Insured
2419
107
3 Warehouse Ground Certified
2420
108
2 PostOffice Ground Certified
2423
145
3 Warehouse Ground Certified
2441
106
5 Warehouse Air
Insured
2455
145
7 Warehouse Ground Insured

For the Location column, BITAND first compares order_status with 1 (binary 001).
Only significant bit values are compared, so any binary value with a 1 in its rightmost
bit (any odd number) will evaluate positively and return 1. Even numbers will return
0. The DECODE function compares the value returned by BITAND with 1. If they are both
1, then the location is "Warehouse". If they are different, then the location is
"PostOffice".
The Method and Receipt columns are calculated similarly. For Method, BITAND
performs the AND operation on order_status and 2 (binary 010). For Receipt, BITAND
performs the AND operation on order_status and 4 (binary 100).

Functions 5-33

CARDINALITY

CARDINALITY
Syntax
CARDINALITY

(

nested_table

)

Purpose
CARDINALITY returns the number of elements in a nested table. The return type is
NUMBER. If the nested table is empty, or is a null collection, then CARDINALITY returns
NULL.

Examples
The following example shows the number of elements in the nested table column ad_
textdocs_ntab of the sample table pm.print_media:
SELECT product_id, CARDINALITY(ad_textdocs_ntab) cardinality
FROM print_media
ORDER BY product_id;
PRODUCT_ID CARDINALITY
---------- ----------2056
3
2268
3
3060
3
3106
3

5-34 Oracle Database SQL Language Reference

CAST

CAST
Syntax
expr
CAST

(

AS
MULTISET

(

subquery

type_name

)

)

Purpose
CAST converts one built-in data type or collection-typed value into another built-in
data type or collection-typed value.
CAST lets you convert built-in data types or collection-typed values of one type into
another built-in data type or collection type. You can cast an unnamed operand (such
as a date or the result set of a subquery) or a named collection (such as a varray or a
nested table) into a type-compatible data type or named collection. The type_name
must be the name of a built-in data type or collection type and the operand must be a
built-in data type or must evaluate to a collection value.
For the operand, expr can be either a built-in data type, a collection type, or an
instance of an ANYDATA type. If expr is an instance of an ANYDATA type, then CAST tries to
extract the value of the ANYDATA instance and return it if it matches the cast target type,
otherwise, null will be returned. MULTISET informs Oracle Database to take the result
set of the subquery and return a collection value. Table 5–1 shows which built-in data
types can be cast into which other built-in data types. (CAST does not support LONG,
LONG RAW, or the Oracle-supplied types.)
CAST does not directly support any of the LOB data types. When you use CAST to
convert a CLOB value into a character data type or a BLOB value into the RAW data type,
the database implicitly converts the LOB value to character or raw data and then
explicitly casts the resulting value into the target data type. If the resulting value is
larger than the target type, then the database returns an error.
When you use CAST ... MULTISET to get a collection value, each select list item in the
query passed to the CAST function is converted to the corresponding attribute type of
the target collection element type.
Table 5–1

Casting Built-In Data Types
from
BINARY_
FLOAT,
BINARY_
DOUBLE

from
CHAR,
VARCHAR2

to BINARY_FLOAT,
BINARY_DOUBLE

X

to CHAR,
VARCHAR2

from

NUMBER

from
DATETIME /
INTERVAL
(Note 1)

RAW

from ROWID, from
UROWID
NCHAR,
(Note 2)
NVARCHAR2

X

X

--

--

--

X

X

X

X

X

X

X

--

to NUMBER

X

X

X

--

--

--

X

to DATE,
TIMESTAMP,
INTERVAL

--

X

--

X

--

--

--

to RAW

from

--

X

--

--

X

--

--

to ROWID, UROWID --

X

--

--

--

X

--

to NCHAR,
NVARCHAR2

--

X

X

X

X

X

X

Functions 5-35

CAST

Note 1: Datetime/interval includes DATE, TIMESTAMP, TIMESTAMP WITH TIMEZONE,
INTERVAL DAY TO SECOND, and INTERVAL YEAR TO MONTH.
Note 2: You cannot cast a UROWID to a ROWID if the UROWID contains the value of a ROWID
of an index-organized table.
If you want to cast a named collection type into another named collection type, then
the elements of both collections must be of the same type.
"Implicit Data Conversion" on page 3-40 for information
on how Oracle Database implicitly converts collection type data into
character data and "Security Considerations for Data Conversion" on
page 3-44

See Also:

If the result set of subquery can evaluate to multiple rows, then you must specify the
MULTISET keyword. The rows resulting from the subquery form the elements of the
collection value into which they are cast. Without the MULTISET keyword, the subquery
is treated as a scalar subquery.

Built-In Data Type Examples
The following examples use the CAST function with scalar data types. The first example
converts text to a timestamp value by applying the format model provided in the
session parameter NLS_TIMESTAMP_FORMAT. If you want to avoid dependency on this
NLS parameter, then you can use the TO_DATE as shown in the second example.
SELECT CAST('22-OCT-1997'
AS TIMESTAMP WITH LOCAL TIME ZONE)
FROM DUAL;
SELECT CAST(TO_DATE('22-Oct-1997', 'DD-Mon-YYYY')
AS TIMESTAMP WITH LOCAL TIME ZONE)
FROM DUAL;

In the preceding example, TO_DATE converts from text to DATE, and CAST converts from
DATE to TIMESTAMP WITH LOCAL TIME ZONE, interpreting the date in the session time zone
(SESSIONTIMEZONE).
SELECT product_id, CAST(ad_sourcetext AS VARCHAR2(30)) text
FROM print_media
ORDER BY product_id;

Collection Examples
The CAST examples that follow build on the cust_address_typ found in the sample
order entry schema, oe.
CREATE TYPE address_book_t AS TABLE OF cust_address_typ;
/
CREATE TYPE address_array_t AS VARRAY(3) OF cust_address_typ;
/
CREATE TABLE cust_address (
custno
NUMBER,
street_address
VARCHAR2(40),
postal_code
VARCHAR2(10),
city
VARCHAR2(30),
state_province
VARCHAR2(10),
country_id
CHAR(2));
CREATE TABLE cust_short (custno NUMBER, name VARCHAR2(31));

5-36 Oracle Database SQL Language Reference

CAST

CREATE TABLE states (state_id NUMBER, addresses address_array_t);

This example casts a subquery:
SELECT s.custno, s.name,
CAST(MULTISET(SELECT ca.street_address,
ca.postal_code,
ca.city,
ca.state_province,
ca.country_id
FROM cust_address ca
WHERE s.custno = ca.custno)
AS address_book_t)
FROM cust_short s
ORDER BY s.custno;

CAST converts a varray type column into a nested table:
SELECT CAST(s.addresses AS address_book_t)
FROM states s
WHERE s.state_id = 111;

The following objects create the basis of the example that follows:
CREATE TABLE projects
(employee_id NUMBER, project_name VARCHAR2(10));
CREATE TABLE emps_short
(employee_id NUMBER, last_name VARCHAR2(10));
CREATE TYPE project_table_typ AS TABLE OF VARCHAR2(10);
/

The following example of a MULTISET expression uses these objects:
SELECT e.last_name,
CAST(MULTISET(SELECT p.project_name
FROM projects p
WHERE p.employee_id = e.employee_id
ORDER BY p.project_name)
AS project_table_typ)
FROM emps_short e
ORDER BY e.last_name;

Functions 5-37

CEIL

CEIL
Syntax
CEIL

(

n

)

Purpose
CEIL returns the smallest integer that is greater than or equal to n. The number n can
always be written as the difference of an integer k and a positive fraction f such that 0
<= f < 1 and n = k - f. The value of CEIL is the integer k. Thus, the value of CEIL is n
itself if and only if n is precisely an integer.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. The function returns the
same data type as the numeric data type of the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion and FLOOR on
page 5-105

Examples
The following example returns the smallest integer greater than or equal to the order
total of a specified order:
SELECT order_total, CEIL(order_total)
FROM orders
WHERE order_id = 2434;
ORDER_TOTAL CEIL(ORDER_TOTAL)
----------- ----------------268651.8
268652

5-38 Oracle Database SQL Language Reference

CHARTOROWID

CHARTOROWID
Syntax
CHARTOROWID

(

char

)

Purpose
CHARTOROWID converts a value from CHAR, VARCHAR2, NCHAR, or NVARCHAR2 data type to
ROWID data type.
This function does not support CLOB data directly. However, CLOBs can be passed in as
arguments through implicit data conversion.
See Also: "Data Type Comparison Rules" on page 3-36 for more
information.

Examples
The following example converts a character rowid representation to a rowid. (The
actual rowid is different for each database instance.)
SELECT last_name
FROM employees
WHERE ROWID = CHARTOROWID('AAAFd1AAFAAAABSAA/');
LAST_NAME
------------------------Greene

Functions 5-39

CHR

CHR
Syntax
USING
CHR

(

NCHAR_CS

n

)

Purpose
CHR returns the character having the binary equivalent to n as a VARCHAR2 value in
either the database character set or, if you specify USING NCHAR_CS, the national
character set.
For single-byte character sets, if n > 256, then Oracle Database returns the binary
equivalent of n mod 256. For multibyte character sets, n must resolve to one entire code
point. Invalid code points are not validated, and the result of specifying invalid code
points is indeterminate.
This function takes as an argument a NUMBER value, or any value that can be implicitly
converted to NUMBER, and returns a character.
Use of the CHR function (either with or without the optional
USING NCHAR_CS clause) results in code that is not portable between
ASCII- and EBCDIC-based machine architectures.
Note:

See Also: NCHR on page 5-155 and Table 3–10, " Implicit Type
Conversion Matrix" on page 3-40 for more information on implicit
conversion

Examples
The following example is run on an ASCII-based machine with the database character
set defined as WE8ISO8859P1:
SELECT CHR(67)||CHR(65)||CHR(84) "Dog"
FROM DUAL;
Dog
--CAT

To produce the same results on an EBCDIC-based machine with the WE8EBCDIC1047
character set, the preceding example would have to be modified as follows:
SELECT CHR(195)||CHR(193)||CHR(227) "Dog"
FROM DUAL;
Dog
--CAT

For multibyte character sets, this sort of concatenation gives different results. For
example, given a multibyte character whose hexadecimal value is a1a2 (a1
representing the first byte and a2 the second byte), you must specify for n the decimal
equivalent of 'a1a2', or 41378:
SELECT CHR(41378)

5-40 Oracle Database SQL Language Reference

CHR

FROM DUAL;

You cannot specify the decimal equivalent of a1 concatenated with the decimal
equivalent of a2, as in the following example:
SELECT CHR(161)||CHR(162)
FROM DUAL;

However, you can concatenate whole multibyte code points, as in the following
example, which concatenates the multibyte characters whose hexadecimal values are
a1a2 and a1a3:
SELECT CHR(41378)||CHR(41379)
FROM DUAL;

The following example assumes that the national character set is UTF16:
SELECT CHR (196 USING NCHAR_CS)
FROM DUAL;
CH
-Ä

Functions 5-41

CLUSTER_ID

CLUSTER_ID
Syntax
schema
CLUSTER_ID

.

(

model

mining_attribute_clause

)

mining_attribute_clause::=
*
,
USING

schema

.
table
AS

.

*

alias

expr

Purpose
This function is for use with clustering models created by the DBMS_DATA_MINING
package or with Oracle Data Miner. It returns the cluster identifier of the predicted
cluster with the highest probability for the set of predictors specified in the mining_
attribute_clause. The value returned is an Oracle NUMBER.
The mining_attribute_clause behaves as described for the PREDICTION function.
Refer to mining_attribute_clause on page 5-192.
See Also:
■

■

Oracle Data Mining Concepts for detailed information about Oracle
Data Mining
Oracle Data Mining Application Developer's Guide for detailed
information about scoring with the Data Mining SQL functions

Examples
The following example lists the clusters into which customers of a given dataset have
been grouped.
This example, and the prerequisite data mining operations, including the creation of
the km_sh_clus_sample model and the mining_data_apply_v view, can be found in the
demo file $ORACLE_HOME/rdbms/demo/dmkmdemo.sql. General information on data
mining demo files is available in Oracle Data Mining Administrator's Guide. The
example is presented here to illustrate the syntactic use of the function.
SELECT CLUSTER_ID(km_sh_clus_sample USING *) AS clus, COUNT(*) AS cnt
FROM mining_data_apply_v
GROUP BY CLUSTER_ID(km_sh_clus_sample USING *)
ORDER BY cnt DESC;
CLUS
CNT
---------- ---------2
580
10
216
6
186
8
115

5-42 Oracle Database SQL Language Reference

CLUSTER_ID

19
12
18
16
17
14

110
101
81
39
38
34

10 rows selected.

Functions 5-43

CLUSTER_PROBABILITY

CLUSTER_PROBABILITY
Syntax
schema
CLUSTER_PROBABILITY

.

,

(

cluster_id

model

mining_attribute_clause

)

mining_attribute_clause::=
*
,
USING

schema

.
table
AS

.

*

alias

expr

Purpose
This function is for use with clustering models created by the DBMS_DATA_MINING
package or with Oracle Data Miner. It returns a measure of the degree of confidence of
membership of an input row in a cluster associated with the specified model.
■

■

For cluster_id, specify the identifier of the cluster in the model. The function
returns the probability for the specified cluster. If you omit this clause, then the
function returns the probability associated with the best predicted cluster. You can
use the form without cluster_id in conjunction with the CLUSTER_ID function to
obtain the best predicted pair of cluster ID and probability.
The mining_attribute_clause behaves as described for the PREDICTION function.
Refer to mining_attribute_clause on page 5-192
See Also:
■

■

Oracle Data Mining Concepts for detailed information about Oracle
Data Mining
Oracle Data Mining Application Developer's Guide for detailed
information about scoring with the Data Mining SQL functions

Examples
The following example determines the ten most representative customers, based on
likelihood, in cluster 2.
This example, and the prerequisite data mining operations, including the creation of
the km_sh_clus_sample model and the mining_data_apply_v view, can be found in the
demo file $ORACLE_HOME/rdbms/demo/dmkmdemo.sql. General information on data
mining demo files is available in Oracle Data Mining Administrator's Guide. The
example is presented here to illustrate the syntactic use of the function.
SELECT *
FROM (SELECT cust_id, CLUSTER_PROBABILITY(km_sh_clus_sample, 2 USING *) prob
FROM mining_data_apply_v
ORDER BY prob DESC)
WHERE ROWNUM < 11;

5-44 Oracle Database SQL Language Reference

CLUSTER_PROBABILITY

CUST_ID
---------100256
100988
100889
101086
101215
100390
100985
101026
100601
100672

PROB
---------.999387471
.99936194
.999335107
.99928882
.999266521
.999264718
.999251722
.999247906
.999242089
.999235711

10 rows selected.

Functions 5-45

CLUSTER_SET

CLUSTER_SET
Syntax
,
schema
CLUSTER_SET

.

,

(

cutoff

topN

model

mining_attribute_clause

)

mining_attribute_clause::=
*
,
USING

schema

.
table
AS

.

*

alias

expr

Purpose
This function is for use with clustering models created by the DBMS_DATA_MINING
package or with Oracle Data Miner. It returns a varray of objects containing all
possible clusters that a given row belongs to. Each object in the varray is a pair of
scalar values containing the cluster ID and the cluster probability. The object fields are
named CLUSTER_ID and PROBABILITY, and both are Oracle NUMBER.
■

■

For the optional topN argument, specify a positive integer. Doing so restricts the
set of predicted clusters to those that have one of the top N probability values. If
you omit topN or set it to NULL, then all clusters are returned in the collection. If
multiple clusters are tied for the Nth value, the database still returns only N values.
For the optional cutoff argument, specify a positive integer to restrict the
returned clusters to those with a probability greater than or equal to the specified
cutoff. You can filter only by cutoff by specifying NULL for topN and the desired
cutoff value for cutoff.

You can specify topN and cutoff together to restrict the returned clusters to those that
are in the top N and have a probability that passes the threshold.
The mining_attribute_clause behaves as described for the PREDICTION function.
Refer to mining_attribute_clause on page 5-192.
See Also:
■

■

Oracle Data Mining Concepts for detailed information about Oracle
Data Mining
Oracle Data Mining Application Developer's Guide for detailed
information about scoring with the Data Mining SQL functions

Examples
The following example lists the most relevant attributes (with confidence > 55%) of
each cluster to which customer 101362 belongs with > 20% likelihood.
This example, and the prerequisite data mining operations, including the creation of
the km_sh_clus_sample model and the views and type, can be found in the demo file

5-46 Oracle Database SQL Language Reference

CLUSTER_SET

$ORACLE_HOME/rdbms/demo/dmkmdemo.sql. General information on data mining demo
files is available in Oracle Data Mining Administrator's Guide. The example is presented
here to illustrate the syntactic use of the function.
WITH
clus_tab AS (
SELECT id,
A.attribute_name aname,
A.conditional_operator op,
NVL(A.attribute_str_value,
ROUND(A.attribute_num_value),4)) val,
A.attribute_support support,
A.attribute_confidence confidence
FROM TABLE(DBMS_DATA_MINING.GET_MODEL_DETAILS_KM('km_sh_clus_sample')) T,
TABLE(T.rule.antecedent) A
WHERE A.attribute_confidence > 0.55
),
clust AS (
SELECT id,
CAST(COLLECT(Cattr(aname, op, TO_CHAR(val), support, confidence))
AS Cattrs) cl_attrs
FROM clus_tab
GROUP BY id
),
custclus AS (
SELECT T.cust_id, S.cluster_id, S.probability
FROM (SELECT cust_id, CLUSTER_SET(km_sh_clus_sample, NULL, 0.2 USING *) pset
FROM mining_data_apply_v
WHERE cust_id = 101362) T,
TABLE(T.pset) S
)
SELECT A.probability prob, A.cluster_id cl_id,
B.attr, B.op, B.val, B.supp, B.conf
FROM custclus A,
(SELECT T.id, C.*
FROM clust T,
TABLE(T.cl_attrs) C) B
WHERE A.cluster_id = B.id
ORDER BY prob DESC, cl_id ASC, conf DESC, attr ASC, val ASC;
PROB
CL_ID
------- ---------.7745
8
.7745
8
.7745
8
.7745
8
.7745
8
.2028
6
.2028
6
.2028
6
8 rows selected.

ATTR
-------------------------HOUSEHOLD_SIZE
CUST_MARITAL_STATUS
CUST_MARITAL_STATUS
CUST_MARITAL_STATUS
CUST_MARITAL_STATUS
AGE
AGE
CUST_MARITAL_STATUS

OP
-IN
IN
IN
IN
IN
>=
<=
IN

VAL
SUPP
CONF
----------- -------- ------9+
124
.7500
Divorc.
116
.6000
NeverM
116
.6000
Separ.
116
.6000
Widowed
116
.6000
17
154
.6667
31.6
154
.6667
NeverM
172
.6667

Functions 5-47

COALESCE

COALESCE
Syntax
,
COALESCE

(

expr

)

Purpose
COALESCE returns the first non-null expr in the expression list. You must specify at least
two expressions. If all occurrences of expr evaluate to null, then the function returns
null.
Oracle Database uses short-circuit evaluation. The database evaluates each expr value
and determines whether it is NULL, rather than evaluating all of the expr values before
determining whether any of them is NULL.
If all occurrences of expr are numeric data type or any nonnumeric data type that can
be implicitly converted to a numeric data type, then Oracle Database determines the
argument with the highest numeric precedence, implicitly converts the remaining
arguments to that data type, and returns that data type.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion and "Numeric
Precedence" on page 3-14 for information on numeric precedence

This function is a generalization of the NVL function.
You can also use COALESCE as a variety of the CASE expression. For example,
COALESCE(expr1, expr2)

is equivalent to:
CASE WHEN expr1 IS NOT NULL THEN expr1 ELSE expr2 END

Similarly,
COALESCE(expr1, expr2, ..., exprn)

where n >= 3, is equivalent to:
CASE WHEN expr1 IS NOT NULL THEN expr1
ELSE COALESCE (expr2, ..., exprn) END

See Also:

NVL on page 5-173 and "CASE Expressions" on page 6-5

Examples
The following example uses the sample oe.product_information table to organize a
clearance sale of products. It gives a 10% discount to all products with a list price. If
there is no list price, then the sale price is the minimum price. If there is no minimum
price, then the sale price is "5":
SELECT product_id, list_price, min_price,
COALESCE(0.9*list_price, min_price, 5) "Sale"
FROM product_information
WHERE supplier_id = 102050
ORDER BY product_id;

5-48 Oracle Database SQL Language Reference

COALESCE

PRODUCT_ID LIST_PRICE MIN_PRICE
Sale
---------- ---------- ---------- ---------1769
48
43.2
1770
73
73
2378
305
247
274.5
2382
850
731
765
3355
5

Functions 5-49

COLLECT

COLLECT
Syntax
DISTINCT
UNIQUE
COLLECT

(

ORDER

BY

expr

column

)

Purpose
COLLECT is an aggregate function that takes as its argument a column of any type and
creates a nested table of the input type out of the rows selected. To get accurate results
from this function you must use it within a CAST function.
If column is itself a collection, then the output of COLLECT is a nested table of
collections. If column is of a user-defined type, then column must have a MAP or ORDER
method defined on it in order for you to use the optional DISTINCT, UNIQUE, and ORDER
BY clauses.
See Also: CAST on page 5-35 and "Aggregate Functions" on
page 5-10

Examples
The following example creates a nested table from the varray column of phone
numbers in the sample table oe.customers. The nested table includes only the phone
numbers of customers with an income level of L: 300,000 and above.
CREATE TYPE phone_book_t AS TABLE OF phone_list_typ;
/
SELECT CAST(COLLECT(phone_numbers) AS phone_book_t) "Income Level L Phone Book"
FROM customers
WHERE income_level = 'L: 300,000 and above';
Income Level L Phone Book
-------------------------------------------------------------------------------PHONE_BOOK_T(PHONE_LIST_TYP('+1 414 123 4307'), PHONE_LIST_TYP('+1 608 123 4344'
), PHONE_LIST_TYP('+1 814 123 4696'), PHONE_LIST_TYP('+1 215 123 4721'), PHONE_L
IST_TYP('+1 814 123 4755'), PHONE_LIST_TYP('+91 11 012 4817', '+91 11 083 4817')
, PHONE_LIST_TYP('+91 172 012 4837'), PHONE_LIST_TYP('+41 31 012 3569', '+41 31
083 3569'))

The following example creates a nested table from the column of warehouse names in
the sample table oe.warehouses. It uses ORDER BY to order the warehouse names.
CREATE TYPE warehouse_name_t AS TABLE OF VARCHAR2(35);
/
SELECT CAST(COLLECT(warehouse_name ORDER BY warehouse_name)
AS warehouse_name_t) "Warehouses"
FROM warehouses;
Warehouses
-------------------------------------------------------------------------------WAREHOUSE_NAME_TYP('Beijing', 'Bombay', 'Mexico City', 'New Jersey', 'San Franci
sco', 'Seattle, Washington', 'Southlake, Texas', 'Sydney', 'Toronto')

5-50 Oracle Database SQL Language Reference

COMPOSE

COMPOSE
Syntax
COMPOSE

(

char

)

Purpose
COMPOSE takes as its argument a string, or an expression that resolves to a string, in any
data type, and returns a Unicode string in the same character set as the input. char can
be any of the data types CHAR, VARCHAR2, NCHAR, NVARCHAR2, CLOB, or NCLOB. For
example, an o code point qualified by an umlaut code point will be returned as the
o-umlaut code point.
COMPOSE returns the string in NFC normal form. For a more exclusive setting, you can
first call DECOMPOSE with the CANONICAL setting and then COMPOSE. This combination
returns the string in NFKC normal form.
CLOB and NCLOB values are supported through implicit conversion. If char is a
character LOB value, then it is converted to a VARCHAR value before the COMPOSE
operation. The operation will fail if the size of the LOB value exceeds the supported
length of the VARCHAR in the particular development environment.
See Also:
■

■

Oracle Database Globalization Support Guide for information on
Unicode character sets and character semantics
DECOMPOSE on page 5-79

Examples
The following example returns the o-umlaut code point:
SELECT COMPOSE( 'o' || UNISTR('\0308') )
FROM DUAL;
CO
-ö

See Also:

UNISTR on page 5-333

Functions 5-51

CONCAT

CONCAT
Syntax
CONCAT

(

char1

,

char2

)

Purpose
CONCAT returns char1 concatenated with char2. Both char1 and char2 can be any of the
data types CHAR, VARCHAR2, NCHAR, NVARCHAR2, CLOB, or NCLOB. The string returned is in
the same character set as char1. Its data type depends on the data types of the
arguments.
In concatenations of two different data types, Oracle Database returns the data type
that results in a lossless conversion. Therefore, if one of the arguments is a LOB, then
the returned value is a LOB. If one of the arguments is a national data type, then the
returned value is a national data type. For example:
■

CONCAT(CLOB, NCLOB) returns NCLOB

■

CONCAT(NCLOB, NCHAR) returns NCLOB

■

CONCAT(NCLOB, CHAR) returns NCLOB

■

CONCAT(NCHAR, CLOB) returns NCLOB

This function is equivalent to the concatenation operator (||).
See Also: "Concatenation Operator" on page 4-4 for information on
the CONCAT operator

Examples
This example uses nesting to concatenate three character strings:
SELECT CONCAT(CONCAT(last_name, '''s job category is '), job_id) "Job"
FROM employees
WHERE employee_id = 152;
Job
-----------------------------------------------------Hall's job category is SA_REP

5-52 Oracle Database SQL Language Reference

CONVERT

CONVERT
Syntax
,
CONVERT

(

char

,

dest_char_set

source_char_set
)

Purpose
CONVERT converts a character string from one character set to another.
■

■

■

The char argument is the value to be converted. It can be any of the data types
CHAR, VARCHAR2, NCHAR, NVARCHAR2, CLOB, or NCLOB.
The dest_char_set argument is the name of the character set to which char is
converted.
The source_char_set argument is the name of the character set in which char is
stored in the database. The default value is the database character set.

The return value for CHAR and VARCHAR2 is VARCHAR2. For NCHAR and NVARCHAR2, it is
NVARCHAR2. For CLOB, it is CLOB, and for NCLOB, it is NCLOB.
Both the destination and source character set arguments can be either literals or
columns containing the name of the character set.
For complete correspondence in character conversion, it is essential that the
destination character set contains a representation of all the characters defined in the
source character set. Where a character does not exist in the destination character set, a
replacement character appears. Replacement characters can be defined as part of a
character set definition.
Note: Oracle discourages the use of the CONVERT function in the
current Oracle Database release. The return value of CONVERT has a
character data type, so it should be either in the database character set
or in the national character set, depending on the data type. Any
dest_char_set that is not one of these two character sets is
unsupported. The char argument and the source_char_set have the
same requirements. Therefore, the only practical use of the function is
to correct data that has been stored in a wrong character set.

Values that are in neither the database nor the national character set
should be processed and stored as RAW or BLOB. Procedures in the
PL/SQL packages UTL_RAW and UTL_I18N—for example, UTL_
RAW.CONVERT—allow limited processing of such values. Procedures
accepting a RAW argument in the packages UTL_FILE, UTL_TCP, UTL_
HTTP, and UTL_SMTP can be used to output the processed data.

Examples
The following example illustrates character set conversion by converting a Latin-1
string to ASCII. The result is the same as importing the same string from a
WE8ISO8859P1 database to a US7ASCII database.
SELECT CONVERT('Ä Ê Í Õ Ø A B C D E ', 'US7ASCII', 'WE8ISO8859P1')
FROM DUAL;
CONVERT('ÄÊÍÕØABCDE'

Functions 5-53

CONVERT

--------------------A E I ? ? A B C D E ?

Common character sets include:
■

US7ASCII: US 7-bit ASCII character set

■

WE8ISO8859P1: ISO 8859-1 West European 8-bit character set

■

EE8MSWIN1250: Microsoft Windows East European Code Page 1250

■

WE8MSWIN1252: Microsoft Windows West European Code Page 1252

■

WE8EBCDIC1047: IBM West European EBCDIC Code Page 1047

■

JA16SJISTILDE: Japanese Shift-JIS Character Set, compatible with MS Code Page
932

■

ZHT16MSWIN950: Microsoft Windows Traditional Chinese Code Page 950

■

UTF8: Unicode 3.0 Universal character set CESU-8 encoding form

■

AL32UTF8: Unicode 5.0 Universal character set UTF-8 encoding form

You can query the V$NLS_VALID_VALUES view to get a listing of valid character sets, as
follows:
SELECT * FROM V$NLS_VALID_VALUES WHERE parameter = 'CHARACTERSET';

See Also: Oracle Database Globalization Support Guide for information
on supported character sets and Oracle Database Reference for
information on the V$NLS_VALID_VALUES view

5-54 Oracle Database SQL Language Reference

CORR

CORR
Syntax
OVER
CORR

(

expr1

,

expr2

(

analytic_clause

)

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions

Purpose
CORR returns the coefficient of correlation of a set of number pairs. You can use it as an
aggregate or analytic function.
This function takes as arguments any numeric data type or any nonnumeric data type
that can be implicitly converted to a numeric data type. Oracle determines the
argument with the highest numeric precedence, implicitly converts the remaining
arguments to that data type, and returns that data type.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion and "Numeric
Precedence" on page 3-14 for information on numeric precedence

Oracle Database applies the function to the set of (expr1, expr2) after eliminating the
pairs for which either expr1 or expr2 is null. Then Oracle makes the following
computation:
COVAR_POP(expr1, expr2) / (STDDEV_POP(expr1) * STDDEV_POP(expr2))

The function returns a value of type NUMBER. If the function is applied to an empty set,
then it returns null.
The CORR function calculates the Pearson's correlation
coefficient, which requires numeric expressions as arguments. Oracle
also provides the CORR_S (Spearman's rho coefficient) and CORR_K
(Kendall's tau-b coefficient) functions to support nonparametric or
rank correlation.
Note:

"Aggregate Functions" on page 5-10, "About SQL
Expressions" on page 6-1 for information on valid forms of expr, and
CORR_* on page 5-57 for information on the CORR_S and CORR_K
functions

See Also:

Aggregate Example
The following example calculates the coefficient of correlation between the list prices
and minimum prices of products by weight class in the sample table oe.product_
information:
SELECT weight_class, CORR(list_price, min_price) "Correlation"
FROM product_information
GROUP BY weight_class
ORDER BY weight_class, "Correlation";

Functions 5-55

CORR

WEIGHT_CLASS Correlation
------------ ----------1 .999149795
2 .999022941
3 .998484472
4 .999359909
5 .999536087

Analytic Example
The following example shows the correlation between duration at the company and
salary by the employee's position. The result set shows the same correlation for each
employee in a given job:
SELECT employee_id, job_id,
TO_CHAR((SYSDATE - hire_date) YEAR TO MONTH ) "Yrs-Mns",
CORR(SYSDATE-hire_date, salary)
OVER(PARTITION BY job_id) AS "Correlation"
FROM employees
WHERE department_id in (50, 80)
ORDER BY job_id, employee_id;
EMPLOYEE_ID
----------145
146
147
148
149
150
151
152
153
154
155
. . .

JOB_ID
---------SA_MAN
SA_MAN
SA_MAN
SA_MAN
SA_MAN
SA_REP
SA_REP
SA_REP
SA_REP
SA_REP
SA_REP

Yrs-Mns
SALARY Correlation
------- ---------- ----------+04-09
14000 .912385598
+04-06
13500 .912385598
+04-04
12000 .912385598
+01-08
11000 .912385598
+01-05
10500 .912385598
+04-05
10000
.80436755
+04-03
9500
.80436755
+03-10
9000
.80436755
+03-03
8000
.80436755
+02-07
7500
.80436755
+01-07
7000
.80436755

5-56 Oracle Database SQL Language Reference

salary,

CORR_*

CORR_*
The CORR_* functions are:
■

CORR_S

■

CORR_K

Syntax
correlation::=
COEFFICIENT
ONE_SIDED_SIG
,

ONE_SIDED_SIG_POS
ONE_SIDED_SIG_NEG
TWO_SIDED_SIG

CORR_K
(

expr1

,

expr2

)

CORR_S

Purpose
The CORR function (see CORR on page 5-55) calculates the Pearson's correlation
coefficient and requires numeric expressions as input. The CORR_* functions support
nonparametric or rank correlation. They let you find correlations between expressions
that are ordinal scaled (where ranking of the values is possible). Correlation
coefficients take on a value ranging from -1 to 1, where 1 indicates a perfect
relationship, -1 a perfect inverse relationship (when one variable increases as the other
decreases), and a value close to 0 means no relationship.
These functions takes as arguments any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. Oracle Database
determines the argument with the highest numeric precedence, implicitly converts the
remaining arguments to that data type, makes the calculation, and returns NUMBER.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion and "Numeric
Precedence" on page 3-14 for information on numeric precedence

expr1 and expr2 are the two variables being analyzed. The third argument is a return
value of type VARCHAR2. If you omit the third argument, then the default is
COEFFICIENT. The meaning of the return values is shown in the table that follows:
Table 5–2

CORR_* Return Values

Return Value

Meaning

COEFFICIENT

Coefficient of correlation

ONE_SIDED_SIG

Positive one-tailed significance of the correlation

ONE_SIDED_SIG_POS

Same as ONE_SIDED_SIG

ONE_SIDED_SIG_NEG

Negative one-tailed significance of the correlation

TWO_SIDED_SIG

Two-tailed significance of the correlation

Functions 5-57

CORR_S

CORR_S
CORR_S calculates the Spearman's rho correlation coefficient. The input expressions
should be a set of (xi, yi) pairs of observations. The function first replaces each value
with a rank. Each value of xi is replaced with its rank among all the other xis in the
sample, and each value of yi is replaced with its rank among all the other yis. Thus,
each xi and yi take on a value from 1 to n, where n is the total number of pairs of
values. Ties are assigned the average of the ranks they would have had if their values
had been slightly different. Then the function calculates the linear correlation
coefficient of the ranks.
Using Spearman's rho correlation coefficient, the following
example derives a coefficient of correlation for each of two different comparisons -salary and commission_pct, and salary and employee_id:
CORR_S Example

SELECT COUNT(*) count,
CORR_S(salary, commission_pct) commission,
CORR_S(salary, employee_id) empid
FROM employees;
COUNT COMMISSION
EMPID
---------- ---------- ---------107 .735837022 -.04473016

5-58 Oracle Database SQL Language Reference

CORR_*

CORR_K
CORR_K calculates the Kendall's tau-b correlation coefficient. As for CORR_S, the input
expressions are a set of (xi, yi) pairs of observations. To calculate the coefficient, the
function counts the number of concordant and discordant pairs. A pair of observations
is concordant if the observation with the larger x also has a larger value of y. A pair of
observations is discordant if the observation with the larger x has a smaller y.
The significance of tau-b is the probability that the correlation indicated by tau-b was
due to chance—a value of 0 to 1. A small value indicates a significant correlation for
positive values of tau-b (or anticorrelation for negative values of tau-b).
Using Kendall's tau-b correlation coefficient, the following
example determines whether a correlation exists between an employee's salary and
commission percent:

CORR_K Example

SELECT CORR_K(salary, commission_pct, 'COEFFICIENT') coefficient,
CORR_K(salary, commission_pct, 'TWO_SIDED_SIG') two_sided_p_value
FROM employees;
COEFFICIENT TWO_SIDED_P_VALUE
----------- ----------------.603079768
3.4702E-07

Functions 5-59

COS

COS
Syntax
COS

(

n

)

Purpose
COS returns the cosine of n (an angle expressed in radians).
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. If the argument is
BINARY_FLOAT, then the function returns BINARY_DOUBLE. Otherwise the function
returns the same numeric data type as the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

Examples
The following example returns the cosine of 180 degrees:
SELECT COS(180 * 3.14159265359/180) "Cosine of 180 degrees"
FROM DUAL;
Cosine of 180 degrees
---------------------1

5-60 Oracle Database SQL Language Reference

COSH

COSH
Syntax
COSH

(

n

)

Purpose
COSH returns the hyperbolic cosine of n.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. If the argument is
BINARY_FLOAT, then the function returns BINARY_DOUBLE. Otherwise the function
returns the same numeric data type as the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

Examples
The following example returns the hyperbolic cosine of zero:
SELECT COSH(0) "Hyperbolic cosine of 0"
FROM DUAL;
Hyperbolic cosine of 0
---------------------1

Functions 5-61

COUNT

COUNT
Syntax
*
COUNT

(

OVER

DISTINCT

(

analytic_clause

)

)

ALL
expr

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions

Purpose
COUNT returns the number of rows returned by the query. You can use it as an
aggregate or analytic function.
If you specify DISTINCT, then you can specify only the query_partition_clause of the
analytic_clause. The order_by_clause and windowing_clause are not allowed.
If you specify expr, then COUNT returns the number of rows where expr is not null. You
can count either all rows, or only distinct values of expr.
If you specify the asterisk (*), then this function returns all rows, including duplicates
and nulls. COUNT never returns null.
See Also: "About SQL Expressions" on page 6-1 for information on
valid forms of expr and "Aggregate Functions" on page 5-10

Aggregate Examples
The following examples use COUNT as an aggregate function:
SELECT COUNT(*) "Total"
FROM employees;
Total
---------107
SELECT COUNT(*) "Allstars"
FROM employees
WHERE commission_pct > 0;
Allstars
--------35
SELECT COUNT(commission_pct) "Count"
FROM employees;
Count
---------35
SELECT COUNT(DISTINCT manager_id) "Managers"
FROM employees;

5-62 Oracle Database SQL Language Reference

COUNT

Managers
---------18

Analytic Example
The following example calculates, for each employee in the employees table, the
moving count of employees earning salaries in the range 50 less than through 150
greater than the employee's salary.
SELECT last_name, salary,
COUNT(*) OVER (ORDER BY salary RANGE BETWEEN 50 PRECEDING AND
150 FOLLOWING) AS mov_count
FROM employees
ORDER BY salary, last_name;
LAST_NAME
SALARY MOV_COUNT
------------------------- ---------- ---------Olson
2100
3
Markle
2200
2
Philtanker
2200
2
Gee
2400
8
Landry
2400
8
Colmenares
2500
10
Marlow
2500
10
Patel
2500
10
. . .

Functions 5-63

COVAR_POP

COVAR_POP
Syntax
OVER
COVAR_POP

(

expr1

,

expr2

(

analytic_clause

)

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions

Purpose
COVAR_POP returns the population covariance of a set of number pairs. You can use it as
an aggregate or analytic function.
This function takes as arguments any numeric data type or any nonnumeric data type
that can be implicitly converted to a numeric data type. Oracle determines the
argument with the highest numeric precedence, implicitly converts the remaining
arguments to that data type, and returns that data type.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion and "Numeric
Precedence" on page 3-14 for information on numeric precedence

Oracle Database applies the function to the set of (expr1, expr2) pairs after eliminating
all pairs for which either expr1 or expr2 is null. Then Oracle makes the following
computation:
(SUM(expr1 * expr2) - SUM(expr2) * SUM(expr1) / n) / n

where n is the number of (expr1, expr2) pairs where neither expr1 nor expr2 is null.
The function returns a value of type NUMBER. If the function is applied to an empty set,
then it returns null.
See Also: "About SQL Expressions" on page 6-1 for information on
valid forms of expr and "Aggregate Functions" on page 5-10

Aggregate Example
The following example calculates the population covariance and sample covariance for
time employed (SYSDATE - hire_date) and salary using the sample table
hr.employees:
SELECT job_id,
COVAR_POP(SYSDATE-hire_date, salary) AS covar_pop,
COVAR_SAMP(SYSDATE-hire_date, salary) AS covar_samp
FROM employees
WHERE department_id in (50, 80)
GROUP BY job_id
ORDER BY job_id, covar_pop, covar_samp;
JOB_ID
COVAR_POP COVAR_SAMP
---------- ----------- ----------SA_MAN
660700
825875
SA_REP
579988.466
600702.34
SH_CLERK
212432.5 223613.158
ST_CLERK
176577.25 185870.789

5-64 Oracle Database SQL Language Reference

COVAR_POP

ST_MAN

436092

545115

Analytic Example
The following example calculates cumulative sample covariance of the list price and
minimum price of the products in the sample schema oe:
SELECT product_id, supplier_id,
COVAR_POP(list_price, min_price)
OVER (ORDER BY product_id, supplier_id)
AS CUM_COVP,
COVAR_SAMP(list_price, min_price)
OVER (ORDER BY product_id, supplier_id)
AS CUM_COVS
FROM product_information p
WHERE category_id = 29
ORDER BY product_id, supplier_id;
PRODUCT_ID SUPPLIER_ID
CUM_COVP
CUM_COVS
---------- ----------- ---------- ---------1774
103088
0
1775
103087
1473.25
2946.5
1794
103096 1702.77778 2554.16667
1825
103093
1926.25 2568.33333
2004
103086
1591.4
1989.25
2005
103086
1512.5
1815
2416
103088 1475.97959 1721.97619
. . .

Functions 5-65

COVAR_SAMP

COVAR_SAMP
Syntax
OVER
COVAR_SAMP

(

expr1

,

expr2

(

analytic_clause

)

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions

Purpose
COVAR_SAMP returns the sample covariance of a set of number pairs. You can use it as
an aggregate or analytic function.
This function takes as arguments any numeric data type or any nonnumeric data type
that can be implicitly converted to a numeric data type. Oracle determines the
argument with the highest numeric precedence, implicitly converts the remaining
arguments to that data type, and returns that data type.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion and "Numeric
Precedence" on page 3-14 for information on numeric precedence

Oracle Database applies the function to the set of (expr1, expr2) pairs after eliminating
all pairs for which either expr1 or expr2 is null. Then Oracle makes the following
computation:
(SUM(expr1 * expr2) - SUM(expr1) * SUM(expr2) / n) / (n-1)

where n is the number of (expr1, expr2) pairs where neither expr1 nor expr2 is null.
The function returns a value of type NUMBER. If the function is applied to an empty set,
then it returns null.
See Also: "About SQL Expressions" on page 6-1 for information on
valid forms of expr and "Aggregate Functions" on page 5-10

Aggregate Example
Refer to the aggregate example for COVAR_POP on page 5-64.

Analytic Example
Refer to the analytic example for COVAR_POP on page 5-64.

5-66 Oracle Database SQL Language Reference

CUBE_TABLE

CUBE_TABLE
Syntax
CUBE_TABLE

HIERARCHY
dimension

hierarchy

HRR
schema
(

.

cube
HIERARCHY

’

dimension
hierarchy

’

)

HRR
schema

.

dimension

Purpose
CUBE_TABLE extracts data from a cube or dimension and returns it in the
two-dimensional format of a relational table, which can be used by SQL-based
applications.
The function takes a single VARCHAR2 argument. The optional hierarchy clause enables
you to specify a dimension hierarchy. A cube can have multiple hierarchy clauses, one
for each dimension.
You can generate these different types of tables:
■

■

A cube table contains a key column for each dimension and a column for each
measure and calculated measure in the cube. To create a cube table, you can
specify the cube with or without a cube hierarchy clause. For a dimension with
multiple hierarchies, this clause limits the return values to the dimension members
and levels in the specified hierarchy. Without a hierarchy clause, all dimension
members and all levels are included.
A dimension table contains a key column, and a column for each level and each
attribute. It also contains a MEMBER_TYPE column, which identifies each member
with one of the following codes:
–

L - Loaded from a table, view, or synonym

–

A - Loaded member and the single root of all hierarchies in the dimension, that
is, the "all" aggregate member

–

C - Calculated member

All dimension members and all levels are included in the table. To create a
dimension table, specify the dimension without a dimension hierarchy clause.
■

A hierarchy table contains all the columns of a dimension table plus a column for
the parent member and a column for each source level. It also contains a MEMBER_
TYPE column, as described for dimension tables. Any dimension members and
levels that are not part of the named hierarchy are excluded from the table. To
create a hierarchy table, specify the dimension with a dimension hierarchy clause.

CUBE_TABLE is a table function and is always used in the context of a SELECT statement
with this syntax:
SELECT ... FROM TABLE(CUBE_TABLE('arg'));

Functions 5-67

CUBE_TABLE

See Also: Oracle OLAP User's Guide for information about
dimensional objects and about the tables generated by CUBE_TABLE.

Examples
The following examples require Oracle Database with the OLAP option and the
GLOBAL sample schema. Refer to Oracle OLAP User's Guide for information on
downloading and installing the GLOBAL sample schema.
The following SELECT statement generates a dimension table of CHANNEL in the GLOBAL
schema.
SELECT dim_key, level_name, long_description, channel_total_id tot_id,
channel_channel_id chan_id, channel_long_description chan_desc,
total_long_description tot_desc
FROM TABLE(CUBE_TABLE('global.channel'));
DIM_KEY
----------CHANNEL_CAT
CHANNEL_DIR
CHANNEL_INT
TOTAL_TOTAL

LEVEL_NAME
---------CHANNEL
CHANNEL
CHANNEL
TOTAL

LONG_DESCRIPTION
---------------Catalog
Direct Sales
Internet
Total Channel

TOT_ID
-----TOTAL
TOTAL
TOTAL
TOTAL

CHAN_ID
------CAT
DIR
INT

CHAN_DESC
-----------Catalog
Direct Sales
Internet

TOT_DESC
------------Total Channel
Total Channel
Total Channel
Total Channel

The next statement generates a cube table of UNITS_CUBE. It restricts the table to the
MARKET and CALENDAR hierarchies.
SELECT sales, units, cost, time, customer, product, channel
FROM TABLE(CUBE_TABLE('global.units_cube HIERARCHY customer market HIERARCHY time calendar'))
WHERE rownum < 20;
SALES
UNITS
COST TIME
CUSTOMER
---------- ---------- ---------- -------------------------- -------------24538587.9
61109 22840853.7 CALENDAR_QUARTER_CY1998.Q1 TOTAL_TOTAL
24993273.3
61320
23147171 CALENDAR_QUARTER_CY1998.Q2 TOTAL_TOTAL
25080541.4
65265 23242535.4 CALENDAR_QUARTER_CY1998.Q3 TOTAL_TOTAL
26258474
66122 24391020.6 CALENDAR_QUARTER_CY1998.Q4 TOTAL_TOTAL
32785170
77589 30607218.1 CALENDAR_QUARTER_CY1999.Q1 TOTAL_TOTAL
. . .

5-68 Oracle Database SQL Language Reference

PRODUCT
----------TOTAL_TOTAL
TOTAL_TOTAL
TOTAL_TOTAL
TOTAL_TOTAL
TOTAL_TOTAL

CHANNEL
----------TOTAL_TOTAL
TOTAL_TOTAL
TOTAL_TOTAL
TOTAL_TOTAL
TOTAL_TOTAL

CUME_DIST

CUME_DIST
Aggregate Syntax
cume_dist_aggregate::=
,
CUME_DIST

(

expr

)

WITHIN

GROUP

,
DESC

FIRST
NULLS

ASC
(

ORDER

BY

LAST

expr

)

Analytic Syntax
cume_dist_analytic::=
query_partition_clause
CUME_DIST

(

)

OVER

(

order_by_clause

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions

Purpose
CUME_DIST calculates the cumulative distribution of a value in a group of values. The
range of values returned by CUME_DIST is >0 to <=1. Tie values always evaluate to the
same cumulative distribution value.
This function takes as arguments any numeric data type or any nonnumeric data type
that can be implicitly converted to a numeric data type. Oracle Database determines
the argument with the highest numeric precedence, implicitly converts the remaining
arguments to that data type, makes the calculation, and returns NUMBER.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion and "Numeric
Precedence" on page 3-14 for information on numeric precedence
■

■

As an aggregate function, CUME_DIST calculates, for a hypothetical row r identified
by the arguments of the function and a corresponding sort specification, the
relative position of row r among the rows in the aggregation group. Oracle makes
this calculation as if the hypothetical row r were inserted into the group of rows to
be aggregated over. The arguments of the function identify a single hypothetical
row within each aggregate group. Therefore, they must all evaluate to constant
expressions within each aggregate group. The constant argument expressions and
the expressions in the ORDER BY clause of the aggregate match by position.
Therefore, the number of arguments must be the same and their types must be
compatible.
As an analytic function, CUME_DIST computes the relative position of a specified
value in a group of values. For a row r, assuming ascending ordering, the CUME_
Functions 5-69

CUME_DIST

DIST of r is the number of rows with values lower than or equal to the value of r,
divided by the number of rows being evaluated (the entire query result set or a
partition).

Aggregate Example
The following example calculates the cumulative distribution of a hypothetical
employee with a salary of $15,500 and commission rate of 5% among the employees in
the sample table oe.employees:
SELECT CUME_DIST(15500, .05) WITHIN GROUP
(ORDER BY salary, commission_pct) "Cume-Dist of 15500"
FROM employees;
Cume-Dist of 15500
-----------------.972222222

Analytic Example
The following example calculates the salary percentile for each employee in the
purchasing division. For example, 40% of clerks have salaries less than or equal to
Himuro.
SELECT job_id, last_name, salary, CUME_DIST()
OVER (PARTITION BY job_id ORDER BY salary) AS cume_dist
FROM employees
WHERE job_id LIKE 'PU%'
ORDER BY job_id, last_name, salary, cume_dist;
JOB_ID
---------PU_CLERK
PU_CLERK
PU_CLERK
PU_CLERK
PU_CLERK
PU_MAN

LAST_NAME
SALARY CUME_DIST
------------------------- ---------- ---------Baida
2900
.8
Colmenares
2500
.2
Himuro
2600
.4
Khoo
3100
1
Tobias
2800
.6
Raphaely
11000
1

5-70 Oracle Database SQL Language Reference

CURRENT_DATE

CURRENT_DATE
Syntax
CURRENT_DATE

Purpose
CURRENT_DATE returns the current date in the session time zone, in a value in the
Gregorian calendar of data type DATE.

Examples
The following example illustrates that CURRENT_DATE is sensitive to the session time
zone:
ALTER SESSION SET TIME_ZONE = '-5:0';
ALTER SESSION SET NLS_DATE_FORMAT = 'DD-MON-YYYY HH24:MI:SS';
SELECT SESSIONTIMEZONE, CURRENT_DATE FROM DUAL;
SESSIONTIMEZONE CURRENT_DATE
--------------- --------------------05:00
29-MAY-2000 13:14:03
ALTER SESSION SET TIME_ZONE = '-8:0';
SELECT SESSIONTIMEZONE, CURRENT_DATE FROM DUAL;
SESSIONTIMEZONE CURRENT_DATE
--------------- --------------------08:00
29-MAY-2000 10:14:33

Functions 5-71

CURRENT_TIMESTAMP

CURRENT_TIMESTAMP
Syntax
(

precision

)

CURRENT_TIMESTAMP

Purpose
CURRENT_TIMESTAMP returns the current date and time in the session time zone, in a
value of data type TIMESTAMP WITH TIME ZONE. The time zone offset reflects the current
local time of the SQL session. If you omit precision, then the default is 6. The difference
between this function and LOCALTIMESTAMP is that CURRENT_TIMESTAMP returns a
TIMESTAMP WITH TIME ZONE value while LOCALTIMESTAMP returns a TIMESTAMP value.
In the optional argument, precision specifies the fractional second precision of the
time value returned.
See Also:

LOCALTIMESTAMP on page 5-140

Examples
The following example illustrates that CURRENT_TIMESTAMP is sensitive to the session
time zone:
ALTER SESSION SET TIME_ZONE = '-5:0';
ALTER SESSION SET NLS_DATE_FORMAT = 'DD-MON-YYYY HH24:MI:SS';
SELECT SESSIONTIMEZONE, CURRENT_TIMESTAMP FROM DUAL;
SESSIONTIMEZONE CURRENT_TIMESTAMP
--------------- ---------------------------------------------------05:00
04-APR-00 01.17.56.917550 PM -05:00
ALTER SESSION SET TIME_ZONE = '-8:0';
SELECT SESSIONTIMEZONE, CURRENT_TIMESTAMP FROM DUAL;
SESSIONTIMEZONE CURRENT_TIMESTAMP
--------------- ----------------------------------------------------08:00
04-APR-00 10.18.21.366065 AM -08:00

When you use the CURRENT_TIMESTAMP with a format mask, take care that the format
mask matches the value returned by the function. For example, consider the following
table:
CREATE TABLE current_test (col1 TIMESTAMP WITH TIME ZONE);

The following statement fails because the mask does not include the TIME ZONE portion
of the type returned by the function:
INSERT INTO current_test VALUES
(TO_TIMESTAMP_TZ(CURRENT_TIMESTAMP, 'DD-MON-RR HH.MI.SSXFF PM'));

The following statement uses the correct format mask to match the return type of
CURRENT_TIMESTAMP:
INSERT INTO current_test VALUES
(TO_TIMESTAMP_TZ(CURRENT_TIMESTAMP, 'DD-MON-RR HH.MI.SSXFF PM TZH:TZM'));

5-72 Oracle Database SQL Language Reference

CV

CV
Syntax
dimension_column
CV

(

)

Purpose
The CV function can be used only in the model_clause of a SELECT statement and then
only on the right-hand side of a model rule. It returns the current value of a dimension
column or a partitioning column carried from the left-hand side to the right-hand side
of a rule. This function is used in the model_clause to provide relative indexing with
respect to the dimension column. The return type is that of the data type of the
dimension column. If you omit the argument, then it defaults to the dimension column
associated with the relative position of the function within the cell reference.
The CV function can be used outside a cell reference. In this case, dimension_column is
required.
model_clause on page 19-28 and "Model Expressions" on
page 6-11 for the syntax and semantics

See Also:

Examples
The following example assigns the sum of the sales of the product represented by the
current value of the dimension column (Mouse Pad or Standard Mouse) for years 1999
and 2000 to the sales of that product for year 2001:
SELECT country, prod, year, s
FROM sales_view_ref
MODEL
PARTITION BY (country)
DIMENSION BY (prod, year)
MEASURES (sale s)
IGNORE NAV
UNIQUE DIMENSION
RULES UPSERT SEQUENTIAL ORDER
(
s[FOR prod IN ('Mouse Pad', 'Standard Mouse'), 2001] =
s[CV( ), 1999] + s[CV( ), 2000]
)
ORDER BY country, prod, year;
COUNTRY
---------France
France
France
France
France
France
France
France
Germany
Germany
Germany
Germany

PROD
----------------------------------Mouse Pad
Mouse Pad
Mouse Pad
Mouse Pad
Standard Mouse
Standard Mouse
Standard Mouse
Standard Mouse
Mouse Pad
Mouse Pad
Mouse Pad
Mouse Pad

YEAR
-------1998
1999
2000
2001
1998
1999
2000
2001
1998
1999
2000
2001

S
--------2509.42
3678.69
3000.72
6679.41
2390.83
2280.45
1274.31
3554.76
5827.87
8346.44
7375.46
15721.9

Functions 5-73

CV

Germany
Germany
Germany
Germany

Standard
Standard
Standard
Standard

Mouse
Mouse
Mouse
Mouse

1998
1999
2000
2001

7116.11
6263.14
2637.31
8900.45

16 rows selected.

The preceding example requires the view sales_view_ref. Refer to "The MODEL
clause: Examples" on page 19-45 to create this view.

5-74 Oracle Database SQL Language Reference

DATAOBJ_TO_PARTITION

DATAOBJ_TO_PARTITION
Syntax
DATAOBJ_TO_PARTITION

(

table

,

partition_id

)

Purpose
DATAOBJ_TO_PARTITION is useful only to Data Cartridge developers who are
performing data maintenance or query operations on system-partitioned tables that
are used to store domain index data. The DML or query operations are triggered by
corresponding operations on the base table of the domain index.
This function takes as arguments the name of the base table and the partition ID of the
base table partition, both of which are passed to the function by the appropriate
ODCIIndex method. The function returns the partition ID of the corresponding
system-partitioned table, which can be used to perform the operation (DML or query)
on that partition of the system-partitioned table.
See Also: Oracle Database Data Cartridge Developer's Guide for
information on the use of this function, including examples

Functions 5-75

DBTIMEZONE

DBTIMEZONE
Syntax
DBTIMEZONE

Purpose
DBTIMEZONE returns the value of the database time zone. The return type is a time zone
offset (a character type in the format '[+|-]TZH:TZM') or a time zone region name,
depending on how the user specified the database time zone value in the most recent
CREATE DATABASE or ALTER DATABASE statement.

Examples
The following example assumes that the database time zone is set to UTC time zone:
SELECT DBTIMEZONE
FROM DUAL;
DBTIME
-----+00:00

5-76 Oracle Database SQL Language Reference

DECODE

DECODE
Syntax
,
DECODE

(

expr

,

search

,

,
result

default
)

Purpose
DECODE compares expr to each search value one by one. If expr is equal to a search,
then Oracle Database returns the corresponding result. If no match is found, then
Oracle returns default. If default is omitted, then Oracle returns null.
The arguments can be any of the numeric types (NUMBER, BINARY_FLOAT, or BINARY_
DOUBLE) or character types.
■

■

If expr and search are character data, then Oracle compares them using
nonpadded comparison semantics. expr, search, and result can be any of the
data types CHAR, VARCHAR2, NCHAR, or NVARCHAR2. The string returned is of VARCHAR2
data type and is in the same character set as the first result parameter.
If the first search-result pair are numeric, then Oracle compares all
search-result expressions and the first expr to determine the argument with the
highest numeric precedence, implicitly converts the remaining arguments to that
data type, and returns that data type.

The search, result, and default values can be derived from expressions. Oracle
Database uses short-circuit evaluation. The database evaluates each search value only
before comparing it to expr, rather than evaluating all search values before comparing
any of them with expr. Consequently, Oracle never evaluates a search if a previous
search is equal to expr.
Oracle automatically converts expr and each search value to the data type of the first
search value before comparing. Oracle automatically converts the return value to the
same data type as the first result. If the first result has the data type CHAR or if the
first result is null, then Oracle converts the return value to the data type VARCHAR2.
In a DECODE function, Oracle considers two nulls to be equivalent. If expr is null, then
Oracle returns the result of the first search that is also null.
The maximum number of components in the DECODE function, including expr,
searches, results, and default, is 255.
See Also:
■

■

■

■

■

"Data Type Comparison Rules" on page 3-36 for information on
comparison semantics
"Data Conversion" on page 3-40 for information on data type
conversion in general
"Floating-Point Numbers" on page 3-12 for information on
floating-point comparison semantics
"Implicit and Explicit Data Conversion" on page 3-40 for
information on the drawbacks of implicit conversion
"COALESCE" on page 5-48 and "CASE Expressions" on page 6-5,
which provide functionality similar to that of DECODE

Functions 5-77

DECODE

Examples
This example decodes the value warehouse_id. If warehouse_id is 1, then the function
returns 'Southlake'; if warehouse_id is 2, then it returns 'San Francisco'; and so forth.
If warehouse_id is not 1, 2, 3, or 4, then the function returns 'Non domestic'.
SELECT product_id,
DECODE (warehouse_id, 1,
2,
3,
4,

'Southlake',
'San Francisco',
'New Jersey',
'Seattle',
'Non domestic') "Location"

FROM inventories
WHERE product_id < 1775
ORDER BY product_id, "Location";

5-78 Oracle Database SQL Language Reference

DECOMPOSE

DECOMPOSE
Syntax
’

CANONICAL

’

COMPATIBILITY

’

,
DECOMPOSE

(

string

’
)

Purpose
DECOMPOSE is valid only for Unicode characters. DECOMPOSE takes as its argument a
string in any data type and returns a Unicode string after decomposition in the same
character set as the input. For example, an o-umlaut code point will be returned as the
"o" code point followed by an umlaut code point.
■

■

■

string can be any of the data types CHAR, VARCHAR2, NCHAR, NVARCHAR2, CLOB, or
NCLOB.
CANONICAL causes canonical decomposition, which allows recomposition (for
example, with the COMPOSE function) to the original string. This is the default and
returns the string in NFD normal form.
COMPATIBILITY causes decomposition in compatibility mode. In this mode,
recomposition is not possible. This mode is useful, for example, when
decomposing half-width and full-width katakana characters, where recomposition
might not be desirable without external formatting or style information. It returns
the string in NFKD normal form.

CLOB and NCLOB values are supported through implicit conversion. If char is a
character LOB value, then it is converted to a VARCHAR value before the COMPOSE
operation. The operation will fail if the size of the LOB value exceeds the supported
length of the VARCHAR in the particular development environment.
See Also:
■

■

Oracle Database Globalization Support Guide for information on
Unicode character sets and character semantics
COMPOSE on page 5-51

Examples
The following example decomposes the string "Châteaux" into its component code
points:
SELECT DECOMPOSE ('Châteaux')
FROM DUAL;
DECOMPOSE
--------Cha^teaux

The results of this example can vary depending on the
character set of your operating system.

Note:

Functions 5-79

DELETEXML

DELETEXML
Syntax
,
DELETEXML

(

XMLType_instance

,

namespace_string

XPath_string

)

Purpose
DELETEXML deletes the node or nodes matched by the XPath expression in the target
XML.
■
■

■

XMLType_instance is an instance of XMLType.
XPath_string is an Xpath expression indicating one or more nodes that are to be
deleted. You can specify an absolute XPath_string with an initial slash or a
relative XPath_string by omitting the initial slash. If you omit the initial slash,
then the context of the relative path defaults to the root node. Any child nodes of
the nodes specified by XPath_string are also deleted.
The optional namespace_string provides namespace information for the XPath_
string. This parameter must be of type VARCHAR2.
See Also: Oracle XML DB Developer's Guide for more information
about this function

Examples
The following example removes the /Owner node from the warehouse_spec of one of
the warehouses modified in the example for APPENDCHILDXML on page 5-21:
UPDATE warehouses
SET warehouse_spec = DELETEXML(warehouse_spec, '/Warehouse/Building/Owner')
WHERE warehouse_id = 2;
SELECT warehouse_id, warehouse_spec
FROM warehouses
WHERE warehouse_id in (2,3);
ID WAREHOUSE_SPEC
---------- ----------------------------------2 

Rented
50000
1
Side load
Y
N
Lot
12 ft

3 

RentedGrandco

85700


5-80 Oracle Database SQL Language Reference

DELETEXML

N
N
Street
11.5 ft


Functions 5-81

DENSE_RANK

DENSE_RANK
Aggregate Syntax
dense_rank_aggregate::=
,
DENSE_RANK

(

expr

)

WITHIN

GROUP

,
DESC

FIRST
NULLS

ASC
(

ORDER

BY

LAST

expr

)

Analytic Syntax
dense_rank_analytic::=
query_partition_clause
DENSE_RANK

(

)

OVER

(

order_by_clause

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions

Purpose
DENSE_RANK computes the rank of a row in an ordered group of rows and returns the
rank as a NUMBER. The ranks are consecutive integers beginning with 1. The largest rank
value is the number of unique values returned by the query. Rank values are not
skipped in the event of ties. Rows with equal values for the ranking criteria receive the
same rank. This function is useful for top-N and bottom-N reporting.
This function accepts as arguments any numeric data type and returns NUMBER.
■

■

As an aggregate function, DENSE_RANK calculates the dense rank of a hypothetical
row identified by the arguments of the function with respect to a given sort
specification. The arguments of the function must all evaluate to constant
expressions within each aggregate group, because they identify a single row
within each group. The constant argument expressions and the expressions in the
order_by_clause of the aggregate match by position. Therefore, the number of
arguments must be the same and types must be compatible.
As an analytic function, DENSE_RANK computes the rank of each row returned from
a query with respect to the other rows, based on the values of the value_exprs in
the order_by_clause.

Aggregate Example
The following example computes the ranking of a hypothetical employee with the
salary $15,500 and a commission of 5% in the sample table oe.employees:
SELECT DENSE_RANK(15500, .05) WITHIN GROUP
(ORDER BY salary DESC, commission_pct) "Dense Rank"
FROM employees;
5-82 Oracle Database SQL Language Reference

DENSE_RANK

Dense Rank
---------3

Analytic Example
The following statement ranks the employees in the sample hr schema in department
60 based on their salaries. Identical salary values receive the same rank. However, no
rank values are skipped. Compare this example with the analytic example for RANK
on page 5-209.
SELECT department_id, last_name, salary,
DENSE_RANK() OVER (PARTITION BY department_id ORDER BY salary) DENSE_RANK
FROM employees WHERE department_id = 60
ORDER BY DENSE_RANK, last_name;
DEPARTMENT_ID
------------60
60
60
60
60

LAST_NAME
SALARY DENSE_RANK
------------------------- ---------- ---------Lorentz
4200
1
Austin
4800
2
Pataballa
4800
2
Ernst
6000
3
Hunold
9000
4

Functions 5-83

DEPTH

DEPTH
Syntax
DEPTH

(

correlation_integer

)

Purpose
DEPTH is an ancillary function used only with the UNDER_PATH and EQUALS_PATH
conditions. It returns the number of levels in the path specified by the UNDER_PATH
condition with the same correlation variable.
The correlation_integer can be any NUMBER integer. Use it to correlate this ancillary
function with its primary condition if the statement contains multiple primary
conditions. Values less than 1 are treated as 1.
See Also: EQUALS_PATH Condition on page 7-20, UNDER_PATH
Condition on page 7-21, and the related function PATH on page 5-179

Examples
The EQUALS_PATH and UNDER_PATH conditions can take two ancillary functions, DEPTH
and PATH. The following example shows the use of both ancillary functions. The
example assumes the existence of the XMLSchema warehouses.xsd (created in "Using
XML in SQL Statements" on page F-8).
SELECT PATH(1), DEPTH(2)
FROM RESOURCE_VIEW
WHERE UNDER_PATH(res, '/sys/schemas/OE', 1)=1
AND UNDER_PATH(res, '/sys/schemas/OE', 2)=1;
PATH(1)
DEPTH(2)
-------------------------------- -------. . .
www.example.com
1
www.example.com/xwarehouses.xsd
2
. . .

5-84 Oracle Database SQL Language Reference

DEREF

DEREF
Syntax
DEREF

(

expr

)

Purpose
DEREF returns the object reference of argument expr, where expr must return a REF to
an object. If you do not use this function in a query, then Oracle Database returns the
object ID of the REF instead, as shown in the example that follows.
See Also:

MAKE_REF on page 5-145

Examples
The sample schema oe contains an object type cust_address_typ. The "REF Constraint
Examples" on page 8-25 create a similar type, cust_address_typ_new, and a table with
one column that is a REF to the type. The following example shows how to insert into
such a column and how to use DEREF to extract information from the column:
INSERT INTO address_table VALUES
('1 First', 'G45 EU8', 'Paris', 'CA', 'US');
INSERT INTO customer_addresses
SELECT 999, REF(a) FROM address_table a;
SELECT address
FROM customer_addresses
ORDER BY address;
ADDRESS
-------------------------------------------------------------------------------000022020876B2245DBE325C5FE03400400B40DCB176B2245DBE305C5FE03400400B40DCB1
SELECT DEREF(address)
FROM customer_addresses;
DEREF(ADDRESS)(STREET_ADDRESS, POSTAL_CODE, CITY, STATE_PROVINCE, COUNTRY_ID)
-------------------------------------------------------------------------------CUST_ADDRESS_TYP_NEW('1 First', 'G45 EU8', 'Paris', 'CA', 'US')

Functions 5-85

DUMP

DUMP
Syntax
,
,
,
DUMP

(

length

start_position

return_fmt

expr

)

Purpose
DUMP returns a VARCHAR2 value containing the data type code, length in bytes, and
internal representation of expr. The returned result is always in the database character
set. For the data type corresponding to each code, see Table 3–1, " Built-in Data Type
Summary" on page 3-6.
The argument return_fmt specifies the format of the return value and can have any of
the following values:
■

8 returns result in octal notation.

■

10 returns result in decimal notation.

■

16 returns result in hexadecimal notation.

■

17 returns each byte printed as a character if and only if it can be interpreted as a
printable character in the character set of the compiler—typically ASCII or
EBCDIC. Some ASCII control characters may be printed in the form ^X as well.
Otherwise the character is printed in hexadecimal notation. All NLS parameters
are ignored. Do not depend on any particular output format for DUMP with return_
fmt 17.

By default, the return value contains no character set information. To retrieve the
character set name of expr, add 1000 to any of the preceding format values. For
example, a return_fmt of 1008 returns the result in octal and provides the character set
name of expr.
The arguments start_position and length combine to determine which portion of
the internal representation to return. The default is to return the entire internal
representation in decimal notation.
If expr is null, then this function returns NULL.
This function does not support CLOB data directly. However, CLOBs can be passed in as
arguments through implicit data conversion.
See Also: "Data Type Comparison Rules" on page 3-36 for more
information

Examples
The following examples show how to extract dump information from a string
expression and a column:
SELECT DUMP('abc', 1016)
FROM DUAL;
DUMP('ABC',1016)
-----------------------------------------Typ=96 Len=3 CharacterSet=WE8DEC: 61,62,63

5-86 Oracle Database SQL Language Reference

DUMP

SELECT DUMP(last_name, 8, 3, 2) "OCTAL"
FROM employees
WHERE last_name = 'Hunold'
ORDER BY employee_id;
OCTAL
------------------------------------------------------------------Typ=1 Len=6: 156,157
SELECT DUMP(last_name, 10, 3, 2) "ASCII"
FROM employees
WHERE last_name = 'Hunold'
ORDER BY employee_id;
ASCII
-------------------------------------------------------------------Typ=1 Len=6: 110,111

Functions 5-87

EMPTY_BLOB, EMPTY_CLOB

EMPTY_BLOB, EMPTY_CLOB
Syntax
empty_LOB::=
EMPTY_BLOB
(

)

EMPTY_CLOB

Purpose
EMPTY_BLOB and EMPTY_CLOB return an empty LOB locator that can be used to initialize
a LOB variable or, in an INSERT or UPDATE statement, to initialize a LOB column or
attribute to EMPTY. EMPTY means that the LOB is initialized, but not populated with
data.
An empty LOB is not the same as a null LOB, and an empty
CLOB is not the same as a LOB containing a string of 0 length. For more
information, see Oracle Database SecureFiles and Large Objects
Developer's Guide.
Note:

You cannot use the locator returned from this function
as a parameter to the DBMS_LOB package or the OCI.

Restriction on LOB Locators

Examples
The following example initializes the ad_photo column of the sample pm.print_media
table to EMPTY:
UPDATE print_media
SET ad_photo = EMPTY_BLOB();

5-88 Oracle Database SQL Language Reference

EXISTSNODE

EXISTSNODE
Note: The EXISTSNODE function is deprecated. It is still supported for
backward compatibility. However, Oracle recommends that you use
the XMLEXISTS function instead. See XMLEXISTS on page 5-361 for
more information.

Syntax
,
EXISTSNODE

(

XMLType_instance

,

namespace_string

XPath_string

)

Purpose
EXISTSNODE determines whether traversal of an XML document using a specified path
results in any nodes. It takes as arguments the XMLType instance containing an XML
document and a VARCHAR2 XPath string designating a path. The optional namespace_
string must resolve to a VARCHAR2 value that specifies a default mapping or
namespace mapping for prefixes, which Oracle Database uses when evaluating the
XPath expression(s).
The namespace_string argument defaults to the namespace of the root element. If you
refer to any subelement in Xpath_string, then you must specify namespace_string,
and you must specify the "who" prefix in both of these arguments.
"Using XML in SQL Statements" on page F-8 for examples
that specify namespace_string and use the "who" prefix.

See Also:

The return value is NUMBER:
■

0 if no nodes remain after applying the XPath traversal on the document

■

1 if any nodes remain

Examples
The following example tests for the existence of the /Warehouse/Dock node in the XML
path of the warehouse_spec column of the sample table oe.warehouses:
SELECT warehouse_id, warehouse_name
FROM warehouses
WHERE EXISTSNODE(warehouse_spec, '/Warehouse/Docks') = 1
ORDER BY warehouse_id;
WAREHOUSE_ID
-----------1
2
4

WAREHOUSE_NAME
----------------------------------Southlake, Texas
San Francisco
Seattle, Washington

Functions 5-89

EXP

EXP
Syntax
EXP

(

n

)

Purpose
EXP returns e raised to the nth power, where e = 2.71828183... . The function returns a
value of the same type as the argument.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. If the argument is
BINARY_FLOAT, then the function returns BINARY_DOUBLE. Otherwise the function
returns the same numeric data type as the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

Examples
The following example returns e to the 4th power:
SELECT EXP(4) "e to the 4th power"
FROM DUAL;
e to the 4th power
-----------------54.59815

5-90 Oracle Database SQL Language Reference

EXTRACT (datetime)

EXTRACT (datetime)
Syntax
extract_datetime::=
YEAR
MONTH
DAY
HOUR
MINUTE
EXTRACT

(

FROM

expr

)

SECOND
TIMEZONE_HOUR
TIMEZONE_MINUTE
TIMEZONE_REGION
TIMEZONE_ABBR

Purpose
EXTRACT extracts and returns the value of a specified datetime field from a datetime or
interval expression. The expr can be any expression that evaluates to a datetime or
interval data type compatible with the requested field:
■

■

■

■

If YEAR or MONTH is requested, then expr must evaluate to an expression of data
type DATE, TIMESTAMP, TIMESTAMP WITH TIME ZONE, TIMESTAMP WITH LOCAL TIME
ZONE, or INTERVAL YEAR TO MONTH.
If DAY is requested, then expr must evaluate to an expression of data type DATE,
TIMESTAMP, TIMESTAMP WITH TIME ZONE, TIMESTAMP WITH LOCAL TIME ZONE, or
INTERVAL DAY TO SECOND.
If HOUR, MINUTE, or SECOND is requested, then expr must evaluate to an expression
of data type TIMESTAMP, TIMESTAMP WITH TIME ZONE, TIMESTAMP WITH LOCAL TIME
ZONE, or INTERVAL DAY TO SECOND. DATE is not valid here, because Oracle Database
treats it as ANSI DATE data type, which has no time fields.
If TIMEZONE_HOUR, TIMEZONE_MINUTE, TIMEZONE_ABBR, TIMEZONE_REGION, or
TIMEZONE_OFFSET is requested, then expr must evaluate to an expression of data
type TIMESTAMP WITH TIME ZONE or TIMESTAMP WITH LOCAL TIME ZONE.

EXTRACT interprets expr as an ANSI datetime data type. For example, EXTRACT treats
DATE not as legacy Oracle DATE but as ANSI DATE, without time elements. Therefore,
you can extract only YEAR, MONTH, and DAY from a DATE value. Likewise, you can extract
TIMEZONE_HOUR and TIMEZONE_MINUTE only from the TIMESTAMP WITH TIME ZONE data
type.
When you specify TIMEZONE_REGION or TIMEZONE_ABBR (abbreviation), the value
returned is a VARCHAR2 string containing the appropriate time zone region name or
abbreviation. When you specify any of the other datetime fields, the value returned is
an integer value of NUMBER data type representing the datetime value in the Gregorian
calendar. When extracting from a datetime with a time zone value, the value returned

Functions 5-91

EXTRACT (datetime)

is in UTC. For a listing of time zone region names and their corresponding
abbreviations, query the V$TIMEZONE_NAMES dynamic performance view.
This function can be very useful for manipulating datetime field values in very large
tables, as shown in the first example below.

Time zone region names are needed by the daylight saving
feature. These names are stored in two types of time zone files: one
large and one small. One of these files is the default file, depending
on your environment and the release of Oracle Database you are
using. For more information regarding time zone files and names,
see Oracle Database Globalization Support Guide.

Note:

Some combinations of datetime field and datetime or interval value expression result
in ambiguity. In these cases, Oracle Database returns UNKNOWN (see the examples that
follow for additional information).
See Also:
■

■

■

Oracle Database Globalization Support Guide for a complete listing of
the time zone region names in both files
"Datetime/Interval Arithmetic" on page 3-20 for a description of
datetime_value_expr and interval_value_expr
Oracle Database Reference for information on the dynamic
performance views

Examples
The following example returns from the oe.orders table the number of orders placed
in each month:
SELECT EXTRACT(month FROM order_date) "Month", COUNT(order_date) "No. of Orders"
FROM orders
GROUP BY EXTRACT(month FROM order_date)
ORDER BY "No. of Orders" DESC, "Month";
Month No. of Orders
---------- ------------11
15
6
14
7
14
3
11
5
10
2
9
9
9
8
7
10
6
1
5
12
4
4
1
12 rows selected.

The following example returns the year 1998.
SELECT EXTRACT(YEAR FROM DATE '1998-03-07')
FROM DUAL;

5-92 Oracle Database SQL Language Reference

EXTRACT (datetime)

EXTRACT(YEARFROMDATE'1998-03-07')
--------------------------------1998

The following example selects from the sample table hr.employees all employees who
were hired after 2007:
SELECT last_name, employee_id, hire_date
FROM employees
WHERE EXTRACT(YEAR FROM TO_DATE(hire_date, 'DD-MON-RR')) > 2007
ORDER BY hire_date;
LAST_NAME
EMPLOYEE_ID HIRE_DATE
------------------------- ----------- --------Johnson
179 04-JAN-08
Grant
199 13-JAN-08
Marvins
164 24-JAN-08
. . .

The following example results in ambiguity, so Oracle returns UNKNOWN:
SELECT EXTRACT(TIMEZONE_REGION FROM TIMESTAMP '1999-01-01 10:00:00 -08:00')
FROM DUAL;
EXTRACT(TIMEZONE_REGIONFROMTIMESTAMP'1999-01-0110:00:00-08:00')
---------------------------------------------------------------UNKNOWN

The ambiguity arises because the time zone numerical offset is provided in the
expression, and that numerical offset may map to more than one time zone region
name.

Functions 5-93

EXTRACT (XML)

EXTRACT (XML)
Note: The EXTRACT (XML) function is deprecated. It is still supported
for backward compatibility. However, Oracle recommends that you
use the XMLQUERY function instead. See XMLQUERY on page 5-368 for
more information.

Syntax
extract_xml::=
,
EXTRACT

(

XMLType_instance

,

namespace_string

XPath_string

)

Purpose
EXTRACT (XML) is similar to the EXISTSNODE function. It applies a VARCHAR2 XPath
string and returns an XMLType instance containing an XML fragment. You can specify
an absolute XPath_string with an initial slash or a relative XPath_string by omitting
the initial slash. If you omit the initial slash, then the context of the relative path
defaults to the root node. The optional namespace_string is required if the XML you
are handling uses a namespace prefix. This argument must resolve to a VARCHAR2 value
that specifies a default mapping or namespace mapping for prefixes, which Oracle
Database uses when evaluating the XPath expression(s).

Examples
The following example extracts the value of the /Warehouse/Dock node of the XML
path of the warehouse_spec column in the sample table oe.warehouses:
SELECT warehouse_name,
EXTRACT(warehouse_spec, '/Warehouse/Docks') "Number of Docks"
FROM warehouses
WHERE warehouse_spec IS NOT NULL
ORDER BY warehouse_name;
WAREHOUSE_NAME
------------------------New Jersey
San Francisco
Seattle, Washington
Southlake, Texas

Number of Docks
------------------------1
3
2

Compare this example with the example for EXTRACTVALUE on page 5-95, which
returns the scalar value of the XML fragment.

5-94 Oracle Database SQL Language Reference

EXTRACTVALUE

EXTRACTVALUE
The EXTRACTVALUE function is deprecated. It is still supported
for backward compatibility. However, Oracle recommends that you
use the XMLTABLE function, or the XMLCAST and XMLQUERY functions
instead. See XMLTABLE on page 5-375, XMLCAST on page 5-351, and
XMLQUERY on page 5-368 for more information.
Note:

Syntax
,
EXTRACTVALUE

(

XMLType_instance

,

XPath_string

namespace_string
)

The EXTRACTVALUE function takes as arguments an XMLType instance and an XPath
expression and returns a scalar value of the resultant node. The result must be a single
node and be either a text node, attribute, or element. If the result is an element, then
the element must have a single text node as its child, and it is this value that the
function returns. You can specify an absolute XPath_string with an initial slash or a
relative XPath_string by omitting the initial slash. If you omit the initial slash, the
context of the relative path defaults to the root node.
If the specified XPath points to a node with more than one child, or if the node pointed
to has a non-text node child, then Oracle returns an error. The optional namespace_
string must resolve to a VARCHAR2 value that specifies a default mapping or
namespace mapping for prefixes, which Oracle uses when evaluating the XPath
expression(s).
For documents based on XML schemas, if Oracle can infer the type of the return value,
then a scalar value of the appropriate type is returned. Otherwise, the result is of type
VARCHAR2. For documents that are not based on XML schemas, the return type is
always VARCHAR2.

Examples
The following example takes as input the same arguments as the example for
EXTRACT (XML) on page 5-94. Instead of returning an XML fragment, as does the
EXTRACT function, it returns the scalar value of the XML fragment:
SELECT warehouse_name, EXTRACTVALUE(e.warehouse_spec, '/Warehouse/Docks') "Docks"
FROM warehouses e
WHERE warehouse_spec IS NOT NULL
ORDER BY warehouse_name;
WAREHOUSE_NAME
-------------------New Jersey
San Francisco
Seattle, Washington
Southlake, Texas

Docks
-----------1
3
2

Functions 5-95

FEATURE_ID

FEATURE_ID
Syntax
schema
FEATURE_ID

.

(

model

mining_attribute_clause

)

mining_attribute_clause:=
*
,
USING

schema

.
table
AS

.

*

alias

expr

Purpose
This function is for use with feature extraction models created by the DBMS_DATA_
MINING package or with Oracle Data Miner. It returns an Oracle NUMBER that is the
identifier of the feature with the highest value in the row.
The mining_attribute_clause behaves as described for the PREDICTION function.
Refer to mining_attribute_clause on page 5-192.
See Also:
■

■

Oracle Data Mining Concepts for detailed information about Oracle
Data Mining
Oracle Data Mining Application Developer's Guide for detailed
information about scoring with the Data Mining SQL functions

Examples
The following example lists the features and corresponding count of customers in a
dataset.
This example and the prerequisite data mining operations, including creation of the
nmf_sh_sample model and nmf_sh_sample_apply_prepared view, can be found in the
demo file $ORACLE_HOME/rdbms/demo/dmnmdemo.sql. General information on data
mining demo files is available in Oracle Data Mining Administrator's Guide. The
example is presented here to illustrate the syntactic use of the function.
SELECT FEATURE_ID(nmf_sh_sample USING *) AS feat, COUNT(*) AS cnt
FROM nmf_sh_sample_apply_prepared
GROUP BY FEATURE_ID(nmf_sh_sample USING *)
ORDER BY cnt DESC, feat DESC;
FEAT
CNT
---------- ---------7
1443
2
49
3
6
6
1
1
1

5-96 Oracle Database SQL Language Reference

FEATURE_SET

FEATURE_SET
Syntax
,
schema
FEATURE_SET

.

,

(

cutoff

topN

model

mining_attribute_clause

)

mining_attribute_clause:=
*
,
USING

schema

.
table
AS

.

*

alias

expr

Purpose
This function is for use with feature extraction models created by the DBMS_DATA_
MINING package or with Oracle Data Miner. It returns a varray of objects containing all
possible features. Each object in the varray is a pair of scalar values containing the
feature ID and the feature value. The object fields are named FEATURE_ID and VALUE,
and both are Oracle NUMBER.
The optional topN argument is a positive integer that restricts the set of features to
those that have one of the top N values. If there is a tie at the Nth value, then the
database still returns only N values. If you omit this argument, then the function
returns all features.
The optional cutoff argument restricts the returned features to only those that have a
feature value greater than or equal to the specified cutoff. To filter only by cutoff,
specify NULL for topN and the desired cutoff for cutoff.
The mining_attribute_clause behaves as described for the PREDICTION function.
Refer to mining_attribute_clause on page 5-192.
See Also:
■

■

Oracle Data Mining Concepts for detailed information about Oracle
Data Mining
Oracle Data Mining Application Developer's Guide for detailed
information about scoring with the Data Mining SQL functions

Examples
The following example lists the top features corresponding to a given customer record
(based on match quality), and determines the top attributes for each feature (based on
coefficient > 0.25).
This example and the prerequisite data mining operations, including the creation of
the model, views, and type, can be found in the demo file $ORACLE_
HOME/rdbms/demo/dmnmdemo.sql. General information on data mining demo files is
available in Oracle Data Mining Administrator's Guide. The example is presented here to
illustrate the syntactic use of the function.
Functions 5-97

FEATURE_SET

WITH
feat_tab AS (
SELECT F.feature_id fid,
A.attribute_name attr,
TO_CHAR(A.attribute_value) val,
A.coefficient coeff
FROM TABLE(DBMS_DATA_MINING.GET_MODEL_DETAILS_NMF('nmf_sh_sample')) F,
TABLE(F.attribute_set) A
WHERE A.coefficient > 0.25
),
feat AS (
SELECT fid,
CAST(COLLECT(Featattr(attr, val, coeff))
AS Featattrs) f_attrs
FROM feat_tab
GROUP BY fid
),
cust_10_features AS (
SELECT T.cust_id, S.feature_id, S.value
FROM (SELECT cust_id, FEATURE_SET(nmf_sh_sample, 10 USING *) pset
FROM nmf_sh_sample_apply_prepared
WHERE cust_id = 100002) T,
TABLE(T.pset) S
)
SELECT A.value, A.feature_id fid,
B.attr, B.val, B.coeff
FROM cust_10_features A,
(SELECT T.fid, F.*
FROM feat T,
TABLE(T.f_attrs) F) B
WHERE A.feature_id = B.fid
ORDER BY A.value DESC, A.feature_id ASC, coeff DESC, attr ASC, val ASC;
VALUE FID ATTR
-------- ---- ------------------------6.8409
7 YRS_RESIDENCE
6.8409
7 BOOKKEEPING_APPLICATION
6.8409
7 CUST_GENDER
6.8409
7 COUNTRY_NAME
6.4975
3 YRS_RESIDENCE
6.4975
3 BOOKKEEPING_APPLICATION
6.4975
3 COUNTRY_NAME
6.4886
2 YRS_RESIDENCE
6.4886
2 CUST_GENDER
6.4886
2 PRINTER_SUPPLIES
6.3953
4 YRS_RESIDENCE
5.9640
6 YRS_RESIDENCE
5.9640
6 HOME_THEATER_PACKAGE
5.2424
5 YRS_RESIDENCE
2.4714
8 YRS_RESIDENCE
2.3559
1 YRS_RESIDENCE
2.3559
1 FLAT_PANEL_MONITOR
17 rows selected.

5-98 Oracle Database SQL Language Reference

VAL
COEFF
------------------------ ------1.3879
.4388
M
.2956
United States of America
.2848
1.2668
.3465
United States of America
.2927
1.3285
M
.2819
.2704
1.2931
1.1585
.2576
1.0067
.3297
.2768
.2593

FEATURE_VALUE

FEATURE_VALUE
Syntax
schema
FEATURE_VALUE

.

(

,

feature_id

model

mining_attribute_clause

)

mining_attribute_clause:=
*
,
USING

schema

.
table
AS

.

*

alias

expr

Purpose
This function is for use with feature extraction models created by the DBMS_DATA_
MINING package or with Oracle Data Miner. It returns the value of a given feature. If
you omit the feature_id argument, then the function returns the highest feature
value. You can use this form in conjunction with the FEATURE_ID function to obtain the
largest feature/value combination.
The mining_attribute_clause behaves as described for the PREDICTION function.
Refer to mining_attribute_clause on page 5-192.
See Also:
■

■

Oracle Data Mining Concepts for detailed information about Oracle
Data Mining
Oracle Data Mining Application Developer's Guide for detailed
information about scoring with the Data Mining SQL functions

Examples
The following example lists the customers that correspond to feature 3, ordered by
match quality.
This example and the prerequisite data mining operations, including the creation of
the model and view, can be found in the demo file $ORACLE_
HOME/rdbms/demo/dmnmdemo.sql. General information on data mining demo files is
available in Oracle Data Mining Administrator's Guide. The example is presented here to
illustrate the syntactic use of the function.
SELECT *
FROM (SELECT cust_id, FEATURE_VALUE(nmf_sh_sample, 3 USING *) match_quality
FROM nmf_sh_sample_apply_prepared
ORDER BY match_quality DESC)
WHERE ROWNUM < 11;
CUST_ID MATCH_QUALITY
---------- ------------100210
19.4101627
100962
15.2482251

Functions 5-99

FEATURE_VALUE

101151
101499
100363
100372
100982
101039
100759
100953

14.5685197
14.4186292
14.4037396
14.3335148
14.1716545
14.1079914
14.0913761
14.0799737

10 rows selected.

5-100 Oracle Database SQL Language Reference

FIRST

FIRST
Syntax
first::=
aggregate_function

KEEP

,
DESC

FIRST
NULLS

ASC
(

DENSE_RANK

FIRST

ORDER

BY

LAST

expr

)

query_partition_clause
OVER

(

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions of the ORDER BY clause and OVER
clause

Purpose
FIRST and LAST are very similar functions. Both are aggregate and analytic functions
that operate on a set of values from a set of rows that rank as the FIRST or LAST with
respect to a given sorting specification. If only one row ranks as FIRST or LAST, then
the aggregate operates on the set with only one element.
If you omit the OVER clause, then the FIRST and LAST functions are treated as aggregate
functions. You can use these functions as analytic functions by specifying the OVER
clause. The query_partition_clause is the only part of the OVER clause valid with
these functions. If you include the OVER clause but omit the query_partition_clause,
then the function is treated as an analytic function, but the window defined for
analysis is the entire table.
These functions take as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. The function returns the
same data type as the numeric data type of the argument.
When you need a value from the first or last row of a sorted group, but the needed
value is not the sort key, the FIRST and LAST functions eliminate the need for self-joins
or views and enable better performance.
■

■

■

The aggregate_function argument is any one of the MIN, MAX, SUM, AVG, COUNT,
VARIANCE, or STDDEV functions. It operates on values from the rows that rank either
FIRST or LAST. If only one row ranks as FIRST or LAST, then the aggregate operates
on a singleton (nonaggregate) set.
The KEEP keyword is for semantic clarity. It qualifies aggregate_function,
indicating that only the FIRST or LAST values of aggregate_function will be
returned.
DENSE_RANK FIRST or DENSE_RANK LAST indicates that Oracle Database will
aggregate over only those rows with the minimum (FIRST) or the maximum (LAST)
dense rank (also called olympic rank).
Functions

5-101

FIRST

See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion and LAST on
page 5-126

Aggregate Example
The following example returns, within each department of the sample table
hr.employees, the minimum salary among the employees who make the lowest
commission and the maximum salary among the employees who make the highest
commission:
SELECT department_id,
MIN(salary) KEEP (DENSE_RANK FIRST ORDER BY commission_pct) "Worst",
MAX(salary) KEEP (DENSE_RANK LAST ORDER BY commission_pct) "Best"
FROM employees
GROUP BY department_id
ORDER BY department_id;
DEPARTMENT_ID
Worst
Best
------------- ---------- ---------10
4400
4400
20
6000
13000
30
2500
11000
40
6500
6500
50
2100
8200
60
4200
9000
70
10000
10000
80
6100
14000
90
17000
24000
100
6900
12008
110
8300
12008
7000
7000

Analytic Example
The next example makes the same calculation as the previous example but returns the
result for each employee within the department:
SELECT last_name, department_id, salary,
MIN(salary) KEEP (DENSE_RANK FIRST ORDER BY commission_pct)
OVER (PARTITION BY department_id) "Worst",
MAX(salary) KEEP (DENSE_RANK LAST ORDER BY commission_pct)
OVER (PARTITION BY department_id) "Best"
FROM employees
ORDER BY department_id, salary, last_name;
LAST_NAME
DEPARTMENT_ID
SALARY
Worst
Best
------------------- ------------- ---------- ---------- ---------Whalen
10
4400
4400
4400
Fay
20
6000
6000
13000
Hartstein
20
13000
6000
13000
. . .
Gietz
110
8300
8300
12008
Higgins
110
12008
8300
12008
Grant
7000
7000
7000

5-102 Oracle Database SQL Language Reference

FIRST_VALUE

FIRST_VALUE
Syntax
RESPECT
NULLS
IGNORE
(

expr

)
RESPECT

FIRST_VALUE

OVER

NULLS

(

analytic_clause

IGNORE
(

expr

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions, including valid forms of expr

Purpose
FIRST_VALUE is an analytic function. It returns the first value in an ordered set of
values. If the first value in the set is null, then the function returns NULL unless you
specify IGNORE NULLS. This setting is useful for data densification.

The two forms of this syntax have the same behavior. The top
branch is the ANSI format, which Oracle recommends. The bottom
branch is deprecated but is supported for backward compatibility.

Note:

{RESPECT | IGNORE} NULLS determines whether null values of expr are included in or
eliminated from the calculation. The default is RESPECT NULLS. If you specify IGNORE
NULLS, then FIRST_VALUE returns the first non-null value in the set, or NULL if all values
are null. Refer to "Using Partitioned Outer Joins: Examples" on page 19-52 for an
example of data densification.
You cannot nest analytic functions by using FIRST_VALUE or any other analytic
function for expr. However, you can use other built-in function expressions for expr.
Refer to "About SQL Expressions" on page 6-1 for information on valid forms of expr.

Examples
The following example selects, for each employee in Department 90, the name of the
employee with the lowest salary.
SELECT department_id, last_name, salary,
FIRST_VALUE(last_name)
OVER (ORDER BY salary ASC ROWS UNBOUNDED PRECEDING) AS lowest_sal
FROM (SELECT * FROM employees
WHERE department_id = 90
ORDER BY employee_id)
ORDER BY last_name;
DEPARTMENT_ID
------------90
90
90

LAST_NAME
SALARY LOWEST_SAL
------------------------- ---------- ------------------------De Haan
17000 Kochhar
King
24000 Kochhar
Kochhar
17000 Kochhar

Functions

5-103

)

FIRST_VALUE

The example illustrates the nondeterministic nature of the FIRST_VALUE function.
Kochhar and DeHaan have the same salary, so are in adjacent rows. Kochhar appears
first because the rows returned by the subquery are ordered by employee_id.
However, if the rows returned by the subquery are ordered by employee_id in
descending order, as in the next example, then the function returns a different value:
SELECT department_id, last_name, salary,
FIRST_VALUE(last_name)
OVER (ORDER BY salary ASC ROWS UNBOUNDED PRECEDING) AS fv
FROM (SELECT * FROM employees
WHERE department_id = 90
ORDER by employee_id DESC)
ORDER BY last_name;
DEPARTMENT_ID
------------90
90
90

LAST_NAME
SALARY FV
------------------------- ---------- ------------------------De Haan
17000 De Haan
King
24000 De Haan
Kochhar
17000 De Haan

The following example shows how to make the FIRST_VALUE function deterministic by
ordering on a unique key.
SELECT department_id, last_name, salary, hire_date,
FIRST_VALUE(last_name)
OVER (ORDER BY salary ASC, hire_date ROWS UNBOUNDED PRECEDING) AS fv
FROM (SELECT * FROM employees
WHERE department_id = 90
ORDER BY employee_id DESC)
ORDER BY last_name;
DEPARTMENT_ID
------------90
90
90

LAST_NAME
SALARY HIRE_DATE FV
--------------- ---------- --------- ------------------------De Haan
17000 13-JAN-01 De Haan
King
24000 17-JUN-03 De Haan
Kochhar
17000 21-SEP-05 De Haan

When you use a logical offset (RANGE instead of ROWS), the function is deterministic.
When duplicates are found for the ORDER BY expression, the FIRST_VALUE is the lowest
value of expr:
SELECT department_id, last_name, salary,
FIRST_VALUE(last_name)
OVER (ORDER BY salary ASC RANGE UNBOUNDED PRECEDING) AS lowest_sal
FROM (SELECT * FROM employees
WHERE department_id = 90
ORDER BY employee_id);
DEPARTMENT_ID
------------90
90
90

LAST_NAME
SALARY LOWEST_SAL
------------------------- ---------- ------------------------De Haan
17000 De Haan
Kochhar
17000 De Haan
King
24000 De Haan

5-104 Oracle Database SQL Language Reference

FLOOR

FLOOR
Syntax
FLOOR

(

n

)

Purpose
FLOOR returns the largest integer equal to or less than n. The number n can always be
written as the sum of an integer k and a positive fraction f such that 0 <= f < 1 and n =
k + f. The value of FLOOR is the integer k. Thus, the value of FLOOR is n itself if and only
if n is precisely an integer.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. The function returns the
same data type as the numeric data type of the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion and CEIL on
page 5-38

Examples
The following example returns the largest integer equal to or less than 15.7:
SELECT FLOOR(15.7) "Floor"
FROM DUAL;
Floor
---------15

Functions

5-105

FROM_TZ

FROM_TZ
Syntax
FROM_TZ

(

timestamp_value

,

time_zone_value

)

Purpose
FROM_TZ converts a timestamp value and a time zone to a TIMESTAMP WITH TIME ZONE
value. time_zone_value is a character string in the format 'TZH:TZM' or a character
expression that returns a string in TZR with optional TZD format.

Examples
The following example returns a timestamp value to TIMESTAMP WITH TIME ZONE:
SELECT FROM_TZ(TIMESTAMP '2000-03-28 08:00:00', '3:00')
FROM DUAL;
FROM_TZ(TIMESTAMP'2000-03-2808:00:00','3:00')
--------------------------------------------------------------28-MAR-00 08.00.000000000 AM +03:00

5-106 Oracle Database SQL Language Reference

GREATEST

GREATEST
Syntax
,
GREATEST

(

expr

)

Purpose
GREATEST returns the greatest of a list of one or more expressions. Oracle Database uses
the first expr to determine the return type. If the first expr is numeric, then Oracle
determines the argument with the highest numeric precedence, implicitly converts the
remaining arguments to that data type before the comparison, and returns that data
type. If the first expr is not numeric, then each expr after the first is implicitly
converted to the data type of the first expr before the comparison.
Oracle Database compares each expr using nonpadded comparison semantics. The
comparison is binary by default and is linguistic if the NLS_COMP parameter is set to
LINGUISTIC and the NLS_SORT parameter has a setting other than BINARY. Character
comparison is based on the numerical codes of the characters in the database character
set and is performed on whole strings treated as one sequence of bytes, rather than
character by character. If the value returned by this function is character data, then its
data type is VARCHAR2 if the first expr is a character data type and NVARCHAR2 if the first
expr is a national character data type.
See Also:
■

■

■

"Data Type Comparison Rules" on page 3-36 for more information
on character comparison
Table 3–10, " Implicit Type Conversion Matrix" on page 3-40 for
more information on implicit conversion and "Floating-Point
Numbers" on page 3-12 for information on binary-float
comparison semantics
"LEAST" on page 5-133, which returns the least of a list of one or
more expressions

Examples
The following statement selects the string with the greatest value:
SELECT GREATEST('HARRY', 'HARRIOT', 'HAROLD') "Greatest"
FROM DUAL;
Greatest
-------HARRY

In the following statement, the first argument is numeric. Oracle Database determines
that the argument with the highest numeric precedence is the second argument,
converts the remaining arguments to the data type of the second argument, and
returns the greatest value as that data type:
SELECT GREATEST (1, '3.925', '2.4') "Greatest"
FROM DUAL;
Greatest

Functions

5-107

GREATEST

-------3.925

5-108 Oracle Database SQL Language Reference

GROUP_ID

GROUP_ID
Syntax
GROUP_ID

(

)

Purpose
GROUP_ID distinguishes duplicate groups resulting from a GROUP BY specification. It is
useful in filtering out duplicate groupings from the query result. It returns an Oracle
NUMBER to uniquely identify duplicate groups. This function is applicable only in a
SELECT statement that contains a GROUP BY clause.
If n duplicates exist for a particular grouping, then GROUP_ID returns numbers in the
range 0 to n-1.

Examples
The following example assigns the value 1 to the duplicate co.country_region
grouping from a query on the sample tables sh.countries and sh.sales:
SELECT co.country_region, co.country_subregion,
SUM(s.amount_sold) "Revenue", GROUP_ID() g
FROM sales s, customers c, countries co
WHERE s.cust_id = c.cust_id
AND c.country_id = co.country_id
AND s.time_id = '1-JAN-00'
AND co.country_region IN ('Americas', 'Europe')
GROUP BY GROUPING SETS ( (co.country_region, co.country_subregion),
(co.country_region, co.country_subregion) )
ORDER BY co.country_region, co.country_subregion, "Revenue", g;
COUNTRY_REGION
-------------------Americas
Americas
Europe
Europe

COUNTRY_SUBREGION
Revenue
G
------------------------------ ---------- ---------Northern America
944.6
0
Northern America
944.6
1
Western Europe
566.39
0
Western Europe
566.39
1

To ensure that only rows with GROUP_ID < 1 are returned, add the following HAVING
clause to the end of the statement :
HAVING GROUP_ID() < 1

Functions

5-109

GROUPING

GROUPING
Syntax
GROUPING

(

expr

)

Purpose
GROUPING distinguishes superaggregate rows from regular grouped rows. GROUP BY
extensions such as ROLLUP and CUBE produce superaggregate rows where the set of all
values is represented by null. Using the GROUPING function, you can distinguish a null
representing the set of all values in a superaggregate row from a null in a regular row.
The expr in the GROUPING function must match one of the expressions in the GROUP BY
clause. The function returns a value of 1 if the value of expr in the row is a null
representing the set of all values. Otherwise, it returns zero. The data type of the value
returned by the GROUPING function is Oracle NUMBER. Refer to the SELECT group_by_
clause on page 19-27 for a discussion of these terms.

Examples
In the following example, which uses the sample tables hr.departments and
hr.employees, if the GROUPING function returns 1 (indicating a superaggregate row
rather than a regular row from the table), then the string "All Jobs" appears in the
"JOB" column instead of the null that would otherwise appear:
SELECT
DECODE(GROUPING(department_name), 1, 'ALL DEPARTMENTS', department_name)
AS department,
DECODE(GROUPING(job_id), 1, 'All Jobs', job_id) AS job,
COUNT(*) "Total Empl",
AVG(salary) * 12 "Average Sal"
FROM employees e, departments d
WHERE d.department_id = e.department_id
GROUP BY ROLLUP (department_name, job_id)
ORDER BY department, job;
DEPARTMENT
-----------------------------ALL DEPARTMENTS
Accounting
Accounting
Accounting
Administration
Administration
Executive
Executive
Executive
Finance
Finance
. . .

5-110 Oracle Database SQL Language Reference

JOB
Total Empl Average Sal
---------- ---------- ----------All Jobs
106 77481.0566
AC_ACCOUNT
1
99600
AC_MGR
1
144096
All Jobs
2
121848
AD_ASST
1
52800
All Jobs
1
52800
AD_PRES
1
288000
AD_VP
2
204000
All Jobs
3
232000
All Jobs
6
103216
FI_ACCOUNT
5
95040

GROUPING_ID

GROUPING_ID
Syntax
,
GROUPING_ID

(

expr

)

Purpose
GROUPING_ID returns a number corresponding to the GROUPING bit vector associated
with a row. GROUPING_ID is applicable only in a SELECT statement that contains a GROUP
BY extension, such as ROLLUP or CUBE, and a GROUPING function. In queries with many
GROUP BY expressions, determining the GROUP BY level of a particular row requires
many GROUPING functions, which leads to cumbersome SQL. GROUPING_ID is useful in
these cases.
GROUPING_ID is functionally equivalent to taking the results of multiple GROUPING
functions and concatenating them into a bit vector (a string of ones and zeros). By
using GROUPING_ID you can avoid the need for multiple GROUPING functions and make
row filtering conditions easier to express. Row filtering is easier with GROUPING_ID
because the desired rows can be identified with a single condition of GROUPING_ID = n.
The function is especially useful when storing multiple levels of aggregation in a
single table.

Examples
The following example shows how to extract grouping IDs from a query of the sample
table sh.sales:
SELECT channel_id, promo_id, sum(amount_sold) s_sales,
GROUPING(channel_id) gc,
GROUPING(promo_id) gp,
GROUPING_ID(channel_id, promo_id) gcp,
GROUPING_ID(promo_id, channel_id) gpc
FROM sales
WHERE promo_id > 496
GROUP BY CUBE(channel_id, promo_id)
ORDER BY channel_id, promo_id, s_sales, gc;
CHANNEL_ID
PROMO_ID
S_SALES
GC
GP
GCP
GPC
---------- ---------- ---------- ---------- ---------- ---------- ---------2
999 25797563.2
0
0
0
0
2
25797563.2
0
1
1
2
3
999 55336945.1
0
0
0
0
3
55336945.1
0
1
1
2
4
999 13370012.5
0
0
0
0
4
13370012.5
0
1
1
2
999 94504520.8
1
0
2
1
94504520.8
1
1
3
3

Functions

5-111

HEXTORAW

HEXTORAW
Syntax
HEXTORAW

(

char

)

Purpose
HEXTORAW converts char containing hexadecimal digits in the CHAR, VARCHAR2, NCHAR, or
NVARCHAR2 data type to a raw value.
This function does not support CLOB data directly. However, CLOBs can be passed in as
arguments through implicit data conversion.
See Also: "Data Type Comparison Rules" on page 3-36 for more
information.

Examples
The following example creates a simple table with a raw column, and inserts a
hexadecimal value that has been converted to RAW:
CREATE TABLE test (raw_col RAW(10));
INSERT INTO test VALUES (HEXTORAW('7D'));

The following example converts hexadecimal digits to a raw value and casts the raw
value to VARCHAR2:
SELECT UTL_RAW.CAST_TO_VARCHAR2(HEXTORAW('4041424344'))
FROM DUAL;
UTL_RAW.CAST_TO_VARCHAR2(HEXTORAW('4041424344'))
-----------------------------------------------@ABCD

See Also: "RAW and LONG RAW Data Types" on page 3-23 and
RAWTOHEX on page 5-212

5-112 Oracle Database SQL Language Reference

INITCAP

INITCAP
Syntax
INITCAP

(

char

)

Purpose
INITCAP returns char, with the first letter of each word in uppercase, all other letters in
lowercase. Words are delimited by white space or characters that are not
alphanumeric.
char can be of any of the data types CHAR, VARCHAR2, NCHAR, or NVARCHAR2. The return
value is the same data type as char. The database sets the case of the initial characters
based on the binary mapping defined for the underlying character set. For
linguistic-sensitive uppercase and lowercase, refer to NLS_INITCAP on page 5-161.
This function does not support CLOB data directly. However, CLOBs can be passed in as
arguments through implicit data conversion.
See Also: "Data Type Comparison Rules" on page 3-36 for more
information.

Examples
The following example capitalizes each word in the string:
SELECT INITCAP('the soap') "Capitals"
FROM DUAL;
Capitals
--------The Soap

Functions

5-113

INSERTCHILDXML

INSERTCHILDXML
Syntax
INSERTCHILDXML

,
(

XMLType_instance

,

XPath_string

,

child_expr

,

namespace_string

value_expr

)

Purpose
INSERTCHILDXML inserts a user-supplied value into the target XML at the node
indicated by the XPath expression. Compare this function with INSERTXMLBEFORE
on page 5-119.
■
■

■
■

■

XMLType_instance is an instance of XMLType.
XPath_string is an Xpath expression indicating one or more nodes into which the
one or more child nodes are to be inserted. You can specify an absolute XPath_
string with an initial slash or a relative XPath_string by omitting the initial slash.
If you omit the initial slash, then the context of the relative path defaults to the
root node.
child_expr specifies the one or more element or attribute nodes to be inserted.
value_expr is an fragment of XMLType that specifies one or more notes being
inserted. It must resolve to a string.
The optional namespace_string provides namespace information for the XPath_
string. This parameter must be of type VARCHAR2.
See Also: Oracle XML DB Developer's Guide for more information
about this function

Examples
The following example adds a second /Owner node to the warehouse_spec of one of
the warehouses updated in the example for APPENDCHILDXML on page 5-21:
UPDATE warehouses
SET warehouse_spec = INSERTCHILDXML(warehouse_spec, '/Warehouse/Building',
'Owner', XMLType('LesserCo'))
WHERE warehouse_id = 3;
SELECT warehouse_spec
FROM warehouses
WHERE warehouse_id = 3;
WAREHOUSE_SPEC
---------------------------------------------------------------------------

Rented
Grandco
LesserCo

85700

N

5-114 Oracle Database SQL Language Reference

INSERTCHILDXML

N
Street
11.5 ft


Functions

5-115

INSERTCHILDXMLAFTER

INSERTCHILDXMLAFTER
Syntax
INSERTCHILDXMLAFTER

,
(

XMLType_instance

,

XPath_string

,

child_expr

,

namespace_string

value_expr

)

Purpose
INSERTXMLCHILDAFTER inserts one or more collection elements as children of target
parent elements. The insertion for each target occurs immediately after a specified
existing collection element. The existing XML document that is the target of the
insertion can be schema-based or non-schema-based.
■
■

■

■

■

XMLType_instance identifies the XML data that is the target of the insertion.
XPath_string locates the parent elements within target-data; child-data is inserted
under each parent element.
child_expr is a relative XPath 1.0 expression that locates the existing child that
will precede the inserted child-data. It must name a child element of the element
indicated by parent-xpath, and it can include a predicate.
value_expr is the XMLType child element data to insert. Each top-level element
node in this argument must have the same data type as the element indicated by
child_expr.
The optional namespace_string specifies the namespace for the parent elements,
existing child element, and child element XML data to be inserted.
See Also: Oracle XML DB Developer's Guide for more information
about this function

Examples
The following example is similar to that for INSERTCHILDXML, but it adds a third /Owner
node after the /Owner node added in the other example. The output of the query has
been formatted for readability.
UPDATE warehouses
SET warehouse_spec = INSERTCHILDXMLAFTER(warehouse_spec, '/Warehouse/Building',
'Owner[2]', XMLType('ThirdOwner'))
WHERE warehouse_id = 3;
SELECT warehouse_name,
EXTRACT(warehouse_spec, '/Warehouse/Building/Owner') "Owners"
FROM warehouses
WHERE warehouse_id = 3;
WAREHOUSE_NAME
Owners
----------------------------------- -----------------------------New Jersey
GrandCo
LesserCo
ThirdOwner

5-116 Oracle Database SQL Language Reference

INSERTCHILDXMLBEFORE

INSERTCHILDXMLBEFORE
Syntax
INSERTCHILDXMLBEFORE

,
(

XMLType_instance

,

XPath_string

,

child_expr

,

namespace_string

value_expr

)

Purpose
INSERTXMLCHILDBEFORE inserts one or more collection elements as children of target
parent elements. The insertion for each target occurs immediately before a specified
existing collection element. The existing XML document that is the target of the
insertion can be schema-based or non-schema-based.
■
■

■

■

■

XMLType_instance identifies the XML data that is the target of the insertion.
XPath_string locates the parent elements within target-data; child-data is inserted
under each parent element.
child_expr is a relative XPath 1.0 expression that locates the existing child that
will follow the inserted child-data. It must name a child element of the element
indicated by parent-xpath, and it can include a predicate.
value_expr is the XMLType child element data to insert. Each top-level element
node in this argument must have the same data type as the element indicated by
child_expr.
The optional namespace_string specifies the namespace for the parent elements,
existing child element, and child element XML data to be inserted.
See Also: Oracle XML DB Developer's Guide for more information
about this function

Examples
The following example is similar to that for INSERTCHILDXML, but it adds a third /Owner
node before the /Owner node added in the other example. The output of the query has
been formatted for readability.
UPDATE warehouses
SET warehouse_spec = INSERTCHILDXMLBEFORE(warehouse_spec, '/Warehouse/Building',
'Owner[2]', XMLType('ThirdOwner'))
WHERE warehouse_id = 3;
SELECT warehouse_name,
EXTRACT(warehouse_spec, '/Warehouse/Building/Owner') "Owners"
FROM warehouses
WHERE warehouse_id = 3;
WAREHOUSE_NAME
Owners
----------------------------------- -----------------------------New Jersey
GrandCo
ThirdOwner
LesserCo

Functions

5-117

INSERTXMLAFTER

INSERTXMLAFTER
Syntax
,
INSERTXMLAFTER

(

XMLType_instance

,

XPath_string

,

namespace_string

value_expr

)

Purpose
INSERTXMLAFTER inserts one or more nodes of any kind immediately after a target node
that is not an attribute node. The XML document that is the target of the insertion can
be schema-based or non-schema-based. This function is similar to insertXMLbefore,
but it inserts after, not before, the target node.
■
■

■

■

XMLType_instance specifies the target node of the of the insertion.
XPath_string is an XPath 1.0 expression that locates in the target node zero or
more nodes of any kind except attribute nodes. XML-data is inserted immediately
after each of these nodes; that is, each node specified becomes the preceding
sibling node of a node specified in value_expr.
value_expr is the XML data to be inserted. You can specify one or more nodes of
any kind. The order of the nodes is preserved after the insertion.
The optional namespace_string is the namespace for the target node.
See Also: Oracle XML DB Developer's Guide for more information
about this function

Examples
The following example is similar to that for INSERTCHILDXML, but it adds a third /Owner
node after the /Owner node added in the other example. The output of the query has
been formatted for readability.
UPDATE warehouses
SET warehouse_spec = INSERTXMLAFTER(warehouse_spec,
'/Warehouse/Building/Owner[1]', XMLType('SecondOwner'))
WHERE warehouse_id = 3;
SELECT warehouse_name,
EXTRACT(warehouse_spec, '/Warehouse/Building/Owner') "Owners"
FROM warehouses
WHERE warehouse_id = 3;
WAREHOUSE_NAME
Owners
----------------------------------- -----------------------------New Jersey
GrandCo
SecondOwner
LesserCo

5-118 Oracle Database SQL Language Reference

INSERTXMLBEFORE

INSERTXMLBEFORE
Syntax
,
INSERTXMLBEFORE

(

XMLType_instance

,

XPath_string

,

namespace_string

value_expr

)

Purpose
INSERTXMLBEFORE inserts a user-supplied value into the target XML before the node
indicated by the XPath expression. This function is similar to INSERTXMLAFTER, but
it inserts before, not after, the target node. Compare this function with
INSERTCHILDXML on page 5-114.
■
■

■

■

XMLType_instance is an instance of XMLType.
XPath_string is an Xpath expression indicating one or more nodes into which one
or more child nodes are to be inserted. You can specify an absolute XPath_string
with an initial slash or a relative XPath_string by omitting the initial slash. If you
omit the initial slash, then the context of the relative path defaults to the root node.
value_expr is a fragment of XMLType that defines one or more nodes being
inserted and their position within the parent node. It must resolve to a string.
The optional namespace_string provides namespace information for the XPath_
string. This parameter must be of type VARCHAR2.
See Also: Oracle XML DB Developer's Guide for more information
about this function

Examples
The following example is similar to that for INSERTCHILDXML on page 5-114, but it
adds a third /Owner node before the /Owner node added in the other example. The
output of the query has been formatted for readability.
UPDATE warehouses
SET warehouse_spec = INSERTXMLBEFORE(warehouse_spec,
'/Warehouse/Building/Owner[2]', XMLType('ThirdOwner'))
WHERE warehouse_id = 3;
SELECT warehouse_name,
EXTRACT(warehouse_spec, '/Warehouse/Building/Owner') "Owners"
FROM warehouses
WHERE warehouse_id = 3;
WAREHOUSE_NAME
Owners
----------------------------------- -----------------------------New Jersey
GrandCo
ThirdOwner
LesserCo

Functions

5-119

INSTR

INSTR
Syntax
INSTR
,

INSTRB
INSTRC

,
(

string

,

occurrence

position

substring

)

INSTR2
INSTR4

Purpose
The INSTR functions search string for substring. The search operation is defined as
comparing the substring argument with substrings of string of the same length for
equality until a match is found or there are no more substrings left. Each consecutive
compared substring of string begins one character to the right (for forward searches)
or one character to the left (for backward searches) from the first character of the
previous compared substring. If a substring that is equal to substring is found, then
the function returns an integer indicating the position of the first character of this
substring. If no such substring is found, then the function returns zero.
■

■

position is an nonzero integer indicating the character of string where Oracle
Database begins the search—that is, the position of the first character of the first
substring to compare with substring. If position is negative, then Oracle counts
backward from the end of string and then searches backward from the resulting
position.
occurrence is an integer indicating which occurrence of substring in string
Oracle should search for. The value of occurrence must be positive. If occurrence
is greater than 1, then the database does not return on the first match but continues
comparing consecutive substrings of string, as described above, until match
number occurrence has been found.

INSTR accepts and returns positions in characters as defined by the input character set,
with the first character of string having position 1. INSTRB uses bytes instead of
characters. INSTRC uses Unicode complete characters. INSTR2 uses UCS2 code points.
INSTR4 uses UCS4 code points.
string can be any of the data types CHAR, VARCHAR2, NCHAR, NVARCHAR2, CLOB, or NCLOB.
The exceptions are INSTRC, INSTR2, and INSTR4, which do not allow string to be a
CLOB or NCLOB.
substring can be any of the data types CHAR, VARCHAR2, NCHAR, NVARCHAR2, CLOB, or
NCLOB.
The value returned is of NUMBER data type.
Both position and occurrence must be of data type NUMBER, or any data type that can
be implicitly converted to NUMBER, and must resolve to an integer. The default values of
both position and occurrence are 1, meaning Oracle begins searching at the first
character of string for the first occurrence of substring. The return value is relative to
the beginning of string, regardless of the value of position.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion
5-120 Oracle Database SQL Language Reference

INSTR

Examples
The following example searches the string CORPORATE FLOOR, beginning with the third
character, for the string "OR". It returns the position in CORPORATE FLOOR at which the
second occurrence of "OR" begins:
SELECT INSTR('CORPORATE FLOOR','OR', 3, 2) "Instring"
FROM DUAL;
Instring
---------14

In the next example, Oracle counts backward from the last character to the third
character from the end, which is the first O in FLOOR. Oracle then searches backward for
the second occurrence of OR, and finds that this second occurrence begins with the
second character in the search string :
SELECT INSTR('CORPORATE FLOOR','OR', -3, 2) "Reversed Instring"
FROM DUAL;
Reversed Instring
----------------2

The next example assumes a double-byte database character set.
SELECT INSTRB('CORPORATE FLOOR','OR',5,2) "Instring in bytes"
FROM DUAL;
Instring in bytes
----------------27

Functions

5-121

ITERATION_NUMBER

ITERATION_NUMBER
Syntax
ITERATION_NUMBER

Purpose
The ITERATION_NUMBER function can be used only in the model_clause of the SELECT
statement and then only when ITERATE(number) is specified in the model_rules_
clause. It returns an integer representing the completed iteration through the model
rules. The ITERATION_NUMBER function returns 0 during the first iteration. For each
subsequent iteration, the ITERATION_NUMBER function returns the equivalent of
iteration_number plus one.
model_clause on page 19-28 and "Model Expressions" on
page 6-11 for the syntax and semantics

See Also:

Examples
The following example assigns the sales of the Mouse Pad for the years 1998 and 1999
to the sales of the Mouse Pad for the years 2001 and 2002 respectively:
SELECT country, prod, year, s
FROM sales_view_ref
MODEL
PARTITION BY (country)
DIMENSION BY (prod, year)
MEASURES (sale s)
IGNORE NAV
UNIQUE DIMENSION
RULES UPSERT SEQUENTIAL ORDER ITERATE(2)
(
s['Mouse Pad', 2001 + ITERATION_NUMBER] =
s['Mouse Pad', 1998 + ITERATION_NUMBER]
)
ORDER BY country, prod, year;
COUNTRY
---------France
France
France
France
France
France
France
France
France
Germany
Germany
Germany
Germany
Germany
Germany
Germany
Germany
Germany

PROD
----------------------------------Mouse Pad
Mouse Pad
Mouse Pad
Mouse Pad
Mouse Pad
Standard Mouse
Standard Mouse
Standard Mouse
Standard Mouse
Mouse Pad
Mouse Pad
Mouse Pad
Mouse Pad
Mouse Pad
Standard Mouse
Standard Mouse
Standard Mouse
Standard Mouse

5-122 Oracle Database SQL Language Reference

YEAR
-------1998
1999
2000
2001
2002
1998
1999
2000
2001
1998
1999
2000
2001
2002
1998
1999
2000
2001

S
--------2509.42
3678.69
3000.72
2509.42
3678.69
2390.83
2280.45
1274.31
2164.54
5827.87
8346.44
7375.46
5827.87
8346.44
7116.11
6263.14
2637.31
6456.13

ITERATION_NUMBER

18 rows selected.

The preceding example requires the view sales_view_ref. Refer to "The MODEL
clause: Examples" on page 19-45 to create this view.

Functions

5-123

LAG

LAG
Syntax
,
,
(

RESPECT

default

NULLS

offset

IGNORE

value_expr

)
RESPECT

LAG

,

NULLS
IGNORE
(

,

default

offset

value_expr

)

query_partition_clause
OVER

(

order_by_clause

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions, including valid forms of value_
expr

Purpose
LAG is an analytic function. It provides access to more than one row of a table at the
same time without a self join. Given a series of rows returned from a query and a
position of the cursor, LAG provides access to a row at a given physical offset prior to
that position.
For the optional offset argument, specify an integer that is greater than zero. If you
do not specify offset, then its default is 1. The optional default value is returned if
the offset goes beyond the scope of the window. If you do not specify default, then its
default is null.
{RESPECT | IGNORE} NULLS determines whether null values of value_expr are included
in or eliminated from the calculation. The default is RESPECT NULLS.
You cannot nest analytic functions by using LAG or any other analytic function for
value_expr. However, you can use other built-in function expressions for value_expr.
See Also: "About SQL Expressions" on page 6-1 for information on
valid forms of expr and LEAD on page 5-131

Examples
The following example provides, for each purchasing clerk in the employees table, the
salary of the employee hired just before:
SELECT hire_date, last_name, salary,
LAG(salary, 1, 0 ) OVER (ORDER BY hire_date) AS prev_sal
FROM employees
WHERE job_id = 'PU_CLERK'
ORDER BY hire_date;
HIRE_DATE
--------18-MAY-03
24-JUL-05
24-DEC-05

LAST_NAME
SALARY
PREV_SAL
------------------------- ---------- ---------Khoo
3100
0
Tobias
2800
3100
Baida
2900
2800

5-124 Oracle Database SQL Language Reference

LAG

15-NOV-06 Himuro
10-AUG-07 Colmenares

2600
2500

2900
2600

Functions

5-125

LAST

LAST
Syntax
last::=
aggregate_function

KEEP

,
DESC

FIRST
NULLS

ASC
(

DENSE_RANK

LAST

ORDER

BY

LAST

expr

)

query_partition_clause
OVER

(

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions of the query_partitioning_clause

Purpose
FIRST and LAST are very similar functions. Both are aggregate and analytic functions
that operate on a set of values from a set of rows that rank as the FIRST or LAST with
respect to a given sorting specification. If only one row ranks as FIRST or LAST, then
the aggregate operates on the set with only one element.
Refer to FIRST on page 5-101 for complete information on this function and for
examples of its use.

5-126 Oracle Database SQL Language Reference

LAST_DAY

LAST_DAY
Syntax
LAST_DAY

(

date

)

Purpose
LAST_DAY returns the date of the last day of the month that contains date. The last day
of the month is defined by the session parameter NLS_CALENDAR. The return type is
always DATE, regardless of the data type of date.

Examples
The following statement determines how many days are left in the current month.
SELECT SYSDATE,
LAST_DAY(SYSDATE) "Last",
LAST_DAY(SYSDATE) - SYSDATE "Days Left"
FROM DUAL;
SYSDATE
Last
Days Left
--------- --------- ---------30-MAY-09 31-MAY-09
1

The following example adds 5 months to the hire date of each employee to give an
evaluation date:
SELECT last_name, hire_date,
TO_CHAR(ADD_MONTHS(LAST_DAY(hire_date), 5)) "Eval Date"
FROM employees
ORDER BY last_name, hire_date;
LAST_NAME
------------------------Abel
Ande
Atkinson
Austin
Baer
Baida
Banda
Bates
. . .

HIRE_DATE
--------11-MAY-04
24-MAR-08
30-OCT-05
25-JUN-05
07-JUN-02
24-DEC-05
21-APR-08
24-MAR-07

Eval Date
--------31-OCT-04
31-AUG-08
31-MAR-06
30-NOV-05
30-NOV-02
31-MAY-06
30-SEP-08
31-AUG-07

Functions

5-127

LAST_VALUE

LAST_VALUE
Syntax
RESPECT
NULLS
IGNORE
(

expr

)
RESPECT

LAST_VALUE

OVER

NULLS

(

analytic_clause

IGNORE
(

expr

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions, including valid forms of expr

Purpose
LAST_VALUE is an analytic function that is useful for data densification. It returns the
last value in an ordered set of values.

The two forms of this syntax have the same behavior. The top
branch is the ANSI format, which Oracle recommends. The bottom
branch is deprecated but is supported for backward compatibility.

Note:

{RESPECT | IGNORE} NULLS determines whether null values of expr are included in or
eliminated from the calculation. The default is RESPECT NULLS. If the last value in the
set is null, then the function returns NULL unless you specify IGNORE NULLS. If you
specify IGNORE NULLS, then LAST_VALUE returns the last non-null value in the set, or
NULL if all values are null. Refer to "Using Partitioned Outer Joins: Examples" on
page 19-52 for an example of data densification.
You cannot nest analytic functions by using LAST_VALUE or any other analytic function
for expr. However, you can use other built-in function expressions for expr. Refer to
"About SQL Expressions" on page 6-1 for information on valid forms of expr.
If you omit the windowing_clause of the analytic_clause, it defaults to RANGE
BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW. This default sometimes returns an
unexpected value, because the last value in the window is at the bottom of the
window, which is not fixed. It keeps changing as the current row changes. For
expected results, specify the windowing_clause as RANGE BETWEEN UNBOUNDED
PRECEDING AND UNBOUNDED FOLLOWING. Alternatively, you can specify the windowing_
clause as RANGE BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING.

Examples
The following example returns, for each row, the hire date of the employee earning the
highest salary:
SELECT last_name, salary, hire_date,
LAST_VALUE(hire_date)
OVER (ORDER BY salary ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED
FOLLOWING) AS lv
FROM (SELECT * FROM employees
WHERE department_id = 90

5-128 Oracle Database SQL Language Reference

)

LAST_VALUE

ORDER BY hire_date);
LAST_NAME
SALARY HIRE_DATE LV
--------------- ---------- --------- --------De Haan
17000 13-JAN-01 17-JUN-03
Kochhar
17000 21-SEP-05 17-JUN-03
King
24000 17-JUN-03 17-JUN-03

This example illustrates the nondeterministic nature of the LAST_VALUE function.
Kochhar and De Haan have the same salary, so they are in adjacent rows. Kochhar
appears first because the rows in the subquery are ordered by hire_date. However, if
the rows are ordered by hire_date in descending order, as in the next example, then
the function returns a different value:
SELECT last_name, salary, hire_date,
LAST_VALUE(hire_date)
OVER (ORDER BY salary ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED
FOLLOWING) AS lv
FROM (SELECT * FROM employees
WHERE department_id = 90
ORDER BY hire_date DESC);
LAST_NAME
SALARY HIRE_DATE LV
--------------- ---------- --------- --------Kochhar
17000 21-SEP-05 17-JUN-03
De Haan
17000 13-JAN-01 17-JUN-03
King
24000 17-JUN-03 17-JUN-03

The following two examples show how to make the LAST_VALUE function deterministic
by ordering on a unique key. By ordering within the function by both salary and
hire_date, you can ensure the same result regardless of the ordering in the subquery.
SELECT last_name, salary, hire_date,
LAST_VALUE(hire_date)
OVER (ORDER BY salary, hire_date ROWS BETWEEN UNBOUNDED PRECEDING AND
UNBOUNDED FOLLOWING) AS lv
FROM (SELECT * FROM employees
WHERE department_id = 90
ORDER BY hire_date)
ORDER BY last_name, salary, hire_date;
LAST_NAME
SALARY HIRE_DATE LV
--------------- ---------- --------- --------De Haan
17000 13-JAN-01 17-JUN-03
King
24000 17-JUN-03 17-JUN-03
Kochhar
17000 21-SEP-05 17-JUN-03
SELECT last_name, salary, hire_date,
LAST_VALUE(hire_date)
OVER (ORDER BY salary, hire_date ROWS BETWEEN UNBOUNDED PRECEDING AND
UNBOUNDED FOLLOWING) AS lv
FROM (SELECT * FROM employees
WHERE department_id = 90
ORDER BY hire_date DESC)
ORDER BY last_name, salary, hire_date;
LAST_NAME
SALARY HIRE_DATE LV
--------------- ---------- --------- --------De Haan
17000 13-JAN-01 17-JUN-03
King
24000 17-JUN-03 17-JUN-03

Functions

5-129

LAST_VALUE

Kochhar

17000 21-SEP-05 17-JUN-03

When you use a logical offset (RANGE instead of ROWS), the function is deterministic.
When duplicates are found for the ORDER BY expression, the LAST_VALUE is the highest
value of expr:
SELECT last_name, salary, hire_date,
LAST_VALUE(hire_date)
OVER (ORDER BY salary RANGE BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED
FOLLOWING) AS lv
FROM (SELECT * FROM employees
WHERE department_id = 90
ORDER BY hire_date DESC);
LAST_NAME
SALARY HIRE_DATE LV
------------------------- ---------- --------- --------De Haan
17000 13-JAN-01 17-JUN-03
Kochhar
17000 21-SEP-05 17-JUN-03
King
24000 17-JUN-03 17-JUN-03

5-130 Oracle Database SQL Language Reference

LEAD

LEAD
Syntax
,
,
(

RESPECT

default

NULLS

offset

IGNORE

value_expr

)
RESPECT

LEAD

,

NULLS
IGNORE
(

,

default

offset

value_expr

)

query_partition_clause
OVER

(

order_by_clause

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions, including valid forms of value_
expr

Purpose
LEAD is an analytic function. It provides access to more than one row of a table at the
same time without a self join. Given a series of rows returned from a query and a
position of the cursor, LEAD provides access to a row at a given physical offset beyond
that position.
If you do not specify offset, then its default is 1. The optional default value is
returned if the offset goes beyond the scope of the table. If you do not specify default,
then its default value is null.
{RESPECT | IGNORE} NULLS determines whether null values of value_expr are included
in or eliminated from the calculation. The default is RESPECT NULLS.
You cannot nest analytic functions by using LEAD or any other analytic function for
value_expr. However, you can use other built-in function expressions for value_expr.
See Also: "About SQL Expressions" on page 6-1 for information on
valid forms of expr and LAG on page 5-124

Examples
The following example provides, for each employee in Department 30 in the
employees table, the hire date of the employee hired just after:
SELECT hire_date, last_name,
LEAD(hire_date, 1) OVER (ORDER BY hire_date) AS "Next Hired"
FROM employees
WHERE department_id = 30
ORDER BY hire_date;
HIRE_DATE
--------07-DEC-02
18-MAY-03
24-JUL-05
24-DEC-05
15-NOV-06

LAST_NAME
------------------------Raphaely
Khoo
Tobias
Baida
Himuro

Next Hired
---------18-MAY-03
24-JUL-05
24-DEC-05
15-NOV-06
10-AUG-07
Functions

5-131

LEAD

10-AUG-07 Colmenares

5-132 Oracle Database SQL Language Reference

LEAST

LEAST
Syntax
,
LEAST

(

expr

)

Purpose
LEAST returns the least of a list of one or more expressions. Oracle Database uses the
first expr to determine the return type. If the first expr is numeric, then Oracle
determines the argument with the highest numeric precedence, implicitly converts the
remaining arguments to that data type before the comparison, and returns that data
type. If the first expr is not numeric, then each expr after the first is implicitly
converted to the data type of the first expr before the comparison.
Oracle Database compares each expr using nonpadded comparison semantics. The
comparison is binary by default and is linguistic if the NLS_COMP parameter is set to
LINGUISTIC and the NLS_SORT parameter has a setting other than BINARY. Character
comparison is based on the numerical codes of the characters in the database character
set and is performed on whole strings treated as one sequence of bytes, rather than
character by character. If the value returned by this function is character data, then its
data type is VARCHAR2 if the first expr is a character data type and NVARCHAR2 if the first
expr is a national character data type.
See Also:
■

■

■

"Data Type Comparison Rules" on page 3-36 for more information
on character comparison
Table 3–10, " Implicit Type Conversion Matrix" on page 3-40 for
more information on implicit conversion and "Floating-Point
Numbers" on page 3-12 for information on binary-float
comparison semantics
"GREATEST" on page 5-107, which returns the greatest of a list of
one or more expressions

Examples
The following statement selects the string with the least value:
SELECT LEAST('HARRY','HARRIOT','HAROLD') "Least"
FROM DUAL;
Least
-----HAROLD

In the following statement, the first argument is numeric. Oracle Database determines
that the argument with the highest numeric precedence is the third argument, converts
the remaining arguments to the data type of the third argument, and returns the least
value as that data type:
SELECT LEAST (1, '2.1', '.000832') "Least"
FROM DUAL;
Least

Functions

5-133

LEAST

------.000832

5-134 Oracle Database SQL Language Reference

LENGTH

LENGTH
Syntax
length::=
LENGTH
LENGTHB
LENGTHC

(

char

)

LENGTH2
LENGTH4

Purpose
The LENGTH functions return the length of char. LENGTH calculates length using
characters as defined by the input character set. LENGTHB uses bytes instead of
characters. LENGTHC uses Unicode complete characters. LENGTH2 uses UCS2 code
points. LENGTH4 uses UCS4 code points.
char can be any of the data types CHAR, VARCHAR2, NCHAR, NVARCHAR2, CLOB, or NCLOB.
The exceptions are LENGTHC, LENGTH2, and LENGTH4, which do not allow char to be a
CLOB or NCLOB. The return value is of data type NUMBER. If char has data type CHAR, then
the length includes all trailing blanks. If char is null, then this function returns null.
Restriction on LENGTHB The LENGTHB function is supported for single-byte LOBs

only. It cannot be used with CLOB and NCLOB data in a multibyte character set.

Examples
The following example uses the LENGTH function using a single-byte database character
set:
SELECT LENGTH('CANDIDE') "Length in characters"
FROM DUAL;
Length in characters
-------------------7

The next example assumes a double-byte database character set.
SELECT LENGTHB ('CANDIDE') "Length in bytes"
FROM DUAL;
Length in bytes
--------------14

Functions

5-135

LISTAGG

LISTAGG
Syntax
,
LISTAGG

OVER

(

’

delimiter

measure_expr

’
)

WITHIN

GROUP

(

order_by_clause

)

query_partition_clause

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions of the ORDER BY clause and OVER
clause

Purpose
For a specified measure, LISTAGG orders data within each group specified in the ORDER
BY clause and then concatenates the values of the measure column.
■

■

■

As a single-set aggregate function, LISTAGG operates on all rows and returns a
single output row.
As a group-set aggregate, the function operates on and returns an output row for
each group defined by the GROUP BY clause.
As an analytic function, LISTAGG partitions the query result set into groups based
on one or more expression in the query_partition_clause.

The arguments to the function are subject to the following rules:
■

■

■

The measure_expr can be any expression. Null values in the measure column are
ignored.
The delimiter_expr designates the string that is to separate the measure values.
This clause is optional and defaults to NULL.
The order_by_clause determines the order in which the concatenated values are
returned. The function is deterministic only if the ORDER BY column list achieved
unique ordering.

The return data type is RAW if the measure column is RAW; otherwise the return value is
VARCHAR2.

Aggregate Examples
The following single-set aggregate example lists all of the employees in Department 30
in the hr.employees table, ordered by hire date and last name:
SELECT LISTAGG(last_name, '; ')
WITHIN GROUP (ORDER BY hire_date, last_name) "Emp_list",
MIN(hire_date) "Earliest"
FROM employees
WHERE department_id = 30;
Emp_list
Earliest
------------------------------------------------------------ --------Raphaely; Khoo; Tobias; Baida; Himuro; Colmenares
07-DEC-02

The following group-set aggregate example lists, for each department ID in the
hr.employees table, the employees in that department in order of their hire date:
5-136 Oracle Database SQL Language Reference

LISTAGG

SELECT department_id "Dept.",
LISTAGG(last_name, '; ') WITHIN GROUP (ORDER BY hire_date) "Employees"
FROM employees
GROUP BY department_id
ORDER BY department_id;
Dept. Employees
------ -----------------------------------------------------------10 Whalen
20 Hartstein; Fay
30 Raphaely; Khoo; Tobias; Baida; Himuro; Colmenares
40 Mavris
50 Kaufling; Ladwig; Rajs; Sarchand; Bell; Mallin; Weiss; Davie
s; Marlow; Bull; Everett; Fripp; Chung; Nayer; Dilly; Bissot
; Vollman; Stiles; Atkinson; Taylor; Seo; Fleaur; Matos; Pat
el; Walsh; Feeney; Dellinger; McCain; Vargas; Gates; Rogers;
Mikkilineni; Landry; Cabrio; Jones; Olson; OConnell; Sulliv
an; Mourgos; Gee; Perkins; Grant; Geoni; Philtanker; Markle
60 Austin; Hunold; Pataballa; Lorentz; Ernst
70 Baer
. . .

Analytic Example
The following analytic example shows, for each employee hired earlier than
September 1, 2003, the employee's department, hire date, and all other employees in
that department also hired before September 1, 2003:
SELECT department_id "Dept", hire_date "Date", last_name "Name",
LISTAGG(last_name, '; ') WITHIN GROUP (ORDER BY hire_date, last_name)
OVER (PARTITION BY department_id) as "Emp_list"
FROM employees
WHERE hire_date < '01-SEP-2003'
ORDER BY "Dept", "Date", "Name";
Dept
----30
30
40
50
50
70
90
90
100
100
110
110

Date
--------07-DEC-02
18-MAY-03
07-JUN-02
01-MAY-03
14-JUL-03
07-JUN-02
13-JAN-01
17-JUN-03
16-AUG-02
17-AUG-02
07-JUN-02
07-JUN-02

Name
--------------Raphaely
Khoo
Mavris
Kaufling
Ladwig
Baer
De Haan
King
Faviet
Greenberg
Gietz
Higgins

Emp_list
--------------------------------------------Raphaely; Khoo
Raphaely; Khoo
Mavris
Kaufling; Ladwig
Kaufling; Ladwig
Baer
De Haan; King
De Haan; King
Faviet; Greenberg
Faviet; Greenberg
Gietz; Higgins
Gietz; Higgins

Functions

5-137

LN

LN
Syntax
LN

(

n

)

Purpose
LN returns the natural logarithm of n, where n is greater than 0.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. If the argument is
BINARY_FLOAT, then the function returns BINARY_DOUBLE. Otherwise the function
returns the same numeric data type as the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

Examples
The following example returns the natural logarithm of 95:
SELECT LN(95) "Natural log of 95"
FROM DUAL;
Natural log of 95
----------------4.55387689

5-138 Oracle Database SQL Language Reference

LNNVL

LNNVL
Syntax
LNNVL

(

condition

)

Purpose
LNNVL provides a concise way to evaluate a condition when one or both operands of
the condition may be null. The function can be used in the WHERE clause of a query, or
as the WHEN condition in a searched CASE expression. It takes as an argument a
condition and returns TRUE if the condition is FALSE or UNKNOWN and FALSE if the
condition is TRUE. LNNVL can be used anywhere a scalar expression can appear, even in
contexts where the IS [NOT] NULL, AND, or OR conditions are not valid but would
otherwise be required to account for potential nulls.
Oracle Database sometimes uses the LNNVL function internally in this way to rewrite
NOT IN conditions as NOT EXISTS conditions. In such cases, output from EXPLAIN PLAN
shows this operation in the plan table output. The condition can evaluate any scalar
values but cannot be a compound condition containing AND, OR, or BETWEEN.
The table that follows shows what LNNVL returns given that a = 2 and b is null.
Condition

Truth of Condition

LNNVL Return Value

a=1

FALSE

TRUE

a=2

TRUE

FALSE

a IS NULL

FALSE

TRUE

b=1

UNKNOWN

TRUE

b IS NULL

TRUE

FALSE

a=b

UNKNOWN

TRUE

Examples
Suppose that you want to know the number of employees with commission rates of
less than 20%, including employees who do not receive commissions. The following
query returns only employees who actually receive a commission of less than 20%:
SELECT COUNT(*)
FROM employees
WHERE commission_pct < .2;
COUNT(*)
---------11

To include the 72 employees who receive no commission at all, you could rewrite the
query using the LNNVL function as follows:
SELECT COUNT(*)
FROM employees
WHERE LNNVL(commission_pct >= .2);
COUNT(*)
---------83
Functions

5-139

LOCALTIMESTAMP

LOCALTIMESTAMP
Syntax
(

timestamp_precision

)

LOCALTIMESTAMP

Purpose
LOCALTIMESTAMP returns the current date and time in the session time zone in a value
of data type TIMESTAMP. The difference between this function and CURRENT_TIMESTAMP
is that LOCALTIMESTAMP returns a TIMESTAMP value while CURRENT_TIMESTAMP returns a
TIMESTAMP WITH TIME ZONE value.
The optional argument timestamp_precision specifies the fractional second precision
of the time value returned.
CURRENT_TIMESTAMP on page 5-72, "TIMESTAMP
Data Type" on page 3-18, and "TIMESTAMP WITH TIME ZONE Data
Type" on page 3-18

See Also:

Examples
This example illustrates the difference between LOCALTIMESTAMP and CURRENT_
TIMESTAMP:
ALTER SESSION SET TIME_ZONE = '-5:00';
SELECT CURRENT_TIMESTAMP, LOCALTIMESTAMP FROM DUAL;
CURRENT_TIMESTAMP
LOCALTIMESTAMP
------------------------------------------------------------------04-APR-00 01.27.18.999220 PM -05:00 04-APR-00 01.27.19 PM
ALTER SESSION SET TIME_ZONE = '-8:00';
SELECT CURRENT_TIMESTAMP, LOCALTIMESTAMP FROM DUAL;
CURRENT_TIMESTAMP
----------------------------------04-APR-00 10.27.45.132474 AM -08:00

LOCALTIMESTAMP
-----------------------------04-APR-00 10.27.451 AM

When you use the LOCALTIMESTAMP with a format mask, take care that the format mask
matches the value returned by the function. For example, consider the following table:
CREATE TABLE local_test (col1 TIMESTAMP WITH LOCAL TIME ZONE);

The following statement fails because the mask does not include the TIME ZONE portion
of the return type of the function:
INSERT INTO local_test
VALUES (TO_TIMESTAMP(LOCALTIMESTAMP, 'DD-MON-RR HH.MI.SSXFF'));

The following statement uses the correct format mask to match the return type of
LOCALTIMESTAMP:
INSERT INTO local_test
VALUES (TO_TIMESTAMP(LOCALTIMESTAMP, 'DD-MON-RR HH.MI.SSXFF PM'));

5-140 Oracle Database SQL Language Reference

LOG

LOG
Syntax
LOG

(

n2

,

n1

)

Purpose
LOG returns the logarithm, base n2, of n1. The base n1 can be any positive value other
than 0 or 1 and n2 can be any positive value.
This function takes as arguments any numeric data type or any nonnumeric data type
that can be implicitly converted to a numeric data type. If any argument is BINARY_
FLOAT or BINARY_DOUBLE, then the function returns BINARY_DOUBLE. Otherwise the
function returns NUMBER.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

Examples
The following example returns the log of 100:
SELECT LOG(10,100) "Log base 10 of 100"
FROM DUAL;
Log base 10 of 100
-----------------2

Functions

5-141

LOWER

LOWER
Syntax
LOWER

(

char

)

Purpose
LOWER returns char, with all letters lowercase. char can be any of the data types CHAR,
VARCHAR2, NCHAR, NVARCHAR2, CLOB, or NCLOB. The return value is the same data type as
char. The database sets the case of the characters based on the binary mapping defined
for the underlying character set. For linguistic-sensitive lowercase, refer to NLS_
LOWER on page 5-162.

Examples
The following example returns a string in lowercase:
SELECT LOWER('MR. SCOTT MCMILLAN') "Lowercase"
FROM DUAL;
Lowercase
-------------------mr. scott mcmillan

5-142 Oracle Database SQL Language Reference

LPAD

LPAD
Syntax
,
LPAD

(

expr1

,

expr2

n

)

Purpose
LPAD returns expr1, left-padded to length n characters with the sequence of characters
in expr2. This function is useful for formatting the output of a query.
Both expr1 and expr2 can be any of the data types CHAR, VARCHAR2, NCHAR, NVARCHAR2,
CLOB, or NCLOB. The string returned is of VARCHAR2 data type if expr1 is a character data
type, NVARCHAR2 if expr1 is a national character data type, and a LOB if expr1 is a LOB
data type. The string returned is in the same character set as expr1. The argument n
must be a NUMBER integer or a value that can be implicitly converted to a NUMBER
integer.
If you do not specify expr2, then the default is a single blank. If expr1 is longer than n,
then this function returns the portion of expr1 that fits in n.
The argument n is the total length of the return value as it is displayed on your
terminal screen. In most character sets, this is also the number of characters in the
return value. However, in some multibyte character sets, the display length of a
character string can differ from the number of characters in the string.

Examples
The following example left-pads a string with the asterisk (*) and period (.) characters:
SELECT LPAD('Page 1',15,'*.') "LPAD example"
FROM DUAL;
LPAD example
--------------*.*.*.*.*Page 1

Functions

5-143

LTRIM

LTRIM
Syntax
,
LTRIM

(

set

char

)

Purpose
LTRIM removes from the left end of char all of the characters contained in set. If you
do not specify set, then it defaults to a single blank. If char is a character literal, then
you must enclose it in single quotation marks. Oracle Database begins scanning char
from its first character and removes all characters that appear in set until reaching a
character not in set and then returns the result.
Both char and set can be any of the data types CHAR, VARCHAR2, NCHAR, NVARCHAR2,
CLOB, or NCLOB. The string returned is of VARCHAR2 data type if char is a character data
type, NVARCHAR2 if char is a national character data type, and a LOB if char is a LOB
data type.
See Also:

RTRIM on page 5-241

Examples
The following example trims all the left-most occurrences of less than sign (<), greater
than sign (>) , and equal sign (=) from a string:
SELECT LTRIM('<=====>BROWNING<=====>', '<>=') "LTRIM Example"
FROM DUAL;
LTRIM Example
--------------BROWNING<=====>

5-144 Oracle Database SQL Language Reference

MAKE_REF

MAKE_REF
Syntax
,
table
MAKE_REF

(

,

key

)

view

Purpose
MAKE_REF creates a REF to a row of an object view or a row in an object table whose
object identifier is primary key based. This function is useful, for example, if you are
creating an object view
See Also: Oracle Database Object-Relational Developer's Guide for more
information about object views and DEREF on page 5-85

Examples
The sample schema oe contains an object view oc_inventories based on inventory_
typ. The object identifier is product_id. The following example creates a REF to the
row in the oc_inventories object view with a product_id of 3003:
SELECT MAKE_REF (oc_inventories, 3003)
FROM DUAL;
MAKE_REF(OC_INVENTORIES,3003)
-----------------------------------------------------------------00004A038A0046857C14617141109EE03408002082543600000014260100010001
00290090606002A00078401FE0000000B03C21F040000000000000000000000000
0000000000

Functions

5-145

MAX

MAX
Syntax
DISTINCT
ALL
MAX

(

OVER
expr

(

analytic_clause

)

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions

Purpose
MAX returns maximum value of expr. You can use it as an aggregate or analytic
function.
See Also: "About SQL Expressions" on page 6-1 for information on
valid forms of expr, "Floating-Point Numbers" on page 3-12 for
information on binary-float comparison semantics, and "Aggregate
Functions" on page 5-10

Aggregate Example
The following example determines the highest salary in the hr.employees table:
SELECT MAX(salary) "Maximum"
FROM employees;
Maximum
---------24000

Analytic Examples
The following example calculates, for each employee, the highest salary of the
employees reporting to the same manager as the employee.
SELECT manager_id, last_name, salary,
MAX(salary) OVER (PARTITION BY manager_id) AS mgr_max
FROM employees
ORDER BY manager_id, last_name, salary;
MANAGER_ID
---------100
100
100
100
100
100
100
. . .

LAST_NAME
SALARY
MGR_MAX
------------------------- ---------- ---------Cambrault
11000
17000
De Haan
17000
17000
Errazuriz
12000
17000
Fripp
8200
17000
Hartstein
13000
17000
Kaufling
7900
17000
Kochhar
17000
17000

If you enclose this query in the parent query with a predicate, then you can determine
the employee who makes the highest salary in each department:
SELECT manager_id, last_name, salary
FROM (SELECT manager_id, last_name, salary,

5-146 Oracle Database SQL Language Reference

MAX

MAX(salary) OVER (PARTITION BY manager_id) AS rmax_sal
FROM employees)
WHERE salary = rmax_sal
ORDER BY manager_id, last_name, salary;
MANAGER_ID
---------100
100
101
101
102
103
108
114
120
120
121
122
123
124
145
146
147
148
149
201
205

LAST_NAME
SALARY
------------------------- ---------De Haan
17000
Kochhar
17000
Greenberg
12008
Higgins
12008
Hunold
9000
Ernst
6000
Faviet
9000
Khoo
3100
Nayer
3200
Taylor
3200
Sarchand
4200
Chung
3800
Bell
4000
Rajs
3500
Tucker
10000
King
10000
Vishney
10500
Ozer
11500
Abel
11000
Fay
6000
Gietz
8300
King
24000

22 rows selected.

Functions

5-147

MEDIAN

MEDIAN
Syntax
OVER
MEDIAN

(

expr

(

query_partition_clause

)

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions

Purpose
MEDIAN is an inverse distribution function that assumes a continuous distribution
model. It takes a numeric or datetime value and returns the middle value or an
interpolated value that would be the middle value once the values are sorted. Nulls
are ignored in the calculation.
This function takes as arguments any numeric data type or any nonnumeric data type
that can be implicitly converted to a numeric data type. If you specify only expr, then
the function returns the same data type as the numeric data type of the argument. If
you specify the OVER clause, then Oracle Database determines the argument with the
highest numeric precedence, implicitly converts the remaining arguments to that data
type, and returns that data type.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion and "Numeric
Precedence" on page 3-14 for information on numeric precedence

The result of MEDIAN is computed by first ordering the rows. Using N as the number of
rows in the group, Oracle calculates the row number (RN) of interest with the formula
RN = (1 + (0.5*(N-1)). The final result of the aggregate function is computed by linear
interpolation between the values from rows at row numbers CRN = CEILING(RN) and
FRN = FLOOR(RN).
The final result will be:
if (CRN = FRN = RN) then
(value of expression from row at RN)
else
(CRN - RN) * (value of expression for row at FRN) +
(RN - FRN) * (value of expression for row at CRN)

You can use MEDIAN as an analytic function. You can specify only the query_
partition_clause in its OVER clause. It returns, for each row, the value that would fall
in the middle among a set of values within each partition.
Compare this function with these functions:
■

■

PERCENTILE_CONT on page 5-182, which returns, for a given percentile, the
value that corresponds to that percentile by way of interpolation. MEDIAN is the
specific case of PERCENTILE_CONT where the percentile value defaults to 0.5.
PERCENTILE_DISC on page 5-185, which is useful for finding values for a given
percentile without interpolation.

5-148 Oracle Database SQL Language Reference

MEDIAN

Aggregate Example
The following query returns the median salary for each department in the
hr.employees table:
SELECT department_id, MEDIAN(salary)
FROM employees
GROUP BY department_id
ORDER BY department_id;
DEPARTMENT_ID MEDIAN(SALARY)
------------- -------------10
4400
20
9500
30
2850
40
6500
50
3100
60
4800
70
10000
80
8900
90
17000
100
8000
110
10154
7000

Analytic Example
The following query returns the median salary for each manager in a subset of
departments in the hr.employees table:
SELECT manager_id, employee_id, salary,
MEDIAN(salary) OVER (PARTITION BY manager_id) "Median by Mgr"
FROM employees
WHERE department_id > 60
ORDER BY manager_id, employee_id;
MANAGER_ID EMPLOYEE_ID
SALARY Median by Mgr
---------- ----------- ---------- ------------100
101
17000
13500
100
102
17000
13500
100
145
14000
13500
100
146
13500
13500
100
147
12000
13500
100
148
11000
13500
100
149
10500
13500
101
108
12008
12008
101
204
10000
12008
101
205
12008
12008
108
109
9000
7800
108
110
8200
7800
108
111
7700
7800
108
112
7800
7800
108
113
6900
7800
145
150
10000
8500
145
151
9500
8500
145
152
9000
8500
. . .

Functions

5-149

MIN

MIN
Syntax
DISTINCT
ALL
MIN

(

OVER
expr

(

analytic_clause

)

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions

Purpose
MIN returns minimum value of expr. You can use it as an aggregate or analytic
function.
See Also: "About SQL Expressions" on page 6-1 for information on
valid forms of expr, "Floating-Point Numbers" on page 3-12 for
information on binary-float comparison semantics, and "Aggregate
Functions" on page 5-10

Aggregate Example
The following statement returns the earliest hire date in the hr.employees table:
SELECT MIN(hire_date) "Earliest"
FROM employees;
Earliest
--------13-JAN-01

Analytic Example
The following example determines, for each employee, the employees who were hired
on or before the same date as the employee. It then determines the subset of
employees reporting to the same manager as the employee, and returns the lowest
salary in that subset.
SELECT manager_id, last_name, hire_date, salary,
MIN(salary) OVER(PARTITION BY manager_id ORDER BY hire_date
RANGE UNBOUNDED PRECEDING) AS p_cmin
FROM employees
ORDER BY manager_id, last_name, hire_date, salary;
MANAGER_ID
---------100
100
100
100
100
100
100
100
100
100
100

LAST_NAME
------------------------Cambrault
De Haan
Errazuriz
Fripp
Hartstein
Kaufling
Kochhar
Mourgos
Partners
Raphaely
Russell

5-150 Oracle Database SQL Language Reference

HIRE_DATE
SALARY
P_CMIN
--------- ---------- ---------15-OCT-07
11000
6500
13-JAN-01
17000
17000
10-MAR-05
12000
7900
10-APR-05
8200
7900
17-FEB-04
13000
7900
01-MAY-03
7900
7900
21-SEP-05
17000
7900
16-NOV-07
5800
5800
05-JAN-05
13500
7900
07-DEC-02
11000
11000
01-OCT-04
14000
7900

MIN

. . .

Functions

5-151

MOD

MOD
Syntax
MOD

(

n2

,

n1

)

Purpose
MOD returns the remainder of n2 divided by n1. Returns n2 if n1 is 0.
This function takes as arguments any numeric data type or any nonnumeric data type
that can be implicitly converted to a numeric data type. Oracle determines the
argument with the highest numeric precedence, implicitly converts the remaining
arguments to that data type, and returns that data type.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion and "Numeric
Precedence" on page 3-14 for information on numeric precedence

Examples
The following example returns the remainder of 11 divided by 4:
SELECT MOD(11,4) "Modulus"
FROM DUAL;
Modulus
---------3

This function behaves differently from the classical mathematical modulus function
when m is negative. The classical modulus can be expressed using the MOD function
with this formula:
n2 - n1 * FLOOR(n2/n1)

The following table illustrates the difference between the MOD function and the classical
modulus:
n2

n1

MOD(n2,n1)

Classical Modulus

11

4

3

3

11

-4

3

-1

-11

4

-3

1

-11

-4

-3

-3

FLOOR on page 5-105 and REMAINDER on page 5-231,
which is similar to MOD, but uses ROUND in its formula instead of FLOOR

See Also:

5-152 Oracle Database SQL Language Reference

MONTHS_BETWEEN

MONTHS_BETWEEN
Syntax
MONTHS_BETWEEN

(

date1

,

date2

)

Purpose
MONTHS_BETWEEN returns number of months between dates date1 and date2. The
month and the last day of the month are defined by the parameter NLS_CALENDAR. If
date1 is later than date2, then the result is positive. If date1 is earlier than date2, then
the result is negative. If date1 and date2 are either the same days of the month or both
last days of months, then the result is always an integer. Otherwise Oracle Database
calculates the fractional portion of the result based on a 31-day month and considers
the difference in time components date1 and date2.

Examples
The following example calculates the months between two dates:
SELECT MONTHS_BETWEEN
(TO_DATE('02-02-1995','MM-DD-YYYY'),
TO_DATE('01-01-1995','MM-DD-YYYY') ) "Months"
FROM DUAL;
Months
---------1.03225806

Functions

5-153

NANVL

NANVL
Syntax
NANVL

(

n2

,

n1

)

Purpose
The NANVL function is useful only for floating-point numbers of type BINARY_FLOAT or
BINARY_DOUBLE. It instructs Oracle Database to return an alternative value n1 if the
input value n2 is NaN (not a number). If n2 is not NaN, then Oracle returns n2.
This function takes as arguments any numeric data type or any nonnumeric data type
that can be implicitly converted to a numeric data type. Oracle determines the
argument with the highest numeric precedence, implicitly converts the remaining
arguments to that data type, and returns that data type.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion, "Floating-Point
Numbers" on page 3-12 for information on binary-float comparison
semantics, and "Numeric Precedence" on page 3-14 for information on
numeric precedence

Examples
Using table float_point_demo created for TO_BINARY_DOUBLE on page 5-297,
insert a second entry into the table:
INSERT INTO float_point_demo
VALUES (0,'NaN','NaN');
SELECT *
FROM float_point_demo;
DEC_NUM BIN_DOUBLE BIN_FLOAT
---------- ---------- ---------1234.56 1.235E+003 1.235E+003
0
Nan
Nan

The following example returns bin_float if it is a number. Otherwise, 0 is returned.
SELECT bin_float, NANVL(bin_float,0)
FROM float_point_demo;
BIN_FLOAT NANVL(BIN_FLOAT,0)
---------- -----------------1.235E+003
1.235E+003
Nan
0

5-154 Oracle Database SQL Language Reference

NCHR

NCHR
Syntax
NCHR

(

number

)

Purpose
NCHR returns the character having the binary equivalent to number in the national
character set. The value returned is always NVARCHAR2. This function is equivalent to
using the CHR function with the USING NCHAR_CS clause.
This function takes as an argument a NUMBER value, or any value that can be implicitly
converted to NUMBER, and returns a character.
See Also:

CHR on page 5-40

Examples
The following examples return the nchar character 187:
SELECT NCHR(187)
FROM DUAL;
N
>
SELECT CHR(187 USING NCHAR_CS)
FROM DUAL;
C
>

Functions

5-155

NEW_TIME

NEW_TIME
Syntax
NEW_TIME

(

date

,

timezone1

,

timezone2

)

Purpose
NEW_TIME returns the date and time in time zone timezone2 when date and time in
time zone timezone1 are date. Before using this function, you must set the NLS_DATE_
FORMAT parameter to display 24-hour time. The return type is always DATE, regardless
of the data type of date.
Note: This function takes as input only a limited number of time
zones. You can have access to a much greater number of time zones by
combining the FROM_TZ function and the datetime expression. See
FROM_TZ on page 5-106 and the example for "Datetime Expressions"
on page 6-8.

The arguments timezone1 and timezone2 can be any of these text strings:
■

AST, ADT: Atlantic Standard or Daylight Time

■

BST, BDT: Bering Standard or Daylight Time

■

CST, CDT: Central Standard or Daylight Time

■

EST, EDT: Eastern Standard or Daylight Time

■

GMT: Greenwich Mean Time

■

HST, HDT: Alaska-Hawaii Standard Time or Daylight Time.

■

MST, MDT: Mountain Standard or Daylight Time

■

NST: Newfoundland Standard Time

■

PST, PDT: Pacific Standard or Daylight Time

■

YST, YDT: Yukon Standard or Daylight Time

Examples
The following example returns an Atlantic Standard time, given the Pacific Standard
time equivalent:
ALTER SESSION SET NLS_DATE_FORMAT = 'DD-MON-YYYY HH24:MI:SS';
SELECT NEW_TIME(TO_DATE('11-10-09 01:23:45', 'MM-DD-YY HH24:MI:SS'), 'AST', 'PST')
"New Date and Time"
FROM DUAL;
New Date and Time
-------------------09-NOV-2009 21:23:45

5-156 Oracle Database SQL Language Reference

NEXT_DAY

NEXT_DAY
Syntax
NEXT_DAY

(

date

,

char

)

Purpose
NEXT_DAY returns the date of the first weekday named by char that is later than the
date date. The return type is always DATE, regardless of the data type of date. The
argument char must be a day of the week in the date language of your session, either
the full name or the abbreviation. The minimum number of letters required is the
number of letters in the abbreviated version. Any characters immediately following
the valid abbreviation are ignored. The return value has the same hours, minutes, and
seconds component as the argument date.

Examples
This example returns the date of the next Tuesday after October 15, 2009:
SELECT NEXT_DAY('15-OCT-2009','TUESDAY') "NEXT DAY"
FROM DUAL;
NEXT DAY
-------------------20-OCT-2009 00:00:00

Functions

5-157

NLS_CHARSET_DECL_LEN

NLS_CHARSET_DECL_LEN
Syntax
NLS_CHARSET_DECL_LEN

(

byte_count

,

char_set_id

)

Purpose
NLS_CHARSET_DECL_LEN returns the declaration length (in number of characters) of an
NCHAR column. The byte_count argument is the width of the column. The char_set_id
argument is the character set ID of the column.

Examples
The following example returns the number of characters that are in a 200-byte column
when you are using a multibyte character set:
SELECT NLS_CHARSET_DECL_LEN(200, nls_charset_id('ja16eucfixed'))
FROM DUAL;
NLS_CHARSET_DECL_LEN(200,NLS_CHARSET_ID('JA16EUCFIXED'))
-------------------------------------------------------100

5-158 Oracle Database SQL Language Reference

NLS_CHARSET_ID

NLS_CHARSET_ID
Syntax
NLS_CHARSET_ID

(

string

)

Purpose
NLS_CHARSET_ID returns the character set ID number corresponding to character set
name string. The string argument is a run-time VARCHAR2 value. The string value
'CHAR_CS' returns the database character set ID number of the server. The string value
'NCHAR_CS' returns the national character set ID number of the server.
Invalid character set names return null.

Examples
The following example returns the character set ID of a character set:
SELECT NLS_CHARSET_ID('ja16euc')
FROM DUAL;
NLS_CHARSET_ID('JA16EUC')
------------------------830

See Also: Oracle Database Globalization Support Guide for a list of
character sets

Functions

5-159

NLS_CHARSET_NAME

NLS_CHARSET_NAME
Syntax
NLS_CHARSET_NAME

(

number

)

Purpose
NLS_CHARSET_NAME returns the name of the character set corresponding to ID number
number. The character set name is returned as a VARCHAR2 value in the database
character set.
If number is not recognized as a valid character set ID, then this function returns null.

Examples
The following example returns the character set corresponding to character set ID
number 2:
SELECT NLS_CHARSET_NAME(2)
FROM DUAL;
NLS_CH
-----WE8DEC

5-160 Oracle Database SQL Language Reference

NLS_INITCAP

NLS_INITCAP
Syntax
,
NLS_INITCAP

(

’

nlsparam

char

’
)

Purpose
NLS_INITCAP returns char, with the first letter of each word in uppercase, all other
letters in lowercase. Words are delimited by white space or characters that are not
alphanumeric.
Both char and 'nlsparam' can be any of the data types CHAR, VARCHAR2, NCHAR, or
NVARCHAR2. The string returned is of VARCHAR2 data type and is in the same character
set as char.
The value of 'nlsparam' can have this form:
'NLS_SORT = sort'

where sort is either a linguistic sort sequence or BINARY. The linguistic sort sequence
handles special linguistic requirements for case conversions. These requirements can
result in a return value of a different length than the char. If you omit 'nlsparam',
then this function uses the default sort sequence for your session.
This function does not support CLOB data directly. However, CLOBs can be passed in as
arguments through implicit data conversion.
See Also: "Data Type Comparison Rules" on page 3-36 for more
information.

Examples
The following examples show how the linguistic sort sequence results in a different
return value from the function:
SELECT NLS_INITCAP('ijsland') "InitCap"
FROM DUAL;
InitCap
------Ijsland
SELECT NLS_INITCAP('ijsland', 'NLS_SORT = XDutch') "InitCap"
FROM DUAL;
InitCap
------IJsland

See Also: Oracle Database Globalization Support Guide for information
on sort sequences

Functions

5-161

NLS_LOWER

NLS_LOWER
Syntax
,
NLS_LOWER

(

’

nlsparam

’

char

)

Purpose
NLS_LOWER returns char, with all letters lowercase.
Both char and 'nlsparam' can be any of the data types CHAR, VARCHAR2, NCHAR,
NVARCHAR2, CLOB, or NCLOB. The string returned is of VARCHAR2 data type if char is a
character data type and a LOB if char is a LOB data type. The return string is in the
same character set as char.
The 'nlsparam' can have the same form and serve the same purpose as in the NLS_
INITCAP function.

Examples
The following statement returns the lowercase form of the character string
'NOKTASINDA' using the XTurkish linguistic sort sequence. The Turkish uppercase I
becoming a small, dotless i.
SELECT NLS_LOWER('NOKTASINDA', 'NLS_SORT = XTurkish') "Lowercase"
FROM DUAL;

5-162 Oracle Database SQL Language Reference

NLS_UPPER

NLS_UPPER
Syntax
,
NLS_UPPER

(

’

nlsparam

’

char

)

Purpose
NLS_UPPER returns char, with all letters uppercase.
Both char and 'nlsparam' can be any of the data types CHAR, VARCHAR2, NCHAR,
NVARCHAR2, CLOB, or NCLOB. The string returned is of VARCHAR2 data type if char is a
character data type and a LOB if char is a LOB data type. The return string is in the
same character set as char.
The 'nlsparam' can have the same form and serve the same purpose as in the NLS_
INITCAP function.

Examples
The following example returns a string with all the letters converted to uppercase:
SELECT NLS_UPPER('große') "Uppercase"
FROM DUAL;
Upper
----GROßE
SELECT NLS_UPPER('große', 'NLS_SORT = XGerman') "Uppercase"
FROM DUAL;
Upperc
-----GROSSE

See Also:

NLS_INITCAP on page 5-161

Functions

5-163

NLSSORT

NLSSORT
Syntax
,
NLSSORT

(

’

nlsparam

char

’
)

Purpose
NLSSORT returns the string of bytes used to sort char.
Both char and 'nlsparam' can be any of the data types CHAR, VARCHAR2, NCHAR, or
NVARCHAR2.
The value of 'nlsparam' can have the form
'NLS_SORT = sort'

where sort is a linguistic sort sequence (collation) or BINARY. If you omit 'nlsparam',
then this function uses the default sort sequence for your session. If you specify
BINARY, then this function returns char.
If you specify 'nlsparam', then you can append to the linguistic sort name the suffix _
ai to request an accent-insensitive sort or _ci to request a case-insensitive sort. Refer
to Oracle Database Globalization Support Guide for more information on accent- and
case-insensitive sorting.
The string returned, also known as the collation key, is of RAW data type. The length of
the collation key resulting from a given char value for a given collation may exceed
2000 bytes, which is the maximum length of the RAW value returned by NLSSORT. In this
case, NLSSORT calculates the collation key for a maximum prefix, or initial substring, of
char so that the calculated result does not exceed 2000 bytes. For monolingual
collations, for example FRENCH, the prefix length is typically 1000 characters. For
multilingual collations, for example GENERIC_M, the prefix is typically 500 characters.
The exact length may be lower or higher depending on the collation and the characters
contained in char.
This behavior implies that two character values whose collation keys (NLSSORT results)
are compared to find the linguistic ordering are considered equal if they do not differ
in the prefix even though they may differ at some further character position. Because
the NLSSORT function is used implicitly to find linguistic ordering for comparison
conditions, the BETWEEN condition, the IN condition, ORDER BY, GROUP BY, and
COUNT(DISTINCT), those operations may return results that are only approximate for
long character values. This is a restriction of the current comparison architecture.
Currently, the only way to guarantee precise linguistic comparison results is to not
compare character values that are longer than 499 characters for monolingual
collations and 249 characters for multilingual collations.
This function does not support CLOB data directly. However, CLOBs can be passed in as
arguments through implicit data conversion.
See Also: "Data Type Comparison Rules" on page 3-36 for more
information.

Examples
This function can be used to specify sorting and comparison operations based on a
linguistic sort sequence rather than on the binary value of a string. The following

5-164 Oracle Database SQL Language Reference

NLSSORT

example creates a test table containing two values and shows how the values returned
can be ordered by the NLSSORT function:
CREATE
INSERT
INSERT
INSERT

TABLE test (name
INTO test VALUES
INTO test VALUES
INTO test VALUES

VARCHAR2(15));
('Gaardiner');
('Gaberd');
('Gaasten');

SELECT *
FROM test
ORDER BY name;
NAME
--------------Gaardiner
Gaasten
Gaberd
SELECT *
FROM test
ORDER BY NLSSORT(name, 'NLS_SORT = XDanish');
NAME
--------------Gaberd
Gaardiner
Gaasten

The following example shows how to use the NLSSORT function in comparison
operations:
SELECT *
FROM test
WHERE name > 'Gaberd'
ORDER BY name;
no rows selected
SELECT *
FROM test
WHERE NLSSORT(name, 'NLS_SORT = XDanish') >
NLSSORT('Gaberd', 'NLS_SORT = XDanish')
ORDER BY name;
NAME
--------------Gaardiner
Gaasten

If you frequently use NLSSORT in comparison operations with the same linguistic sort
sequence, then consider this more efficient alternative: Set the NLS_COMP parameter
(either for the database or for the current session) to LINGUISTIC, and set the NLS_SORT
parameter for the session to the desired sort sequence. Oracle Database will use that
sort sequence by default for all sorting and comparison operations during the current
session:
ALTER SESSION SET NLS_COMP = 'LINGUISTIC';
ALTER SESSION SET NLS_SORT = 'XDanish';
SELECT *
FROM test

Functions

5-165

NLSSORT

WHERE name > 'Gaberd'
ORDER BY name;
NAME
--------------Gaardiner
Gaasten

See Also: Oracle Database Globalization Support Guide for information
on sort sequences

5-166 Oracle Database SQL Language Reference

NTH_VALUE

NTH_VALUE
Syntax
FIRST

RESPECT

LAST

IGNORE

FROM
NTH_VALUE

OVER

(

(

measure_expr

analytic_clause

,

n

NULLS

)

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions of the analytic_clause

Purpose
NTH_VALUE returns the measure_expr value of the nth row in the window defined by
the analytic_clause. The returned value has the data type of the measure_expr.
■

■

■

{RESPECT | IGNORE} NULLS determines whether null values of measure_expr are
included in or eliminated from the calculation. The default is RESPECT NULLS.
n determines the nth row for which the measure value is to be returned. n can be a
constant, bind variable, column, or an expression involving them, as long as it
resolves to a positive integer. The function returns NULL if the data source window
has fewer than n rows. If n is null, then the function returns an error.
FROM {FIRST | LAST} determines whether the calculation begins at the first or last
row of the window. The default is FROM FIRST.

If you omit the windowing_clause of the analytic_clause, it defaults to RANGE
BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW. This default sometimes returns an
unexpected value for NTH_VALUE ... FROM LAST ... , because the last value in the window
is at the bottom of the window, which is not fixed. It keeps changing as the current
row changes. For expected results, specify the windowing_clause as RANGE BETWEEN
UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING. Alternatively, you can specify the
windowing_clause as RANGE BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING.
See Also: Oracle Database Data Warehousing Guide for more
information on the use of this function

Examples
The following example shows the minimum amount_sold value for the second
channel_id in ascending order for each prod_id between 13 and 16:
SELECT prod_id, channel_id, MIN(amount_sold),
NTH_VALUE(MIN(amount_sold), 2) OVER (PARTITION BY prod_id ORDER BY channel_id
ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING) nv
FROM sales
WHERE prod_id BETWEEN 13 and 16
GROUP BY prod_id, channel_id;
PROD_ID CHANNEL_ID MIN(AMOUNT_SOLD)
NV
---------- ---------- ---------------- ---------13
2
907.34
906.2
13
3
906.2
906.2
13
4
842.21
906.2

Functions

5-167

NTH_VALUE

14
14
14
15
15
15
16
16
16
16

2
3
4
2
3
4
2
3
4
9

13 rows selected.

5-168 Oracle Database SQL Language Reference

1015.94
1036.72
935.79
871.19
871.19
871.19
266.84
266.84
266.84
11.99

1036.72
1036.72
1036.72
871.19
871.19
871.19
266.84
266.84
266.84
266.84

NTILE

NTILE
Syntax
query_partition_clause
NTILE

(

expr

)

OVER

(

order_by_clause

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions, including valid forms of expr

Purpose
NTILE is an analytic function. It divides an ordered data set into a number of buckets
indicated by expr and assigns the appropriate bucket number to each row. The buckets
are numbered 1 through expr. The expr value must resolve to a positive constant for
each partition. Oracle Database expects an integer, and if expr is a noninteger constant,
then Oracle truncates the value to an integer. The return value is NUMBER.
The number of rows in the buckets can differ by at most 1. The remainder values (the
remainder of number of rows divided by buckets) are distributed one for each bucket,
starting with bucket 1.
If expr is greater than the number of rows, then a number of buckets equal to the
number of rows will be filled, and the remaining buckets will be empty.
You cannot nest analytic functions by using NTILE or any other analytic function for
expr. However, you can use other built-in function expressions for expr.
See Also: "About SQL Expressions" on page 6-1 for information on
valid forms of expr and Table 3–10, " Implicit Type Conversion
Matrix" on page 3-40 for more information on implicit conversion

Examples
The following example divides into 4 buckets the values in the salary column of the
oe.employees table from Department 100. The salary column has 6 values in this
department, so the two extra values (the remainder of 6 / 4) are allocated to buckets 1
and 2, which therefore have one more value than buckets 3 or 4.
SELECT last_name, salary, NTILE(4) OVER (ORDER BY salary DESC) AS quartile
FROM employees
WHERE department_id = 100
ORDER BY last_name, salary, quartile;
LAST_NAME
SALARY
QUARTILE
------------------------- ---------- ---------Chen
8200
2
Faviet
9000
1
Greenberg
12008
1
Popp
6900
4
Sciarra
7700
3
Urman
7800
2

Functions

5-169

NULLIF

NULLIF
Syntax
NULLIF

(

expr1

,

expr2

)

Purpose
NULLIF compares expr1 and expr2. If they are equal, then the function returns null. If
they are not equal, then the function returns expr1. You cannot specify the literal NULL
for expr1.
If both arguments are numeric data types, then Oracle Database determines the
argument with the higher numeric precedence, implicitly converts the other argument
to that data type, and returns that data type. If the arguments are not numeric, then
they must be of the same data type, or Oracle returns an error.
The NULLIF function is logically equivalent to the following CASE expression:
CASE WHEN expr1 = expr2 THEN NULL ELSE expr1 END

See Also:

"CASE Expressions" on page 6-5

Examples
The following example selects those employees from the sample schema hr who have
changed jobs since they were hired, as indicated by a job_id in the job_history table
different from the current job_id in the employees table:
SELECT e.last_name, NULLIF(j.job_id, e.job_id) "Old Job ID"
FROM employees e, job_history j
WHERE e.employee_id = j.employee_id
ORDER BY last_name, "Old Job ID";
LAST_NAME
------------------------De Haan
Hartstein
Kaufling
Kochhar
Kochhar
Raphaely
Taylor
Taylor
Whalen
Whalen

Old Job ID
---------IT_PROG
MK_REP
ST_CLERK
AC_ACCOUNT
AC_MGR
ST_CLERK
SA_MAN
AC_ACCOUNT

5-170 Oracle Database SQL Language Reference

NUMTODSINTERVAL

NUMTODSINTERVAL
Syntax
NUMTODSINTERVAL

(

n

,

’

interval_unit

’

)

Purpose
NUMTODSINTERVAL converts n to an INTERVAL DAY TO SECOND literal. The argument n can
be any NUMBER value or an expression that can be implicitly converted to a NUMBER
value. The argument interval_unit can be of CHAR, VARCHAR2, NCHAR, or NVARCHAR2
data type. The value for interval_unit specifies the unit of n and must resolve to one
of the following string values:
■

'DAY'

■

'HOUR'

■

'MINUTE'

■

'SECOND'

interval_unit is case insensitive. Leading and trailing values within the parentheses
are ignored. By default, the precision of the return is 9.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

Examples
The following example uses NUMTODSINTERVAL in a COUNT analytic function to calculate,
for each employee, the number of employees hired by the same manager within the
past 100 days from his or her hire date. Refer to "Analytic Functions" on page 5-11 for
more information on the syntax of the analytic functions.
SELECT manager_id, last_name, hire_date,
COUNT(*) OVER (PARTITION BY manager_id ORDER BY hire_date
RANGE NUMTODSINTERVAL(100, 'day') PRECEDING) AS t_count
FROM employees
ORDER BY last_name, hire_date;
MANAGER_ID
---------149
147
121
103
. . .
124
100
101
100

LAST_NAME
------------------------Abel
Ande
Atkinson
Austin

HIRE_DATE
T_COUNT
--------- ---------11-MAY-04
1
24-MAR-08
3
30-OCT-05
2
25-JUN-05
1

Walsh
Weiss
Whalen
Zlotkey

24-APR-06
18-JUL-04
17-SEP-03
29-JAN-08

2
1
1
2

Functions

5-171

NUMTOYMINTERVAL

NUMTOYMINTERVAL
Syntax
NUMTOYMINTERVAL

(

n

,

’

interval_unit

’

)

Purpose
NUMTOYMINTERVAL converts number n to an INTERVAL YEAR TO MONTH literal. The
argument n can be any NUMBER value or an expression that can be implicitly converted
to a NUMBER value. The argument interval_unit can be of CHAR, VARCHAR2, NCHAR, or
NVARCHAR2 data type. The value for interval_unit specifies the unit of n and must
resolve to one of the following string values:
■

'YEAR'

■

'MONTH'

interval_unit is case insensitive. Leading and trailing values within the parentheses
are ignored. By default, the precision of the return is 9.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

Examples
The following example uses NUMTOYMINTERVAL in a SUM analytic function to calculate,
for each employee, the total salary of employees hired in the past one year from his or
her hire date. Refer to "Analytic Functions" on page 5-11 for more information on the
syntax of the analytic functions.
SELECT last_name, hire_date, salary,
SUM(salary) OVER (ORDER BY hire_date
RANGE NUMTOYMINTERVAL(1,'year') PRECEDING) AS t_sal
FROM employees
ORDER BY last_name, hire_date;
LAST_NAME
------------------------Abel
Ande
Atkinson
Austin
. . .
Walsh
Weiss
Whalen
Zlotkey

HIRE_DATE
SALARY
T_SAL
--------- ---------- ---------11-MAY-04
11000
90300
24-MAR-08
6400
112500
30-OCT-05
2800
177000
25-JUN-05
4800
134700
24-APR-06
18-JUL-04
17-SEP-03
29-JAN-08

5-172 Oracle Database SQL Language Reference

3100
8000
4400
10500

186200
70900
54000
119000

NVL

NVL
Syntax
NVL

(

expr1

,

expr2

)

Purpose
NVL lets you replace null (returned as a blank) with a string in the results of a query. If
expr1 is null, then NVL returns expr2. If expr1 is not null, then NVL returns expr1.
The arguments expr1 and expr2 can have any data type. If their data types are
different, then Oracle Database implicitly converts one to the other. If they cannot be
converted implicitly, then the database returns an error. The implicit conversion is
implemented as follows:
■

■

If expr1 is character data, then Oracle Database converts expr2 to the data type of
expr1 before comparing them and returns VARCHAR2 in the character set of expr1.
If expr1 is numeric, then Oracle Database determines which argument has the
highest numeric precedence, implicitly converts the other argument to that data
type, and returns that data type.
See Also:
■

■

Table 3–10, " Implicit Type Conversion Matrix" on page 3-40 for
more information on implicit conversion and "Numeric
Precedence" on page 3-14 for information on numeric precedence
"COALESCE" on page 5-48 and "CASE Expressions" on page 6-5,
which provide functionality similar to that of NVL

Examples
The following example returns a list of employee names and commissions,
substituting "Not Applicable" if the employee receives no commission:
SELECT last_name, NVL(TO_CHAR(commission_pct), 'Not Applicable') commission
FROM employees
WHERE last_name LIKE 'B%'
ORDER BY last_name;
LAST_NAME
------------------------Baer
Baida
Banda
Bates
Bell
Bernstein
Bissot
Bloom
Bull

COMMISSION
---------------------------------------Not Applicable
Not Applicable
.1
.15
Not Applicable
.25
Not Applicable
.2
Not Applicable

Functions

5-173

NVL2

NVL2
Syntax
NVL2

(

expr1

,

expr2

,

expr3

)

Purpose
NVL2 lets you determine the value returned by a query based on whether a specified
expression is null or not null. If expr1 is not null, then NVL2 returns expr2. If expr1 is
null, then NVL2 returns expr3.
The argument expr1 can have any data type. The arguments expr2 and expr3 can have
any data types except LONG.
If the data types of expr2 and expr3 are different, then Oracle Database implicitly
converts one to the other. If they cannot be converted implicitly, then the database
returns an error. If expr2 is character or numeric data, then the implicit conversion is
implemented as follows:
■

■

If expr2 is character data, then Oracle Database converts expr3 to the data type of
expr2 before returning a value unless expr3 is a null constant. In that case, a data
type conversion is not necessary, and the database returns VARCHAR2 in the
character set of expr2.
If expr2 is numeric data, then Oracle Database determines which argument has
the highest numeric precedence, implicitly converts the other argument to that
data type, and returns that data type.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion and "Numeric
Precedence" on page 3-14 for information on numeric precedence

Examples
The following example shows whether the income of some employees is made up of
salary plus commission, or just salary, depending on whether the commission_pct
column of employees is null or not.
SELECT last_name, salary,
NVL2(commission_pct, salary + (salary * commission_pct), salary) income
FROM employees
WHERE last_name like 'B%'
ORDER BY last_name;
LAST_NAME
SALARY
INCOME
------------------------- ---------- ---------Baer
10000
10000
Baida
2900
2900
Banda
6200
6820
Bates
7300
8395
Bell
4000
4000
Bernstein
9500
11875
Bissot
3300
3300
Bloom
10000
12000
Bull
4100
4100

5-174 Oracle Database SQL Language Reference

ORA_DST_AFFECTED

ORA_DST_AFFECTED
Syntax
ORA_DST_AFFECTED

(

datetime_expr

)

Purpose
ORA_DST_AFFECTED is useful when you are changing the time zone data file for your
database. The function takes as an argument a datetime expression that resolves to a
TIMESTAMP WITH TIME ZONE value or a VARRAY object that contains TIMESTAMP WITH TIME
ZONE values. The function returns 1 if the datetime value is affected by or will result in
a "nonexisting time" or "duplicate time" error with the new time zone data. Otherwise,
it returns 0.
This function can be issued only when changing the time zone data file of the database
and upgrading the timestamp with the time zone data, and only between the
execution of the DBMS_DST.BEGIN_PREPARE and the DBMS_DST.END_PREPARE procedures
or between the execution of the DBMS_DST.BEGIN_UPGRADE and the DBMS_DST.END_
UPGRADE procedures.
See Also: Oracle Database Globalization Support Guide for more
information on time zone datafiles and on how Oracle Database
handles daylight saving time, and Oracle Database PL/SQL Packages and
Types Reference for information on the DBMS_DST package

Functions

5-175

ORA_DST_CONVERT

ORA_DST_CONVERT
Syntax
,
,
ORA_DST_CONVERT

(

datetime_expr

integer

integer
)

Purpose
ORA_DST_CONVERT is useful when you are changing the time zone data file for your
database. The function lets you specify error handling for a specified datetime
expression.
■

■

■

For datetime_expr, specify a datetime expression that resolves to a TIMESTAMP
WITH TIME ZONE value or a VARRAY object that contains TIMESTAMP WITH TIME ZONE
values.
The optional second argument specifies handling of "duplicate time" errors.
Specify 0 (false) to suppress the error by returning the source datetime value. This
is the default. Specify 1 (true) to allow the database to return the duplicate time
error.
The optional third argument specifies handling of "nonexisting time" errors.
Specify 0 (false) to suppress the error by returning the source datetime value. This
is the default. Specify 1 (true) to allow the database to return the nonexisting time
error.

If no error occurs, this function returns a value of the same data type as datetime_expr
(a TIMESTAMP WITH TIME ZONE value or a VARRAY object that contains TIMESTAMP WITH
TIME ZONE values). The returned datetime value when interpreted with the new time
zone file corresponds to datetime_expr interpreted with the old time zone file.
This function can be issued only when changing the time zone data file of the database
and upgrading the timestamp with the time zone data, and only between the
execution of the DBMS_DST.BEGIN_UPGRADE and the DBMS_DST.END_UPGRADE procedures.
See Also: Oracle Database Globalization Support Guide for more
information on time zone datafiles and on how Oracle Database
handles daylight saving time, and Oracle Database PL/SQL Packages and
Types Reference for information on the DBMS_DST package

5-176 Oracle Database SQL Language Reference

ORA_DST_ERROR

ORA_DST_ERROR
Syntax
ORA_DST_ERROR

(

datetime_expr

)

Purpose
ORA_DST_ERROR is useful when you are changing the time zone data file for your
database. The function takes as an argument a datetime expression that resolves to a
TIMESTAMP WITH TIME ZONE value or a VARRAY object that contains TIMESTAMP WITH TIME
ZONE values, and indicates whether the datetime value will result in an error with the
new time zone data. The return values are:
■

0: the datetime value does not result in an error with the new time zone data.

■

1878: the datetime value results in a "nonexisting time" error.

■

1883: the datetime value results in a "duplicate time" error.

This function can be issued only when changing the time zone data file of the database
and upgrading the timestamp with the time zone data, and only between the
execution of the DBMS_DST.BEGIN_PREPARE and the DBMS_DST.END_PREPARE procedures
or between the execution of the DBMS_DST.BEGIN_UPGRADE and the DBMS_DST.END_
UPGRADE procedures.
See Also: Oracle Database Globalization Support Guide for more
information on time zone datafiles and on how Oracle Database
handles daylight saving time, and Oracle Database PL/SQL Packages and
Types Reference for information on the DBMS_DST package

Functions

5-177

ORA_HASH

ORA_HASH
Syntax
,
,
ORA_HASH

(

seed_value

max_bucket

expr

)

Purpose
ORA_HASH is a function that computes a hash value for a given expression. This
function is useful for operations such as analyzing a subset of data and generating a
random sample.
■

■

■

The expr argument determines the data for which you want Oracle Database to
compute a hash value. There are no restrictions on the length of data represented
by expr, which commonly resolves to a column name. The expr cannot be a LONG
or LOB type. It cannot be a user-defined object type unless it is a nested table type.
The hash value for nested table types does not depend on the order of elements in
the collection. All other data types are supported for expr.
The optional max_bucket argument determines the maximum bucket value
returned by the hash function. You can specify any value between 0 and
4294967295. The default is 4294967295.
The optional seed_value argument enables Oracle to produce many different
results for the same set of data. Oracle applies the hash function to the
combination of expr and seed_value. You can specify any value between 0 and
4294967295. The default is 0.

The function returns a NUMBER value.

Examples
The following example creates a hash value for each combination of customer ID and
product ID in the sh.sales table, divides the hash values into a maximum of 100
buckets, and returns the sum of the amount_sold values in the first bucket (bucket 0).
The third argument (5) provides a seed value for the hash function. You can obtain
different hash results for the same query by changing the seed value.
SELECT SUM(amount_sold)
FROM sales
WHERE ORA_HASH(CONCAT(cust_id, prod_id), 99, 5) = 0;
SUM(AMOUNT_SOLD)
---------------989431.14

5-178 Oracle Database SQL Language Reference

PATH

PATH
Syntax
PATH

(

correlation_integer

)

Purpose
PATH is an ancillary function used only with the UNDER_PATH and EQUALS_PATH
conditions. It returns the relative path that leads to the resource specified in the parent
condition.
The correlation_integer can be any NUMBER integer and is used to correlate this
ancillary function with its primary condition. Values less than 1 are treated as 1.
See Also: EQUALS_PATH Condition on page 7-20 and UNDER_
PATH Condition on page 7-21

Examples
Refer to the related function DEPTH on page 5-84 for an example using both of these
ancillary functions of the EQUALS_PATH and UNDER_PATH conditions.

Functions

5-179

PERCENT_RANK

PERCENT_RANK
Aggregate Syntax
percent_rank_aggregate::=
,
PERCENT_RANK

(

expr

)

WITHIN

GROUP

,
DESC

FIRST
NULLS

ASC
(

ORDER

BY

LAST

expr

)

Analytic Syntax
percent_rank_analytic::=
query_partition_clause
PERCENT_RANK

(

)

OVER

(

order_by_clause

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions

Purpose
PERCENT_RANK is similar to the CUME_DIST (cumulative distribution) function. The range
of values returned by PERCENT_RANK is 0 to 1, inclusive. The first row in any set has a
PERCENT_RANK of 0. The return value is NUMBER.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion
■

As an aggregate function, PERCENT_RANK calculates, for a hypothetical row r
identified by the arguments of the function and a corresponding sort specification,
the rank of row r minus 1 divided by the number of rows in the aggregate group.
This calculation is made as if the hypothetical row r were inserted into the group
of rows over which Oracle Database is to aggregate.
The arguments of the function identify a single hypothetical row within each
aggregate group. Therefore, they must all evaluate to constant expressions within
each aggregate group. The constant argument expressions and the expressions in
the ORDER BY clause of the aggregate match by position. Therefore the number of
arguments must be the same and their types must be compatible.

■

As an analytic function, for a row r, PERCENT_RANK calculates the rank of r minus 1,
divided by 1 less than the number of rows being evaluated (the entire query result
set or a partition).

5-180 Oracle Database SQL Language Reference

PERCENT_RANK

Aggregate Example
The following example calculates the percent rank of a hypothetical employee in the
sample table hr.employees with a salary of $15,500 and a commission of 5%:
SELECT PERCENT_RANK(15000, .05) WITHIN GROUP
(ORDER BY salary, commission_pct) "Percent-Rank"
FROM employees;
Percent-Rank
-----------.971962617

Analytic Example
The following example calculates, for each employee, the percent rank of the
employee's salary within the department:
SELECT department_id, last_name, salary, PERCENT_RANK()
OVER (PARTITION BY department_id ORDER BY salary DESC) AS pr
FROM employees
ORDER BY pr, salary, last_name;
DEPARTMENT_ID
------------10
40

LAST_NAME
SALARY
PR
------------------------- ---------- ---------Whalen
4400
0
Mavris
6500
0
Grant
7000
0

. . .
80 Vishney
80 Zlotkey
30 Khoo

10500 .181818182
10500 .181818182
3100
.2

. . .
50 Markle
50 Philtanker
50 Olson

2200 .954545455
2200 .954545455
2100
1

. . .

Functions

5-181

PERCENTILE_CONT

PERCENTILE_CONT
Syntax
DESC
ASC
PERCENTILE_CONT

OVER

(

(

expr

)

query_partition_clause

WITHIN

GROUP

(

ORDER

BY

expr

)

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions of the OVER clause

Purpose
PERCENTILE_CONT is an inverse distribution function that assumes a continuous
distribution model. It takes a percentile value and a sort specification, and returns an
interpolated value that would fall into that percentile value with respect to the sort
specification. Nulls are ignored in the calculation.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. The function returns the
same data type as the numeric data type of the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

The first expr must evaluate to a numeric value between 0 and 1, because it is a
percentile value. This expr must be constant within each aggregation group. The ORDER
BY clause takes a single expression that must be a numeric or datetime value, as these
are the types over which Oracle can perform interpolation.
The result of PERCENTILE_CONT is computed by linear interpolation between values
after ordering them. Using the percentile value (P) and the number of rows (N) in the
aggregation group, you can compute the row number you are interested in after
ordering the rows with respect to the sort specification. This row number (RN) is
computed according to the formula RN = (1+(P*(N-1)). The final result of the
aggregate function is computed by linear interpolation between the values from rows
at row numbers CRN = CEILING(RN) and FRN = FLOOR(RN).
The final result will be:
If (CRN = FRN = RN) then the result is
(value of expression from row at RN)
Otherwise the result is
(CRN - RN) * (value of expression for row at FRN) +
(RN - FRN) * (value of expression for row at CRN)

You can use the PERCENTILE_CONT function as an analytic function. You can specify
only the query_partitioning_clause in its OVER clause. It returns, for each row, the
value that would fall into the specified percentile among a set of values within each
partition.
The MEDIAN function is a specific case of PERCENTILE_CONT where the percentile value
defaults to 0.5. For more information, refer to MEDIAN on page 5-148.

5-182 Oracle Database SQL Language Reference

PERCENTILE_CONT

Aggregate Example
The following example computes the median salary in each department:
SELECT department_id,
PERCENTILE_CONT(0.5) WITHIN GROUP (ORDER BY salary DESC) "Median cont",
PERCENTILE_DISC(0.5) WITHIN GROUP (ORDER BY salary DESC) "Median disc"
FROM employees
GROUP BY department_id
ORDER BY department_id;
DEPARTMENT_ID Median cont Median disc
------------- ----------- ----------10
4400
4400
20
9500
13000
30
2850
2900
40
6500
6500
50
3100
3100
60
4800
4800
70
10000
10000
80
8900
9000
90
17000
17000
100
8000
8200
110
10154
12008
7000
7000

PERCENTILE_CONT and PERCENTILE_DISC may return different results. PERCENTILE_CONT
returns a computed result after doing linear interpolation. PERCENTILE_DISC simply
returns a value from the set of values that are aggregated over. When the percentile
value is 0.5, as in this example, PERCENTILE_CONT returns the average of the two
middle values for groups with even number of elements, whereas PERCENTILE_DISC
returns the value of the first one among the two middle values. For aggregate groups
with an odd number of elements, both functions return the value of the middle
element.

Analytic Example
In the following example, the median for Department 60 is 4800, which has a
corresponding percentile (Percent_Rank) of 0.5. None of the salaries in Department 30
have a percentile of 0.5, so the median value must be interpolated between 2900
(percentile 0.4) and 2800 (percentile 0.6), which evaluates to 2850.
SELECT last_name, salary, department_id,
PERCENTILE_CONT(0.5) WITHIN GROUP (ORDER BY salary DESC)
OVER (PARTITION BY department_id) "Percentile_Cont",
PERCENT_RANK()
OVER (PARTITION BY department_id ORDER BY salary DESC) "Percent_Rank"
FROM employees
WHERE department_id IN (30, 60)
ORDER BY last_name, salary, department_id;
LAST_NAME
SALARY DEPARTMENT_ID Percentile_Cont Percent_Rank
------------------------- ---------- ------------- --------------- -----------Austin
4800
60
4800
.5
Baida
2900
30
2850
.4
Colmenares
2500
30
2850
1
Ernst
6000
60
4800
.25
Himuro
2600
30
2850
.8
Hunold
9000
60
4800
0
Khoo
3100
30
2850
.2
Lorentz
4200
60
4800
1

Functions

5-183

PERCENTILE_CONT

Pataballa
Raphaely
Tobias

5-184 Oracle Database SQL Language Reference

4800
11000
2800

60
30
30

4800
2850
2850

.5
0
.6

PERCENTILE_DISC

PERCENTILE_DISC
Syntax
DESC
ASC
PERCENTILE_DISC

OVER

(

(

expr

)

query_partition_clause

WITHIN

GROUP

(

ORDER

BY

expr

)

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions of the OVER clause

Purpose
PERCENTILE_DISC is an inverse distribution function that assumes a discrete
distribution model. It takes a percentile value and a sort specification and returns an
element from the set. Nulls are ignored in the calculation.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. The function returns the
same data type as the numeric data type of the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

The first expr must evaluate to a numeric value between 0 and 1, because it is a
percentile value. This expression must be constant within each aggregate group. The
ORDER BY clause takes a single expression that can be of any type that can be sorted.
For a given percentile value P, PERCENTILE_DISC sorts the values of the expression in
the ORDER BY clause and returns the value with the smallest CUME_DIST value (with
respect to the same sort specification) that is greater than or equal to P.

Aggregate Example
See aggregate example for PERCENTILE_CONT on page 5-182.

Analytic Example
The following example calculates the median discrete percentile of the salary of each
employee in the sample table hr.employees:
SELECT last_name, salary, department_id,
PERCENTILE_DISC(0.5) WITHIN GROUP (ORDER BY salary DESC)
OVER (PARTITION BY department_id) "Percentile_Disc",
CUME_DIST() OVER (PARTITION BY department_id
ORDER BY salary DESC) "Cume_Dist"
FROM employees
WHERE department_id in (30, 60)
ORDER BY last_name, salary, department_id;
LAST_NAME
SALARY DEPARTMENT_ID Percentile_Disc Cume_Dist
------------------------- ---------- ------------- --------------- ---------Austin
4800
60
4800
.8
Baida
2900
30
2900
.5

Functions

5-185

PERCENTILE_DISC

Colmenares
Ernst
Himuro
Hunold
Khoo
Lorentz
Pataballa
Raphaely
Tobias

2500
6000
2600
9000
3100
4200
4800
11000
2800

30
60
30
60
30
60
60
30
30

2900
4800
2900
4800
2900
4800
4800
2900
2900

1
.4
.833333333
.2
.333333333
1
.8
.166666667
.666666667

The median value for Department 30 is 2900, which is the value whose corresponding
percentile (Cume_Dist) is the smallest value greater than or equal to 0.5. The median
value for Department 60 is 4800, which is the value whose corresponding percentile is
the smallest value greater than or equal to 0.5.

5-186 Oracle Database SQL Language Reference

POWER

POWER
Syntax
POWER

(

n2

,

n1

)

Purpose
POWER returns n2 raised to the n1 power. The base n2 and the exponent n1 can be any
numbers, but if n2 is negative, then n1 must be an integer.
This function takes as arguments any numeric data type or any nonnumeric data type
that can be implicitly converted to a numeric data type. If any argument is BINARY_
FLOAT or BINARY_DOUBLE, then the function returns BINARY_DOUBLE. Otherwise, the
function returns NUMBER.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

Examples
The following example returns 3 squared:
SELECT POWER(3,2) "Raised"
FROM DUAL;
Raised
---------9

Functions

5-187

POWERMULTISET

POWERMULTISET
Syntax
POWERMULTISET

(

expr

)

Purpose
POWERMULTISET takes as input a nested table and returns a nested table of nested tables
containing all nonempty subsets (called submultisets) of the input nested table.
■

expr can be any expression that evaluates to a nested table.

■

If expr resolves to null, then Oracle Database returns NULL.

■

If expr resolves to a nested table that is empty, then Oracle returns an error.

■

The element types of the nested table must be comparable. Refer to "Comparison
Conditions" on page 7-4 for information on the comparability of nonscalar types.
Note:

This function is not supported in PL/SQL.

Examples
First, create a data type that is a nested table of the cust_address_tab_type data type:
CREATE TYPE cust_address_tab_tab_typ
AS TABLE OF cust_address_tab_typ;
/

Now, select the nested table column cust_address_ntab from the customers_demo
table using the POWERMULTISET function:
SELECT CAST(POWERMULTISET(cust_address_ntab) AS cust_address_tab_tab_typ)
FROM customers_demo;
CAST(POWERMULTISET(CUST_ADDRESS_NTAB) AS CUST_ADDRESS_TAB_TAB_TYP)
(STREET_ADDRESS, POSTAL_CODE, CITY, STATE_PROVINCE, COUNTRY_ID)
-----------------------------------------------------------------CUST_ADDRESS_TAB_TAB_TYP(CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP
('514 W Superior St', '46901', 'Kokomo', 'IN', 'US')))
CUST_ADDRESS_TAB_TAB_TYP(CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP
('2515 Bloyd Ave', '46218', 'Indianapolis', 'IN', 'US')))
CUST_ADDRESS_TAB_TAB_TYP(CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP
('8768 N State Rd 37', '47404', 'Bloomington', 'IN', 'US')))
CUST_ADDRESS_TAB_TAB_TYP(CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP
('6445 Bay Harbor Ln', '46254', 'Indianapolis', 'IN', 'US')))
. . .

The preceding example requires the customers_demo table and a nested table column
containing data. Refer to "Multiset Operators" on page 4-6 to create this table and
nested table columns.

5-188 Oracle Database SQL Language Reference

POWERMULTISET_BY_CARDINALITY

POWERMULTISET_BY_CARDINALITY
Syntax
POWERMULTISET_BY_CARDINALITY

(

expr

,

cardinality

)

Purpose
POWERMULTISET_BY_CARDINALITY takes as input a nested table and a cardinality and
returns a nested table of nested tables containing all nonempty subsets (called
submultisets) of the nested table of the specified cardinality.
■

expr can be any expression that evaluates to a nested table.

■

cardinality can be any positive integer.

■

If expr resolves to null, then Oracle Database returns NULL.

■

If expr resolves to a nested table that is empty, then Oracle returns an error.

■

The element types of the nested table must be comparable. Refer to "Comparison
Conditions" on page 7-4 for information on the comparability of nonscalar types.
Note:

This function is not supported in PL/SQL.

Examples
First, create a data type that is a nested table of the cust_address_tab_type data type:
CREATE TYPE cust_address_tab_tab_typ
AS TABLE OF cust_address_tab_typ;
/

Next, duplicate the elements in all the nested table rows to increase the cardinality of
the nested table rows to 2:
UPDATE customers_demo
SET cust_address_ntab = cust_address_ntab MULTISET UNION cust_address_ntab;

Now, select the nested table column cust_address_ntab from the customers_demo
table using the POWERMULTISET_BY_CARDINALITY function:
SELECT CAST(POWERMULTISET_BY_CARDINALITY(cust_address_ntab, 2)
AS cust_address_tab_tab_typ)
FROM customers_demo;
CAST(POWERMULTISET_BY_CARDINALITY(CUST_ADDRESS_NTAB,2) AS CUST_ADDRESS_TAB_TAB_TYP)
(STREET_ADDRESS, POSTAL_CODE, CITY, STATE_PROVINCE, COUNTRY_ID)
---------------------------------------------------------------------------------------CUST_ADDRESS_TAB_TAB_TYP(CUST_ADDRESS_TAB_TYP
(CUST_ADDRESS_TYP('514 W Superior St', '46901', 'Kokomo', 'IN', 'US'),
CUST_ADDRESS_TYP('514 W Superior St', '46901', 'Kokomo', 'IN', 'US')))
CUST_ADDRESS_TAB_TAB_TYP(CUST_ADDRESS_TAB_TYP
(CUST_ADDRESS_TYP('2515 Bloyd Ave', '46218', 'Indianapolis', 'IN', 'US'),
CUST_ADDRESS_TYP('2515 Bloyd Ave', '46218', 'Indianapolis', 'IN', 'US')))
CUST_ADDRESS_TAB_TAB_TYP(CUST_ADDRESS_TAB_TYP
(CUST_ADDRESS_TYP('8768 N State Rd 37', '47404', 'Bloomington', 'IN', 'US'),
CUST_ADDRESS_TYP('8768 N State Rd 37', '47404', 'Bloomington', 'IN', 'US')))
. . .

Functions

5-189

POWERMULTISET_BY_CARDINALITY

The preceding example requires the customers_demo table and a nested table column
containing data. Refer to "Multiset Operators" on page 4-6 to create this table and
nested table columns.

5-190 Oracle Database SQL Language Reference

PREDICTION

PREDICTION
Syntax
schema
PREDICTION

.

cost_matrix_clause

(

model

mining_attribute_clause

)

cost_matrix_clause::=
AUTO
MODEL
,
COST
,
(

,

class_value

)

VALUES

(

(

cost_value

)

)

mining_attribute_clause::=
*
,
USING

schema

.
table
AS

.

*

alias

expr

Purpose
This function is for use with mining models created by the DBMS_DATA_MINING package
or with Oracle Data Miner. It returns the best prediction for the model. The data type
returned depends on the target value type used during the build of the model. For
regression models, this function returns the expected value.
cost_matrix_clause
■

■

■

The COST clause is relevant for all classification models.

Specify COST MODEL to indicate that the scoring should be performed by taking into
account the scoring cost matrix associated with the model. If no such scoring cost
matrix exists, then the database returns an error.
Specify COST MODEL AUTO if the existence of a cost matrix is unknown. In this case:
–

If the stored cost matrix exists, then the function returns the lowest cost
prediction using the stored cost matrix.

–

If no stored cost matrix exists, then the function returns the highest probability
prediction.

Use the VALUES clause (the bottom branch of the cost_matrix_clause) to specify
an inline cost matrix. You can use an inline cost matrix regardless of whether the
model has an associated scoring cost matrix. Refer to Oracle Data Mining
Application Developer's Guide for an example of an inline cost matrix

If you omit the cost_matrix_clause clause, then the best prediction is the target class
with the highest probability. If two or more classes are tied with the highest
probability, the database chooses one class.

Functions

5-191

PREDICTION

mining_attribute_clause This maps the predictors that were provided when the
model was built. Specifying USING * maps to all to the columns and expressions that
can be retrieved from the underlying inputs (tables, views, and so on).
■

■

■

If you specify more predictors in the mining_attribute_clause than there are
predictors used by the model, then the extra expressions are silently ignored.
If you specify fewer predictors than are used during the build, then the operation
proceeds with the subset of predictors you specify and returns information on a
best-effort basis. All types of models will return a result regardless of the number
of predictors you specify in this clause.
If you specify a predictor with the same name as was used during the build but a
different data type, then the database implicitly converts to produce a predictor
value of the same type as the original build.
See Also:
■

■

Oracle Data Mining Concepts for detailed information about Oracle
Data Mining
Oracle Data Mining Application Developer's Guide for detailed
information about scoring with the Data Mining SQL functions

Example
The following example returns by gender the average age of customers who are likely
to use an affinity card. The PREDICTION function takes into account only the cust_
marital_status, education, and household_size predictors.
This example, and the prerequisite data mining operations, including the creation of
the view, can be found in the demo file $ORACLE_HOME/rdbms/demo/dmdtdemo.sql.
General information on data mining demo files is available in Oracle Data Mining
Administrator's Guide. The example is presented here to illustrate the syntactic use of
the function.
SELECT cust_gender, COUNT(*) AS cnt, ROUND(AVG(age)) AS avg_age
FROM mining_data_apply_v
WHERE PREDICTION(DT_SH_Clas_sample COST MODEL
USING cust_marital_status, education, household_size) = 1
GROUP BY cust_gender
ORDER BY cust_gender;
C
CNT
AVG_AGE
- ---------- ---------F
170
38
M
685
42

5-192 Oracle Database SQL Language Reference

PREDICTION_BOUNDS

PREDICTION_BOUNDS
Syntax
PREDICTION_BOUNDS

,
schema

.

,

(

class_value

confidence_level

model

mining_attribute_clause

)

mining_attribute_clause::=
*
,
USING

schema

.
table
AS

.

*

alias

expr

Purpose
The PREDICTION_BOUNDS function is for use with generalized linear models (GLM)
created by the DBMS_DATA_MINING package or with Oracle Data Miner. It returns an
object with two NUMBER fields LOWER and UPPER. For a regression mining function, the
bounds apply to value of the prediction. For a classification mining function, the
bounds apply to the probability value. If the GLM was built using ridge regression, or
if the covariance matrix is found to be singular during the build, then this function
returns NULL for both fields.
■

■

For confidence_level, specify a number in the range (0,1). If you omit this clause,
then the default value is 0.95.
The class_value argument is valid for classification models but not for regression
models. By default, the function returns the bounds for the prediction with the
highest probability. You can use the class_value argument to filter out the bounds
value specific to a target value.
You can specify class_value while leaving confidence_level at its default by
specifying NULL for confidence_level.

■

The mining_attribute_clause has the same behavior for PREDICTION_BOUNDS that
it has for PREDICTION. Refer to mining_attribute_clause on page 5-192.
See Also:
■

■

Oracle Data Mining Concepts for detailed information about Oracle
Data Mining and about generalized linear models
Oracle Data Mining Application Developer's Guide for detailed
information about scoring with the Data Mining SQL functions

Example
The following example returns the distribution of customers whose ages are predicted
to be between 25 and 45 years with 98% confidence.

Functions

5-193

PREDICTION_BOUNDS

This example and the prerequisite data mining operations can be found in the demo
file $ORACLE_HOME/rdbms/demo/dmglcdem.sql. The example is presented here to
illustrate the syntactic use of the function. General information on data mining demo
files is available in Oracle Data Mining Administrator's Guide.
SELECT count(cust_id) cust_count, cust_marital_status
FROM (SELECT cust_id, cust_marital_status
FROM mining_data_apply_v
WHERE PREDICTION_BOUNDS(glmr_sh_regr_sample,0.98 USING *).LOWER > 24 AND
PREDICTION_BOUNDS(glmr_sh_regr_sample,0.98 USING *).UPPER < 46)
GROUP BY cust_marital_status;
CUST_COUNT
-------------46
7
5
35
72

CUST_MARITAL_STATUS
-------------------NeverM
Mabsent
Separ.
Divorc.
Married

5-194 Oracle Database SQL Language Reference

PREDICTION_COST

PREDICTION_COST
Syntax
schema
PREDICTION_COST

.

,

(

class

model

cost_matrix_clause

mining_attribute_clause

)

cost_matrix_clause::=
AUTO
MODEL
,
COST
,
(

,

class_value

)

VALUES

(

(

cost_value

)

)

mining_attribute_clause::=
*
,
USING

schema

.
table
AS

.

*

alias

expr

Purpose
This function is for use with classification models created by the DBMS_DATA_MINING
package or with Oracle Data Miner. It returns a measure of cost for a given prediction
as an Oracle NUMBER.
If you specify the optional class parameter, then the function returns the cost for the
specified class. If you omit the class parameter, then the function returns the cost
associated with the best prediction. You can use this form in conjunction with the
PREDICTION function to obtain the best pair of prediction value and cost.
The COST clause is relevant for all classification models.
■

■

■

Specify COST MODEL to indicate that the scoring should be performed by taking into
account the scoring cost matrix associated with the model. If no such scoring cost
matrix exists, then the database returns an error.
Specify COST MODEL AUTO if the existence of a cost matrix is unknown. In this case:
–

If the stored cost matrix exists, then the function returns the cost using the
stored cost matrix.

–

If no stored cost matrix exists, then the function applies the unit cost matrix
(0's on the diagonal and 1's everywhere else). This is equivalent to one minus
probability for the given class.

Use the VALUES clause (the bottom branch of the cost_matrix_clause) to specify
an inline cost matrix. You can use an inline cost matrix regardless of whether the
model has an associated scoring cost matrix. Refer to Oracle Data Mining
Application Developer's Guide for an example of an inline cost matrix
Functions

5-195

PREDICTION_COST

The mining_attribute_clause behaves as described for the PREDICTION function.
Refer to mining_attribute_clause on page 5-192.
See Also:
■

■

Oracle Data Mining Concepts for detailed information about Oracle
Data Mining in general and about costs in particular
Oracle Data Mining Application Developer's Guide for detailed
information about scoring with the Data Mining SQL functions

Example
The following example finds the ten customers living in Italy who are least expensive
to convince to use an affinity card.
This example and the prerequisite data mining operations can be found in the demo
file $ORACLE_HOME/rdbms/demo/dmdtdemo.sql. General information on data mining
demo files is available in Oracle Data Mining Administrator's Guide. The example is
presented here to illustrate the syntactic use of the function.
WITH
cust_italy AS (
SELECT cust_id
FROM mining_data_apply_v
WHERE country_name = 'Italy'
ORDER BY PREDICTION_COST(DT_SH_Clas_sample, 1 COST MODEL USING *) ASC, 1
)
SELECT cust_id
FROM cust_italy
WHERE rownum < 11;
CUST_ID
---------100081
100179
100185
100324
100344
100554
100662
100733
101250
101306
10 rows selected.

5-196 Oracle Database SQL Language Reference

PREDICTION_DETAILS

PREDICTION_DETAILS
Syntax
schema
PREDICTION_DETAILS

.

(

model

mining_attribute_clause

)

mining_attribute_clause::=
*
,
USING

schema

.
table
AS

.

*

alias

expr

Purpose
This function is for use with decision tree models created by the DBMS_DATA_MINING
package or with Oracle Data Miner. It returns an XML string containing model-specific
information related to the scoring of the input row. In this release, the return value
takes the following form:


where integer is the identifier of a data mining tree node. The form of the output is
subject to change. It may be enhanced to provide additional prediction information in
future releases.
The mining_attribute_clause behaves as described for the PREDICTION function.
Refer to mining_attribute_clause on page 5-192.
See Also:
■

■

Oracle Data Mining Concepts for detailed information about Oracle
Data Mining
Oracle Data Mining Application Developer's Guide for detailed
information about scoring with the Data Mining SQL functions

Example
The following example uses all attributes from the mining_data_apply_v view that are
relevant predictors for the DT_SH_Clas_sample decision tree model. For customers who
work in technical support and are under age 25, it returns the tree node that results
from scoring those records with the DT_SH_Clas_sample model.
This example, and the prerequisite data mining operations, including the creation of
the view, can be found in the demo files $ORACLE_HOME/rdbms/demo/dmdtdemo.sql.
General information on data mining demo files is available in Oracle Data Mining
Administrator's Guide. The example is presented here to illustrate the syntactic use of
the function.
SELECT cust_id, education,
PREDICTION_DETAILS(DT_SH_Clas_sample using *) treenode
FROM mining_data_apply_v

Functions

5-197

PREDICTION_DETAILS

WHERE occupation = 'TechSup' AND age < 25
ORDER BY cust_id;
CUST_ID
---------100234
100320
100349
100419
100583
100657
101171
101225
101338

EDUCATION
--------------------< Bach.
< Bach.
< Bach.
< Bach.
< Bach.
HS-grad
< Bach.
< Bach.
< Bach.

9 rows selected.

5-198 Oracle Database SQL Language Reference

TREENODE
------------------------









PREDICTION_PROBABILITY

PREDICTION_PROBABILITY
Syntax
schema
PREDICTION_PROBABILITY

.

,

(

class

model

mining_attribute_clause

)

mining_attribute_clause::=
*
,
USING

schema

.
table
AS

.

*

alias

expr

Purpose
This function is for use with classification models created by the DBMS_DATA_MINING
package or with Oracle Data Miner. It is not valid with other types of models. It
returns the probability for a given prediction as an Oracle NUMBER.
If you specify the optional class parameter, then the function returns the probability
for the specified class. This is equivalent to the probability associated with choosing
the given target class value.
If you omit the class parameter, then the function returns the probability associated
with the best prediction. You can use this form in conjunction with the PREDICTION
function to obtain the best pair of prediction value and probability.
The mining_attribute_clause behaves as described for the PREDICTION function.
Refer to mining_attribute_clause on page 5-192.
See Also:
■

■

Oracle Data Mining Concepts for detailed information about Oracle
Data Mining
Oracle Data Mining Application Developer's Guide for detailed
information about scoring with the Data Mining SQL functions

Example
The following example returns the 10 customers living in Italy who are most likely to
use an affinity card.
This example, and the prerequisite data mining operations, including the creation of
the view, can be found in the demo files $ORACLE_HOME/rdbms/demo/dmdtdemo.sql.
General information on data mining demo files is available in Oracle Data Mining
Administrator's Guide. The example is presented here to illustrate the syntactic use of
the function.
SELECT cust_id FROM (
SELECT cust_id
FROM mining_data_apply_v
WHERE country_name = 'Italy'
ORDER BY PREDICTION_PROBABILITY(DT_SH_Clas_sample, 1 USING *)

Functions

5-199

PREDICTION_PROBABILITY

DESC, cust_id)
WHERE rownum < 11;
CUST_ID
---------100081
100179
100185
100324
100344
100554
100662
100733
101250
101306
10 rows selected.

5-200 Oracle Database SQL Language Reference

PREDICTION_SET

PREDICTION_SET
Syntax
,
schema
PREDICTION_SET

.

,

(

cutoff

bestN

model

cost_matrix_clause
mining_attribute_clause

)

cost_matrix_clause::=
AUTO
MODEL
,
COST
,
(

,

class_value

)

VALUES

(

(

cost_value

)

)

mining_attribute_clause::=
*
,
USING

schema

.
table
AS

.

*

alias

expr

Purpose
This function is for use with classification models created by the DBMS_DATA_MINING
package or with Oracle Data Miner. It is not valid with other types of models. It
returns a varray of objects containing all classes in a multiclass classification scenario.
The object fields are named PREDICTION, PROBABILITY, and COST. The data type of the
PREDICTION field depends on the target value type used during the build of the model.
The other two fields are both Oracle NUMBER. The elements are returned in the order of
best prediction to worst prediction.
■

■

For bestN, specify a positive integer to restrict the returned target classes to the N
having the highest probability, or lowest cost if cost matrix clause is specified. If
multiple classes are tied in the Nth value, then the database still returns only N
values. If you want to filter only by cutoff, specify NULL for this parameter.
For cutoff, specify a NUMBER value to restrict the returned target classes to those
with a probability greater than or equal to (or a cost less than or equal to if cost
matrix clause is specified) to the specified cutoff value. You can filter solely by
cutoff by specifying NULL for bestN.
When you specify values for both bestN and cutoff, you restrict the returned
predictions to only those that are the bestN and have a probability (or cost when
the cost_matrix_clause is specified) surpassing the threshold.

Functions

5-201

PREDICTION_SET

The cost_matrix_clause clause is relevant for all classification models. When you
specify this clause, both bestN and cutoff are treated with respect to the prediction
cost, not the prediction probability. The value of bestN restricts the result to the target
classes having the N best (lowest) costs, and cutoff restricts the target classes to those
with a cost less than or equal to the specified cutoff.
When you specify this clause, each object in the collection is a triplet of scalar values
containing the prediction value (the data type of which depends on the target value
type used during model build), the prediction probability, and the prediction cost
(both Oracle NUMBER).
If you omit this clause, then each object in the varray is a pair of scalars containing the
prediction value and prediction probability. The data types returned are as described
in the preceding paragraph.
■

■

■

Specify COST MODEL to indicate that the scoring should be performed by taking into
account the scoring cost matrix associated with the model. If no such cost matrix
exists, then the database returns an error.
Specify COST MODEL AUTO if the existence of a cost matrix is unknown. In this case:
–

If the stored cost matrix exists, then the result is the same as with COST MODEL.

–

If no stored cost matrix exists, then the result is almost the same as without the
cost_matrix_clause, except the object in the collection is a triplet and the cost
value is computed based on the unit cost matrix (0's on the diagonal and 1's
everywhere else). This is equivalent to one minus probability for the given
class. The cutoff parameter is ignored if no stored cost matrix exists.

Use the VALUES clause (the bottom branch of the cost_matrix_clause) to specify
an inline cost matrix. You can use an inline cost matrix regardless of whether the
model has an associated scoring cost matrix. Refer to Oracle Data Mining
Application Developer's Guide for an example of an inline cost matrix

The mining_attribute_clause behaves as described for the PREDICTION function.
Refer to mining_attribute_clause on page 5-192.
See Also:
■

■

Oracle Data Mining Concepts for detailed information about Oracle
Data Mining
Oracle Data Mining Application Developer's Guide for detailed
information about scoring with the Data Mining SQL functions

Example
The following example lists, for ten customers, the likelihood and cost of using or
rejecting an affinity card. This example has a binary target, but such a query is also
useful in multiclass classification such as Low, Med, and High.
This example and the prerequisite data mining operations can be found in the demo
file $ORACLE_HOME/rdbms/demo/dmdtdemo.sql. General information on data mining
demo files is available in Oracle Data Mining Administrator's Guide. The example is
presented here to illustrate the syntactic use of the function.
SELECT T.cust_id, S.prediction, S.probability, S.cost
FROM (SELECT cust_id,
PREDICTION_SET(dt_sh_clas_sample COST MODEL USING *) pset
FROM mining_data_apply_v
WHERE cust_id < 100011) T,
TABLE(T.pset) S
ORDER BY cust_id, S.prediction;
5-202 Oracle Database SQL Language Reference

PREDICTION_SET

CUST_ID PREDICTION PROBABILITY COST
---------- ---------- ----------- ----100001
0
.96682
.27
100001
1
.03318
.97
100002
0
.74038 2.08
100002
1
.25962
.74
100003
0
.90909
.73
100003
1
.09091
.91
100004
0
.90909
.73
100004
1
.09091
.91
100005
0
.27236 5.82
100005
1
.72764
.27
100006
0
1.00000
.00
100006
1
.00000 1.00
100007
0
.90909
.73
100007
1
.09091
.91
100008
0
.90909
.73
100008
1
.09091
.91
100009
0
.27236 5.82
100009
1
.72764
.27
100010
0
.80808 1.54
100010
1
.19192
.81
20 rows selected.

Functions

5-203

PRESENTNNV

PRESENTNNV
Syntax
PRESENTNNV

(

cell_reference

,

expr1

,

expr2

)

Purpose
The PRESENTNNV function can be used only in the model_clause of the SELECT
statement and then only on the right-hand side of a model rule. It returns expr1 when
cell_reference exists prior to the execution of the model_clause and is not null when
PRESENTNNV is evaluated. Otherwise it returns expr2. This function differs from NVL2 in
that NVL2 evaluates the data at the time it is executed, rather than evaluating the data
as it was prior to the execution of the model_clause.
See Also:
■

■

model_clause on page 19-28 and "Model Expressions" on page 6-11
for the syntax and semantics
NVL2 on page 5-174 for comparison

Examples
In the following example, if a row containing sales for the Mouse Pad for the year 2002
exists, and the sales value is not null, then the sales value remains unchanged. If the
row exists and the sales value is null, then the sales value is set to 10. If the row does
not exist, then the row is created with the sales value set to 10.
SELECT country, prod, year, s
FROM sales_view_ref
MODEL
PARTITION BY (country)
DIMENSION BY (prod, year)
MEASURES (sale s)
IGNORE NAV
UNIQUE DIMENSION
RULES UPSERT SEQUENTIAL ORDER
( s['Mouse Pad', 2002] =
PRESENTNNV(s['Mouse Pad', 2002], s['Mouse Pad', 2002], 10)
)
ORDER BY country, prod, year;
COUNTRY
---------France
France
France
France
France
France
France
France
France
Germany
Germany
Germany
Germany
Germany

PROD
----------------------------------Mouse Pad
Mouse Pad
Mouse Pad
Mouse Pad
Mouse Pad
Standard Mouse
Standard Mouse
Standard Mouse
Standard Mouse
Mouse Pad
Mouse Pad
Mouse Pad
Mouse Pad
Mouse Pad

5-204 Oracle Database SQL Language Reference

YEAR
-------1998
1999
2000
2001
2002
1998
1999
2000
2001
1998
1999
2000
2001
2002

S
--------2509.42
3678.69
3000.72
3269.09
10
2390.83
2280.45
1274.31
2164.54
5827.87
8346.44
7375.46
9535.08
10

PRESENTNNV

Germany
Germany
Germany
Germany

Standard
Standard
Standard
Standard

Mouse
Mouse
Mouse
Mouse

1998
1999
2000
2001

7116.11
6263.14
2637.31
6456.13

18 rows selected.

The preceding example requires the view sales_view_ref. Refer to "Examples" on
page 19-36 to create this view.

Functions

5-205

PRESENTV

PRESENTV
Syntax
PRESENTV

(

cell_reference

,

expr1

,

expr2

)

Purpose
The PRESENTV function can be used only within the model_clause of the SELECT
statement and then only on the right-hand side of a model rule. It returns expr1 when,
prior to the execution of the model_clause, cell_reference exists. Otherwise it
returns expr2.
model_clause on page 19-28 and "Model Expressions" on
page 6-11 for the syntax and semantics

See Also:

Examples
In the following example, if a row containing sales for the Mouse Pad for the year 2000
exists, then the sales value for the Mouse Pad for the year 2001 is set to the sales value
for the Mouse Pad for the year 2000. If the row does not exist, then a row is created
with the sales value for the Mouse Pad for year 20001 set to 0.
SELECT country, prod, year, s
FROM sales_view_ref
MODEL
PARTITION BY (country)
DIMENSION BY (prod, year)
MEASURES (sale s)
IGNORE NAV
UNIQUE DIMENSION
RULES UPSERT SEQUENTIAL ORDER
(
s['Mouse Pad', 2001] =
PRESENTV(s['Mouse Pad', 2000], s['Mouse Pad', 2000], 0)
)
ORDER BY country, prod, year;
COUNTRY
---------France
France
France
France
France
France
France
France
Germany
Germany
Germany
Germany
Germany
Germany
Germany
Germany

PROD
----------------------------------Mouse Pad
Mouse Pad
Mouse Pad
Mouse Pad
Standard Mouse
Standard Mouse
Standard Mouse
Standard Mouse
Mouse Pad
Mouse Pad
Mouse Pad
Mouse Pad
Standard Mouse
Standard Mouse
Standard Mouse
Standard Mouse

16 rows selected.

5-206 Oracle Database SQL Language Reference

YEAR
-------1998
1999
2000
2001
1998
1999
2000
2001
1998
1999
2000
2001
1998
1999
2000
2001

S
--------2509.42
3678.69
3000.72
3000.72
2390.83
2280.45
1274.31
2164.54
5827.87
8346.44
7375.46
7375.46
7116.11
6263.14
2637.31
6456.13

PRESENTV

The preceding example requires the view sales_view_ref. Refer to "The MODEL
clause: Examples" on page 19-45 to create this view.

Functions

5-207

PREVIOUS

PREVIOUS
Syntax
PREVIOUS

(

cell_reference

)

Purpose
The PREVIOUS function can be used only in the model_clause of the SELECT statement
and then only in the ITERATE ... [ UNTIL ] clause of the model_rules_clause. It returns
the value of cell_reference at the beginning of each iteration.
model_clause on page 19-28 and "Model Expressions" on
page 6-11 for the syntax and semantics

See Also:

Examples
The following example repeats the rules, up to 1000 times, until the difference between
the values of cur_val at the beginning and at the end of an iteration is less than one:
SELECT dim_col, cur_val, num_of_iterations
FROM (SELECT 1 AS dim_col, 10 AS cur_val FROM dual)
MODEL
DIMENSION BY (dim_col)
MEASURES (cur_val, 0 num_of_iterations)
IGNORE NAV
UNIQUE DIMENSION
RULES ITERATE (1000) UNTIL (PREVIOUS(cur_val[1]) - cur_val[1] < 1)
(
cur_val[1] = cur_val[1]/2,
num_of_iterations[1] = num_of_iterations[1] + 1
);
DIM_COL
CUR_VAL NUM_OF_ITERATIONS
---------- ---------- ----------------1
.625
4

5-208 Oracle Database SQL Language Reference

RANK

RANK
Aggregate Syntax
rank_aggregate::=
,
RANK

(

expr

)

WITHIN

GROUP

,
DESC

FIRST
NULLS

ASC
(

ORDER

BY

LAST

expr

)

Analytic Syntax
rank_analytic::=
query_partition_clause
RANK

(

)

OVER

(

order_by_clause

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions

Purpose
RANK calculates the rank of a value in a group of values. The return type is NUMBER.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion and "Numeric
Precedence" on page 3-14 for information on numeric precedence

Rows with equal values for the ranking criteria receive the same rank. Oracle Database
then adds the number of tied rows to the tied rank to calculate the next rank.
Therefore, the ranks may not be consecutive numbers. This function is useful for top-N
and bottom-N reporting.
■

■

As an aggregate function, RANK calculates the rank of a hypothetical row identified
by the arguments of the function with respect to a given sort specification. The
arguments of the function must all evaluate to constant expressions within each
aggregate group, because they identify a single row within each group. The
constant argument expressions and the expressions in the ORDER BY clause of the
aggregate match by position. Therefore, the number of arguments must be the
same and their types must be compatible.
As an analytic function, RANK computes the rank of each row returned from a
query with respect to the other rows returned by the query, based on the values of
the value_exprs in the order_by_clause.

Functions

5-209

RANK

Aggregate Example
The following example calculates the rank of a hypothetical employee in the sample
table hr.employees with a salary of $15,500 and a commission of 5%:
SELECT RANK(15500, .05) WITHIN GROUP
(ORDER BY salary, commission_pct) "Rank"
FROM employees;
Rank
---------105

Similarly, the following query returns the rank for a $15,500 salary among the
employee salaries:
SELECT RANK(15500) WITHIN GROUP
(ORDER BY salary DESC) "Rank of 15500"
FROM employees;
Rank of 15500
-------------4

Analytic Example
The following statement ranks the employees in the sample hr schema in department
60 based on their salaries. Identical salary values receive the same rank and cause
nonconsecutive ranks. Compare this example with the analytic example for DENSE_
RANK on page 5-82.
SELECT department_id, last_name, salary,
RANK() OVER (PARTITION BY department_id ORDER BY salary) RANK
FROM employees WHERE department_id = 60
ORDER BY RANK, last_name;
DEPARTMENT_ID
------------60
60
60
60
60

LAST_NAME
SALARY
RANK
------------------------- ---------- ---------Lorentz
4200
1
Austin
4800
2
Pataballa
4800
2
Ernst
6000
4
Hunold
9000
5

5-210 Oracle Database SQL Language Reference

RATIO_TO_REPORT

RATIO_TO_REPORT
Syntax
query_partition_clause
RATIO_TO_REPORT

(

expr

)

OVER

(

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions, including valid forms of expr

Purpose
RATIO_TO_REPORT is an analytic function. It computes the ratio of a value to the sum of
a set of values. If expr evaluates to null, then the ratio-to-report value also evaluates to
null.
The set of values is determined by the query_partition_clause. If you omit that
clause, then the ratio-to-report is computed over all rows returned by the query.
You cannot nest analytic functions by using RATIO_TO_REPORT or any other analytic
function for expr. However, you can use other built-in function expressions for expr.
Refer to "About SQL Expressions" on page 6-1 for information on valid forms of expr.

Examples
The following example calculates the ratio-to-report value of each purchasing clerk's
salary to the total of all purchasing clerks' salaries:
SELECT last_name, salary, RATIO_TO_REPORT(salary) OVER () AS rr
FROM employees
WHERE job_id = 'PU_CLERK'
ORDER BY last_name, salary, rr;
LAST_NAME
SALARY
RR
------------------------- ---------- ---------Baida
2900 .208633094
Colmenares
2500 .179856115
Himuro
2600 .18705036
Khoo
3100 .223021583
Tobias
2800 .201438849

Functions

5-211

RAWTOHEX

RAWTOHEX
Syntax
RAWTOHEX

(

raw

)

Purpose
RAWTOHEX converts raw to a character value containing its hexadecimal representation.
As a SQL built-in function, RAWTOHEX accepts an argument of any scalar data type other
than LONG, LONG RAW, CLOB, BLOB, or BFILE. It returns a VARCHAR2 value with the
hexadecimal representation of bytes that make up the value of raw. Each byte is
represented by two hexadecimal digits.
RAWTOHEX behaves differently when used as a PL/SQL built-in
function. Refer to Oracle Database PL/SQL Language Reference for more
information.

Note:

Examples
The following hypothetical example returns the hexadecimal equivalent of a RAW
column value:
SELECT RAWTOHEX(raw_column) "Graphics"
FROM graphics;
Graphics
-------7D

See Also: "RAW and LONG RAW Data Types" on page 3-23 and
HEXTORAW on page 5-112

5-212 Oracle Database SQL Language Reference

RAWTONHEX

RAWTONHEX
Syntax
RAWTONHEX

(

raw

)

Purpose
RAWTONHEX converts raw to a character value containing its hexadecimal representation.
RAWTONHEX (raw) is equivalent to TO_NCHAR(RAWTOHEX(raw)). The value returned is
always in the national character set.

Examples
The following hypothetical example returns the hexadecimal equivalent of a RAW
column value:
SELECT RAWTONHEX(raw_column),
DUMP ( RAWTONHEX (raw_column) ) "DUMP"
FROM graphics;
RAWTONHEX(RA)
DUMP
----------------------- -----------------------------7D
Typ=1 Len=4: 0,55,0,68

See Also:

RAWTOHEX on page 5-212

Functions

5-213

REF

REF
Syntax
REF

(

correlation_variable

)

Purpose
REF takes as its argument a correlation variable (table alias) associated with a row of an
object table or an object view. A REF value is returned for the object instance that is
bound to the variable or row.

Examples
The sample schema oe contains a type called cust_address_typ, described as follows:
Attribute
Type
----------------------------- ---------------STREET_ADDRESS
VARCHAR2(40)
POSTAL_CODE
VARCHAR2(10)
CITY
VARCHAR2(30)
STATE_PROVINCE
VARCHAR2(10)
COUNTRY_ID
CHAR(2)

The following example creates a table based on the sample type oe.cust_address_typ,
inserts a row into the table, and retrieves a REF value for the object instance of the type
in the addresses table:
CREATE TABLE addresses OF cust_address_typ;
INSERT INTO addresses VALUES (
'123 First Street', '4GF H1J', 'Our Town', 'Ourcounty', 'US');
SELECT REF(e) FROM addresses e;
REF(E)
----------------------------------------------------------------------------------00002802097CD1261E51925B60E0340800208254367CD1261E51905B60E034080020825436010101820000

See Also: Oracle Database Object-Relational Developer's Guide for
information on REFs

5-214 Oracle Database SQL Language Reference

REFTOHEX

REFTOHEX
Syntax
REFTOHEX

(

expr

)

Purpose
REFTOHEX converts argument expr to a character value containing its hexadecimal
equivalent. expr must return a REF.

Examples
The sample schema oe contains a warehouse_typ. The following example builds on
that type to illustrate how to convert the REF value of a column to a character value
containing its hexadecimal equivalent:
CREATE TABLE warehouse_table OF warehouse_typ
(PRIMARY KEY (warehouse_id));
CREATE TABLE location_table
(location_number NUMBER, building REF warehouse_typ
SCOPE IS warehouse_table);
INSERT INTO warehouse_table VALUES (1, 'Downtown', 99);
INSERT INTO location_table SELECT 10, REF(w) FROM warehouse_table w;
SELECT REFTOHEX(building) FROM location_table;
REFTOHEX(BUILDING)
-------------------------------------------------------------------------0000220208859B5E9255C31760E034080020825436859B5E9255C21760E034080020825436

Functions

5-215

REGEXP_COUNT

REGEXP_COUNT
Syntax
,
,
REGEXP_COUNT

(

source_char

,

match_param

position

pattern

)

Purpose
REGEXP_COUNT complements the functionality of the REGEXP_INSTR function by
returning the number of times a pattern occurs in a source string. The function
evaluates strings using characters as defined by the input character set. It returns an
integer indicating the number of occurrences of pattern. If no match is found, then the
function returns 0.
■

■

source_char is a character expression that serves as the search value. It is
commonly a character column and can be of any of the data types CHAR, VARCHAR2,
NCHAR, NVARCHAR2, CLOB, or NCLOB.
pattern is the regular expression. It is usually a text literal and can be of any of the
data types CHAR, VARCHAR2, NCHAR, or NVARCHAR2. It can contain up to 512 bytes. If
the data type of pattern is different from the data type of source_char, then
Oracle Database converts pattern to the data type of source_char.
REGEXP_COUNT ignores subexpression parentheses in pattern. For example, the
pattern '(123(45))' is equivalent to '12345'. For a listing of the operators you
can specify in pattern, refer to Appendix D, "Oracle Regular Expression Support".

■

■

position is a positive integer indicating the character of source_char where
Oracle should begin the search. The default is 1, meaning that Oracle begins the
search at the first character of source_char. After finding the first occurrence of
pattern, the database searches for a second occurrence beginning with the first
character following the first occurrence.
match_param is a text literal that lets you change the default matching behavior of
the function. You can specify one or more of the following values for match_param:
–

'i' specifies case-insensitive matching.

–

'c' specifies case-sensitive matching.

–

'n' allows the period (.), which is the match-any-character character, to match
the newline character. If you omit this parameter, then the period does not
match the newline character.

–

'm' treats the source string as multiple lines. Oracle interprets the caret (^) and
dollar sign ($) as the start and end, respectively, of any line anywhere in the
source string, rather than only at the start or end of the entire source string. If
you omit this parameter, then Oracle treats the source string as a single line.

–

'x' ignores whitespace characters. By default, whitespace characters match
themselves.

If you specify multiple contradictory values, then Oracle uses the last value. For
example, if you specify 'ic', then Oracle uses case-sensitive matching. If you
specify a character other than those shown above, then Oracle returns an error.
If you omit match_param, then:

5-216 Oracle Database SQL Language Reference

REGEXP_COUNT

–

The default case sensitivity is determined by the value of the NLS_SORT
parameter.

–

A period (.) does not match the newline character.

–

The source string is treated as a single line.

Examples
The following example shows that subexpressions parentheses in pattern are ignored:
SELECT REGEXP_COUNT('123123123123123', '(12)3', 1, 'i') REGEXP_COUNT
FROM DUAL;
REGEXP_COUNT
-----------5

In the following example, the function begins to evaluate the source string at the third
character, so skips over the first occurrence of pattern:
SELECT REGEXP_COUNT('123123123123', '123', 3, 'i') COUNT FROM DUAL;
COUNT
---------3

Functions

5-217

REGEXP_INSTR

REGEXP_INSTR
Syntax
REGEXP_INSTR

(

source_char

,

pattern

,
,
,
,
,

subexpr

match_param

return_opt

occurrence

position
)

Purpose
REGEXP_INSTR extends the functionality of the INSTR function by letting you search a
string for a regular expression pattern. The function evaluates strings using characters
as defined by the input character set. It returns an integer indicating the beginning or
ending position of the matched substring, depending on the value of the return_
option argument. If no match is found, then the function returns 0.
This function complies with the POSIX regular expression standard and the Unicode
Regular Expression Guidelines. For more information, refer to Appendix D, "Oracle
Regular Expression Support".
■

■

■

■

■

source_char is a character expression that serves as the search value. It is
commonly a character column and can be of any of the data types CHAR, VARCHAR2,
NCHAR, NVARCHAR2, CLOB, or NCLOB.
pattern is the regular expression. It is usually a text literal and can be of any of the
data types CHAR, VARCHAR2, NCHAR, or NVARCHAR2. It can contain up to 512 bytes. If
the data type of pattern is different from the data type of source_char, then
Oracle Database converts pattern to the data type of source_char. For a listing of
the operators you can specify in pattern, refer to Appendix D, "Oracle Regular
Expression Support".
position is a positive integer indicating the character of source_char where
Oracle should begin the search. The default is 1, meaning that Oracle begins the
search at the first character of source_char.
occurrence is a positive integer indicating which occurrence of pattern in
source_char Oracle should search for. The default is 1, meaning that Oracle
searches for the first occurrence of pattern. If occurrence is greater than 1, then
the database searches for the second occurrence beginning with the first character
following the first occurrence of pattern, and so forth. This behavior is different
from the INSTR function, which begins its search for the second occurrence at the
second character of the first occurrence.
return_option lets you specify what Oracle should return in relation to the
occurrence:
–

If you specify 0, then Oracle returns the position of the first character of the
occurrence. This is the default.

–

If you specify 1, then Oracle returns the position of the character following the
occurrence.

5-218 Oracle Database SQL Language Reference

REGEXP_INSTR

■

■

match_parameter is a text literal that lets you change the default matching
behavior of the function. The behavior of this parameter is the same for this
function as for REGEXP_COUNT. Refer to REGEXP_COUNT on page 5-216 for
detailed information.
For a pattern with subexpressions, subexpr is an integer from 0 to 9 indicating
which subexpression in pattern is the target of the function. The subexpr is a
fragment of pattern enclosed in parentheses. Subexpressions can be nested.
Subexpressions are numbered in order in which their left parentheses appear in
pattern. For example, consider the following expression:
0123(((abc)(de)f)ghi)45(678)

This expression has five subexpressions in the following order: "abcdefghi"
followed by "abcdef", "abc", "de" and "678".
If subexpr is zero, then the position of the entire substring that matches the
pattern is returned. If subexpr is greater than zero, then the position of the
substring fragment that corresponds to subexpression number subexpr in the
matched substring is returned. If pattern does not have at least subexpr
subexpressions, the function returns zero. A null subexpr value returns NULL. The
default value for subexpr is zero.
See Also:
■
■

INSTR on page 5-120 and REGEXP_SUBSTR on page 5-224
REGEXP_REPLACE on page 5-221 and REGEXP_LIKE Condition
on page 7-18

Examples
The following example examines the string, looking for occurrences of one or more
non-blank characters. Oracle begins searching at the first character in the string and
returns the starting position (default) of the sixth occurrence of one or more non-blank
characters.
SELECT
REGEXP_INSTR('500 Oracle Parkway, Redwood Shores, CA',
'[^ ]+', 1, 6) "REGEXP_INSTR"
FROM DUAL;
REGEXP_INSTR
-----------37

The following example examines the string, looking for occurrences of words
beginning with s, r, or p, regardless of case, followed by any six alphabetic characters.
Oracle begins searching at the third character in the string and returns the position in
the string of the character following the second occurrence of a seven-letter word
beginning with s, r, or p, regardless of case.
SELECT
REGEXP_INSTR('500 Oracle Parkway, Redwood Shores, CA',
'[s|r|p][[:alpha:]]{6}', 3, 2, 1, 'i') "REGEXP_INSTR"
FROM DUAL;
REGEXP_INSTR
-----------28

Functions

5-219

REGEXP_INSTR

The following examples use the subexpr argument to search for a particular
subexpression in pattern. The first statement returns the position in the source string
of the first character in the first subexpression, which is '123':
SELECT REGEXP_INSTR('1234567890', '(123)(4(56)(78))', 1, 1, 0, 'i', 1)
"REGEXP_INSTR" FROM DUAL;
REGEXP_INSTR
------------------1

The next statement returns the position in the source string of the first character in the
second subexpression, which is '45678':
SELECT REGEXP_INSTR('1234567890', '(123)(4(56)(78))', 1, 1, 0, 'i', 2)
"REGEXP_INSTR" FROM DUAL;
REGEXP_INSTR
------------------4

The next statement returns the position in the source string of the first character in the
fourth subexpression, which is '78':
SELECT REGEXP_INSTR('1234567890', '(123)(4(56)(78))', 1, 1, 0, 'i', 4)
"REGEXP_INSTR" FROM DUAL;
REGEXP_INSTR
------------------7

5-220 Oracle Database SQL Language Reference

REGEXP_REPLACE

REGEXP_REPLACE
Syntax
REGEXP_REPLACE

(

source_char

,

pattern

,
,
,
,

match_param

occurrence

position

replace_string
)

Purpose
REGEXP_REPLACE extends the functionality of the REPLACE function by letting you
search a string for a regular expression pattern. By default, the function returns
source_char with every occurrence of the regular expression pattern replaced with
replace_string. The string returned is in the same character set as source_char. The
function returns VARCHAR2 if the first argument is not a LOB and returns CLOB if the
first argument is a LOB.
This function complies with the POSIX regular expression standard and the Unicode
Regular Expression Guidelines. For more information, refer to Appendix D, "Oracle
Regular Expression Support".
■

■

■

■

■

source_char is a character expression that serves as the search value. It is
commonly a character column and can be of any of the data types CHAR, VARCHAR2,
NCHAR, NVARCHAR2, CLOB or NCLOB.
pattern is the regular expression. It is usually a text literal and can be of any of the
data types CHAR, VARCHAR2, NCHAR, or NVARCHAR2. It can contain up to 512 bytes. If
the data type of pattern is different from the data type of source_char, then
Oracle Database converts pattern to the data type of source_char. For a listing of
the operators you can specify in pattern, refer to Appendix D, "Oracle Regular
Expression Support".
replace_string can be of any of the data types CHAR, VARCHAR2, NCHAR, NVARCHAR2,
CLOB, or NCLOB. If replace_string is a CLOB or NCLOB, then Oracle truncates
replace_string to 32K. The replace_string can contain up to 500 backreferences
to subexpressions in the form \n, where n is a number from 1 to 9. If you want to
include a backslash (\) in replace_string, then you must precede it with the
escape character, which is also a backslash. For example, to replace \2 you would
enter \\2. For more information on backreference expressions, refer to the notes to
"Oracle Regular Expression Support", Table D–1 on page D-1.
position is a positive integer indicating the character of source_char where
Oracle should begin the search. The default is 1, meaning that Oracle begins the
search at the first character of source_char.
occurrence is a nonnegative integer indicating the occurrence of the replace
operation:
–

If you specify 0, then Oracle replaces all occurrences of the match.

–

If you specify a positive integer n, then Oracle replaces the nth occurrence.

If occurrence is greater than 1, then the database searches for the second
occurrence beginning with the first character following the first occurrence of
Functions

5-221

REGEXP_REPLACE

pattern, and so forth. This behavior is different from the INSTR function, which
begins its search for the second occurrence at the second character of the first
occurrence.
■

match_parameter is a text literal that lets you change the default matching
behavior of the function. The behavior of this parameter is the same for this
function as for REGEXP_COUNT. Refer to REGEXP_COUNT on page 5-216 for
detailed information.
See Also:
■
■

REPLACE on page 5-232
REGEXP_INSTR on page 5-218, REGEXP_SUBSTR on page 5-224,
and REGEXP_LIKE Condition on page 7-18

Examples
The following example examines phone_number, looking for the pattern xxx.xxx.xxxx.
Oracle reformats this pattern with (xxx) xxx-xxxx.
SELECT
REGEXP_REPLACE(phone_number,
'([[:digit:]]{3})\.([[:digit:]]{3})\.([[:digit:]]{4})',
'(\1) \2-\3') "REGEXP_REPLACE"
FROM employees
ORDER BY "REGEXP_REPLACE";
REGEXP_REPLACE
-------------------------------------------------------------------------------(515) 123-4444
(515) 123-4567
(515) 123-4568
(515) 123-4569
(515) 123-5555
. . .

The following example examines country_name. Oracle puts a space after each
non-null character in the string.
SELECT
REGEXP_REPLACE(country_name, '(.)', '\1 ') "REGEXP_REPLACE"
FROM countries;
REGEXP_REPLACE
-------------------------------------------------------------------------------A r g e n t i n a
A u s t r a l i a
B e l g i u m
B r a z i l
C a n a d a
. . .

The following example examines the string, looking for two or more spaces. Oracle
replaces each occurrence of two or more spaces with a single space.
SELECT
REGEXP_REPLACE('500
Oracle
Parkway,
Redwood
'( ){2,}', ' ') "REGEXP_REPLACE"
FROM DUAL;
REGEXP_REPLACE

5-222 Oracle Database SQL Language Reference

Shores, CA',

REGEXP_REPLACE

-------------------------------------500 Oracle Parkway, Redwood Shores, CA

Functions

5-223

REGEXP_SUBSTR

REGEXP_SUBSTR
Syntax
REGEXP_SUBSTR

(

source_char

,

pattern

subexpr
,
,
,

match_param

occurrence

position
)

Purpose
REGEXP_SUBSTR extends the functionality of the SUBSTR function by letting you search a
string for a regular expression pattern. It is also similar to REGEXP_INSTR, but instead of
returning the position of the substring, it returns the substring itself. This function is
useful if you need the contents of a match string but not its position in the source
string. The function returns the string as VARCHAR2 or CLOB data in the same character
set as source_char.
This function complies with the POSIX regular expression standard and the Unicode
Regular Expression Guidelines. For more information, refer to Appendix D, "Oracle
Regular Expression Support".
■

■

■

■

source_char is a character expression that serves as the search value. It is
commonly a character column and can be of any of the data types CHAR, VARCHAR2,
NCHAR, NVARCHAR2, CLOB, or NCLOB.
pattern is the regular expression. It is usually a text literal and can be of any of the
data types CHAR, VARCHAR2, NCHAR, or NVARCHAR2. It can contain up to 512 bytes. If
the data type of pattern is different from the data type of source_char, then
Oracle Database converts pattern to the data type of source_char. For a listing of
the operators you can specify in pattern, refer to Appendix D, "Oracle Regular
Expression Support".
position is a positive integer indicating the character of source_char where
Oracle should begin the search. The default is 1, meaning that Oracle begins the
search at the first character of source_char.
occurrence is a positive integer indicating which occurrence of pattern in
source_char Oracle should search for. The default is 1, meaning that Oracle
searches for the first occurrence of pattern.
If occurrence is greater than 1, then the database searches for the second
occurrence beginning with the first character following the first occurrence of
pattern, and so forth. This behavior is different from the SUBSTR function, which
begins its search for the second occurrence at the second character of the first
occurrence.

■

■

match_parameter is a text literal that lets you change the default matching
behavior of the function. The behavior of this parameter is the same for this
function as for REGEXP_COUNT. Refer to REGEXP_COUNT on page 5-216 for
detailed information.
For a pattern with subexpressions, subexpr is a nonnegative integer from 0 to 9
indicating which subexpression in pattern is to be returned by the function. This

5-224 Oracle Database SQL Language Reference

REGEXP_SUBSTR

parameter has the same semantics that it has for the REGEXP_INSTR function. Refer
to REGEXP_INSTR on page 5-218 for more information.
See Also:
■
■

SUBSTR on page 5-274 and REGEXP_INSTR on page 5-218
REGEXP_REPLACE on page 5-221, and REGEXP_LIKE Condition
on page 7-18

Examples
The following example examines the string, looking for the first substring bounded by
commas. Oracle Database searches for a comma followed by one or more occurrences
of non-comma characters followed by a comma. Oracle returns the substring,
including the leading and trailing commas.
SELECT
REGEXP_SUBSTR('500 Oracle Parkway, Redwood Shores, CA',
',[^,]+,') "REGEXPR_SUBSTR"
FROM DUAL;
REGEXPR_SUBSTR
----------------, Redwood Shores,

The following example examines the string, looking for http:// followed by a
substring of one or more alphanumeric characters and optionally, a period (.). Oracle
searches for a minimum of three and a maximum of four occurrences of this substring
between http:// and either a slash (/) or the end of the string.
SELECT
REGEXP_SUBSTR('http://www.example.com/products',
'http://([[:alnum:]]+\.?){3,4}/?') "REGEXP_SUBSTR"
FROM DUAL;
REGEXP_SUBSTR
---------------------http://www.example.com/

The next two examples use the subexpr argument to return a specific subexpression of
pattern. The first statement returns the first subexpression in pattern:
SELECT REGEXP_SUBSTR('1234567890', '(123)(4(56)(78))', 1, 1, 'i', 1)
"REGEXP_SUBSTR" FROM DUAL;
REGEXP_SUBSTR
------------------123

The next statement returns the fourth subexpression in pattern:
SELECT REGEXP_SUBSTR('1234567890', '(123)(4(56)(78))', 1, 1, 'i', 4)
"REGEXP_SUBSTR" FROM DUAL;
REGEXP_SUBSTR
------------------78

Functions

5-225

REGR_ (Linear Regression) Functions

REGR_ (Linear Regression) Functions
The linear regression functions are:
■

REGR_SLOPE

■

REGR_INTERCEPT

■

REGR_COUNT

■

REGR_R2

■

REGR_AVGX

■

REGR_AVGY

■

REGR_SXX

■

REGR_SYY

■

REGR_SXY

Syntax
linear_regr::=
REGR_SLOPE
REGR_INTERCEPT
REGR_COUNT
REGR_R2
REGR_AVGX

OVER
(

expr1

,

expr2

(

analytic_clause

)

)

REGR_AVGY
REGR_SXX
REGR_SYY
REGR_SXY

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions

Purpose
The linear regression functions fit an ordinary-least-squares regression line to a set of
number pairs. You can use them as both aggregate and analytic functions.
See Also: "Aggregate Functions" on page 5-10 and "About SQL
Expressions" on page 6-1 for information on valid forms of expr

These functions take as arguments any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. Oracle determines the
argument with the highest numeric precedence, implicitly converts the remaining
arguments to that data type, and returns that data type.

5-226 Oracle Database SQL Language Reference

REGR_ (Linear Regression) Functions

See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion and "Numeric
Precedence" on page 3-14 for information on numeric precedence

Oracle applies the function to the set of (expr1, expr2) pairs after eliminating all pairs
for which either expr1 or expr2 is null. Oracle computes all the regression functions
simultaneously during a single pass through the data.
expr1 is interpreted as a value of the dependent variable (a y value), and expr2 is
interpreted as a value of the independent variable (an x value).
■

REGR_SLOPE returns the slope of the line. The return value is a numeric data type
and can be null. After the elimination of null (expr1, expr2) pairs, it makes the
following computation:
COVAR_POP(expr1, expr2) / VAR_POP(expr2)

■

REGR_INTERCEPT returns the y-intercept of the regression line. The return value is a
numeric data type and can be null. After the elimination of null (expr1, expr2)
pairs, it makes the following computation:
AVG(expr1) - REGR_SLOPE(expr1, expr2) * AVG(expr2)

■

■

REGR_COUNT returns an integer that is the number of non-null number pairs used to
fit the regression line.
REGR_R2 returns the coefficient of determination (also called R-squared or
goodness of fit) for the regression. The return value is a numeric data type and can
be null. VAR_POP(expr1) and VAR_POP(expr2) are evaluated after the elimination of
null pairs. The return values are:
NULL if VAR_POP(expr2)

= 0

1 if VAR_POP(expr1) = 0 and
VAR_POP(expr2) != 0
POWER(CORR(expr1,expr),2) if VAR_POP(expr1) > 0 and
VAR_POP(expr2 != 0

All of the remaining regression functions return a numeric data type and can be null:
■

REGR_AVGX evaluates the average of the independent variable (expr2) of the
regression line. It makes the following computation after the elimination of null
(expr1, expr2) pairs:
AVG(expr2)

■

REGR_AVGY evaluates the average of the dependent variable (expr1) of the
regression line. It makes the following computation after the elimination of null
(expr1, expr2) pairs:
AVG(expr1)

REGR_SXY, REGR_SXX, REGR_SYY are auxiliary functions that are used to compute various
diagnostic statistics.
■

REGR_SXX makes the following computation after the elimination of null (expr1,
expr2) pairs:
REGR_COUNT(expr1, expr2) * VAR_POP(expr2)

Functions

5-227

REGR_ (Linear Regression) Functions

■

REGR_SYY makes the following computation after the elimination of null (expr1,
expr2) pairs:
REGR_COUNT(expr1, expr2) * VAR_POP(expr1)

■

REGR_SXY makes the following computation after the elimination of null (expr1,
expr2) pairs:
REGR_COUNT(expr1, expr2) * COVAR_POP(expr1, expr2)

The following examples are based on the sample tables sh.sales and sh.products.

General Linear Regression Example
The following example provides a comparison of the various linear regression
functions used in their analytic form. The analytic form of these functions can be
useful when you want to use regression statistics for calculations such as finding the
salary predicted for each employee by the model. The sections that follow on the
individual linear regression functions contain examples of the aggregate form of these
functions.
SELECT job_id, employee_id ID, salary,
REGR_SLOPE(SYSDATE-hire_date, salary)
OVER (PARTITION BY job_id) slope,
REGR_INTERCEPT(SYSDATE-hire_date, salary)
OVER (PARTITION BY job_id) intcpt,
REGR_R2(SYSDATE-hire_date, salary)
OVER (PARTITION BY job_id) rsqr,
REGR_COUNT(SYSDATE-hire_date, salary)
OVER (PARTITION BY job_id) count,
REGR_AVGX(SYSDATE-hire_date, salary)
OVER (PARTITION BY job_id) avgx,
REGR_AVGY(SYSDATE-hire_date, salary)
OVER (PARTITION BY job_id) avgy
FROM employees
WHERE department_id in (50, 80)
ORDER BY job_id, employee_id;
JOB_ID
ID
SALARY SLOPE
INTCPT RSQR COUNT
AVGX
AVGY
---------- ----- ---------- ----- --------- ----- ------ ---------- --------SA_MAN
145
14000 .355 -1707.035 .832
5 12200.000 2626.589
SA_MAN
146
13500 .355 -1707.035 .832
5 12200.000 2626.589
SA_MAN
147
12000 .355 -1707.035 .832
5 12200.000 2626.589
SA_MAN
148
11000 .355 -1707.035 .832
5 12200.000 2626.589
SA_MAN
149
10500 .355 -1707.035 .832
5 12200.000 2626.589
SA_REP
150
10000 .257
404.763 .647
29
8396.552 2561.244
SA_REP
151
9500 .257
404.763 .647
29
8396.552 2561.244
SA_REP
152
9000 .257
404.763 .647
29
8396.552 2561.244
SA_REP
153
8000 .257
404.763 .647
29
8396.552 2561.244
SA_REP
154
7500 .257
404.763 .647
29
8396.552 2561.244
SA_REP
155
7000 .257
404.763 .647
29
8396.552 2561.244
SA_REP
156
10000 .257
404.763 .647
29
8396.552 2561.244
. . .

REGR_SLOPE and REGR_INTERCEPT Examples
The following example calculates the slope and regression of the linear regression
model for time employed (SYSDATE - hire_date) and salary using the sample table
hr.employees. Results are grouped by job_id.
SELECT job_id,
REGR_SLOPE(SYSDATE-hire_date, salary) slope,
5-228 Oracle Database SQL Language Reference

REGR_ (Linear Regression) Functions

REGR_INTERCEPT(SYSDATE-hire_date, salary) intercept
FROM employees
WHERE department_id in (50,80)
GROUP BY job_id
ORDER BY job_id;
JOB_ID
SLOPE
INTERCEPT
---------- ----- -----------SA_MAN
.355 -1707.030762
SA_REP
.257
404.767151
SH_CLERK
.745
159.015293
ST_CLERK
.904
134.409050
ST_MAN
.479 -570.077291

REGR_COUNT Examples
The following example calculates the count of by job_id for time employed (SYSDATE
- hire_date) and salary using the sample table hr.employees. Results are grouped by
job_id.
SELECT job_id,
REGR_COUNT(SYSDATE-hire_date, salary) count
FROM employees
WHERE department_id in (30, 50)
GROUP BY job_id
ORDER BY job_id, count;
JOB_ID
COUNT
---------- ---------PU_CLERK
5
PU_MAN
1
SH_CLERK
20
ST_CLERK
20
ST_MAN
5

REGR_R2 Examples
The following example calculates the coefficient of determination the linear regression
of time employed (SYSDATE - hire_date) and salary using the sample table
hr.employees:
SELECT job_id,
REGR_R2(SYSDATE-hire_date, salary) Regr_R2
FROM employees
WHERE department_id in (80, 50)
GROUP by job_id
ORDER BY job_id, Regr_R2;
JOB_ID
---------SA_MAN
SA_REP
SH_CLERK
ST_CLERK
ST_MAN

REGR_R2
---------.83244748
.647007156
.879799698
.742808493
.69418508

REGR_AVGY and REGR_AVGX Examples
The following example calculates the average values for time employed (SYSDATE hire_date) and salary using the sample table hr.employees. Results are grouped by
job_id:

Functions

5-229

REGR_ (Linear Regression) Functions

SELECT job_id,
REGR_AVGY(SYSDATE-hire_date, salary) avgy,
REGR_AVGX(SYSDATE-hire_date, salary) avgx
FROM employees
WHERE department_id in (30,50)
GROUP BY job_id
ORDER BY job_id, avgy, avgx;
JOB_ID
AVGY
AVGX
---------- ---------- ---------PU_CLERK
2950.3778
2780
PU_MAN
4026.5778
11000
SH_CLERK
2773.0778
3215
ST_CLERK
2872.7278
2785
ST_MAN
3140.1778
7280

REGR_SXY, REGR_SXX, and REGR_SYY Examples
The following example calculates three types of diagnostic statistics for the linear
regression of time employed (SYSDATE - hire_date) and salary using the sample table
hr.employees:
SELECT job_id,
REGR_SXY(SYSDATE-hire_date, salary) regr_sxy,
REGR_SXX(SYSDATE-hire_date, salary) regr_sxx,
REGR_SYY(SYSDATE-hire_date, salary) regr_syy
FROM employees
WHERE department_id in (80, 50)
GROUP BY job_id
ORDER BY job_id;
JOB_ID
REGR_SXY
REGR_SXX
REGR_SYY
---------- ---------- ----------- ---------SA_MAN
3303500
9300000.0
1409642
SA_REP
16819665.5 65489655.2 6676562.55
SH_CLERK
4248650
5705500.0
3596039
ST_CLERK
3531545
3905500.0 4299084.55
ST_MAN
2180460
4548000.0 1505915.2

5-230 Oracle Database SQL Language Reference

REMAINDER

REMAINDER
Syntax
REMAINDER

(

n2

,

n1

)

Purpose
REMAINDER returns the remainder of n2 divided by n1.
This function takes as arguments any numeric data type or any nonnumeric data type
that can be implicitly converted to a numeric data type. Oracle determines the
argument with the highest numeric precedence, implicitly converts the remaining
arguments to that data type, and returns that data type.
The MOD function is similar to REMAINDER except that it uses FLOOR in its formula,
whereas REMAINDER uses ROUND. Refer to MOD on page 5-152.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion and "Numeric
Precedence" on page 3-14 for information on numeric precedence
■

■

■

If n1 = 0 or n2 = infinity, then Oracle returns
–

An error if the arguments are of type NUMBER

–

NaN if the arguments are BINARY_FLOAT or BINARY_DOUBLE.

If n1 != 0, then the remainder is n2 - (n1*N) where N is the integer nearest n2/n1. If
n2/n1 equals x.5, then N is the nearest even integer.
If n2 is a floating-point number, and if the remainder is 0, then the sign of the
remainder is the sign of n2. Remainders of 0 are unsigned for NUMBER values.

Examples
Using table float_point_demo, created for the TO_BINARY_DOUBLE "Examples" on
page 5-297, the following example divides two floating-point numbers and returns the
remainder of that operation:
SELECT bin_float, bin_double, REMAINDER(bin_float, bin_double)
FROM float_point_demo;
BIN_FLOAT BIN_DOUBLE REMAINDER(BIN_FLOAT,BIN_DOUBLE)
---------- ---------- ------------------------------1.235E+003 1.235E+003
5.859E-005

Functions

5-231

REPLACE

REPLACE
Syntax
,
REPLACE

(

char

,

replacement_string

search_string

)

Purpose
REPLACE returns char with every occurrence of search_string replaced with
replacement_string. If replacement_string is omitted or null, then all occurrences of
search_string are removed. If search_string is null, then char is returned.
Both search_string and replacement_string, as well as char, can be any of the data
types CHAR, VARCHAR2, NCHAR, NVARCHAR2, CLOB, or NCLOB. The string returned is in the
same character set as char. The function returns VARCHAR2 if the first argument is not a
LOB and returns CLOB if the first argument is a LOB.
REPLACE provides functionality related to that provided by the TRANSLATE function.
TRANSLATE provides single-character, one-to-one substitution. REPLACE lets you
substitute one string for another as well as to remove character strings.
See Also:

TRANSLATE on page 5-324

Examples
The following example replaces occurrences of J with BL:
SELECT REPLACE('JACK and JUE','J','BL') "Changes"
FROM DUAL;
Changes
-------------BLACK and BLUE

5-232 Oracle Database SQL Language Reference

ROUND (date)

ROUND (date)
Syntax
round_date::=
,
ROUND

(

date

fmt
)

Purpose
ROUND returns date rounded to the unit specified by the format model fmt. This
function is not sensitive to the NLS_CALENDAR session parameter. It operates according
to the rules of the Gregorian calendar. The value returned is always of data type DATE,
even if you specify a different datetime data type for date. If you omit fmt, then date
is rounded to the nearest day. The date expression must resolve to a DATE value.
"ROUND and TRUNC Date Functions" on page 5-379 for
the permitted format models to use in fmt

See Also:

Examples
The following example rounds a date to the first day of the following year:
SELECT ROUND (TO_DATE ('27-OCT-00'),'YEAR')
"New Year" FROM DUAL;
New Year
--------01-JAN-01

Functions

5-233

ROUND (number)

ROUND (number)
Syntax
round_number::=
,
ROUND

(

integer

n

)

Purpose
ROUND returns n rounded to integer places to the right of the decimal point. If you
omit integer, then n is rounded to zero places. If integer is negative, then n is
rounded off to the left of the decimal point.
n can be any numeric data type or any nonnumeric data type that can be implicitly
converted to a numeric data type. If you omit integer, then the function returns the
value ROUND(n, 0) in the same data type as the numeric data type of n. If you include
integer, then the function returns NUMBER.
ROUND is implemented using the following rules:
1.

If n is 0, then ROUND always returns 0 regardless of integer.

2.

If n is negative, then ROUND(n, integer) returns -ROUND(-n, integer).

3.

If n is positive, then
ROUND(n, integer) = FLOOR(n * POWER(10, integer) + 0.5) * POWER(10, -integer)

ROUND applied to a NUMBER value may give a slightly different result from ROUND applied
to the same value expressed in floating-point. The different results arise from
differences in internal representations of NUMBER and floating point values. The
difference will be 1 in the rounded digit if a difference occurs.
See Also:
■

■

■

Table 3–10, " Implicit Type Conversion Matrix" on page 3-40 for
more information on implicit conversion
"Floating-Point Numbers" on page 3-12 for more information on
how Oracle Database handles BINARY_FLOAT and BINARY_DOUBLE
values
FLOOR on page 5-105 and CEIL on page 5-38, TRUNC (number)
on page 5-330 and MOD on page 5-152 for information on
functions that perform related operations

Examples
The following example rounds a number to one decimal point:
SELECT ROUND(15.193,1) "Round" FROM DUAL;
Round
---------15.2

The following example rounds a number one digit to the left of the decimal point:

5-234 Oracle Database SQL Language Reference

ROUND (number)

SELECT ROUND(15.193,-1) "Round" FROM DUAL;
Round
---------20

Functions

5-235

ROW_NUMBER

ROW_NUMBER
Syntax
query_partition_clause
ROW_NUMBER

(

)

OVER

(

order_by_clause

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions

Purpose
ROW_NUMBER is an analytic function. It assigns a unique number to each row to which it
is applied (either each row in the partition or each row returned by the query), in the
ordered sequence of rows specified in the order_by_clause, beginning with 1.
By nesting a subquery using ROW_NUMBER inside a query that retrieves the ROW_NUMBER
values for a specified range, you can find a precise subset of rows from the results of
the inner query. This use of the function lets you implement top-N, bottom-N, and
inner-N reporting. For consistent results, the query must ensure a deterministic sort
order.
You cannot nest analytic functions by using ROW_NUMBER or any other analytic function
for expr. However, you can use other built-in function expressions for expr. Refer to
"About SQL Expressions" on page 6-1 for information on valid forms of expr.

Examples
The following example finds the three highest paid employees in each department in
the hr.employees table. Fewer than three rows are returned for departments with
fewer than three employees.
SELECT department_id, first_name, last_name, salary
FROM
(
SELECT
department_id, first_name, last_name, salary,
ROW_NUMBER() OVER (PARTITION BY department_id ORDER BY salary desc) rn
FROM employees
)
WHERE rn <= 3
ORDER BY department_id, salary DESC, last_name;

The following example is a join query on the sh.sales table. It finds the sales amounts
in 2000 of the five top-selling products in 1999 and compares the difference between
2000 and 1999. The ten top-selling products are calculated within each distribution
channel.
SELECT sales_2000.channel_desc, sales_2000.prod_name,
sales_2000.amt amt_2000, top_5_prods_1999_year.amt amt_1999,
sales_2000.amt - top_5_prods_1999_year.amt amt_diff
FROM
/* The first subquery finds the 5 top-selling products per channel in year 1999. */
(SELECT channel_desc, prod_name, amt
FROM
(
SELECT channel_desc, prod_name, sum(amount_sold) amt,
ROW_NUMBER () OVER (PARTITION BY channel_desc

5-236 Oracle Database SQL Language Reference

ROW_NUMBER

ORDER BY SUM(amount_sold) DESC) rn
FROM sales, times, channels, products
WHERE sales.time_id = times.time_id
AND times.calendar_year = 1999
AND channels.channel_id = sales.channel_id
AND products.prod_id = sales.prod_id
GROUP BY channel_desc, prod_name
)
WHERE rn <= 5
) top_5_prods_1999_year,
/* The next subquery finds sales per product and per channel in 2000. */
(SELECT channel_desc, prod_name, sum(amount_sold) amt
FROM sales, times, channels, products
WHERE sales.time_id = times.time_id
AND times.calendar_year = 2000
AND channels.channel_id = sales.channel_id
AND products.prod_id = sales.prod_id
GROUP BY channel_desc, prod_name
) sales_2000
WHERE sales_2000.channel_desc = top_5_prods_1999_year.channel_desc
AND sales_2000.prod_name = top_5_prods_1999_year.prod_name
ORDER BY sales_2000.channel_desc, sales_2000.prod_name
;
CHANNEL_DESC
PROD_NAME
AMT_2000
--------------- --------------==-------------------------------- ---------Direct Sales
17" LCD w/built-in HDTV Tuner
628855.7
Direct Sales
Envoy 256MB - 40GB
502938.54
Direct Sales
Envoy Ambassador
2259566.96
Direct Sales
Home Theatre Package with DVD-Audio/Video Play 1235674.15
Direct Sales
Mini DV Camcorder with 3.5" Swivel LCD
775851.87
Internet
17" LCD w/built-in HDTV Tuner
31707.48
Internet
8.3 Minitower Speaker
404090.32
Internet
Envoy 256MB - 40GB
28293.87
Internet
Home Theatre Package with DVD-Audio/Video Play
155405.54
Internet
Mini DV Camcorder with 3.5" Swivel LCD
39726.23
Partners
17" LCD w/built-in HDTV Tuner
269973.97
Partners
Envoy Ambassador
1213063.59
Partners
Home Theatre Package with DVD-Audio/Video Play
700266.58
Partners
Mini DV Camcorder with 3.5" Swivel LCD
404265.85
Partners
Unix/Windows 1-user pack
374002.51

AMT_1999
---------1163645.78
843377.88
1770349.25
1260791.44
1326302.51
160974.7
155235.25
154072.02
153175.04
189921.97
325504.75
614857.93
520166.26
520544.11
340123.02

AMT_DIFF
----------534790.08
-340439.34
489217.71
-25117.29
-550450.64
-129267.22
248855.07
-125778.15
2230.5
-150195.74
-55530.78
598205.66
180100.32
-116278.26
33879.49

15 rows selected.

Functions

5-237

ROWIDTOCHAR

ROWIDTOCHAR
Syntax
ROWIDTOCHAR

(

rowid

)

Purpose
ROWIDTOCHAR converts a rowid value to VARCHAR2 data type. The result of this
conversion is always 18 characters long.

Examples
The following example converts a rowid value in the employees table to a character
value. (Results vary for each build of the sample database.)
SELECT ROWID FROM employees
WHERE ROWIDTOCHAR(ROWID) LIKE '%JAAB%'
ORDER BY ROWID;
ROWID
-----------------AAAFfIAAFAAAABSAAb

5-238 Oracle Database SQL Language Reference

ROWIDTONCHAR

ROWIDTONCHAR
Syntax
ROWIDTONCHAR

(

rowid

)

Purpose
ROWIDTONCHAR converts a rowid value to NVARCHAR2 data type. The result of this
conversion is always in the national character set and is 18 characters long.

Examples
The following example converts a rowid value to an NVARCHAR2 string:
SELECT LENGTHB( ROWIDTONCHAR(ROWID) ) Length, ROWIDTONCHAR(ROWID)
FROM employees
ORDER BY length;
LENGTH
---------36
36
. . .

ROWIDTONCHAR(ROWID
-----------------AAAL52AAFAAAABSABD
AAAL52AAFAAAABSABV

Functions

5-239

RPAD

RPAD
Syntax
,
RPAD

(

expr1

,

expr2

n

)

Purpose
RPAD returns expr1, right-padded to length n characters with expr2, replicated as many
times as necessary. This function is useful for formatting the output of a query.
Both expr1 and expr2 can be any of the data types CHAR, VARCHAR2, NCHAR, NVARCHAR2,
CLOB, or NCLOB. The string returned is of VARCHAR2 data type if expr1 is a character data
type, NVARCHAR2 if expr1 is a national character data type, and a LOB if expr1 is a LOB
data type. The string returned is in the same character set as expr1. The argument n
must be a NUMBER integer or a value that can be implicitly converted to a NUMBER
integer.
expr1 cannot be null. If you do not specify expr2, then it defaults to a single blank. If
expr1 is longer than n, then this function returns the portion of expr1 that fits in n.
The argument n is the total length of the return value as it is displayed on your
terminal screen. In most character sets, this is also the number of characters in the
return value. However, in some multibyte character sets, the display length of a
character string can differ from the number of characters in the string.

Examples
The following example creates a simple chart of salary amounts by padding a single
space with asterisks:
SELECT last_name, RPAD(' ', salary/1000/1, '*') "Salary"
FROM employees
WHERE department_id = 80
ORDER BY last_name, "Salary";
LAST_NAME
Salary
------------------------- --------------Abel
**********
Ande
*****
Banda
*****
Bates
******
Bernstein
********
Bloom
*********
Cambrault
**********
Cambrault
******
Doran
******
Errazuriz
***********
Fox
********
Greene
********
Hall
********
Hutton
*******
Johnson
*****
King
*********
. . .

5-240 Oracle Database SQL Language Reference

RTRIM

RTRIM
Syntax
,
RTRIM

(

char

set
)

Purpose
RTRIM removes from the right end of char all of the characters that appear in set. This
function is useful for formatting the output of a query.
If you do not specify set, then it defaults to a single blank. If char is a character literal,
then you must enclose it in single quotation marks. RTRIM works similarly to LTRIM.
Both char and set can be any of the data types CHAR, VARCHAR2, NCHAR, NVARCHAR2,
CLOB, or NCLOB. The string returned is of VARCHAR2 data type if char is a character data
type, NVARCHAR2 if char is a national character data type, and a LOB if char is a LOB
data type.
See Also:

LTRIM on page 5-144

Examples
The following example trims all the right-most occurrences of less than sign (<),
greater than sign (>) , and equal sign (=) from a string:
SELECT RTRIM('<=====>BROWNING<=====>', '<>=') "RTRIM Example"
FROM DUAL;
RTRIM Example
--------------<=====>BROWNING

Functions

5-241

SCN_TO_TIMESTAMP

SCN_TO_TIMESTAMP
Syntax
SCN_TO_TIMESTAMP

(

number

)

Purpose
SCN_TO_TIMESTAMP takes as an argument a number that evaluates to a system change
number (SCN), and returns the approximate timestamp associated with that SCN. The
returned value is of TIMESTAMP data type. This function is useful any time you want to
know the timestamp associated with an SCN. For example, it can be used in
conjunction with the ORA_ROWSCN pseudocolumn to associate a timestamp with the
most recent change to a row.
Notes:
■
■

The usual precision of the result value is 3 seconds.
The association between an SCN and a timestamp when the SCN
is generated is remembered by the database for a limited period of
time. This period is the maximum of the auto-tuned undo
retention period, if the database runs in the Automatic Undo
Management mode, and the retention times of all flashback
archives in the database, but no less than 120 hours. The time for
the association to become obsolete elapses only when the database
is open. An error is returned if the SCN specified for the argument
to SCN_TO_TIMESTAMP is too old.

ORA_ROWSCN Pseudocolumn on page 2-8 and
TIMESTAMP_TO_SCN on page 5-296

See Also:

Examples
The following example uses the ORA_ROWSCN pseudocolumn to determine the system
change number of the last update to a row and uses SCN_TO_TIMESTAMP to convert that
SCN to a timestamp:
SELECT SCN_TO_TIMESTAMP(ORA_ROWSCN) FROM employees
WHERE employee_id = 188;

You could use such a query to convert a system change number to a timestamp for use
in an Oracle Flashback Query:
SELECT salary FROM employees WHERE employee_id = 188;
SALARY
---------3800
UPDATE employees SET salary = salary*10 WHERE employee_id = 188;
COMMIT;
SELECT salary FROM employees WHERE employee_id = 188;
SALARY
---------38000
SELECT SCN_TO_TIMESTAMP(ORA_ROWSCN) FROM employees

5-242 Oracle Database SQL Language Reference

SCN_TO_TIMESTAMP

WHERE employee_id = 188;
SCN_TO_TIMESTAMP(ORA_ROWSCN)
--------------------------------------------------------------------------28-AUG-03 01.58.01.000000000 PM
FLASHBACK TABLE employees TO TIMESTAMP
TO_TIMESTAMP('28-AUG-03 01.00.00.000000000 PM');
SELECT salary FROM employees WHERE employee_id = 188;
SALARY
---------3800

Functions

5-243

SESSIONTIMEZONE

SESSIONTIMEZONE
Syntax
SESSIONTIMEZONE

Purpose
SESSIONTIMEZONE returns the time zone of the current session. The return type is a time
zone offset (a character type in the format '[+|-]TZH:TZM') or a time zone region
name, depending on how the user specified the session time zone value in the most
recent ALTER SESSION statement.
Note: The default client session time zone is an offset even if the
client operating system uses a named time zone. If you want the
default session time zone to use a named time zone, then set the ORA_
SDTZ variable in the client environment to an Oracle time zone region
name. Refer to Oracle Database Globalization Support Guide for more
information on this variable.

Examples
The following example returns the time zone of the current session:
SELECT SESSIONTIMEZONE FROM DUAL;
SESSION
-------08:00

5-244 Oracle Database SQL Language Reference

SET

SET
Syntax
SET

(

nested_table

)

Purpose
SET converts a nested table into a set by eliminating duplicates. The function returns a
nested table whose elements are distinct from one another. The returned nested table is
of the same type as the input nested table.
The element types of the nested table must be comparable. Refer to "Comparison
Conditions" on page 7-4 for information on the comparability of nonscalar types.

Examples
The following example selects from the customers_demo table the unique elements of
the cust_address_ntab nested table column:
SELECT customer_id, SET(cust_address_ntab) address
FROM customers_demo
ORDER BY customer_id;
CUSTOMER_ID
----------101
102
103
104
105
. . .

ADDRESS(STREET_ADDRESS, POSTAL_CODE, CITY, STATE_PROVINCE, COUNTRY_ID)
-----------------------------------------------------------------------CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP('514 W Superior St', '46901', 'Kokomo', 'IN', 'US'))
CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP('2515 Bloyd Ave', '46218', 'Indianapolis', 'IN', 'US'))
CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP('8768 N State Rd 37', '47404', 'Bloomington', 'IN', 'US'))
CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP('6445 Bay Harbor Ln', '46254', 'Indianapolis', 'IN', 'US'))
CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP('4019 W 3Rd St', '47404', 'Bloomington', 'IN', 'US'))

The preceding example requires the table customers_demo and a nested table column
containing data. Refer to "Multiset Operators" on page 4-1 to create this table and
nested table column.

Functions

5-245

SIGN

SIGN
Syntax
SIGN

(

n

)

Purpose
SIGN returns the sign of n. This function takes as an argument any numeric data type,
or any nonnumeric data type that can be implicitly converted to NUMBER, and returns
NUMBER.
For value of NUMBER type, the sign is:
■

-1 if n<0

■

0 if n=0

■

1 if n>0

For binary floating-point numbers (BINARY_FLOAT and BINARY_DOUBLE), this function
returns the sign bit of the number. The sign bit is:
■

-1 if n<0

■

+1 if n>=0 or n=NaN

Examples
The following example indicates that the argument of the function (-15) is <0:
SELECT SIGN(-15) "Sign" FROM DUAL;
Sign
----------1

5-246 Oracle Database SQL Language Reference

SIN

SIN
Syntax
SIN

(

n

)

Purpose
SIN returns the sine of n (an angle expressed in radians).
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. If the argument is
BINARY_FLOAT, then the function returns BINARY_DOUBLE. Otherwise the function
returns the same numeric data type as the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

Examples
The following example returns the sine of 30 degrees:
SELECT SIN(30 * 3.14159265359/180)
"Sine of 30 degrees" FROM DUAL;
Sine of 30 degrees
-----------------.5

Functions

5-247

SINH

SINH
Syntax
SINH

(

n

)

Purpose
SINH returns the hyperbolic sine of n.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. If the argument is
BINARY_FLOAT, then the function returns BINARY_DOUBLE. Otherwise the function
returns the same numeric data type as the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

Examples
The following example returns the hyperbolic sine of 1:
SELECT SINH(1) "Hyperbolic sine of 1" FROM DUAL;
Hyperbolic sine of 1
-------------------1.17520119

5-248 Oracle Database SQL Language Reference

SOUNDEX

SOUNDEX
Syntax
SOUNDEX

(

char

)

Purpose
SOUNDEX returns a character string containing the phonetic representation of char. This
function lets you compare words that are spelled differently, but sound alike in
English.
The phonetic representation is defined in The Art of Computer Programming, Volume 3:
Sorting and Searching, by Donald E. Knuth, as follows:
1.

Retain the first letter of the string and remove all other occurrences of the
following letters: a, e, h, i, o, u, w, y.

2.

Assign numbers to the remaining letters (after the first) as follows:
b, f, p, v = 1
c, g, j, k, q, s, x, z = 2
d, t = 3
l = 4
m, n = 5
r = 6

3.

If two or more letters with the same number were adjacent in the original name
(before step 1), or adjacent except for any intervening h and w, then retain the first
letter and omit rest of all the adjacent letters with same number.

4.

Return the first four bytes padded with 0.

char can be of any of the data types CHAR, VARCHAR2, NCHAR, or NVARCHAR2. The return
value is the same data type as char.
This function does not support CLOB data directly. However, CLOBs can be passed in as
arguments through implicit data conversion.
See Also: "Data Type Comparison Rules" on page 3-36 for more
information.

Examples
The following example returns the employees whose last names are a phonetic
representation of "Smyth":
SELECT last_name, first_name
FROM hr.employees
WHERE SOUNDEX(last_name)
= SOUNDEX('SMYTHE')
ORDER BY last_name, first_name;
LAST_NAME
---------Smith
Smith

FIRST_NAME
---------Lindsey
William

Functions

5-249

SQRT

SQRT
Syntax
SQRT

(

n

)

Purpose
SQRT returns the square root of n.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. The function returns the
same data type as the numeric data type of the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion
■

■

If n resolves to a NUMBER, then the value n cannot be negative. SQRT returns a real
number.
If n resolves to a binary floating-point number (BINARY_FLOAT or BINARY_DOUBLE):
–

If n >= 0, then the result is positive.

–

If n = -0, then the result is -0.

–

If n < 0, then the result is NaN.

Examples
The following example returns the square root of 26:
SELECT SQRT(26) "Square root" FROM DUAL;
Square root
----------5.09901951

5-250 Oracle Database SQL Language Reference

STATS_BINOMIAL_TEST

STATS_BINOMIAL_TEST
Syntax
TWO_SIDED_PROB
EXACT_PROB
,
ONE_SIDED_PROB_OR_MORE
ONE_SIDED_PROB_OR_LESS
STATS_BINOMIAL_TEST

(

expr1

,

expr2

,

p

)

Purpose
STATS_BINOMIAL_TEST is an exact probability test used for dichotomous variables,
where only two possible values exist. It tests the difference between a sample
proportion and a given proportion. The sample size in such tests is usually small.
This function takes four arguments: expr1 is the sample being examined. expr2
contains the values for which the proportion is expected to be, and p is a proportion to
test against. The fourth argument is a return value of type VARCHAR2. If you omit the
fourth argument, then the default is TWO_SIDED_PROB. The meaning of the return
values is shown in Table 5–3.
Table 5–3

STATS_BINOMIAL Return Values

Return Value

Meaning

TWO_SIDED_PROB

The probability that the given population proportion, p,
could result in the observed proportion or a more extreme
one.

EXACT_PROB

The probability that the given population proportion, p,
could result in exactly the observed proportion.

ONE_SIDED_PROB_OR_MORE

The probability that the given population proportion, p,
could result in the observed proportion or a larger one.

ONE_SIDED_PROB_OR_LESS

The probability that the given population proportion, p,
could result in the observed proportion or a smaller one.

EXACT_PROB gives the probability of getting exactly proportion p. In cases where you
want to test whether the proportion found in the sample is significantly different from
a 50-50 split, p would normally be 0.50. If you want to test only whether the proportion
is different, then use the return value TWO_SIDED_PROB. If your test is whether the
proportion is more than the value of expr2, then use the return value ONE_SIDED_PROB_
OR_MORE. If the test is to determine whether the proportion of expr2 is less, then use the
return value ONE_SIDED_PROB_OR_LESS.
STATS_BINOMIAL_TEST Example The following example determines the probability
that reality exactly matches the number of men observed under the assumption that
69% of the population is composed of men:
SELECT AVG(DECODE(cust_gender, 'M', 1, 0)) real_proportion,
STATS_BINOMIAL_TEST
(cust_gender, 'M', 0.68, 'EXACT_PROB') exact,
STATS_BINOMIAL_TEST
(cust_gender, 'M', 0.68, 'ONE_SIDED_PROB_OR_LESS') prob_or_less
FROM sh.customers;

Functions

5-251

STATS_CROSSTAB

STATS_CROSSTAB
Syntax
CHISQ_OBS
CHISQ_SIG
CHISQ_DF
,

PHI_COEFFICIENT
CRAMERS_V
CONT_COEFFICIENT
COHENS_K

STATS_CROSSTAB

(

expr1

,

expr2

)

Purpose
Crosstabulation (commonly called crosstab) is a method used to analyze two nominal
variables. The STATS_CROSSTAB function takes three arguments: two expressions and a
return value of type VARCHAR2. expr1 and expr2 are the two variables being analyzed.
The function returns one number, determined by the value of the third argument. If
you omit the third argument, then the default is CHISQ_SIG. The meaning of the return
values is shown in Table 5–4.
Table 5–4

STATS_CROSSTAB Return Values

Return Value

Meaning

CHISQ_OBS

Observed value of chi-squared

CHISQ_SIG

Significance of observed chi-squared

CHISQ_DF

Degree of freedom for chi-squared

PHI_COEFFICIENT

Phi coefficient

CRAMERS_V

Cramer's V statistic

CONT_COEFFICIENT

Contingency coefficient

COHENS_K

Cohen's kappa

The following example determines the strength of the
association between gender and income level:

STATS_CROSSTAB Example

SELECT STATS_CROSSTAB
(cust_gender, cust_income_level, 'CHISQ_OBS') chi_squared,
STATS_CROSSTAB
(cust_gender, cust_income_level, 'CHISQ_SIG') p_value,
STATS_CROSSTAB
(cust_gender, cust_income_level, 'PHI_COEFFICIENT') phi_coefficient
FROM sh.customers;
CHI_SQUARED
P_VALUE PHI_COEFFICIENT
----------- ---------- --------------251.690705 1.2364E-47
.067367056

5-252 Oracle Database SQL Language Reference

STATS_F_TEST

STATS_F_TEST
Syntax
STATISTIC
DF_NUM
,

expr3

DF_DEN
,

ONE_SIDED_SIG
TWO_SIDED_SIG

STATS_F_TEST

(

expr1

,

expr2

)

Purpose
STATS_F_TEST tests whether two variances are significantly different. The observed
value of f is the ratio of one variance to the other, so values very different from 1
usually indicate significant differences.
This function takes three arguments: expr1 is the grouping or independent variable
and expr2 is the sample of values. The function returns one number, determined by
the value of the third argument. If you omit the third argument, then the default is
TWO_SIDED_SIG. The meaning of the return values is shown in Table 5–5.
Table 5–5

STATS_F_TEST Return Values

Return Value

Meaning

STATISTIC

The observed value of f

DF_NUM

Degree of freedom for the numerator

DF_DEN

Degree of freedom for the denominator

ONE_SIDED_SIG

One-tailed significance of f

TWO_SIDED_SIG

Two-tailed significance of f

The one-tailed significance is always in relation to the upper tail. The final argument,
expr3, indicates which of the two groups specified by expr1 is the high value or
numerator (the value whose rejection region is the upper tail).
The observed value of f is the ratio of the variance of one group to the variance of the
second group. The significance of the observed value of f is the probability that the
variances are different just by chance—a number between 0 and 1. A small value for
the significance indicates that the variances are significantly different. The degree of
freedom for each of the variances is the number of observations in the sample minus 1.
STATS_F_TEST Example The following example determines whether the variance in
credit limit between men and women is significantly different. The results, a p_value
not close to zero, and an f_statistic close to 1, indicate that the difference between
credit limits for men and women are not significant.
SELECT VARIANCE(DECODE(cust_gender, 'M', cust_credit_limit, null)) var_men,
VARIANCE(DECODE(cust_gender, 'F', cust_credit_limit, null)) var_women,
STATS_F_TEST(cust_gender, cust_credit_limit, 'STATISTIC', 'F') f_statistic,
STATS_F_TEST(cust_gender, cust_credit_limit) two_sided_p_value
FROM sh.customers;

Functions

5-253

STATS_F_TEST

VAR_MEN VAR_WOMEN F_STATISTIC TWO_SIDED_P_VALUE
---------- ---------- ----------- ----------------12879896.7
13046865 1.01296348
.311928071

5-254 Oracle Database SQL Language Reference

STATS_KS_TEST

STATS_KS_TEST
Syntax
STATISTIC
,
SIG
STATS_KS_TEST

(

expr1

,

expr2

)

Purpose
STATS_KS_TEST is a Kolmogorov-Smirnov function that compares two samples to test
whether they are from the same population or from populations that have the same
distribution. It does not assume that the population from which the samples were
taken is normally distributed.
This function takes three arguments: two expressions and a return value of type
VARCHAR2. expr1 classifies the data into the two samples. expr2 contains the values for
each of the samples. If expr1 classifies the rows into only one sample or into more than
two samples, then an error is raised.The function returns one value determined by the
third argument. If you omit the third argument, then the default is SIG. The meaning
of the return values is shown in Table 5–6.
Table 5–6

STATS_KS_TEST Return Values

Return Value

Meaning

STATISTIC

Observed value of D

SIG

Significance of D

Using the Kolmogorov Smirnov test, the following
example determines whether the distribution of sales between men and women is due
to chance:

STATS_KS_TEST Example

SELECT stats_ks_test(cust_gender, amount_sold, 'STATISTIC') ks_statistic,
stats_ks_test(cust_gender, amount_sold) p_value
FROM sh.customers c, sh.sales s
WHERE c.cust_id = s.cust_id;
KS_STATISTIC
P_VALUE
------------ ---------.003841396 .004080006

Functions

5-255

STATS_MODE

STATS_MODE
Syntax
STATS_MODE

(

expr

)

Purpose
STATS_MODE takes as its argument a set of values and returns the value that occurs with
the greatest frequency. If more than one mode exists, then Oracle Database chooses one
and returns only that one value.
To obtain multiple modes (if multiple modes exist), you must use a combination of
other functions, as shown in the hypothetical query:
SELECT x FROM (SELECT x, COUNT(x) AS cnt1
FROM t GROUP BY x)
WHERE cnt1 =
(SELECT MAX(cnt2) FROM (SELECT COUNT(x) AS cnt2 FROM t GROUP BY x));

Examples
The following example returns the mode of salary per department in the
hr.employees table:
SELECT department_id, STATS_MODE(salary) FROM employees
GROUP BY department_id
ORDER BY department_id, stats_mode(salary);
DEPARTMENT_ID STATS_MODE(SALARY)
------------- -----------------10
4400
20
6000
30
2500
40
6500
50
2500
60
4800
70
10000
80
9500
90
17000
100
6900
110
8300
7000

If you need to retrieve all of the modes (in cases with multiple modes), you can do so
using a combination of other functions, as shown in the next example:
SELECT commission_pct FROM
(SELECT commission_pct, COUNT(commission_pct) AS cnt1 FROM employees
GROUP BY commission_pct)
WHERE cnt1 =
(SELECT MAX (cnt2) FROM
(SELECT COUNT(commission_pct) AS cnt2
FROM employees GROUP BY commission_pct))
ORDER BY commission_pct;
COMMISSION_PCT
-------------.2
.3
5-256 Oracle Database SQL Language Reference

STATS_MW_TEST

STATS_MW_TEST
Syntax
STATISTIC
U_STATISTIC
,
ONE_SIDED_SIG

,

expr3

TWO_SIDED_SIG
STATS_MW_TEST

(

expr1

,

expr2

)

Purpose
A Mann Whitney test compares two independent samples to test the null hypothesis
that two populations have the same distribution function against the alternative
hypothesis that the two distribution functions are different.
The STATS_MW_TEST does not assume that the differences between the samples are
normally distributed, as do the STATS_T_TEST_* functions. This function takes three
arguments and a return value of type VARCHAR2. expr1 classifies the data into groups.
expr2 contains the values for each of the groups. The function returns one value,
determined by the third argument. If you omit the third argument, then the default is
TWO_SIDED_SIG. The meaning of the return values is shown in the table that follows.
The significance of the observed value of Z or U is the probability that the variances
are different just by chance—a number between 0 and 1. A small value for the
significance indicates that the variances are significantly different. The degree of
freedom for each of the variances is the number of observations in the sample minus 1.
Table 5–7

STATS_MW_TEST Return Values

Return Value

Meaning

STATISTIC

The observed value of Z

U_STATISTIC

The observed value of U

ONE_SIDED_SIG

One-tailed significance of Z

TWO_SIDED_SIG

Two-tailed significance of Z

The one-tailed significance is always in relation to the upper tail. The final argument,
expr3, indicates which of the two groups specified by expr1 is the high value (the
value whose rejection region is the upper tail).
STATS_MW_TEST computes the probability that the samples are from the same
distribution by checking the differences in the sums of the ranks of the values. If the
samples come from the same distribution, then the sums should be close in value.
Using the Mann Whitney test, the following example
determines whether the distribution of sales between men and women is due to
chance:

STATS_MW_TEST Example

SELECT STATS_MW_TEST
(cust_gender, amount_sold, 'STATISTIC') z_statistic,
STATS_MW_TEST
(cust_gender, amount_sold, 'ONE_SIDED_SIG', 'F') one_sided_p_value
FROM sh.customers c, sh.sales s

Functions

5-257

STATS_MW_TEST

WHERE c.cust_id = s.cust_id;
Z_STATISTIC ONE_SIDED_P_VALUE
----------- -----------------1.4011509
.080584471

5-258 Oracle Database SQL Language Reference

STATS_ONE_WAY_ANOVA

STATS_ONE_WAY_ANOVA
Syntax
SUM_SQUARES_BETWEEN
SUM_SQUARES_WITHIN
DF_BETWEEN
DF_WITHIN
,
MEAN_SQUARES_BETWEEN
MEAN_SQUARES_WITHIN
F_RATIO
SIG
STATS_ONE_WAY_ANOVA

(

expr1

,

expr2

)

Purpose
The one-way analysis of variance function (STATS_ONE_WAY_ANOVA) tests differences in
means (for groups or variables) for statistical significance by comparing two different
estimates of variance. One estimate is based on the variances within each group or
category. This is known as the mean squares within or mean square error. The other
estimate is based on the variances among the means of the groups. This is known as
the mean squares between. If the means of the groups are significantly different, then
the mean squares between will be larger than expected and will not match the mean
squares within. If the mean squares of the groups are consistent, then the two variance
estimates will be about the same.
STATS_ONE_WAY_ANOVA takes three arguments: two expressions and a return value of
type VARCHAR2. expr1 is an independent or grouping variable that divides the data into
a set of groups. expr2 is a dependent variable (a numeric expression) containing the
values corresponding to each member of a group. The function returns one number,
determined by the value of the third argument. If you omit the third argument, then
the default is SIG. The meaning of the return values is shown in Table 5–8.
Table 5–8

STATS_ONE_WAY_ANOVA Return Values

Return Value

Meaning

SUM_SQUARES_BETEEN

Sum of squares between groups

SUM_SQUARES_WITHIN

Sum of squares within groups

DF_BETWEEN

Degree of freedom between groups

DF_WITHIN

Degree of freedom within groups

MEAN_SQUARES_BETWEEN

Mean squares between groups

MEAN_SQUARES_WITHIN

Mean squares within groups

F_RATIO

Ratio of the mean squares between to the mean squares within
(MSB/MSW)

SIG

Significance

The significance of one-way analysis of variance is determined by obtaining the
one-tailed significance of an f-test on the ratio of the mean squares between and the
Functions

5-259

STATS_ONE_WAY_ANOVA

mean squares within. The f-test should use one-tailed significance, because the mean
squares between can be only equal to or larger than the mean squares within.
Therefore, the significance returned by STATS_ONE_WAY_ANOVA is the probability that
the differences between the groups happened by chance—a number between 0 and 1.
The smaller the number, the greater the significance of the difference between the
groups. Refer to the STATS_F_TEST on page 5-253 for information on performing an
f-test.
STATS_ONE_WAY_ANOVA Example The following example determines the
significance of the differences in mean sales within an income level and differences in
mean sales between income levels. The results, p_values close to zero, indicate that, for
both men and women, the difference in the amount of goods sold across different
income levels is significant.
SELECT cust_gender,
STATS_ONE_WAY_ANOVA(cust_income_level, amount_sold, 'F_RATIO') f_ratio,
STATS_ONE_WAY_ANOVA(cust_income_level, amount_sold, 'SIG') p_value
FROM sh.customers c, sh.sales s
WHERE c.cust_id = s.cust_id
GROUP BY cust_gender
ORDER BY cust_gender;
C
F_RATIO
P_VALUE
- ---------- ---------F 5.59536943 4.7840E-09
M 9.2865001 6.7139E-17

5-260 Oracle Database SQL Language Reference

STATS_T_TEST_*

STATS_T_TEST_*
The t-test functions are:
■

STATS_T_TEST_ONE: A one-sample t-test

■

STATS_T_TEST_PAIRED: A two-sample, paired t-test (also known as a crossed t-test)

■

■

STATS_T_TEST_INDEP: A t-test of two independent groups with the same variance
(pooled variances)
STATS_T_TEST_INDEPU: A t-test of two independent groups with unequal variance
(unpooled variances)

Syntax
stats_t_test::=
STATISTIC
,

expr3

ONE_SIDED_SIG
,
STATS_T_TEST_ONE

(

expr2

,

expr1

TWO_SIDED_SIG
DF

STATS_T_TEST_PAIRED

)

STATS_T_TEST_INDEP

(

expr1

,

expr2

STATS_T_TEST_INDEPU

Purpose
The t-test measures the significance of a difference of means. You can use it to compare
the means of two groups or the means of one group with a constant. The one-sample
and two-sample STATS_T_TEST_* functions take three arguments: two expressions and
a return value of type VARCHAR2. The functions return one number, determined by the
value of the third argument. If you omit the third argument, then the default is TWO_
SIDED_SIG. The meaning of the return values is shown in Table 5–9.
Table 5–9

STATS_T_TEST_* Return Values

Return Value

Meaning

STATISTIC

The observed value of t

DF

Degree of freedom

ONE_SIDED_SIG

One-tailed significance of t

TWO_SIDED_SIG

Two-tailed significance of t

The two independent STATS_T_TEST_* functions can take a fourth argument (expr3) if
the third argument is specified as STATISTIC or ONE_SIDED_SIG. In this case, expr3
indicates which value of expr1 is the high value, or the value whose rejection region is
the upper tail.
The significance of the observed value of t is the probability that the value of t would
have been obtained by chance—a number between 0 and 1. The smaller the value, the
more significant the difference between the means. One-sided significance is always

Functions

5-261

STATS_T_TEST_*

respect to the upper tail. For one-sample and paired t-test, the high value is the first
expression. For independent t-test, the high value is the one specified by expr3.
The degree of freedom depends on the type of t-test that resulted in the observed
value of t. For example, for a one-sample t-test (STATS_T_TEST_ONE), the degree of
freedom is the number of observations in the sample minus 1.

5-262 Oracle Database SQL Language Reference

STATS_T_TEST_*

STATS_T_TEST_ONE
In the STATS_T_TEST_ONE function, expr1 is the sample and expr2 is the constant mean
against which the sample mean is compared. For this t-test only, expr2 is optional; the
constant mean defaults to 0. This function obtains the value of t by dividing the
difference between the sample mean and the known mean by the standard error of the
mean (rather than the standard error of the difference of the means, as for STATS_T_
TEST_PAIRED).
The following example determines the significance of
the difference between the average list price and the constant value 60:

STATS_T_TEST_ONE Example

SELECT AVG(prod_list_price) group_mean,
STATS_T_TEST_ONE(prod_list_price, 60, 'STATISTIC') t_observed,
STATS_T_TEST_ONE(prod_list_price, 60) two_sided_p_value
FROM sh.products;
GROUP_MEAN T_OBSERVED TWO_SIDED_P_VALUE
---------- ---------- ----------------139.545556 2.32107746
.023158537

Functions

5-263

STATS_T_TEST_PAIRED

STATS_T_TEST_PAIRED
In the STATS_T_TEST_PAIRED function, expr1 and expr2 are the two samples whose
means are being compared. This function obtains the value of t by dividing the
difference between the sample means by the standard error of the difference of the
means (rather than the standard error of the mean, as for STATS_T_TEST_ONE).

5-264 Oracle Database SQL Language Reference

STATS_T_TEST_*

STATS_T_TEST_INDEP and STATS_T_TEST_INDEPU
In the STATS_T_TEST_INDEP and STATS_T_TEST_INDEPU functions, expr1 is the
grouping column and expr2 is the sample of values. The pooled variances version
(STATS_T_TEST_INDEP) tests whether the means are the same or different for two
distributions that have similar variances. The unpooled variances version (STATS_T_
TEST_INDEPU) tests whether the means are the same or different even if the two
distributions are known to have significantly different variances.
Before using these functions, it is advisable to determine whether the variances of the
samples are significantly different. If they are, then the data may come from
distributions with different shapes, and the difference of the means may not be very
useful. You can perform an f-test to determine the difference of the variances. If they
are not significantly different, use STATS_T_TEST_INDEP. If they are significantly
different, use STATS_T_TEST_INDEPU. Refer to STATS_F_TEST on page 5-253 for
information on performing an f-test.
The following example determines the significance
of the difference between the average sales to men and women where the distributions
are assumed to have similar (pooled) variances:

STATS_T_TEST_INDEP Example

SELECT SUBSTR(cust_income_level, 1, 22) income_level,
AVG(DECODE(cust_gender, 'M', amount_sold, null)) sold_to_men,
AVG(DECODE(cust_gender, 'F', amount_sold, null)) sold_to_women,
STATS_T_TEST_INDEP(cust_gender, amount_sold, 'STATISTIC', 'F') t_observed,
STATS_T_TEST_INDEP(cust_gender, amount_sold) two_sided_p_value
FROM sh.customers c, sh.sales s
WHERE c.cust_id = s.cust_id
GROUP BY ROLLUP(cust_income_level)
ORDER BY income_level, sold_to_men, sold_to_women, t_observed;
INCOME_LEVEL
SOLD_TO_MEN SOLD_TO_WOMEN T_OBSERVED TWO_SIDED_P_VALUE
---------------------- ----------- ------------- ---------- ----------------A: Below 30,000
105.28349
99.4281447 -1.9880629
.046811482
B: 30,000 - 49,999
102.59651
109.829642 3.04330875
.002341053
C: 50,000 - 69,999
105.627588
110.127931 2.36148671
.018204221
D: 70,000 - 89,999
106.630299
110.47287 2.28496443
.022316997
E: 90,000 - 109,999
103.396741
101.610416 -1.2544577
.209677823
F: 110,000 - 129,999
106.76476
105.981312 -.60444998
.545545304
G: 130,000 - 149,999
108.877532
107.31377 -.85298245
.393671218
H: 150,000 - 169,999
110.987258
107.152191 -1.9062363
.056622983
I: 170,000 - 189,999
102.808238
107.43556 2.18477851
.028908566
J: 190,000 - 249,999
108.040564
115.343356 2.58313425
.009794516
K: 250,000 - 299,999
112.377993
108.196097 -1.4107871
.158316973
L: 300,000 and above
120.970235
112.216342 -2.0642868
.039003862
107.121845
113.80441 .686144393
.492670059
106.663769
107.276386 1.08013499
.280082357
14 rows selected.

STATS_T_TEST_INDEPU Example The following example determines the
significance of the difference between the average sales to men and women where the
distributions are known to have significantly different (unpooled) variances:
SELECT SUBSTR(cust_income_level, 1, 22) income_level,
AVG(DECODE(cust_gender, 'M', amount_sold, null)) sold_to_men,
AVG(DECODE(cust_gender, 'F', amount_sold, null)) sold_to_women,
STATS_T_TEST_INDEPU(cust_gender, amount_sold, 'STATISTIC', 'F') t_observed,
STATS_T_TEST_INDEPU(cust_gender, amount_sold) two_sided_p_value
FROM sh.customers c, sh.sales s

Functions

5-265

STATS_T_TEST_INDEP and STATS_T_TEST_INDEPU

WHERE c.cust_id = s.cust_id
GROUP BY ROLLUP(cust_income_level)
ORDER BY income_level, sold_to_men, sold_to_women, t_observed;
INCOME_LEVEL
SOLD_TO_MEN SOLD_TO_WOMEN T_OBSERVED TWO_SIDED_P_VALUE
---------------------- ----------- ------------- ---------- ----------------A: Below 30,000
105.28349
99.4281447 -2.0542592
.039964704
B: 30,000 - 49,999
102.59651
109.829642 2.96922332
.002987742
C: 50,000 - 69,999
105.627588
110.127931 2.3496854
.018792277
D: 70,000 - 89,999
106.630299
110.47287 2.26839281
.023307831
E: 90,000 - 109,999
103.396741
101.610416 -1.2603509
.207545662
F: 110,000 - 129,999
106.76476
105.981312 -.60580011
.544648553
G: 130,000 - 149,999
108.877532
107.31377 -.85219781
.394107755
H: 150,000 - 169,999
110.987258
107.152191 -1.9451486
.051762624
I: 170,000 - 189,999
102.808238
107.43556 2.14966921
.031587875
J: 190,000 - 249,999
108.040564
115.343356 2.54749867
.010854966
K: 250,000 - 299,999
112.377993
108.196097 -1.4115514
.158091676
L: 300,000 and above
120.970235
112.216342 -2.0726194
.038225611
107.121845
113.80441 .689462437
.490595765
106.663769
107.276386 1.07853782
.280794207
14 rows selected.

5-266 Oracle Database SQL Language Reference

STATS_WSR_TEST

STATS_WSR_TEST
Syntax
STATISTIC
,

ONE_SIDED_SIG
TWO_SIDED_SIG

STATS_WSR_TEST

(

expr1

,

expr2

)

Purpose
STATS_WSR_TEST is a Wilcoxon Signed Ranks test of paired samples to determine
whether the median of the differences between the samples is significantly different
from zero. The absolute values of the differences are ordered and assigned ranks. Then
the null hypothesis states that the sum of the ranks of the positive differences is equal
to the sum of the ranks of the negative differences.
This function takes three arguments: expr1 and expr2 are the two samples being
analyzed, and the third argument is a return value of type VARCHAR2. If you omit the
third argument, then the default is TWO_SIDED_SIG. The meaning of the return values is
shown in Table 5–10.
Table 5–10

STATS_WSR_TEST_* Return Values

Return Value

Meaning

STATISTIC

The observed value of Z

ONE_SIDED_SIG

One-tailed significance of Z

TWO_SIDED_SIG

Two-tailed significance of Z

One-sided significance is always with respect to the upper tail. The high value (the
value whose rejection region is the upper tail) is expr1.

Functions

5-267

STDDEV

STDDEV
Syntax
DISTINCT
ALL
STDDEV

(

OVER
expr

(

analytic_clause

)

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions

Purpose
STDDEV returns the sample standard deviation of expr, a set of numbers. You can use it
as both an aggregate and analytic function. It differs from STDDEV_SAMP in that STDDEV
returns zero when it has only 1 row of input data, whereas STDDEV_SAMP returns null.
Oracle Database calculates the standard deviation as the square root of the variance
defined for the VARIANCE aggregate function.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. The function returns the
same data type as the numeric data type of the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

If you specify DISTINCT, then you can specify only the query_partition_clause of the
analytic_clause. The order_by_clause and windowing_clause are not allowed.
See Also:
■

■

"Aggregate Functions" on page 5-10, VARIANCE on page 5-344,
and STDDEV_SAMP on page 5-272
"About SQL Expressions" on page 6-1 for information on valid
forms of expr

Aggregate Examples
The following example returns the standard deviation of the salaries in the sample
hr.employees table:
SELECT STDDEV(salary) "Deviation"
FROM employees;
Deviation
---------3909.36575

Analytic Examples
The query in the following example returns the cumulative standard deviation of the
salaries in Department 80 in the sample table hr.employees, ordered by hire_date:
SELECT last_name, salary,
STDDEV(salary) OVER (ORDER BY hire_date) "StdDev"
FROM employees
WHERE department_id = 30
5-268 Oracle Database SQL Language Reference

STDDEV

ORDER BY last_name, salary, "StdDev";
LAST_NAME
SALARY
StdDev
------------------------- ---------- ---------Baida
2900 4035.26125
Colmenares
2500 3362.58829
Himuro
2600 3649.2465
Khoo
3100 5586.14357
Raphaely
11000
0
Tobias
2800 4650.0896

Functions

5-269

STDDEV_POP

STDDEV_POP
Syntax
OVER
STDDEV_POP

(

expr

(

analytic_clause

)

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions

Purpose
STDDEV_POP computes the population standard deviation and returns the square root
of the population variance. You can use it as both an aggregate and analytic function.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. The function returns the
same data type as the numeric data type of the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

This function is the same as the square root of the VAR_POP function. When VAR_POP
returns null, this function returns null.
See Also:
■
■

"Aggregate Functions" on page 5-10 and VAR_POP on page 5-341
"About SQL Expressions" on page 6-1 for information on valid
forms of expr

Aggregate Example
The following example returns the population and sample standard deviations of the
amount of sales in the sample table sh.sales:
SELECT STDDEV_POP(amount_sold) "Pop",
STDDEV_SAMP(amount_sold) "Samp"
FROM sales;
Pop
Samp
---------- ---------896.355151 896.355592

Analytic Example
The following example returns the population standard deviations of salaries in the
sample hr.employees table by department:
SELECT department_id, last_name, salary,
STDDEV_POP(salary) OVER (PARTITION BY department_id) AS pop_std
FROM employees
ORDER BY department_id, last_name, salary, pop_std;
DEPARTMENT_ID
------------10
20

LAST_NAME
SALARY
POP_STD
------------------------- ---------- ---------Whalen
4400
0
Fay
6000
3500

5-270 Oracle Database SQL Language Reference

STDDEV_POP

20 Hartstein
30 Baida

13000
2900

3500
3069.6091

. . .
100 Urman
110 Gietz
110 Higgins
Grant

7800 1644.18166
8300
1850
12000
1850
7000
0

Functions

5-271

STDDEV_SAMP

STDDEV_SAMP
Syntax
OVER
STDDEV_SAMP

(

expr

(

analytic_clause

)

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions

Purpose
STDDEV_SAMP computes the cumulative sample standard deviation and returns the
square root of the sample variance. You can use it as both an aggregate and analytic
function.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. The function returns the
same data type as the numeric data type of the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

This function is same as the square root of the VAR_SAMP function. When VAR_SAMP
returns null, this function returns null.
See Also:
■

■

"Aggregate Functions" on page 5-10 and VAR_SAMP on
page 5-343
"About SQL Expressions" on page 6-1 for information on valid
forms of expr

Aggregate Example
Refer to the aggregate example for STDDEV_POP on page 5-270.

Analytic Example
The following example returns the sample standard deviation of salaries in the
employees table by department:
SELECT department_id, last_name, hire_date, salary,
STDDEV_SAMP(salary) OVER (PARTITION BY department_id
ORDER BY hire_date
ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) AS cum_sdev
FROM employees
ORDER BY department_id, last_name, hire_date, salary, cum_sdev;
DEPARTMENT_ID
------------10
20
20
30
30
30
30

LAST_NAME
--------------Whalen
Fay
Hartstein
Baida
Colmenares
Himuro
Khoo

5-272 Oracle Database SQL Language Reference

HIRE_DATE
SALARY
CUM_SDEV
--------- ---------- ---------17-SEP-03
4400
17-AUG-05
6000 4949.74747
17-FEB-04
13000
24-DEC-05
2900 4035.26125
10-AUG-07
2500 3362.58829
15-NOV-06
2600 3649.2465
18-MAY-03
3100 5586.14357

STDDEV_SAMP

30 Raphaely

07-DEC-02

11000

17-AUG-02
07-DEC-07
30-SEP-05
07-MAR-06
07-JUN-02
07-JUN-02
24-MAY-07

12008
6900
7700
7800
8300
12008
7000

. . .
100
100
100
100
110
110

Greenberg
Popp
Sciarra
Urman
Gietz
Higgins
Grant

2126.9772
1804.13155
1929.76233
1788.92504
2621.95194

Functions

5-273

SUBSTR

SUBSTR
Syntax
substr::=
SUBSTR
SUBSTRB
SUBSTRC

,
(

char

,

position

substring_length
)

SUBSTR2
SUBSTR4

Purpose
The SUBSTR functions return a portion of char, beginning at character position,
substring_length characters long. SUBSTR calculates lengths using characters as
defined by the input character set. SUBSTRB uses bytes instead of characters. SUBSTRC
uses Unicode complete characters. SUBSTR2 uses UCS2 code points. SUBSTR4 uses UCS4
code points.
■
■

■
■

If position is 0, then it is treated as 1.
If position is positive, then Oracle Database counts from the beginning of char to
find the first character.
If position is negative, then Oracle counts backward from the end of char.
If substring_length is omitted, then Oracle returns all characters to the end of
char. If substring_length is less than 1, then Oracle returns null.

char can be any of the data types CHAR, VARCHAR2, NCHAR, NVARCHAR2, CLOB, or NCLOB.
The exceptions are SUBSTRC, SUBSTR2, and SUBSTR4, which do not allow char to be a
CLOB or NCLOB. Both position and substring_length must be of data type NUMBER, or
any data type that can be implicitly converted to NUMBER, and must resolve to an
integer. The return value is the same data type as char. Floating-point numbers passed
as arguments to SUBSTR are automatically converted to integers.
See Also: Oracle Database Globalization Support Guide for more
information about SUBSTR functions and length semantics in different
locales

Examples
The following example returns several specified substrings of "ABCDEFG":
SELECT SUBSTR('ABCDEFG',3,4) "Substring"
FROM DUAL;
Substring
--------CDEF
SELECT SUBSTR('ABCDEFG',-5,4) "Substring"
FROM DUAL;
Substring

5-274 Oracle Database SQL Language Reference

SUBSTR

--------CDEF

Assume a double-byte database character set:
SELECT SUBSTRB('ABCDEFG',5,4.2) "Substring with bytes"
FROM DUAL;
Substring with bytes
-------------------CD

Functions

5-275

SUM

SUM
Syntax
DISTINCT
ALL
SUM

(

OVER
expr

(

analytic_clause

)

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions

Purpose
SUM returns the sum of values of expr. You can use it as an aggregate or analytic
function.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. The function returns the
same data type as the numeric data type of the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

If you specify DISTINCT, then you can specify only the query_partition_clause of the
analytic_clause. The order_by_clause and windowing_clause are not allowed.
See Also: "About SQL Expressions" on page 6-1 for information on
valid forms of expr and "Aggregate Functions" on page 5-10

Aggregate Example
The following example calculates the sum of all salaries in the sample hr.employees
table:
SELECT SUM(salary) "Total"
FROM employees;
Total
---------691400

Analytic Example
The following example calculates, for each manager in the sample table hr.employees,
a cumulative total of salaries of employees who answer to that manager that are equal
to or less than the current salary. You can see that Raphaely and Cambrault have the
same cumulative total. This is because Raphaely and Cambrault have the identical
salaries, so Oracle Database adds together their salary values and applies the same
cumulative total to both rows.
SELECT manager_id, last_name, salary,
SUM(salary) OVER (PARTITION BY manager_id ORDER BY salary
RANGE UNBOUNDED PRECEDING) l_csum
FROM employees
ORDER BY manager_id, last_name, salary, l_csum;
MANAGER_ID LAST_NAME

5-276 Oracle Database SQL Language Reference

SALARY

L_CSUM

SUM

---------MANAGER_ID
---------100
100
100
100
100
100
100
100
100
100
100
. . .
149
149
149
149
201
205

--------------- ---------- ---------LAST_NAME
SALARY
L_CSUM
------------------------- ---------- ---------Cambrault
11000
68900
De Haan
17000
155400
Errazuriz
12000
80900
Fripp
8200
36400
Hartstein
13000
93900
Kaufling
7900
20200
Kochhar
17000
155400
Mourgos
5800
5800
Partners
13500
107400
Raphaely
11000
68900
Russell
14000
121400
Hutton
Johnson
Livingston
Taylor
Fay
Gietz
King

8800
6200
8400
8600
6000
8300
24000

39000
6200
21600
30200
6000
8300
24000

Functions

5-277

SYS_CONNECT_BY_PATH

SYS_CONNECT_BY_PATH
Syntax
SYS_CONNECT_BY_PATH

(

column

,

char

)

Purpose
SYS_CONNECT_BY_PATH is valid only in hierarchical queries. It returns the path of a
column value from root to node, with column values separated by char for each row
returned by CONNECT BY condition.
Both column and char can be any of the data types CHAR, VARCHAR2, NCHAR, or
NVARCHAR2. The string returned is of VARCHAR2 data type and is in the same character
set as column.
See Also: "Hierarchical Queries" on page 9-3 for more information
about hierarchical queries and CONNECT BY conditions

Examples
The following example returns the path of employee names from employee Kochhar to
all employees of Kochhar (and their employees):
SELECT LPAD(' ', 2*level-1)||SYS_CONNECT_BY_PATH(last_name, '/') "Path"
FROM employees
START WITH last_name = 'Kochhar'
CONNECT BY PRIOR employee_id = manager_id;
Path
-----------------------------/Kochhar/Greenberg/Chen
/Kochhar/Greenberg/Faviet
/Kochhar/Greenberg/Popp
/Kochhar/Greenberg/Sciarra
/Kochhar/Greenberg/Urman
/Kochhar/Higgins/Gietz
/Kochhar/Baer
/Kochhar/Greenberg
/Kochhar/Higgins
/Kochhar/Mavris
/Kochhar/Whalen
/Kochhar

5-278 Oracle Database SQL Language Reference

SYS_CONTEXT

SYS_CONTEXT
Syntax
,
SYS_CONTEXT

(

’

namespace

’

,

’

parameter

length

’

)

Purpose
SYS_CONTEXT returns the value of parameter associated with the context namespace at
the current instant. You can use this function in both SQL and PL/SQL statements.
SYS_CONTEXT must be executed locally.
For namespace and parameter, you can specify either a string or an expression that
resolves to a string designating a namespace or an attribute. If you specify literal
arguments for namespace and parameter, and you are using SYS_CONTEXT explicitly in
a SQL statement—rather than in a PL/SQL function that in turn is in mentioned in a
SQL statement—then Oracle Database evaluates SYS_CONTEXT only once per SQL
statement execution for each call site that invokes the SYS_CONTEXT function.
The context namespace must already have been created, and the associated parameter
and its value must also have been set using the DBMS_SESSION.set_context procedure.
The namespace must be a valid SQL identifier. The parameter name can be any string.
It is not case sensitive, but it cannot exceed 30 bytes in length.
The data type of the return value is VARCHAR2. The default maximum size of the return
value is 256 bytes. You can override this default by specifying the optional length
parameter, which must be a NUMBER or a value that can be implicitly converted to
NUMBER. The valid range of values is 1 to 4000 bytes. If you specify an invalid value,
then Oracle Database ignores it and uses the default.
Oracle provides the following built-in namespaces:
■

■

USERENV - Describes the current session. The predefined parameters of namespace
USERENV are listed in Table 5–11 on page 5-281.
SYS_SESSION_ROLES - Indicates whether a specified role is currently enabled for the
session. This namespace is available starting with Oracle Database 11g Release 2
(11.2.0.4).
See Also:
■

■

■

Oracle Database Security Guide for information on using the
application context feature in your application development
CREATE CONTEXT on page 14-9 for information on creating
user-defined context namespaces
Oracle Database PL/SQL Packages and Types Reference for
information on the DBMS_SESSION.set_context procedure

Examples
The following statement returns the name of the user who logged onto the database:
CONNECT OE
Enter password: password
SELECT SYS_CONTEXT ('USERENV', 'SESSION_USER')
FROM DUAL;

Functions

5-279

SYS_CONTEXT

SYS_CONTEXT ('USERENV', 'SESSION_USER')
-----------------------------------------------------OE

The following example queries the SESSION_ROLES data dictionary view to show that
RESOURCE is the only role currently enabled for the session. It then uses the SYS_
CONTEXT function to show that the RESOURCE role is currently enabled for the session
and the DBA role is not.
CONNECT OE
Enter password: password
SELECT role FROM session_roles;
ROLE
-------RESOURCE
SELECT SYS_CONTEXT('SYS_SESSION_ROLES', 'RESOURCE')
FROM DUAL
SYS_CONTEXT('SYS_SESSION_ROLES','RESOURCE')
-------------------------------------TRUE
SELECT SYS_CONTEXT('SYS_SESSION_ROLES', 'DBA')
FROM DUAL;
SYS_CONTEXT('SYS_SESSION_ROLES','DBA')
-------------------------------------FALSE

For simplicity in demonstrating this feature, these examples
do not perform the password management techniques that a deployed
system normally uses. In a production environment, follow the Oracle
Database password management guidelines, and disable any sample
accounts. See Oracle Database Security Guide for password management
guidelines and other security recommendations.

Note:

The following hypothetical example returns the group number that was set as the
value for the attribute group_no in the PL/SQL package that was associated with the
context hr_apps when hr_apps was created:
SELECT SYS_CONTEXT ('hr_apps', 'group_no') "User Group"
FROM DUAL;

5-280 Oracle Database SQL Language Reference

SYS_CONTEXT

Table 5–11

Predefined Parameters of Namespace USERENV

Parameter

Return Value

ACTION

Identifies the position in the module (application name) and is set through the
DBMS_APPLICATION_INFO package or OCI.

AUDITED_CURSORID

Returns the cursor ID of the SQL that triggered the audit. This parameter is not
valid in a fine-grained auditing environment. If you specify it in such an
environment, then Oracle Database always returns NULL.

AUTHENTICATED_IDENTITY

Returns the identity used in authentication. In the list that follows, the type of user
is followed by the value returned:
■
■

Kerberos-authenticated external user : kerberos principal name; same as the
schema name

■

SSL-authenticated enterprise user: the DN in the user's PKI certificate

■

SSL-authenticated external user: the DN in the user's PKI certificate

■

Password-authenticated enterprise user: nickname; same as the login name

■

Password-authenticated database user: the database username; same as the
schema name

■

OS-authenticated external user: the external operating system user name

■

Radius-authenticated external user: the schema name

■

Proxy with DN : Oracle Internet Directory DN of the client

■

Proxy with certificate: certificate DN of the client

■

AUTHENTICATION_DATA

Kerberos-authenticated enterprise user: kerberos principal name

Proxy with username: database user name if client is a local database user;
nickname if client is an enterprise user.

■

SYSDBA/SYSOPER using Password File: login name

■

SYSDBA/SYSOPER using OS authentication: operating system user name

Data being used to authenticate the login user. For X.503 certificate authenticated
sessions, this field returns the context of the certificate in HEX2 format.
Note: You can change the return value of the AUTHENTICATION_DATA attribute using
the length parameter of the syntax. Values of up to 4000 are accepted. This is the
only attribute of USERENV for which Oracle Database implements such a change.

AUTHENTICATION_METHOD

Returns the method of authentication. In the list that follows, the type of user is
followed by the method returned:
■

Password-authenticated enterprise user, local database user, or
SYSDBA/SYSOPER using Password File; proxy with username using
password: PASSWORD

■

Kerberos-authenticated enterprise or external user: KERBEROS

■

SSL-authenticated enterprise or external user: SSL

■

Radius-authenticated external user: RADIUS

■

OS-authenticated external user or SYSDBA/SYSOPER: OS

■

Proxy with certificate, DN, or username without using password: NONE

■

Background process (job queue slave process): JOB

■

Parallel Query Slave process: PQ_SLAVE

You can use IDENTIFICATION_TYPE to distinguish between external and enterprise
users when the authentication method is Password, Kerberos, or SSL.
BG_JOB_ID

Job ID of the current session if it was established by an Oracle Database
background process. Null if the session was not established by a background
process.

Functions

5-281

SYS_CONTEXT

Table 5–11 (Cont.) Predefined Parameters of Namespace USERENV
Parameter

Return Value

CLIENT_IDENTIFIER

Returns an identifier that is set by the application through the DBMS_SESSION.SET_
IDENTIFIER procedure, the OCI attribute OCI_ATTR_CLIENT_IDENTIFIER, or Oracle
Dynamic Monitoring Service (DMS). This attribute is used by various database
components to identify lightweight application users who authenticate as the same
database user.

CLIENT_INFO

Returns up to 64 bytes of user session information that can be stored by an
application using the DBMS_APPLICATION_INFO package.

CURRENT_BIND

The bind variables for fine-grained auditing. You can specify this attribute only
inside the event handler for the fine-grained auditing feature.

CURRENT_EDITION_ID

The identifier of the current edition.

CURRENT_EDITION_NAME

The name of the current edition.

CURRENT_SCHEMA

The name of the currently active default schema. This value may change during
the duration of a session through use of an ALTER SESSION SET CURRENT_SCHEMA
statement. This may also change during the duration of a session to reflect the
owner of any active definer's rights object. When used directly in the body of a
view definition, this returns the default schema used when executing the cursor
that is using the view; it does not respect views used in the cursor as being
definer's rights.
Note: Oracle recommends against issuing the SQL statement ALTER SESSION SET
CURRENT_SCHEMA from within all types of stored PL/SQL units except logon
triggers.

CURRENT_SCHEMAID

Identifier of the currently active default schema.

CURRENT_SQL

CURRENT_SQL returns the first 4K bytes of the current SQL that triggered the
fine-grained auditing event. The CURRENT_SQLn attributes return subsequent
4K-byte increments, where n can be an integer from 1 to 7, inclusive. CURRENT_SQL1
returns bytes 4K to 8K; CURRENT_SQL2 returns bytes 8K to 12K, and so forth. You
can specify these attributes only inside the event handler for the fine-grained
auditing feature.

CURRENT_SQLn

CURRENT_SQL_LENGTH

The length of the current SQL statement that triggers fine-grained audit or
row-level security (RLS) policy functions or event handlers. You can specify this
attribute only inside the event handler for the fine-grained auditing feature.

CURRENT_USER

The name of the database user whose privileges are currently active. This may
change during the duration of a session to reflect the owner of any active definer's
rights object. When no definer's rights object is active, CURRENT_USER returns the
same value as SESSION_USER. When used directly in the body of a view definition,
this returns the user that is executing the cursor that is using the view; it does not
respect views used in the cursor as being definer's rights.

CURRENT_USERID

The identifier of the database user whose privileges are currently active.

DATABASE_ROLE

The database role using the SYS_CONTEXT function with the USERENV namespace.
The role is one of the following: PRIMARY, PHYSICAL STANDBY, LOGICAL STANDBY,
SNAPSHOT STANDBY.

DB_DOMAIN

Domain of the database as specified in the DB_DOMAIN initialization parameter.

DB_NAME

Name of the database as specified in the DB_NAME initialization parameter.

DB_UNIQUE_NAME

Name of the database as specified in the DB_UNIQUE_NAME initialization parameter.

5-282 Oracle Database SQL Language Reference

SYS_CONTEXT

Table 5–11 (Cont.) Predefined Parameters of Namespace USERENV
Parameter

Return Value

DBLINK_INFO

Returns the source of a database link session. Specifically, it returns a string of the
form:
SOURCE_GLOBAL_NAME=dblink_src_global_name, DBLINK_NAME=dblink_name,
SOURCE_AUDIT_SESSIONID=dblink_src_audit_sessionid
where:
■

dblink_src_global_name is the unique global name of the source database

■

dblink_name is the name of the database link on the source database

■

dblink_src_audit_sessionid is the audit session ID of the session on the
source database that initiated the connection to the remote database using
dblink_name

ENTRYID

The current audit entry number. The audit entryid sequence is shared between
fine-grained audit records and regular audit records. You cannot use this attribute
in distributed SQL statements. The correct auditing entry identifier can be seen
only through an audit handler for standard or fine-grained audit.

ENTERPRISE_IDENTITY

Returns the user's enterprise-wide identity:
■
■

■

For enterprise users: the Oracle Internet Directory DN.
For external users: the external identity (Kerberos principal name, Radius
schema names, OS user name, Certificate DN).
For local users and SYSDBA/SYSOPER logins: NULL.

The value of the attribute differs by proxy method:
■
■

■

For a proxy with DN: the Oracle Internet Directory DN of the client
For a proxy with certificate: the certificate DN of the client for external users;
the Oracle Internet Directory DN for global users
For a proxy with username: the Oracle Internet Directory DN if the client is an
enterprise users; NULL if the client is a local database user.

FG_JOB_ID

Job ID of the current session if it was established by a client foreground process.
Null if the session was not established by a foreground process.

GLOBAL_CONTEXT_MEMORY

Returns the number being used in the System Global Area by the globally accessed
context.

GLOBAL_UID

Returns the global user ID from Oracle Internet Directory for Enterprise User
Security (EUS) logins; returns null for all other logins.

HOST

Name of the host machine from which the client has connected.

IDENTIFICATION_TYPE

Returns the way the user's schema was created in the database. Specifically, it
reflects the IDENTIFIED clause in the CREATE/ALTER USER syntax. In the list that
follows, the syntax used during schema creation is followed by the identification
type returned:
■

IDENTIFIED BY password: LOCAL

■

IDENTIFIED EXTERNALLY: EXTERNAL

■

IDENTIFIED GLOBALLY: GLOBAL SHARED

■

IDENTIFIED GLOBALLY AS DN: GLOBAL PRIVATE

INSTANCE

The instance identification number of the current instance.

INSTANCE_NAME

The name of the instance.

IP_ADDRESS

IP address of the machine from which the client is connected. If the client and
server are on the same machine and the connection uses IPv6 addressing, then ::1
is returned.

ISDBA

Returns TRUE if the user has been authenticated as having DBA privileges either
through the operating system or through a password file.

Functions

5-283

SYS_CONTEXT

Table 5–11 (Cont.) Predefined Parameters of Namespace USERENV
Parameter

Return Value

LANG

The abbreviated name for the language, a shorter form than the existing 'LANGUAGE'
parameter.

LANGUAGE

The language and territory currently used by your session, along with the
database character set, in this form:
language_territory.characterset

MODULE

The application name (module) set through the DBMS_APPLICATION_INFO package
or OCI.

NETWORK_PROTOCOL

Network protocol being used for communication, as specified in the
'PROTOCOL=protocol' portion of the connect string.

NLS_CALENDAR

The current calendar of the current session.

NLS_CURRENCY

The currency of the current session.

NLS_DATE_FORMAT

The date format for the session.

NLS_DATE_LANGUAGE

The language used for expressing dates.

NLS_SORT

BINARY or the linguistic sort basis.

NLS_TERRITORY

The territory of the current session.

OS_USER

Operating system user name of the client process that initiated the database
session.

POLICY_INVOKER

The invoker of row-level security (RLS) policy functions.

PROXY_ENTERPRISE_IDENTITY

Returns the Oracle Internet Directory DN when the proxy user is an enterprise
user.

PROXY_USER

Name of the database user who opened the current session on behalf of SESSION_
USER.

PROXY_USERID

Identifier of the database user who opened the current session on behalf of
SESSION_USER.

SERVER_HOST

The host name of the machine on which the instance is running.

SERVICE_NAME

The name of the service to which a given session is connected.

SESSION_EDITION_ID

The identifier of the session edition.

SESSION_EDITION_NAME

The name of the session edition.

SESSION_USER

The name of the database user at logon. For enterprise users, returns the schema.
For other users, returns the database user name. This value remains the same
throughout the duration of the session.

SESSION_USERID

The identifier of the database user at logon.

SESSIONID

The auditing session identifier. You cannot use this attribute in distributed SQL
statements.

SID

The session ID.

STATEMENTID

The auditing statement identifier. STATEMENTID represents the number of SQL
statements audited in a given session. You cannot use this attribute in distributed
SQL statements. The correct auditing statement identifier can be seen only through
an audit handler for standard or fine-grained audit.

TERMINAL

The operating system identifier for the client of the current session. In distributed
SQL statements, this attribute returns the identifier for your local session. In a
distributed environment, this is supported only for remote SELECT statements, not
for remote INSERT, UPDATE, or DELETE operations. (The return length of this
parameter may vary by operating system.)

5-284 Oracle Database SQL Language Reference

SYS_CONTEXT

Table 5–12 lists the parameters of namespace USERENV that have been deprecated. Do
not specify any of these parameters. Instead use the alternatives suggested in the
Comments column.
Table 5–12

Deprecated Parameters of Namespace USERENV

Parameter

Comments

AUTHENTICATION_TYPE

This parameter returned a value indicating how the user was authenticated. The same
information is now available from the new AUTHENTICATION_METHOD parameter
combined with IDENTIFICATION_TYPE.

EXTERNAL_NAME

This parameter returned the external name of the user. More complete information
can now be obtained from the AUTHENTICATED_IDENTITY and ENTERPRISE_IDENTITY
parameter.

Functions

5-285

SYS_DBURIGEN

SYS_DBURIGEN
Syntax
,
column
SYS_DBURIGEN

rowid

,

(

’

text

(

)

’
)

attribute

Purpose
SYS_DBURIGen takes as its argument one or more columns or attributes, and optionally
a rowid, and generates a URL of data type DBURIType to a particular column or row
object. You can then use the URL to retrieve an XML document from the database.
All columns or attributes referenced must reside in the same table. They must perform
the function of a primary key. They need not actually match the primary key of the
table, but they must reference a unique value. If you specify multiple columns, then all
but the final column identify the row in the database, and the last column specified
identifies the column within the row.
By default the URL points to a formatted XML document. If you want the URL to
point only to the text of the document, then specify the optional 'text()'.
In this XML context, the lowercase text is a keyword, not a
syntactic placeholder.

Note:

If the table or view containing the columns or attributes does not have a schema
specified in the context of the query, then Oracle Database interprets the table or view
name as a public synonym.
Oracle XML DB Developer's Guide for information on the
URIType data type and XML documents in the database
See Also:

Examples
The following example uses the SYS_DBURIGen function to generate a URL of data type
DBURIType to the email column of the row in the sample table hr.employees where the
employee_id = 206:
SELECT SYS_DBURIGEN(employee_id, email)
FROM employees
WHERE employee_id = 206;
SYS_DBURIGEN(EMPLOYEE_ID,EMAIL)(URL, SPARE)
-------------------------------------------------------------------DBURITYPE('/PUBLIC/EMPLOYEES/ROW[EMPLOYEE_ID=''206'']/EMAIL', NULL)

5-286 Oracle Database SQL Language Reference

SYS_EXTRACT_UTC

SYS_EXTRACT_UTC
Syntax
SYS_EXTRACT_UTC

(

datetime_with_timezone

)

Purpose
SYS_EXTRACT_UTC extracts the UTC (Coordinated Universal Time—formerly
Greenwich Mean Time) from a datetime value with time zone offset or time zone
region name. If a time zone is not specified, then the datetime is associated with the
session time zone.

Examples
The following example extracts the UTC from a specified datetime:
SELECT SYS_EXTRACT_UTC(TIMESTAMP '2000-03-28 11:30:00.00 -08:00')
FROM DUAL;
SYS_EXTRACT_UTC(TIMESTAMP'2000-03-2811:30:00.00-08:00')
----------------------------------------------------------------28-MAR-00 07.30.00 PM

Functions

5-287

SYS_GUID

SYS_GUID
Syntax
SYS_GUID

(

)

Purpose
SYS_GUID generates and returns a globally unique identifier (RAW value) made up of 16
bytes. On most platforms, the generated identifier consists of a host identifier, a
process or thread identifier of the process or thread invoking the function, and a
nonrepeating value (sequence of bytes) for that process or thread.

Examples
The following example adds a column to the sample table hr.locations, inserts
unique identifiers into each row, and returns the 32-character hexadecimal
representation of the 16-byte RAW value of the global unique identifier:
ALTER TABLE locations ADD (uid_col RAW(16));
UPDATE locations SET uid_col = SYS_GUID();
SELECT location_id, uid_col FROM locations
ORDER BY location_id, uid_col;
LOCATION_ID
----------1000
1100
1200
1300
1400
1500
. . .

UID_COL
---------------------------------------------------------------09F686761827CF8AE040578CB20B7491
09F686761828CF8AE040578CB20B7491
09F686761829CF8AE040578CB20B7491
09F68676182ACF8AE040578CB20B7491
09F68676182BCF8AE040578CB20B7491
09F68676182CCF8AE040578CB20B7491

5-288 Oracle Database SQL Language Reference

SYS_TYPEID

SYS_TYPEID
Syntax
SYS_TYPEID

(

object_type_value

)

Purpose
SYS_TYPEID returns the typeid of the most specific type of the operand. This value is
used primarily to identify the type-discriminant column underlying a substitutable
column. For example, you can use the value returned by SYS_TYPEID to build an index
on the type-discriminant column.
You can use this function only on object type operands. All final root object
types—final types not belonging to a type hierarchy—have a null typeid. Oracle
Database assigns to all types belonging to a type hierarchy a unique non-null typeid.
See Also: Oracle Database Object-Relational Developer's Guide for more
information on typeids

Examples
The following examples use the tables persons and books, which are created in
"Substitutable Table and Column Examples" on page 16-71. The first query returns the
most specific types of the object instances stored in the persons table.
SELECT name, SYS_TYPEID(VALUE(p)) "Type_id" FROM persons p;
NAME
------------------------Bob
Joe
Tim

Type_id
-------------------------------01
02
03

The next query returns the most specific types of authors stored in the table books:
SELECT b.title, b.author.name, SYS_TYPEID(author)
"Type_ID" FROM books b;
TITLE
------------------------An Autobiography
Business Rules
Mixing School and Work

AUTHOR.NAME
-------------------Bob
Joe
Tim

Type_ID
------------------01
02
03

You can use the SYS_TYPEID function to create an index on the type-discriminant
column of a table. For an example, see "Indexing on Substitutable Columns: Examples"
on page 14-85.

Functions

5-289

SYS_XMLAGG

SYS_XMLAGG
Syntax
,
SYS_XMLAGG

(

expr

fmt
)

Purpose
SYS_XMLAgg aggregates all of the XML documents or fragments represented by expr
and produces a single XML document. It adds a new enclosing element with a default
name ROWSET. If you want to format the XML document differently, then specify fmt,
which is an instance of the XMLFormat object.
See Also: SYS_XMLGEN on page 5-291 and "XML Format Model"
on page 3-70 for using the attributes of the XMLFormat type to format
SYS_XMLAgg results

Examples
The following example uses the SYS_XMLGen function to generate an XML document
for each row of the sample table employees where the employee's last name begins
with the letter R, and then aggregates all of the rows into a single XML document in
the default enclosing element ROWSET:
SELECT SYS_XMLAGG(SYS_XMLGEN(last_name)) XMLAGG
FROM employees
WHERE last_name LIKE 'R%'
ORDER BY xmlagg;
XMLAGG
-------------------------------------------------------------------------------

Rajs
Raphaely
Rogers
Russell


5-290 Oracle Database SQL Language Reference

SYS_XMLGEN

SYS_XMLGEN
Syntax
,
SYS_XMLGEN

(

expr

fmt
)

Purpose
SYS_XMLGen takes an expression that evaluates to a particular row and column of the
database, and returns an instance of type XMLType containing an XML document. The
expr can be a scalar value, a user-defined type, or an XMLType instance.
■

■

■

If expr is a scalar value, then the function returns an XML element containing the
scalar value.
If expr is a type, then the function maps the user-defined type attributes to XML
elements.
If expr is an XMLType instance, then the function encloses the document in an XML
element whose default tag name is ROW.

By default the elements of the XML document match the elements of expr. For
example, if expr resolves to a column name, then the enclosing XML element will be
the same column name. If you want to format the XML document differently, then
specify fmt, which is an instance of the XMLFormat object.
"XML Format Model" on page 3-70 for a description of the
XMLFormat type and how to use its attributes to format SYS_XMLGen
results
See Also:

Examples
The following example retrieves the employee email ID from the sample table
oe.employees where the employee_id value is 205, and generates an instance of an
XMLType containing an XML document with an EMAIL element.
SELECT SYS_XMLGEN(email)
FROM employees
WHERE employee_id = 205;
SYS_XMLGEN(EMAIL)
------------------------------------------------------------------
SHIGGINS

Functions

5-291

SYSDATE

SYSDATE
Syntax
SYSDATE

Purpose
SYSDATE returns the current date and time set for the operating system on which the
database server resides. The data type of the returned value is DATE, and the format
returned depends on the value of the NLS_DATE_FORMAT initialization parameter. The
function requires no arguments. In distributed SQL statements, this function returns
the date and time set for the operating system of your local database. You cannot use
this function in the condition of a CHECK constraint.

Note: The FIXED_DATE initialization parameter enables you to set a
constant date and time that SYSDATE will always return instead of the
current date and time. This parameter is useful primarily for testing.
Refer to Oracle Database Reference for more information on the FIXED_
DATE initialization parameter.

Examples
The following example returns the current operating system date and time:
SELECT TO_CHAR
(SYSDATE, 'MM-DD-YYYY HH24:MI:SS') "NOW"
FROM DUAL;
NOW
------------------04-13-2001 09:45:51

5-292 Oracle Database SQL Language Reference

SYSTIMESTAMP

SYSTIMESTAMP
Syntax
SYSTIMESTAMP

Purpose
SYSTIMESTAMP returns the system date, including fractional seconds and time zone, of
the system on which the database resides. The return type is TIMESTAMP WITH TIME
ZONE.

Examples
The following example returns the system timestamp:
SELECT SYSTIMESTAMP FROM DUAL;
SYSTIMESTAMP
-----------------------------------------------------------------28-MAR-00 12.38.55.538741 PM -08:00

The following example shows how to explicitly specify fractional seconds:
SELECT TO_CHAR(SYSTIMESTAMP, 'SSSSS.FF') FROM DUAL;
TO_CHAR(SYSTIME
--------------55615.449255

The following example returns the current timestamp in a specified time zone:
SELECT SYSTIMESTAMP AT TIME ZONE 'UTC' FROM DUAL;
SYSTIMESTAMPATTIMEZONE'UTC'
--------------------------------------------------------------------------08-07-21 20:39:52,743557 UTC

The output format in this example depends on the NLS_TIMESTAMP_TZ_FORMAT for the
session.

Functions

5-293

TAN

TAN
Syntax
TAN

(

n

)

Purpose
TAN returns the tangent of n (an angle expressed in radians).
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. If the argument is
BINARY_FLOAT, then the function returns BINARY_DOUBLE. Otherwise the function
returns the same numeric data type as the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

Examples
The following example returns the tangent of 135 degrees:
SELECT TAN(135 * 3.14159265359/180)
"Tangent of 135 degrees" FROM DUAL;
Tangent of 135 degrees
---------------------- 1

5-294 Oracle Database SQL Language Reference

TANH

TANH
Syntax
TANH

(

n

)

Purpose
TANH returns the hyperbolic tangent of n.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. If the argument is
BINARY_FLOAT, then the function returns BINARY_DOUBLE. Otherwise the function
returns the same numeric data type as the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

Examples
The following example returns the hyperbolic tangent of .5:
SELECT TANH(.5) "Hyperbolic tangent of .5"
FROM DUAL;
Hyperbolic tangent of .5
-----------------------.462117157

Functions

5-295

TIMESTAMP_TO_SCN

TIMESTAMP_TO_SCN
Syntax
TIMESTAMP_TO_SCN

(

timestamp

)

Purpose
TIMESTAMP_TO_SCN takes as an argument a timestamp value and returns the
approximate system change number (SCN) associated with that timestamp. The
returned value is of data type NUMBER. This function is useful any time you want to
know the SCN associated with a particular timestamp.
The association between an SCN and a timestamp when the
SCN is generated is remembered by the database for a limited period
of time. This period is the maximum of the auto-tuned undo retention
period, if the database runs in the Automatic Undo Management
mode, and the retention times of all flashback archives in the database,
but no less than 120 hours. The time for the association to become
obsolete elapses only when the database is open. An error is returned
if the timestamp specified for the argument to TIMESTAMP_TO_SCN is
too old.
Note:

See Also: SCN_TO_TIMESTAMP on page 5-242 for information on
converting SCNs to timestamp

Examples
The following example inserts a row into the oe.orders table and then uses
TIMESTAMP_TO_SCN to determine the system change number of the insert operation.
(The actual SCN returned will differ on each system.)
INSERT INTO orders (order_id, order_date, customer_id, order_total)
VALUES (5000, SYSTIMESTAMP, 188, 2345);
1 row created.
COMMIT;
Commit complete.
SELECT TIMESTAMP_TO_SCN(order_date) FROM orders
WHERE order_id = 5000;
TIMESTAMP_TO_SCN(ORDER_DATE)
---------------------------574100

5-296 Oracle Database SQL Language Reference

TO_BINARY_DOUBLE

TO_BINARY_DOUBLE
Syntax
,
,
TO_BINARY_DOUBLE

(

’

nlsparam

’

fmt

expr

)

Purpose
TO_BINARY_DOUBLE returns a double-precision floating-point number.
■

■

expr can be a character string or a numeric value of type NUMBER, BINARY_FLOAT, or
BINARY_DOUBLE. If expr is BINARY_DOUBLE, then the function returns expr.
The optional 'fmt' and 'nlsparam' arguments are valid only if expr is a character
string. They serve the same purpose as for the TO_CHAR (number) function.
–

The case-insensitive string 'INF' is converted to positive infinity.

–

The case-insensitive string '-INF' is converted to negative identity.

–

The case-insensitive string 'NaN' is converted to NaN (not a number).

You cannot use a floating-point number format element (F, f, D, or d) in a character
string expr.
Conversions from character strings or NUMBER to BINARY_DOUBLE can be inexact,
because the NUMBER and character types use decimal precision to represent the numeric
value, and BINARY_DOUBLE uses binary precision.
Conversions from BINARY_FLOAT to BINARY_DOUBLE are exact.
TO_CHAR (number) on page 5-305 and "Floating-Point
Numbers" on page 3-12

See Also:

Examples
The examples that follow are based on a table with three columns, each with a
different numeric data type:
CREATE TABLE float_point_demo
(dec_num NUMBER(10,2), bin_double BINARY_DOUBLE, bin_float BINARY_FLOAT);
INSERT INTO float_point_demo
VALUES (1234.56,1234.56,1234.56);
SELECT * FROM float_point_demo;
DEC_NUM BIN_DOUBLE BIN_FLOAT
---------- ---------- ---------1234.56 1.235E+003 1.235E+003

The following example converts a value of data type NUMBER to a value of data type
BINARY_DOUBLE:
SELECT dec_num, TO_BINARY_DOUBLE(dec_num)
FROM float_point_demo;
DEC_NUM TO_BINARY_DOUBLE(DEC_NUM)
---------- -------------------------

Functions

5-297

TO_BINARY_DOUBLE

1234.56

1.235E+003

The following example compares extracted dump information from the dec_num and
bin_double columns:
SELECT DUMP(dec_num) "Decimal",
DUMP(bin_double) "Double"
FROM float_point_demo;
Decimal
Double
--------------------------- --------------------------------------------Typ=2 Len=4: 194,13,35,57
Typ=101 Len=8: 192,147,74,61,112,163,215,10

5-298 Oracle Database SQL Language Reference

TO_BINARY_FLOAT

TO_BINARY_FLOAT
Syntax
,
,
TO_BINARY_FLOAT

(

’

nlsparam

’

fmt

expr

)

Purpose
TO_BINARY_FLOAT returns a single-precision floating-point number.
■

■

expr can be a character string or a numeric value of type NUMBER, BINARY_FLOAT, or
BINARY_DOUBLE. If expr is BINARY_FLOAT, then the function returns expr.
The optional 'fmt' and 'nlsparam' arguments are valid only if expr is a character
string. They serve the same purpose as for the TO_CHAR (number) function.
–

The incase-sensitive string 'INF' is converted to positive infinity.

–

The incase-sensitive string '-INF' is converted to negative identity.

–

The incase-sensitive string 'NaN' is converted to NaN (not a number).

You cannot use a floating-point number format element (F, f, D, or d) in a character
string expr.
Conversions from character strings or NUMBER to BINARY_FLOAT can be inexact, because
the NUMBER and character types use decimal precision to represent the numeric value
and BINARY_FLOAT uses binary precision.
Conversions from BINARY_DOUBLE to BINARY_FLOAT are inexact if the BINARY_DOUBLE
value uses more bits of precision than supported by the BINARY_FLOAT.
TO_CHAR (number) on page 5-305 and "Floating-Point
Numbers" on page 3-12

See Also:

Examples
Using table float_point_demo created for TO_BINARY_DOUBLE on page 5-297, the
following example converts a value of data type NUMBER to a value of data type
BINARY_FLOAT:
SELECT dec_num, TO_BINARY_FLOAT(dec_num)
FROM float_point_demo;
DEC_NUM TO_BINARY_FLOAT(DEC_NUM)
---------- -----------------------1234.56
1.235E+003

Functions

5-299

TO_BLOB

TO_BLOB
Syntax
to_blob::=
TO_BLOB

(

raw_value

)

Purpose
TO_BLOB converts LONG RAW and RAW values to BLOB values.
From within a PL/SQL package, you can use TO_BLOB to convert RAW and BLOB values
to BLOB.

Examples
The following hypothetical example returns the BLOB of a RAW column value:
SELECT TO_BLOB(raw_column) blob FROM raw_table;
BLOB
----------------------00AADD343CDBBD

5-300 Oracle Database SQL Language Reference

TO_CHAR (character)

TO_CHAR (character)
Syntax
to_char_char::=
nchar
TO_CHAR

(

clob

)

nclob

Purpose
TO_CHAR (character) converts NCHAR, NVARCHAR2, CLOB, or NCLOB data to the database
character set. The value returned is always VARCHAR2.
When you use this function to convert a character LOB into the database character set,
if the LOB value to be converted is larger than the target type, then the database
returns an error.

Examples
The following example interprets a simple string as character data:
SELECT TO_CHAR('01110') FROM DUAL;
TO_CH
----01110

Compare this example with the first example for TO_CHAR (number) on page 5-305.
The following example converts some CLOB data from the pm.print_media table to the
database character set:
SELECT TO_CHAR(ad_sourcetext) FROM print_media
WHERE product_id = 2268;
TO_CHAR(AD_SOURCETEXT)
-------------------------------------------------------------------******************************
TIGER2 2268...Standard Hayes Compatible Modem
Product ID: 2268
The #1 selling modem in the universe! Tiger2's modem includes call management
and Internet voicing. Make real-time full duplex phone calls at the same time
you're online.
**********************************

Functions

5-301

TO_CHAR (datetime)

TO_CHAR (datetime)
Syntax
to_char_date::=
,
datetime
TO_CHAR

,

’

nlsparam

’

fmt

(

)
interval

Purpose
TO_CHAR (datetime) converts a datetime or interval value of DATE, TIMESTAMP,
TIMESTAMP WITH TIME ZONE, TIMESTAMP WITH LOCAL TIME ZONE, INTERVAL DAY TO SECOND,
or INTERVAL YEAR TO MONTH data type to a value of VARCHAR2 data type in the format
specified by the date format fmt. If you omit fmt, then date is converted to a VARCHAR2
value as follows:
■
■

■

■

DATE values are converted to values in the default date format.
TIMESTAMP and TIMESTAMP WITH LOCAL TIME ZONE values are converted to values in
the default timestamp format.
TIMESTAMP WITH TIME ZONE values are converted to values in the default timestamp
with time zone format.
Interval values are converted to the numeric representation of the interval literal.

Refer to "Format Models" on page 3-56 for information on datetime formats.
The 'nlsparam' argument specifies the language in which month and day names and
abbreviations are returned. This argument can have this form:
'NLS_DATE_LANGUAGE = language'

If you omit 'nlsparam', then this function uses the default date language for your
session.
See Also: "Security Considerations for Data Conversion" on
page 3-44

You can use this function in conjunction with any of the XML functions to generate a
date in the database format rather than the XML Schema standard format.
See Also:
■

■

Oracle XML DB Developer's Guide for information about formatting
of XML dates and timestamps, including examples
"XML Functions" on page 5-8 for a listing of the XML functions

Examples
The following example uses this table:
CREATE TABLE date_tab (
ts_col
TIMESTAMP,
tsltz_col
TIMESTAMP WITH LOCAL TIME ZONE,
tstz_col
TIMESTAMP WITH TIME ZONE);

5-302 Oracle Database SQL Language Reference

TO_CHAR (datetime)

The example shows the results of applying TO_CHAR to different TIMESTAMP data types.
The result for a TIMESTAMP WITH LOCAL TIME ZONE column is sensitive to session time
zone, whereas the results for the TIMESTAMP and TIMESTAMP WITH TIME ZONE columns
are not sensitive to session time zone:
ALTER SESSION SET TIME_ZONE = '-8:00';
INSERT INTO date_tab VALUES (
TIMESTAMP'1999-12-01 10:00:00',
TIMESTAMP'1999-12-01 10:00:00',
TIMESTAMP'1999-12-01 10:00:00');
INSERT INTO date_tab VALUES (
TIMESTAMP'1999-12-02 10:00:00 -8:00',
TIMESTAMP'1999-12-02 10:00:00 -8:00',
TIMESTAMP'1999-12-02 10:00:00 -8:00');
SELECT TO_CHAR(ts_col, 'DD-MON-YYYY HH24:MI:SSxFF') AS ts_date,
TO_CHAR(tstz_col, 'DD-MON-YYYY HH24:MI:SSxFF TZH:TZM') AS tstz_date
FROM date_tab
ORDER BY ts_date, tstz_date;
TS_DATE
-----------------------------01-DEC-1999 10:00:00.000000
02-DEC-1999 10:00:00.000000

TSTZ_DATE
------------------------------------01-DEC-1999 10:00:00.000000 -08:00
02-DEC-1999 10:00:00.000000 -08:00

SELECT SESSIONTIMEZONE,
TO_CHAR(tsltz_col, 'DD-MON-YYYY HH24:MI:SSxFF') AS tsltz
FROM date_tab
ORDER BY sessiontimezone, tsltz;
SESSIONTIM
----------08:00
-08:00

TSLTZ
-----------------------------01-DEC-1999 10:00:00.000000
02-DEC-1999 10:00:00.000000

ALTER SESSION SET TIME_ZONE = '-5:00';
SELECT TO_CHAR(ts_col, 'DD-MON-YYYY HH24:MI:SSxFF') AS ts_col,
TO_CHAR(tstz_col, 'DD-MON-YYYY HH24:MI:SSxFF TZH:TZM') AS tstz_col
FROM date_tab
ORDER BY ts_col, tstz_col;
TS_COL
-----------------------------01-DEC-1999 10:00:00.000000
02-DEC-1999 10:00:00.000000

TSTZ_COL
------------------------------------01-DEC-1999 10:00:00.000000 -08:00
02-DEC-1999 10:00:00.000000 -08:00

SELECT SESSIONTIMEZONE,
TO_CHAR(tsltz_col, 'DD-MON-YYYY HH24:MI:SSxFF') AS tsltz_col
FROM date_tab
ORDER BY sessiontimezone, tsltz_col;
2
3
4
SESSIONTIM TSLTZ_COL
---------- ------------------------------05:00
01-DEC-1999 13:00:00.000000
-05:00
02-DEC-1999 13:00:00.000000

The following example converts an interval literal into a text literal:
SELECT TO_CHAR(INTERVAL '123-2' YEAR(3) TO MONTH) FROM DUAL;

Functions

5-303

TO_CHAR (datetime)

TO_CHAR
------+123-02

5-304 Oracle Database SQL Language Reference

TO_CHAR (number)

TO_CHAR (number)
Syntax
to_char_number::=
,
,
TO_CHAR

(

’

nlsparam

’

fmt

n

)

Purpose
TO_CHAR (number) converts n to a value of VARCHAR2 data type, using the optional
number format fmt. The value n can be of type NUMBER, BINARY_FLOAT, or BINARY_
DOUBLE. If you omit fmt, then n is converted to a VARCHAR2 value exactly long enough
to hold its significant digits.
If n is negative, then the sign is applied after the format is applied. Thus TO_CHAR(-1,
'$9') returns -$1, rather than $-1.
Refer to "Format Models" on page 3-56 for information on number formats.
The 'nlsparam' argument specifies these characters that are returned by number
format elements:
■

Decimal character

■

Group separator

■

Local currency symbol

■

International currency symbol

This argument can have this form:
'NLS_NUMERIC_CHARACTERS = ''dg''
NLS_CURRENCY = ''text''
NLS_ISO_CURRENCY = territory '

The characters d and g represent the decimal character and group separator,
respectively. They must be different single-byte characters. Within the quoted string,
you must use two single quotation marks around the parameter values. Ten characters
are available for the currency symbol.
If you omit 'nlsparam' or any one of the parameters, then this function uses the
default parameter values for your session.
See Also: "Security Considerations for Data Conversion" on
page 3-44

Examples
The following statement uses implicit conversion to combine a string and a number
into a number:
SELECT TO_CHAR('01110' + 1) FROM DUAL;
TO_C
---1111

Functions

5-305

TO_CHAR (number)

Compare this example with the first example for TO_CHAR (character) on page 5-301.
In the next example, the output is blank padded to the left of the currency symbol.
SELECT TO_CHAR(-10000,'L99G999D99MI') "Amount"
FROM DUAL;
Amount
-------------$10,000.00SELECT TO_CHAR(-10000,'L99G999D99MI',
'NLS_NUMERIC_CHARACTERS = '',.''
NLS_CURRENCY = ''AusDollars'' ') "Amount"
FROM DUAL;
Amount
------------------AusDollars10.000,00-

In the optional number format fmt, L designates local currency symbol and MI
designates a trailing minus sign. See Table 3–17, " Matching Character Data and
Format Models with the FX Format Model Modifier" on page 3-68 for a complete
listing of number format elements.

5-306 Oracle Database SQL Language Reference

TO_CLOB

TO_CLOB
Syntax
lob_column
TO_CLOB

(

)
char

Purpose
TO_CLOB converts NCLOB values in a LOB column or other character strings to CLOB
values. char can be any of the data types CHAR, VARCHAR2, NCHAR, NVARCHAR2, CLOB, or
NCLOB. Oracle Database executes this function by converting the underlying LOB data
from the national character set to the database character set.
From within a PL/SQL package, you can use the TO_CLOB function to convert RAW,
CHAR, VARCHAR, VARCHAR2, NCHAR, NVARCHAR2, CLOB, or NCLOB values to CLOB or NCLOB
values.

Examples
The following statement converts NCLOB data from the sample pm.print_media table to
CLOB and inserts it into a CLOB column, replacing existing data in that column.
UPDATE PRINT_MEDIA
SET AD_FINALTEXT = TO_CLOB (AD_FLTEXTN);

Functions

5-307

TO_DATE

TO_DATE
Syntax
,
,
TO_DATE

(

’

nlsparam

’

fmt

char

)

Purpose
TO_DATE converts char of CHAR, VARCHAR2, NCHAR, or NVARCHAR2 data type to a value of
DATE data type.
This function does not convert data to any of the other
datetime data types. For information on other datetime conversions,
refer to TO_TIMESTAMP on page 5-320, TO_TIMESTAMP_TZ on
page 5-321, TO_DSINTERVAL on page 5-310, and TO_YMINTERVAL
on page 5-323.

Note:

The fmt is a datetime model format specifying the format of char. If you omit fmt,
then char must be in the default date format. The default date format is determined
implicitly by the NLS_TERRITORY initialization parameter or can be set explicitly by the
NLS_DATE_FORMAT parameter. If fmt is J, for Julian, then char must be an integer.
Caution: It is good practice always to specify a format mask (fmt)
with TO_DATE, as shown in the examples in the section that follows.
When it is used without a format mask, the function is valid only if
char uses the same format as is determined by the NLS_TERRITORY or
NLS_DATE_FORMAT parameters. Furthermore, the function may not be
stable across databases unless the explicit format mask is specified to
avoid dependencies.

The 'nlsparam' argument specifies the language of the text string that is being
converted to a date. This argument can have this form:
'NLS_DATE_LANGUAGE = language'

Do not use the TO_DATE function with a DATE value for the char argument. The first
two digits of the returned DATE value can differ from the original char, depending on
fmt or the default date format.
This function does not support CLOB data directly. However, CLOBs can be passed in as
arguments through implicit data conversion.
"Datetime Format Models" on page 3-60 and "Data Type
Comparison Rules" on page 3-36 for more information

See Also:

Examples
The following example converts a character string into a date:
SELECT TO_DATE(
'January 15, 1989, 11:00 A.M.',
'Month dd, YYYY, HH:MI A.M.',

5-308 Oracle Database SQL Language Reference

TO_DATE

'NLS_DATE_LANGUAGE = American')
FROM DUAL;
TO_DATE('
--------15-JAN-89

The value returned reflects the default date format if the NLS_TERRITORY parameter is
set to 'AMERICA'. Different NLS_TERRITORY values result in different default date
formats:
ALTER SESSION SET NLS_TERRITORY = 'KOREAN';
SELECT TO_DATE(
'January 15, 1989, 11:00 A.M.',
'Month dd, YYYY, HH:MI A.M.',
'NLS_DATE_LANGUAGE = American')
FROM DUAL;
TO_DATE(
-------89/01/15

Functions

5-309

TO_DSINTERVAL

TO_DSINTERVAL
Syntax
sql_format
TO_DSINTERVAL

(

’

’

)

ds_iso_format

sql_format::=
+
–

.
days

hours

:

minutes

:

frac_secs

seconds

ds_iso_format::=
–

days

D

P

.
hours

H

minutes

M

frac_secs

seconds

S

T

In earlier releases, the TO_DSINTERVAL function accepted an
optional nlsparam clause. This clause is still accepted for backward
compatibility, but has no effect.

Note:

Purpose
TO_DSINTERVAL converts a character string of CHAR, VARCHAR2, NCHAR, or NVARCHAR2 data
type to an INTERVAL DAY TO SECOND type.
TO_DSINTERVAL accepts argument in one of the two formats:
■

SQL interval format compatible with the SQL standard (ISO/IEC 9075:2003)

■

ISO duration format compatible with the ISO 8601:2004 standard

In the SQL format, days is an integer between 0 and 999999999, hours is an integer
between 0 and 23, and minutes and seconds are integers between 0 and 59. frac_secs
is the fractional part of seconds between .0 and .999999999. One or more blanks
separate days from hours. Additional blanks are allowed between format elements.
In the ISO format, days, hours, minutes and seconds are integers between 0 and
999999999. frac_secs is the fractional part of seconds between .0 and .999999999. No
blanks are allowed in the value. If you specify T, then you must specify at least one of
the hours, minutes, or seconds values.

Examples
The following example uses the SQL format to select from the hr.employees table the
employees who had worked for the company for at least 100 days on November 1,
2002:

5-310 Oracle Database SQL Language Reference

TO_DSINTERVAL

SELECT employee_id, last_name FROM employees
WHERE hire_date + TO_DSINTERVAL('100 00:00:00')
<= DATE '2002-11-01'
ORDER BY employee_id;
EMPLOYEE_ID
----------102
203
204
205
206

LAST_NAME
--------------De Haan
Mavris
Baer
Higgins
Giet

The following example uses the ISO format to display the timestamp 100 days and 5
hours after the beginning of the year 2009:
SELECT TO_CHAR(TIMESTAMP '2009-01-01 00:00:00' + TO_DSINTERVAL('P100DT05H'),
'YYYY-MM-DD HH24:MI:SS') "Time Stamp"
FROM DUAL;
Time Stamp
------------------2009-04-11 05:00:00

Functions

5-311

TO_LOB

TO_LOB
Syntax
TO_LOB

(

long_column

)

Purpose
TO_LOB converts LONG or LONG RAW values in the column long_column to LOB values.
You can apply this function only to a LONG or LONG RAW column, and only in the select
list of a subquery in an INSERT statement.
Before using this function, you must create a LOB column to receive the converted
LONG values. To convert LONG values, create a CLOB column. To convert LONG RAW
values, create a BLOB column.
You cannot use the TO_LOB function to convert a LONG column to a LOB column in the
subquery of a CREATE TABLE ... AS SELECT statement if you are creating an
index-organized table. Instead, create the index-organized table without the LONG
column, and then use the TO_LOB function in an INSERT ... AS SELECT statement.
You cannot use this function within a PL/SQL package. Instead use the TO_CLOB or TO_
BLOB functions.
See Also:
■

■

the modify_col_properties clause of ALTER TABLE on page 12-2
for an alternative method of converting LONG columns to LOB
INSERT on page 18-54 for information on the subquery of an
INSERT statement

Examples
The following syntax shows how to use the TO_LOB function on your LONG data in a
hypothetical table old_table:
CREATE TABLE new_table (col1, col2, ... lob_col CLOB);
INSERT INTO new_table (select o.col1, o.col2, ... TO_LOB(o.old_long_col)
FROM old_table o;

5-312 Oracle Database SQL Language Reference

TO_MULTI_BYTE

TO_MULTI_BYTE
Syntax
TO_MULTI_BYTE

(

char

)

Purpose
TO_MULTI_BYTE returns char with all of its single-byte characters converted to their
corresponding multibyte characters. char can be of data type CHAR, VARCHAR2, NCHAR, or
NVARCHAR2. The value returned is in the same data type as char.
Any single-byte characters in char that have no multibyte equivalents appear in the
output string as single-byte characters. This function is useful only if your database
character set contains both single-byte and multibyte characters.
This function does not support CLOB data directly. However, CLOBs can be passed in as
arguments through implicit data conversion.
See Also: "Data Type Comparison Rules" on page 3-36 for more
information.

Examples
The following example illustrates converting from a single byte A to a multibyte A in
UTF8:
SELECT dump(TO_MULTI_BYTE( 'A')) FROM DUAL;
DUMP(TO_MULTI_BYTE('A'))
-----------------------Typ=1 Len=3: 239,188,161

Functions

5-313

TO_NCHAR (character)

TO_NCHAR (character)
Syntax
to_nchar_char::=
char
TO_NCHAR

(

clob

)

nclob

Purpose
TO_NCHAR (character) converts a character string, CHAR, VARCHAR2, CLOB, or NCLOB value
to the national character set. The value returned is always NVARCHAR2. This function is
equivalent to the TRANSLATE ... USING function with a USING clause in the national
character set.
"Data Conversion" on page 3-40 and TRANSLATE ...
USING on page 5-325

See Also:

Examples
The following example converts VARCHAR2 data from the oe.customers table to the
national character set:
SELECT TO_NCHAR(cust_last_name) FROM customers
WHERE customer_id=103;
TO_NCHAR(CUST_LAST_NAME)
-------------------------------------------------Taylor

5-314 Oracle Database SQL Language Reference

TO_NCHAR (datetime)

TO_NCHAR (datetime)
Syntax
to_nchar_date::=
,
datetime
TO_NCHAR

,

’

nlsparam

’

fmt

(

)
interval

Purpose
TO_NCHAR (datetime) converts a datetime or interval value of DATE, TIMESTAMP,
TIMESTAMP WITH TIME ZONE, TIMESTAMP WITH LOCAL TIME ZONE, INTERVAL MONTH TO YEAR,
or INTERVAL DAY TO SECOND data type from the database character set to the national
character set.
See Also: "Security Considerations for Data Conversion" on
page 3-44

Examples
The following example converts the order_date of all orders whose status is 9 to the
national character set:
SELECT TO_NCHAR(ORDER_DATE) AS order_date
FROM ORDERS
WHERE ORDER_STATUS > 9
ORDER BY order_date;
ORDER_DATE
-------------------------------------------------------------------------06-DEC-99 02.22.34.225609 PM
13-SEP-99 10.19.00.654279 AM
14-SEP-99 09.53.40.223345 AM
26-JUN-00 10.19.43.190089 PM
27-JUN-00 09.53.32.335522 PM

Functions

5-315

TO_NCHAR (number)

TO_NCHAR (number)
Syntax
to_nchar_number::=
,
,
TO_NCHAR

(

’

nlsparam

’

fmt

n

)

Purpose
TO_NCHAR (number) converts n to a string in the national character set. The value n can
be of type NUMBER, BINARY_FLOAT, or BINARY_DOUBLE. The function returns a value of
the same type as the argument. The optional fmt and 'nlsparam' corresponding to n
can be of DATE, TIMESTAMP, TIMESTAMP WITH TIME ZONE, TIMESTAMP WITH LOCAL TIME
ZONE, INTERVAL MONTH TO YEAR, or INTERVAL DAY TO SECOND data type.
See Also: "Security Considerations for Data Conversion" on
page 3-44

Examples
The following example converts the customer_id values from the sample table
oe.orders to the national character set:
SELECT TO_NCHAR(customer_id) "NCHAR_Customer_ID"
WHERE order_status > 9
ORDER BY "NCHAR_Customer_ID";
NCHAR_Customer_ID
---------------------------------------102
103
148
148
149

5-316 Oracle Database SQL Language Reference

FROM orders

TO_NCLOB

TO_NCLOB
Syntax
lob_column
TO_NCLOB

(

)
char

Purpose
TO_NCLOB converts CLOB values in a LOB column or other character strings to NCLOB
values. char can be any of the data types CHAR, VARCHAR2, NCHAR, NVARCHAR2, CLOB, or
NCLOB. Oracle Database implements this function by converting the character set of
char from the database character set to the national character set.

Examples
The following example inserts some character data into an NCLOB column of the
pm.print_media table by first converting the data with the TO_NCLOB function:
INSERT INTO print_media (product_id, ad_id, ad_fltextn)
VALUES (3502, 31001,
TO_NCLOB('Placeholder for new product description'));

Functions

5-317

TO_NUMBER

TO_NUMBER
Syntax
,
,
TO_NUMBER

(

’

nlsparam

’

fmt

expr

)

Purpose
TO_NUMBER converts expr to a value of NUMBER data type. The expr can be a number
value of CHAR, VARCHAR2, NCHAR, NVARCHAR2, BINARY_FLOAT, or BINARY_DOUBLE data
type.
■

■

If you specify an expr of CHAR, VARCHAR2, NCHAR, or NVARCHAR2 data type, then you
can optionally specify the format model fmt.
If you specify an expr of BINARY_FLOAT or BINARY_DOUBLE data type, then you
cannot specify a format model because a BINARY_FLOAT or BINARY_DOUBLE can be
interpreted only by its internal representation.

Refer to "Format Models" on page 3-56 for information on format models.
The 'nlsparam' argument in this function has the same purpose as it does in the TO_
CHAR function for number conversions. Refer to TO_CHAR (number) on page 5-305 for
more information.
This function does not support CLOB data directly. However, CLOBs can be passed in as
arguments through implicit data conversion.
See Also: "Data Type Comparison Rules" on page 3-36 for more
information.

Examples
The following examples convert character string data into a number:
UPDATE employees SET salary = salary +
TO_NUMBER('100.00', '9G999D99')
WHERE last_name = 'Perkins';
SELECT TO_NUMBER('-AusDollars100','L9G999D99',
' NLS_NUMERIC_CHARACTERS = '',.''
NLS_CURRENCY
= ''AusDollars''
') "Amount"
FROM DUAL;
Amount
----------100

5-318 Oracle Database SQL Language Reference

TO_SINGLE_BYTE

TO_SINGLE_BYTE
Syntax
TO_SINGLE_BYTE

(

char

)

Purpose
TO_SINGLE_BYTE returns char with all of its multibyte characters converted to their
corresponding single-byte characters. char can be of data type CHAR, VARCHAR2, NCHAR,
or NVARCHAR2. The value returned is in the same data type as char.
Any multibyte characters in char that have no single-byte equivalents appear in the
output as multibyte characters. This function is useful only if your database character
set contains both single-byte and multibyte characters.
This function does not support CLOB data directly. However, CLOBs can be passed in as
arguments through implicit data conversion.
See Also: "Data Type Comparison Rules" on page 3-36 for more
information.

Examples
The following example illustrates going from a multibyte A in UTF8 to a single byte
ASCII A:
SELECT TO_SINGLE_BYTE( CHR(15711393)) FROM DUAL;
T
A

Functions

5-319

TO_TIMESTAMP

TO_TIMESTAMP
Syntax
,
,
TO_TIMESTAMP

(

char

’

nlsparam

’

fmt
)

Purpose
TO_TIMESTAMP converts char of CHAR, VARCHAR2, NCHAR, or NVARCHAR2 data type to a
value of TIMESTAMP data type.
The optional fmt specifies the format of char. If you omit fmt, then char must be in the
default format of the TIMESTAMP data type, which is determined by the NLS_
TIMESTAMP_FORMAT initialization parameter. The optional 'nlsparam' argument has the
same purpose in this function as in the TO_CHAR function for date conversion.
This function does not support CLOB data directly. However, CLOBs can be passed in as
arguments through implicit data conversion.
See Also: "Data Type Comparison Rules" on page 3-36 for more
information.

Examples
The following example converts a character string to a timestamp. The character string
is not in the default TIMESTAMP format, so the format mask must be specified:
SELECT TO_TIMESTAMP ('10-Sep-02 14:10:10.123000', 'DD-Mon-RR HH24:MI:SS.FF')
FROM DUAL;
TO_TIMESTAMP('10-SEP-0214:10:10.123000','DD-MON-RRHH24:MI:SS.FF')
--------------------------------------------------------------------------10-SEP-02 02.10.10.123000000 PM

See Also: NLS_TIMESTAMP_FORMAT parameter for information on the
default TIMESTAMP format and "Datetime Format Models" on page 3-60
for information on specifying the format mask

5-320 Oracle Database SQL Language Reference

TO_TIMESTAMP_TZ

TO_TIMESTAMP_TZ
Syntax
,
,
TO_TIMESTAMP_TZ

(

’

nlsparam

’

fmt

char

)

Purpose
TO_TIMESTAMP_TZ converts char of CHAR, VARCHAR2, NCHAR, or NVARCHAR2 data type to a
value of TIMESTAMP WITH TIME ZONE data type.
This function does not convert character strings to TIMESTAMP
WITH LOCAL TIME ZONE. To do this, use a CAST function, as shown in
CAST on page 5-35.
Note:

The optional fmt specifies the format of char. If you omit fmt, then char must be in the
default format of the TIMESTAMP WITH TIME ZONE data type. The optional 'nlsparam'
has the same purpose in this function as in the TO_CHAR function for date conversion.
Examples
The following example converts a character string to a value of TIMESTAMP WITH TIME
ZONE:
SELECT TO_TIMESTAMP_TZ('1999-12-01 11:00:00 -8:00',
'YYYY-MM-DD HH:MI:SS TZH:TZM') FROM DUAL;
TO_TIMESTAMP_TZ('1999-12-0111:00:00-08:00','YYYY-MM-DDHH:MI:SSTZH:TZM')
-------------------------------------------------------------------01-DEC-99 11.00.00.000000000 AM -08:00

The following example casts a null column in a UNION operation as TIMESTAMP WITH
LOCAL TIME ZONE using the sample tables oe.order_items and oe.orders:
SELECT order_id, line_item_id,
CAST(NULL AS TIMESTAMP WITH LOCAL TIME ZONE) order_date
FROM order_items
UNION
SELECT order_id, to_number(null), order_date
FROM orders;
ORDER_ID LINE_ITEM_ID ORDER_DATE
---------- ------------ ----------------------------------2354
1
2354
2
2354
3
2354
4
2354
5
2354
6
2354
7
2354
8
2354
9
2354
10
2354
11
2354
12

Functions

5-321

TO_TIMESTAMP_TZ

2354
2354
2355
2355

13
14-JUL-00 05.18.23.234567 PM
1
2

. . .

5-322 Oracle Database SQL Language Reference

TO_YMINTERVAL

TO_YMINTERVAL
Syntax
+
–
years
TO_YMINTERVAL

(

–

months

’

’

)

ym_iso_format

ym_iso_format::=
–

years

Y

months

M

days

D

P

.
hours

H

minutes

M

frac_secs

seconds

S

T

Purpose
TO_YMINTERVAL converts a character string of CHAR, VARCHAR2, NCHAR, or NVARCHAR2 data
type to an INTERVAL YEAR TO MONTH type.
TO_YMINTERVAL accepts argument in one of the two formats:
■

SQL interval format compatible with the SQL standard (ISO/IEC 9075:2003)

■

ISO duration format compatible with the ISO 8601:2004 standard

In the SQL format, years is an integer between 0 and 999999999, and months is an
integer between 0 and 11. Additional blanks are allowed between format elements.
In the ISO format, years and months are integers between 0 and 999999999. Days,
hours, minutes, seconds, and frac_secs are non-negative integers, and are ignored, if
specified. No blanks are allowed in the value. If you specify T, then you must specify
at least one of the hours, minutes, or seconds values.

Examples
The following example calculates for each employee in the sample hr.employees table
a date one year two months after the hire date:
SELECT hire_date, hire_date + TO_YMINTERVAL('01-02') "14 months"
FROM employees;
HIRE_DATE
--------17-JUN-03
21-SEP-05
13-JAN-01
20-MAY-08
21-MAY-07

14 months
--------17-AUG-04
21-NOV-06
13-MAR-02
20-JUL-09
21-JUL-08

. . .

The following example makes the same calculation using the ISO format:
SELECT hire_date, hire_date + TO_YMINTERVAL('P1Y2M') FROM employees;

Functions

5-323

TRANSLATE

TRANSLATE
Syntax
TRANSLATE

(

expr

,

from_string

,

to_string

)

Purpose
TRANSLATE returns expr with all occurrences of each character in from_string replaced
by its corresponding character in to_string. Characters in expr that are not in from_
string are not replaced. The argument from_string can contain more characters than
to_string. In this case, the extra characters at the end of from_string have no
corresponding characters in to_string. If these extra characters appear in expr, then
they are removed from the return value.
If a character appears multiple times in from_string, then the to_string mapping
corresponding to the first occurrence is used.
You cannot use an empty string for to_string to remove all characters in from_string
from the return value. Oracle Database interprets the empty string as null, and if this
function has a null argument, then it returns null. To remove all characters in from_
string, concatenate another character to the beginning of from_string and specify
this character as the to_string. For example, TRANSLATE(expr, 'x0123456789', 'x')
removes all digits from expr.
TRANSLATE provides functionality related to that provided by the REPLACE function.
REPLACE lets you substitute a single string for another single string, as well as remove
character strings. TRANSLATE lets you make several single-character, one-to-one
substitutions in one operation.
This function does not support CLOB data directly. However, CLOBs can be passed in as
arguments through implicit data conversion.
See Also: "Data Type Comparison Rules" on page 3-36 for more
information and REPLACE on page 5-232

Examples
The following statement translates a book title into a string that could be used (for
example) as a filename. The from_string contains four characters: a space, asterisk,
slash, and apostrophe (with an extra apostrophe as the escape character). The to_
string contains only three underscores. This leaves the fourth character in the from_
string without a corresponding replacement, so apostrophes are dropped from the
returned value.
SELECT TRANSLATE('SQL*Plus User''s Guide', ' */''', '___') FROM DUAL;
TRANSLATE('SQL*PLUSU
-------------------SQL_Plus_Users_Guide

5-324 Oracle Database SQL Language Reference

TRANSLATE ... USING

TRANSLATE ... USING
Syntax
CHAR_CS
TRANSLATE

(

char

USING

)
NCHAR_CS

Purpose
TRANSLATE ... USING converts char into the character set specified for conversions
between the database character set and the national character set.
Note: The TRANSLATE ... USING function is supported primarily for
ANSI compatibility. Oracle recommends that you use the TO_CHAR and
TO_NCHAR functions, as appropriate, for converting data to the
database or national character set. TO_CHAR and TO_NCHAR can take as
arguments a greater variety of data types than TRANSLATE ... USING,
which accepts only character data.

The char argument is the expression to be converted.
■

■

Specifying the USING CHAR_CS argument converts char into the database character
set. The output data type is VARCHAR2.
Specifying the USING NCHAR_CS argument converts char into the national character
set. The output data type is NVARCHAR2.

This function is similar to the Oracle CONVERT function, but must be used instead of
CONVERT if either the input or the output data type is being used as NCHAR or
NVARCHAR2. If the input contains UCS2 code points or backslash characters (\), then use
the UNISTR function.
See Also:

CONVERT on page 5-53 and UNISTR on page 5-333

Examples
The following statements use data from the sample table oe.product_descriptions to
show the use of the TRANSLATE ... USING function:
CREATE TABLE translate_tab (char_col VARCHAR2(100),
nchar_col NVARCHAR2(50));
INSERT INTO translate_tab
SELECT NULL, translated_name
FROM product_descriptions
WHERE product_id = 3501;
SELECT * FROM translate_tab;
CHAR_COL
NCHAR_COL
-------------------- -------------------------------------------------. . .
C pre SPNIX4.0 - Sys
C pro SPNIX4.0 - Sys
C til SPNIX4.0 - Sys
C voor SPNIX4.0 - Sys
. . .

Functions

5-325

TRANSLATE ... USING

UPDATE translate_tab
SET char_col = TRANSLATE (nchar_col USING CHAR_CS);
SELECT * FROM translate_tab;
CHAR_COL
------------------------. . .
C per a SPNIX4.0 - Sys
C pro SPNIX4.0 - Sys
C for SPNIX4.0 - Sys
C til SPNIX4.0 - Sys
. . .

NCHAR_COL
------------------------C
C
C
C

5-326 Oracle Database SQL Language Reference

per
pro
for
til

a SPNIX4.0
SPNIX4.0 SPNIX4.0 SPNIX4.0 -

- Sys
Sys
Sys
Sys

TREAT

TREAT
Syntax
REF
TREAT

(

expr

schema

AS

.
type

)

Purpose
TREAT changes the declared type of an expression.
You must have the EXECUTE object privilege on type to use this function.
■

■
■

type must be some supertype or subtype of the declared type of expr. If the most
specific type of expr is type (or some subtype of type), then TREAT returns expr. If
the most specific type of expr is not type (or some subtype of type), then TREAT
returns NULL.
You can specify REF only if the declared type of expr is a REF type.
If the declared type of expr is a REF to a source type of expr, then type must be
some subtype or supertype of the source type of expr. If the most specific type of
DEREF(expr) is type (or a subtype of type), then TREAT returns expr. If the most
specific type of DEREF(expr) is not type (or a subtype of type), then TREAT returns
NULL.

This function does not support CLOB data directly. However, CLOBs can be passed in as
arguments through implicit data conversion.
See Also: "Data Type Comparison Rules" on page 3-36 for more
information

Examples
The following statement uses the table oe.persons, which is created in "Substitutable
Table and Column Examples" on page 16-71. The example retrieves the salary attribute
of all people in the persons table, the value being null for instances of people that are
not employees.
SELECT name, TREAT(VALUE(p) AS employee_t).salary salary
FROM persons p;
NAME
SALARY
------------------------- ---------Bob
Joe
100000
Tim
1000

You can use the TREAT function to create an index on the subtype attributes of a
substitutable column. For an example, see "Indexing on Substitutable Columns:
Examples" on page 14-85.

Functions

5-327

TRIM

TRIM
Syntax
LEADING

trim_character

TRAILING
BOTH

FROM

trim_character
TRIM

(

trim_source

)

Purpose
TRIM enables you to trim leading or trailing characters (or both) from a character
string. If trim_character or trim_source is a character literal, then you must enclose
it in single quotation marks.
■

■

■

■
■

■

■

If you specify LEADING, then Oracle Database removes any leading characters
equal to trim_character.
If you specify TRAILING, then Oracle removes any trailing characters equal to
trim_character.
If you specify BOTH or none of the three, then Oracle removes leading and trailing
characters equal to trim_character.
If you do not specify trim_character, then the default value is a blank space.
If you specify only trim_source, then Oracle removes leading and trailing blank
spaces.
The function returns a value with data type VARCHAR2. The maximum length of the
value is the length of trim_source.
If either trim_source or trim_character is null, then the TRIM function returns
null.

Both trim_character and trim_source can be VARCHAR2 or any data type that can be
implicitly converted to VARCHAR2. The string returned is a VARCHAR2 (NVARCHAR2) data
type if trim_source is a CHAR or VARCHAR2 (NCHAR or NVARCHAR2) data type, and a CLOB
if trim_source is a CLOB data type. The return string is in the same character set as
trim_source.

Examples
This example trims leading zeros from the hire date of the employees in the hr schema:
SELECT employee_id,
TO_CHAR(TRIM(LEADING 0 FROM hire_date))
FROM employees
WHERE department_id = 60
ORDER BY employee_id;
EMPLOYEE_ID
----------103
104
105
106
107

TO_CHAR(T
--------20-MAY-08
21-MAY-07
25-JUN-05
5-FEB-06
7-FEB-07

5-328 Oracle Database SQL Language Reference

TRUNC (date)

TRUNC (date)
Syntax
trunc_date::=
,
TRUNC

(

date

fmt
)

Purpose
The TRUNC (date) function returns date with the time portion of the day truncated to
the unit specified by the format model fmt. This function is not sensitive to the NLS_
CALENDAR session parameter. It operates according to the rules of the Gregorian
calendar. The value returned is always of data type DATE, even if you specify a
different datetime data type for date. If you omit fmt, then the default format model
'DD' is used and the value returned is date truncated to the day with a time of
midnight. Refer to "ROUND and TRUNC Date Functions" on page 5-379 for the
permitted format models to use in fmt.

Examples
The following example truncates a date:
SELECT TRUNC(TO_DATE('27-OCT-92','DD-MON-YY'), 'YEAR')
"New Year" FROM DUAL;
New Year
--------01-JAN-92

Functions

5-329

TRUNC (number)

TRUNC (number)
Syntax
trunc_number::=
,
TRUNC

(

n2

n1

)

Purpose
The TRUNC (number) function returns n1 truncated to n2 decimal places. If n2 is
omitted, then n1 is truncated to 0 places. n2 can be negative to truncate (make zero) n2
digits left of the decimal point.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. If you omit n2, then the
function returns the same data type as the numeric data type of the argument. If you
include n2, then the function returns NUMBER.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

Examples
The following examples truncate numbers:
SELECT TRUNC(15.79,1) "Truncate" FROM DUAL;
Truncate
---------15.7
SELECT TRUNC(15.79,-1) "Truncate" FROM DUAL;
Truncate
---------10

5-330 Oracle Database SQL Language Reference

TZ_OFFSET

TZ_OFFSET
Syntax
’

time_zone_name

’

+
’
TZ_OFFSET

(

hh

:

mi

–

’
)

SESSIONTIMEZONE
DBTIMEZONE

Purpose
TZ_OFFSET returns the time zone offset corresponding to the argument based on the
date the statement is executed. You can enter a valid time zone region name, a time
zone offset from UTC (which simply returns itself), or the keyword SESSIONTIMEZONE
or DBTIMEZONE. For a listing of valid values for time_zone_name, query the TZNAME
column of the V$TIMEZONE_NAMES dynamic performance view.

Time zone region names are needed by the daylight saving
feature. These names are stored in two types of time zone files: one
large and one small. One of these files is the default file, depending
on your environment and the release of Oracle Database you are
using. For more information regarding time zone files and names,
see Oracle Database Globalization Support Guide.

Note:

Examples
The following example returns the time zone offset of the US/Eastern time zone from
UTC:
SELECT TZ_OFFSET('US/Eastern') FROM DUAL;
TZ_OFFS
-------04:00

Functions

5-331

UID

UID
Syntax
UID

Purpose
UID returns an integer that uniquely identifies the session user (the user who logged
on).

Examples
The following example returns the UID of the current user:
SELECT UID FROM DUAL;

5-332 Oracle Database SQL Language Reference

UNISTR

UNISTR
Syntax
UNISTR

(

string

)

Purpose
UNISTR takes as its argument a text literal or an expression that resolves to character
data and returns it in the national character set. The national character set of the
database can be either AL16UTF16 or UTF8. UNISTR provides support for Unicode
string literals by letting you specify the Unicode encoding value of characters in the
string. This is useful, for example, for inserting data into NCHAR columns.
The Unicode encoding value has the form '\xxxx' where 'xxxx' is the hexadecimal
value of a character in UCS-2 encoding format. Supplementary characters are encoded
as two code units, the first from the high-surrogates range (U+D800 to U+DBFF), and
the second from the low-surrogates range (U+DC00 to U+DFFF). To include the
backslash in the string itself, precede it with another backslash (\\).
For portability and data preservation, Oracle recommends that in the UNISTR string
argument you specify only ASCII characters and the Unicode encoding values.
See Also: Oracle Database Globalization Support Guide for information
on Unicode and national character sets

Examples
The following example passes both ASCII characters and Unicode encoding values to
the UNISTR function, which returns the string in the national character set:
SELECT UNISTR('abc\00e5\00f1\00f6') FROM DUAL;
UNISTR
-----abcåñö

Functions

5-333

UPDATEXML

UPDATEXML
Syntax
,
UPDATEXML

(

XMLType_instance

,

XPath_string

,
,

value_expr

namespace_string
)

Purpose
UPDATEXML takes as arguments an XMLType instance and an XPath-value pair and
returns an XMLType instance with the updated value. If XPath_string is an XML
element, then the corresponding value_expr must be an XMLType instance. If XPath_
string is an attribute or text node, then the value_expr can be any scalar data type.
You can specify an absolute XPath_string with an initial slash or a relative XPath_
string by omitting the initial slash. If you omit the initial slash, then the context of the
relative path defaults to the root node.
The data types of the target of each XPath_string and its corresponding value_expr
must match. The optional namespace_string must resolve to a VARCHAR2 value that
specifies a default mapping or namespace mapping for prefixes, which Oracle
Database uses when evaluating the XPath expression(s).
If you update an XML element to null, then Oracle removes the attributes and children
of the element, and the element becomes empty. If you update the text node of an
element to null, Oracle removes the text value of the element, and the element itself
remains but is empty.
In most cases, this function materializes an XML document in memory and updates
the value. However, UPDATEXML is optimized for UPDATE statements on object-relational
columns so that the function updates the value directly in the column. This
optimization requires the following conditions:
■

■

The XMLType_instance must be the same as the column in the UPDATE ... SET
clause.
The XPath_string must resolve to scalar content.

Examples
The following example updates to 4 the number of docks in the San Francisco
warehouse in the sample schema OE, which has a warehouse_spec column of type
XMLType:
SELECT warehouse_name,
EXTRACT(warehouse_spec, '/Warehouse/Docks')
"Number of Docks"
FROM warehouses
WHERE warehouse_name = 'San Francisco';
WAREHOUSE_NAME
Number of Docks
-------------------- -------------------San Francisco
1
UPDATE warehouses SET warehouse_spec =
UPDATEXML(warehouse_spec,
'/Warehouse/Docks/text()',4)
WHERE warehouse_name = 'San Francisco';

5-334 Oracle Database SQL Language Reference

UPDATEXML

1 row updated.
SELECT warehouse_name,
EXTRACT(warehouse_spec, '/Warehouse/Docks')
"Number of Docks"
FROM warehouses
WHERE warehouse_name = 'San Francisco';
WAREHOUSE_NAME
Number of Docks
-------------------- -------------------San Francisco
4

Functions

5-335

UPPER

UPPER
Syntax
UPPER

(

char

)

Purpose
UPPER returns char, with all letters uppercase. char can be any of the data types CHAR,
VARCHAR2, NCHAR, NVARCHAR2, CLOB, or NCLOB. The return value is the same data type as
char. The database sets the case of the characters based on the binary mapping defined
for the underlying character set. For linguistic-sensitive uppercase, refer to NLS_
UPPER on page 5-163.

Examples
The following example returns each employee's last name in uppercase:
SELECT UPPER(last_name) "Uppercase"
FROM employees;

5-336 Oracle Database SQL Language Reference

USER

USER
Syntax
USER

Purpose
USER returns the name of the session user (the user who logged on) with the data type
VARCHAR2. Oracle Database compares values of this function with blank-padded
comparison semantics.
In a distributed SQL statement, the UID and USER functions together identify the user
on your local database. You cannot use these functions in the condition of a CHECK
constraint.

Examples
The following example returns the current user and the user's UID:
SELECT USER, UID FROM DUAL;

Functions

5-337

USERENV

USERENV
Syntax
USERENV

(

’

parameter

’

)

Purpose
USERENV is a legacy function that is retained for backward
compatibility. Oracle recommends that you use the SYS_CONTEXT
function with the built-in USERENV namespace for current functionality.
See SYS_CONTEXT on page 5-279 for more information.
Note:

USERENV returns information about the current session. This information can be useful
for writing an application-specific audit trail table or for determining the
language-specific characters currently used by your session. You cannot use USERENV
in the condition of a CHECK constraint. Table 5–13 describes the values for the
parameter argument.
All calls to USERENV return VARCHAR2 data except for calls with the SESSIONID, SID, and
ENTRYID parameters, which return NUMBER.
Table 5–13

Parameters of the USERENV Function

Parameter

Return Value

CLIENT_INFO

CLIENT_INFO returns up to 64 bytes of user session information that
can be stored by an application using the DBMS_APPLICATION_INFO
package.
Caution: Some commercial applications may be using this context
value. Refer to the applicable documentation for those applications to
determine what restrictions they may impose on use of this context
area.
See Also:
■

■

Oracle Database Security Guide for more information on application
context
CREATE CONTEXT on page 14-9 and SYS_CONTEXT on
page 5-279

ENTRYID

The current audit entry number. The audit entryid sequence is shared
between fine-grained audit records and regular audit records. You
cannot use this attribute in distributed SQL statements.

ISDBA

ISDBA returns 'TRUE' if the user has been authenticated as having DBA
privileges either through the operating system or through a password
file.

LANG

LANG returns the ISO abbreviation for the language name, a shorter
form than the existing 'LANGUAGE' parameter.

LANGUAGE

LANGUAGE returns the language and territory used by the current
session along with the database character set in this form:
language_territory.characterset

SESSIONID

SESSIONID returns the auditing session identifier. You cannot specify
this parameter in distributed SQL statements.

5-338 Oracle Database SQL Language Reference

USERENV

Table 5–13 (Cont.) Parameters of the USERENV Function
Parameter

Return Value

SID

SID returns the session ID.

TERMINAL

TERMINAL returns the operating system identifier for the terminal of the
current session. In distributed SQL statements, this parameter returns
the identifier for your local session. In a distributed environment, this
parameter is supported only for remote SELECT statements, not for
remote INSERT, UPDATE, or DELETE operations.

Examples
The following example returns the LANGUAGE parameter of the current session:
SELECT USERENV('LANGUAGE') "Language" FROM DUAL;
Language
----------------------------------AMERICAN_AMERICA.WE8ISO8859P1

Functions

5-339

VALUE

VALUE
Syntax
VALUE

(

correlation_variable

)

Purpose
VALUE takes as its argument a correlation variable (table alias) associated with a row of
an object table and returns object instances stored in the object table. The type of the
object instances is the same type as the object table.

Examples
The following example uses the sample table oe.persons, which is created in
"Substitutable Table and Column Examples" on page 16-71:
SELECT VALUE(p) FROM persons p;
VALUE(P)(NAME, SSN)
------------------------------------------------------------PERSON_T('Bob', 1234)
EMPLOYEE_T('Joe', 32456, 12, 100000)
PART_TIME_EMP_T('Tim', 5678, 13, 1000, 20)

See Also: "IS OF type Condition" on page 7-25 for information on
using IS OF type conditions with the VALUE function

5-340 Oracle Database SQL Language Reference

VAR_POP

VAR_POP
Syntax
OVER
VAR_POP

(

expr

(

analytic_clause

)

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions

Purpose
VAR_POP returns the population variance of a set of numbers after discarding the nulls
in this set. You can use it as both an aggregate and analytic function.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. The function returns the
same data type as the numeric data type of the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

If the function is applied to an empty set, then it returns null. The function makes the
following calculation:
SUM((expr - (SUM(expr) / COUNT(expr)))2) / COUNT(expr)

See Also: "About SQL Expressions" on page 6-1 for information on
valid forms of expr and "Aggregate Functions" on page 5-10

Aggregate Example
The following example returns the population variance of the salaries in the employees
table:
SELECT VAR_POP(salary) FROM employees;
VAR_POP(SALARY)
--------------15141964.9

Analytic Example
The following example calculates the cumulative population and sample variances in
the sh.sales table of the monthly sales in 1998:
SELECT t.calendar_month_desc,
VAR_POP(SUM(s.amount_sold))
OVER (ORDER BY t.calendar_month_desc) "Var_Pop",
VAR_SAMP(SUM(s.amount_sold))
OVER (ORDER BY t.calendar_month_desc) "Var_Samp"
FROM sales s, times t
WHERE s.time_id = t.time_id AND t.calendar_year = 1998
GROUP BY t.calendar_month_desc
ORDER BY t.calendar_month_desc, "Var_Pop", "Var_Samp";
CALENDAR
Var_Pop
Var_Samp
-------- ---------- ----------

Functions

5-341

VAR_POP

1998-01
1998-02
1998-03
1998-04
1998-05
1998-06
1998-07
1998-08
1998-09
1998-10
1998-11
1998-12

0
2269111326
5.5849E+10
4.8252E+10
6.0020E+10
5.4091E+10
4.7150E+10
4.1345E+10
3.9591E+10
3.9995E+10
3.6870E+10
4.0216E+10

4538222653
8.3774E+10
6.4336E+10
7.5025E+10
6.4909E+10
5.5009E+10
4.7252E+10
4.4540E+10
4.4439E+10
4.0558E+10
4.3872E+10

5-342 Oracle Database SQL Language Reference

VAR_SAMP

VAR_SAMP
Syntax
OVER
VAR_SAMP

(

expr

(

analytic_clause

)

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions

Purpose
VAR_SAMP returns the sample variance of a set of numbers after discarding the nulls in
this set. You can use it as both an aggregate and analytic function.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. The function returns the
same data type as the numeric data type of the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

If the function is applied to an empty set, then it returns null. The function makes the
following calculation:
(SUM(expr - (SUM(expr) / COUNT(expr)))2) / (COUNT(expr) - 1)

This function is similar to VARIANCE, except that given an input set of one element,
VARIANCE returns 0 and VAR_SAMP returns null.
See Also: "About SQL Expressions" on page 6-1 for information on
valid forms of expr and "Aggregate Functions" on page 5-10

Aggregate Example
The following example returns the sample variance of the salaries in the sample
employees table.
SELECT VAR_SAMP(salary) FROM employees;
VAR_SAMP(SALARY)
---------------15284813.7

Analytic Example
Refer to the analytic example for VAR_POP on page 5-341.

Functions

5-343

VARIANCE

VARIANCE
Syntax
DISTINCT
ALL
VARIANCE

(

OVER
expr

(

analytic_clause

)

)

See Also: "Analytic Functions" on page 5-11 for information on
syntax, semantics, and restrictions

Purpose
VARIANCE returns the variance of expr. You can use it as an aggregate or analytic
function.
Oracle Database calculates the variance of expr as follows:
■

0 if the number of rows in expr = 1

■

VAR_SAMP if the number of rows in expr > 1

If you specify DISTINCT, then you can specify only the query_partition_clause of the
analytic_clause. The order_by_clause and windowing_clause are not allowed.
This function takes as an argument any numeric data type or any nonnumeric data
type that can be implicitly converted to a numeric data type. The function returns the
same data type as the numeric data type of the argument.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion, "About SQL
Expressions" on page 6-1 for information on valid forms of expr and
"Aggregate Functions" on page 5-10

Aggregate Example
The following example calculates the variance of all salaries in the sample employees
table:
SELECT VARIANCE(salary) "Variance"
FROM employees;
Variance
---------15283140.5

Analytic Example
The following example returns the cumulative variance of salary values in Department
30 ordered by hire date.
SELECT last_name, salary, VARIANCE(salary)
OVER (ORDER BY hire_date) "Variance"
FROM employees
WHERE department_id = 30
ORDER BY last_name, salary, "Variance";
LAST_NAME
SALARY
Variance
------------------------- ---------- ---------Baida
2900 16283333.3
5-344 Oracle Database SQL Language Reference

VARIANCE

Colmenares
Himuro
Khoo
Raphaely
Tobias

2500
11307000
2600
13317000
3100
31205000
11000
0
2800 21623333.3

Functions

5-345

VSIZE

VSIZE
Syntax
VSIZE

(

expr

)

Purpose
VSIZE returns the number of bytes in the internal representation of expr. If expr is null,
then this function returns null.
This function does not support CLOB data directly. However, CLOBs can be passed in as
arguments through implicit data conversion.
See Also: "Data Type Comparison Rules" on page 3-36 for more
information

Examples
The following example returns the number of bytes in the last_name column of the
employees in department 10:
SELECT last_name, VSIZE (last_name) "BYTES"
FROM employees
WHERE department_id = 10
ORDER BY employee_id;
LAST_NAME
BYTES
--------------- ---------Whalen
6

5-346 Oracle Database SQL Language Reference

WIDTH_BUCKET

WIDTH_BUCKET
Syntax
WIDTH_BUCKET

(

expr

,

min_value

,

max_value

,

num_buckets

)

Purpose
WIDTH_BUCKET lets you construct equiwidth histograms, in which the histogram range
is divided into intervals that have identical size. (Compare this function with NTILE,
which creates equiheight histograms.) Ideally each bucket is a closed-open interval of
the real number line. For example, a bucket can be assigned to scores between 10.00
and 19.999 ... to indicate that 10 is included in the interval and 20 is excluded. This is
sometimes denoted [10, 20).
For a given expression, WIDTH_BUCKET returns the bucket number into which the value
of this expression would fall after being evaluated.
■

■

■

expr is the expression for which the histogram is being created. This expression
must evaluate to a numeric or datetime value or to a value that can be implicitly
converted to a numeric or datetime value. If expr evaluates to null, then the
expression returns null.
min_value and max_value are expressions that resolve to the end points of the
acceptable range for expr. Both of these expressions must also evaluate to numeric
or datetime values, and neither can evaluate to null.
num_buckets is an expression that resolves to a constant indicating the number of
buckets. This expression must evaluate to a positive integer.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion

When needed, Oracle Database creates an underflow bucket numbered 0 and an
overflow bucket numbered num_buckets+1. These buckets handle values less than
min_value and more than max_value and are helpful in checking the reasonableness of
endpoints.

Examples
The following example creates a ten-bucket histogram on the credit_limit column
for customers in Switzerland in the sample table oe.customers and returns the bucket
number ("Credit Group") for each customer. Customers with credit limits greater than
or equal to the maximum value are assigned to the overflow bucket, 11:
SELECT customer_id, cust_last_name, credit_limit,
WIDTH_BUCKET(credit_limit, 100, 5000, 10) "Credit Group"
FROM customers WHERE nls_territory = 'SWITZERLAND'
ORDER BY "Credit Group", customer_id, cust_last_name, credit_limit;
CUSTOMER_ID
----------825
826
827
853
843
844

CUST_LAST_NAME
CREDIT_LIMIT Credit Group
-------------------- ------------ -----------Dreyfuss
500
1
Barkin
500
1
Siegel
500
1
Palin
400
1
Oates
700
2
Julius
700
2

Functions

5-347

WIDTH_BUCKET

835
836
837
840
841
842
848
849
828
829
850
851
852
830
831
832
833
834
838
839
845
846
847

Eastwood
Berenger
Stanton
Elliott
Boyer
Stern
Olmos
Kaurusmdki
Minnelli
Hunter
Finney
Brown
Tanner
Dutt
Bel Geddes
Spacek
Moranis
Idle
Nicholson
Johnson
Fawcett
Brando
Streep

5-348 Oracle Database SQL Language Reference

1200
1200
1200
1400
1400
1400
1800
1800
2300
2300
2300
2300
2300
3500
3500
3500
3500
3500
3500
3500
5000
5000
5000

3
3
3
3
3
3
4
4
5
5
5
5
5
7
7
7
7
7
7
7
11
11
11

XMLAGG

XMLAGG
Syntax
order_by_clause
XMLAGG

(

XMLType_instance

)

Purpose
XMLAgg is an aggregate function. It takes a collection of XML fragments and returns an
aggregated XML document. Any arguments that return null are dropped from the
result.
XMLAgg is similar to SYS_XMLAgg except that XMLAgg returns a collection of nodes but it
does not accept formatting using the XMLFormat object. Also, XMLAgg does not enclose
the output in an element tag as does SYS_XMLAgg.
Within the order_by_clause, Oracle Database does not interpret number literals as
column positions, as it does in other uses of this clause, but simply as number literals.
See Also: XMLELEMENT on page 5-358 and SYS_XMLAGG on
page 5-290

Examples
The following example produces a Department element containing Employee elements
with employee job ID and last name as the contents of the elements:
SELECT XMLELEMENT("Department",
XMLAGG(XMLELEMENT("Employee",
e.job_id||' '||e.last_name)
ORDER BY last_name))
as "Dept_list"
FROM employees e
WHERE e.department_id = 30;
Dept_list
------------------------------------------------------------
PU_CLERK Baida
PU_CLERK Colmenares
PU_CLERK Himuro
PU_CLERK Khoo
PU_MAN Raphaely
PU_CLERK Tobias


The result is a single row, because XMLAgg aggregates the rows. You can use the GROUP
BY clause to group the returned set of rows into multiple groups:
SELECT XMLELEMENT("Department",
XMLAGG(XMLELEMENT("Employee", e.job_id||' '||e.last_name)))
AS "Dept_list"
FROM employees e
GROUP BY e.department_id;
Dept_list
--------------------------------------------------------

Functions

5-349

XMLAGG

AD_ASST Whalen


MK_MAN Hartstein
MK_REP Fay


PU_MAN Raphaely
PU_CLERK Khoo
PU_CLERK Tobias
PU_CLERK Baida
PU_CLERK Colmenares
PU_CLERK Himuro

. . .

5-350 Oracle Database SQL Language Reference

XMLCAST

XMLCAST
Syntax
XMLCAST

(

value_expression

AS

datatype

)

Purpose
XMLCast casts value_expression to the scalar SQL data type specified by datatype.
The value_expression argument is a SQL expression that is evaluated. The datatype
argument can be of data type NUMBER, VARCHAR2, CHAR, CLOB, BLOB, REF XMLTYPE, and
any of the datetime data types.
See Also: Oracle XML DB Developer's Guide for more information on
uses for this function and examples

Functions

5-351

XMLCDATA

XMLCDATA
Syntax
XMLCDATA

(

value_expr

)

Purpose
XMLCData generates a CDATA section by evaluating value_expr. The value_expr must
resolve to a string. The value returned by the function takes the following form:


If the resulting value is not a valid XML CDATA section, then the function returns an
error.
The following conditions apply to XMLCData:
■

The value_expr cannot contain the substring ]]>.

■

If value_expr evaluates to null, then the function returns null.
See Also: Oracle XML DB Developer's Guide for more information on
this function

Examples
The following statement uses the DUAL table to illustrate the syntax of XMLCData:
SELECT XMLELEMENT("PurchaseOrder",
XMLAttributes(dummy as "pono"),
XMLCdata('




]>')) "XMLCData" FROM DUAL;
XMLCData
---------------------------------------------------------




]>
]]>


5-352 Oracle Database SQL Language Reference

XMLCOLATTVAL

XMLCOLATTVAL
Syntax
,
c_alias
AS
EVALNAME
XMLCOLATTVAL

(

value_expr

value_expr

)

Purpose
XMLColAttVal creates an XML fragment and then expands the resulting XML so that
each XML fragment has the name column with the attribute name.
You can use the AS clause to change the value of the name attribute to something other
than the column name. You can do this by specifying c_alias, which is a string literal,
or by specifying EVALNAME value_expr. In the latter case, the value expression is
evaluated and the result, which must be a string literal, is used as the alias. The alias
can be up to 4000 characters.
You must specify a value for value_expr. If value_expr is null, then no element is
returned.
Restriction on XMLColAttVal You cannot specify an object type column for value_

expr.

Examples
The following example creates an Emp element for a subset of employees, with nested
employee_id, last_name, and salary elements as the contents of Emp. Each nested
element is named column and has a name attribute with the column name as the
attribute value:
SELECT XMLELEMENT("Emp",
XMLCOLATTVAL(e.employee_id, e.last_name, e.salary)) "Emp Element"
FROM employees e
WHERE employee_id = 204;
Emp Element
-------------------------------------------------------------------
204
Baer
10000


Refer to the example for XMLFOREST on page 5-362 to compare the output of these
two functions.

Functions

5-353

XMLCOMMENT

XMLCOMMENT
Syntax
XMLCOMMENT

(

value_expr

)

Purpose
XMLComment generates an XML comment using an evaluated result of value_expr. The
value_expr must resolve to a string. It cannot contain two consecutive dashes
(hyphens). The value returned by the function takes the following form:


If value_expr resolves to null, then the function returns null.
See Also: Oracle XML DB Developer's Guide for more information on
this function

Examples
The following example uses the DUAL table to illustrate the XMLComment syntax:
SELECT XMLCOMMENT('OrderAnalysisComp imported, reconfigured, disassembled')
AS "XMLCOMMENT" FROM DUAL;
XMLCOMMENT
-------------------------------------------------------------------------------

5-354 Oracle Database SQL Language Reference

XMLCONCAT

XMLCONCAT
Syntax
,
XMLCONCAT

(

XMLType_instance

)

Purpose
XMLConcat takes as input a series of XMLType instances, concatenates the series of
elements for each row, and returns the concatenated series. XMLConcat is the inverse of
XMLSequence.
Null expressions are dropped from the result. If all the value expressions are null, then
the function returns null.
See Also:

XMLSEQUENCE on page 5-371

Examples
The following example creates XML elements for the first and last names of a subset of
employees, and then concatenates and returns those elements:
SELECT XMLCONCAT(XMLELEMENT("First", e.first_name),
XMLELEMENT("Last", e.last_name)) AS "Result"
FROM employees e
WHERE e.employee_id > 202;
Result
---------------------------------------------------------------Susan
Mavris
Hermann
Baer
Shelley
Higgins
William
Gietz
4 rows selected.

Functions

5-355

XMLDIFF

XMLDIFF
Syntax
XMLDiff

,
(

XMLType_document

,

XMLType_document

integer

,

string
)

Purpose
The XMLDiff function is the SQL interface for the XmlDiff C API. This function
compares two XML documents and captures the differences in XML conforming to an
Xdiff schema. The diff document is returned as an XMLType document.
■
■

■

For the first two arguments, specify the names of two XMLType documents.
For the integer, specify a number representing the hashLevel for a C function
XmlDiff. If you do not want hashing, set this argument to 0 or omit it entirely. If
you do not want hashing, but you want to specify flags, then you must set this
argument to 0.
For string, specify the flags that control the behavior of the function. These flags
are specified by one or more names separated by semicolon. The names are the
same as the names of constants for XmlDiff function.
Oracle XML Developer's Kit Programmer's Guide for more
information on using this function, including examples, and Oracle
Database XML C API Reference for information on the XML APIs for C

See Also:

Examples
The following example compares two XML documents and returns the difference as
an XMLType document:
SELECT XMLDIFF(
XMLTYPE('




Chapter 1.




Chapter 2.



'),
XMLTYPE('




Chapter 1.


5-356 Oracle Database SQL Language Reference

XMLDIFF




')
)
FROM DUAL;

Functions

5-357

XMLELEMENT

XMLELEMENT
Syntax
ENTITYESCAPING
NAME
NOENTITYESCAPING
XMLELEMENT

identifier

(
EVALNAME

value_expr

AS
c_alias
,

XML_attributes_clause

,

value_expr
)

XML_attributes_clause::=
XMLATTRIBUTES

ENTITYESCAPING

SCHEMACHECK

NOENTITYESCAPING

NOSCHEMACHECK

(

,
AS
c_alias
AS

EVALNAME

value_expr

value_expr

)

Purpose
XMLElement takes an element name for identifier or evaluates an element name for
EVALNAME value_expr, an optional collection of attributes for the element, and
arguments that make up the content of the element. It returns an instance of type
XMLType. XMLElement is similar to SYS_XMLGen except that XMLElement can include
attributes in the XML returned, but it does not accept formatting using the XMLFormat
object.
The XMLElement function is typically nested to produce an XML document with a
nested structure, as in the example in the following section.
For an explanation of the ENTITYESCAPING and NONENTITYESCAPING keywords, refer to
Oracle XML DB Developer's Guide.
You must specify a value for Oracle Database to use an the enclosing tag. You can do
this by specifying identifier, which is a string literal, or by specifying EVALNAME
value_expr. In the latter case, the value expression is evaluated and the result, which
must be a string literal, is used as the identifier. The identifier can be up to 4000
characters and does not have to be a column name or column reference. It cannot be an
expression or null.
The objects that make up the element content follow the XMLATTRIBUTES keyword. In
the XML_attributes_clause, if the value_expr is null, then no attribute is created for
5-358 Oracle Database SQL Language Reference

XMLELEMENT

that value expression. The type of value_expr cannot be an object type or collection. If
you specify an alias for value_expr using the AS clause, then the c_alias or the
evaluated value expression (EVALNAME value_expr) can be up to 4000 characters.
For the optional value_expr that follows the XML_attributes_clause in the diagram:
■

■

■

If value_expr is a scalar expression, then you can omit the AS clause, and Oracle
uses the column name as the element name.
If value_expr is an object type or collection, then the AS clause is mandatory, and
Oracle uses the specified c_alias as the enclosing tag.
If value_expr is null, then no element is created for that value expression.
See Also:

SYS_XMLGEN on page 5-291

Examples
The following example produces an Emp element for a series of employees, with nested
elements that provide the employee's name and hire date:
SELECT XMLELEMENT("Emp", XMLELEMENT("Name",
e.job_id||' '||e.last_name),
XMLELEMENT("Hiredate", e.hire_date)) as "Result"
FROM employees e WHERE employee_id > 200;
Result
------------------------------------------------------------------
MK_MAN Hartstein
2004-02-17


MK_REP Fay
2005-08-17


HR_REP Mavris
2002-06-07


PR_REP Baer
2002-06-07


AC_MGR Higgins
2002-06-07


AC_ACCOUNT Gietz
2002-06-07

6 rows selected.

Functions

5-359

XMLELEMENT

The following similar example uses the XMLElement function with the XML_
attributes_clause to create nested XML elements with attribute values for the
top-level element:
SELECT XMLELEMENT("Emp",
XMLATTRIBUTES(e.employee_id AS "ID", e.last_name),
XMLELEMENT("Dept", e.department_id),
XMLELEMENT("Salary", e.salary)) AS "Emp Element"
FROM employees e
WHERE e.employee_id = 206;
Emp Element
--------------------------------------------------------------
110
8300


Notice that the AS identifier clause was not specified for the last_name column. As a
result, the XML returned uses the column name last_name as the default.
Finally, the next example uses a subquery within the XML_attributes_clause to
retrieve information from another table into the attributes of an element:
SELECT XMLELEMENT("Emp", XMLATTRIBUTES(e.employee_id, e.last_name),
XMLELEMENT("Dept", XMLATTRIBUTES(e.department_id,
(SELECT d.department_name FROM departments d
WHERE d.department_id = e.department_id) as "Dept_name")),
XMLELEMENT("salary", e.salary),
XMLELEMENT("Hiredate", e.hire_date)) AS "Emp Element"
FROM employees e
WHERE employee_id = 205;
Emp Element
------------------------------------------------------------------

12008
2002-06-07


5-360 Oracle Database SQL Language Reference

XMLEXISTS

XMLEXISTS
Syntax
XML_passing_clause
XMLEXISTS

(

XQuery_string

)

XML_passing_clause::=
,
BY
PASSING

VALUE

AS

identifier

expr

Purpose
XMLExists checks whether a given XQuery expression returns a nonempty XQuery
sequence. If so, the function returns TRUE; otherwise, it returns FALSE. The argument
XQuery_string is a literal string, but it can contain XQuery variables that you bind
using the XML_passing_clause.
The expr in the XML_passing_clause is an expression returning an XMLType or an
instance of a SQL scalar data type that is used as the context for evaluating the XQuery
expression. You can specify only one expr in the PASSING clause without an identifier.
The result of evaluating each expr is bound to the corresponding identifier in the
XQuery_string. If any expr that is not followed by an AS clause, then the result of
evaluating that expression is used as the context item for evaluating the XQuery_
string.
See Also: Oracle XML DB Developer's Guide for more information on
uses for this function and examples

Functions

5-361

XMLFOREST

XMLFOREST
Syntax
,
c_alias
AS
EVALNAME
XMLFOREST

(

value_expr

value_expr

)

Purpose
XMLForest converts each of its argument parameters to XML, and then returns an XML
fragment that is the concatenation of these converted arguments.
■

■

If value_expr is a scalar expression, then you can omit the AS clause, and Oracle
Database uses the column name as the element name.
If value_expr is an object type or collection, then the AS clause is mandatory, and
Oracle uses the specified expression as the enclosing tag.
You can do this by specifying c_alias, which is a string literal, or by specifying
EVALNAME value_expr. In the latter case, the value expression is evaluated and the
result, which must be a string literal, is used as the identifier. The identifier can be
up to 4000 characters and does not have to be a column name or column reference.
It cannot be an expression or null.

■

If value_expr is null, then no element is created for that value_expr.

Examples
The following example creates an Emp element for a subset of employees, with nested
employee_id, last_name, and salary elements as the contents of Emp:
SELECT XMLELEMENT("Emp",
XMLFOREST(e.employee_id, e.last_name, e.salary))
"Emp Element"
FROM employees e WHERE employee_id = 204;
Emp Element
---------------------------------------------------------------
204
Baer
10000


Refer to the example for XMLCOLATTVAL on page 5-353 to compare the output of
these two functions.

5-362 Oracle Database SQL Language Reference

XMLISVALID

XMLISVALID
Syntax
,
,
XMLISVALID

(

element

XMLSchema_URL

XMLType_instance

)

Purpose
XMLISVALID checks whether the input XMLType_instance conforms to the relevant
XML schema. It does not change the validation status recorded for XMLType_instance.
If the input XML document is determined to be valid, then XMLISVALID returns 1;
otherwise, it returns 0. If you provide XMLSchema_URL as an argument, then that is used
to check conformance. Otherwise, the XML schema specified by the XML document is
used to check conformance.
■
■

■

XMLType_instance is the XMLType instance to be validated.
XMLSchema_URL is the URL of the XML schema against which to check
conformance.
element is the element of the specified schema against which to check
conformance. Use this if you have an XML schema that defines more than one top
level element, and you want to check conformance against a specific one of those
elements.
See Also: Oracle XML DB Developer's Guide for information on the
use of this function, including examples

Functions

5-363

XMLPARSE

XMLPARSE
Syntax
WELLFORMED

DOCUMENT
XMLPARSE

(

value_expr

)

CONTENT

Purpose
XMLParse parses and generates an XML instance from the evaluated result of value_
expr. The value_expr must resolve to a string. If value_expr resolves to null, then the
function returns null.
■

■
■

If you specify DOCUMENT, then value_expr must resolve to a singly rooted XML
document.
If you specify CONTENT, then value_expr must resolve to a valid XML value.
When you specify WELLFORMED, you are guaranteeing that value_expr resolves to a
well-formed XML document, so the database does not perform validity checks to
ensure that the input is well formed.
See Also: Oracle XML DB Developer's Guide for more information on
this function

Examples
The following example uses the DUAL table to illustrate the syntax of XMLParse:
SELECT XMLPARSE(CONTENT '124 
 Acme Enterprises
32987457
'
WELLFORMED) AS PO FROM DUAL;
PO
----------------------------------------------------------------124 
 Acme Enterprises
32987457


5-364 Oracle Database SQL Language Reference

XMLPATCH

XMLPATCH
Syntax
XMLPatch

(

XMLType_document

,

XMLType_document

)

Purpose
The XMLPatch function is the SQL interface for the XmlPatch C API. This function
patches an XML document with the changes specified. A patched XMLType document
is returned.
■

For the first argument, specify the name of the input XMLType document
For the second argument, specify the XMLType document containing the changes
to be applied to the first document. The changes should conform to the Xdiff
XML schema

■

For string, specify the flags that control the behavior of the function. These flags
are specified by one or more names separated by semicolon. The names are the
same as the names of constants for XmlPatch C function.
Oracle XML Developer's Kit Programmer's Guide for more
information on using this function, including examples, and Oracle
Database XML C API Reference for information on the XML APIs for C

See Also:

Examples
The following example patches an XMLType document with the changes specified in
another XMLType and returns a patched XMLType document:
SELECT XMLPATCH(
XMLTYPE('




Chapter 1.




Chapter 2.



'),
XMLTYPE('



')
)

Functions

5-365

XMLPATCH

FROM DUAL;

5-366 Oracle Database SQL Language Reference

XMLPI

XMLPI
Syntax
NAME
identifier
XMLPI

(

,

value_expr
)

EVALNAME

value_expr

Purpose
XMLPI generates an XML processing instruction using identifier and optionally the
evaluated result of value_expr. A processing instruction is commonly used to provide
to an application information that is associated with all or part of an XML document.
The application uses the processing instruction to determine how best to process the
XML document.
You must specify a value for Oracle Database to use an the enclosing tag. You can do
this by specifying identifier, which is a string literal, or by specifying EVALNAME
value_expr. In the latter case, the value expression is evaluated and the result, which
must be a string literal, is used as the identifier. The identifier can be up to 4000
characters and does not have to be a column name or column reference. It cannot be an
expression or null.
The optional value_expr must resolve to a string. If you omit the optional value_expr,
then a zero-length string is the default. The value returned by the function takes this
form:


XMLPI is subject to the following restrictions:
■

The identifier must be a valid target for a processing instruction.

■

You cannot specify xml in any case combination for identifier.

■

The identifier cannot contain the consecutive characters ?>.
See Also: Oracle XML DB Developer's Guide for more information on
this function

Examples
The following statement uses the DUAL table to illustrate the use of the XMLPI syntax:
SELECT XMLPI(NAME "Order analysisComp", 'imported, reconfigured, disassembled')
AS "XMLPI" FROM DUAL;
XMLPI
-------------------------------------------------------------------------------

Functions

5-367

XMLQUERY

XMLQUERY
Syntax
XMLQUERY

XML_passing_clause
(

NULL

XQuery_string

RETURNING

ON

EMPTY

CONTENT

)

XML_passing_clause::=
,
BY

VALUE

PASSING

AS

identifier

expr

Purpose
XMLQUERY lets you query XML data in SQL statements. It takes an XQuery expression
as a string literal, an optional context item, and other bind variables and returns the
result of evaluating the XQuery expression using these input values.
■
■

■

■

XQuery_string is a complete XQuery expression, including prolog.
The expr in the XML_passing_clause is an expression returning an XMLType or an
instance of a SQL scalar data type that is used as the context for evaluating the
XQuery expression. You can specify only one expr in the PASSING clause without
an identifier. The result of evaluating each expr is bound to the corresponding
identifier in the XQuery_string. If any expr that is not followed by an AS clause,
then the result of evaluating that expression is used as the context item for
evaluating the XQuery_string.
RETURNING CONTENT indicates that the result from the XQuery evaluation is either
an XML 1.0 document or a document fragment conforming to the XML 1.0
semantics.
If the result set is empty, then the function returns the SQL NULL value. The NULL ON
EMPTY keywords are implemented by default and are shown for syntactic clarity.
See Also: Oracle XML DB Developer's Guide for more information on
this function

Examples
The following statement specifies the warehouse_spec column of the oe.warehouses
table in the XML_passing_clause as a context item. The statement returns specific
information about the warehouses with area greater than 50K.
SELECT warehouse_name,
EXTRACTVALUE(warehouse_spec, '/Warehouse/Area'),
XMLQuery(
'for $i in /Warehouse
where $i/Area > 50000
return 



{
if ($i/RailAccess = "Y") then "true" else "false"

5-368 Oracle Database SQL Language Reference

XMLQUERY

}


' PASSING warehouse_spec RETURNING CONTENT) "Big_warehouses"
FROM warehouses;
WAREHOUSE_ID Area
------------ --------1
25000
2
50000
3
85700
4
103000
. . .

Big_warehouses
--------------------------------------------------------


false


true


Functions

5-369

XMLROOT

XMLROOT
Syntax
YES
,

STANDALONE

value_expr
XMLROOT

(

value_expr

,

NO
NO

VALUE

VERSION

)
NO

VALUE

Purpose
XMLROOT lets you create a new XML value by providing version and standalone
properties in the XML root information (prolog) of an existing XML value. If the
value_expr already has a prolog, then the database returns an error. If the input is
null, then the function returns null.
The value returned takes the following form:

■

■

■

The first value_expr specifies the XML value for which you are providing prolog
information.
In the VERSION clause, value_expr must resolve to a string representing a valid
XML version. If you specify NO VALUE for VERSION, then the version defaults to 1.0.
If you omit the optional STANDALONE clause, or if you specify it with NO VALUE, then
the standalone property is absent from the value returned by the function.

Examples
The following statement uses the DUAL table to illustrate the syntax of XMLROOT:
SELECT XMLROOT ( XMLType('143598'), VERSION '1.0', STANDALONE YES)
AS "XMLROOT" FROM DUAL;
XMLROOT
-------------------------------------------------------------------------------
143598

5-370 Oracle Database SQL Language Reference

XMLSEQUENCE

XMLSEQUENCE
The XMLSEQUENCE function is deprecated. It is still supported
for backward compatibility. However, Oracle recommends that you
use the XMLTABLE function instead. See XMLTABLE on page 5-375 for
more information.
Note:

Syntax
XMLType_instance
XMLSEQUENCE

(

,

fmt

)

sys_refcursor_instance

Purpose
XMLSequence has two forms:
■

■

The first form takes as input an XMLType instance and returns a varray of the
top-level nodes in the XMLType. This form is effectively superseded by the
SQL/XML standard function XMLTable, which provides for more readable SQL
code. Prior to Oracle Database 10g Release 2, XMLSequence was used with SQL
function TABLE to do some of what can now be done better with the XMLTable
function.
The second form takes as input a REFCURSOR instance, with an optional instance of
the XMLFormat object, and returns as an XMLSequence type an XML document for
each row of the cursor.

Because XMLSequence returns a collection of XMLType, you can use this function in a
TABLE clause to unnest the collection values into multiple rows, which can in turn be
further processed in the SQL query.
See Also: Oracle XML DB Developer's Guide for more information on
this function, and XMLTABLE on page 5-375

Examples
The following example shows how XMLSequence divides up an XML document with
multiple elements into VARRAY single-element documents. In this example, the TABLE
keyword instructs Oracle Database to consider the collection a table value that can be
used in the FROM clause of the subquery:
SELECT EXTRACT(warehouse_spec, '/Warehouse') as "Warehouse"
FROM warehouses WHERE warehouse_name = 'San Francisco';
Warehouse
-----------------------------------------------------------
Rented
50000
1
Side load
Y
N
Lot

Functions

5-371

XMLSEQUENCE

12 ft

1 row selected.
SELECT VALUE(p)
FROM warehouses w,
TABLE(XMLSEQUENCE(EXTRACT(warehouse_spec, '/Warehouse/*'))) p
WHERE w.warehouse_name = 'San Francisco';
VALUE(P)
---------------------------------------------------------------Rented
50000
1
Side load
Y
N
Lot
12 ft
8 rows selected.

5-372 Oracle Database SQL Language Reference

XMLSERIALIZE

XMLSERIALIZE
Syntax
AS

DOCUMENT
XMLSERIALIZE

(

datatype

value_expr
CONTENT

ENCODING

NO

xml_encoding_spec

string_literal

INDENT
SIZE

INDENT

VERSION

=

number

HIDE
DEFAULTS
SHOW
)

Purpose
XMLSerialize creates a string or LOB containing the contents of value_expr.
■
■

■

■

■

■

■

If you specify DOCUMENT, then the value_expr must be a valid XML document.
If you specify CONTENT, then the value_expr need not be a singly rooted XML
document. However it must be valid XML content.
The datatype specified can be a string type (VARCHAR2 or VARCHAR, but not
NVARCHAR2), BLOB, or CLOB. The default is CLOB.
If datatype is BLOB, then you can specify the ENCODING clause to use the specified
encoding in the prolog. The xml_encoding_spec is an XML encoding declaration
(encoding="...").
Specify the VERSION clause to use the version you provide as string_literal in
the XML declaration ().
Specify NO INDENT to strip all insignificant whitespace from the output. Specify
INDENT SIZE = N, where N is a whole number, for output that is pretty-printed
using a relative indentation of N spaces. If N is 0, then pretty-printing inserts a
newline character after each element, placing each element on a line by itself, but
omitting all other insignificant whitespace in the output. If INDENT is present
without a SIZE specification, then 2-space indenting is used. If you omit this
clause, then the behavior (pretty-printing or not) is indeterminate.
HIDE DEFAULTS and SHOW DEFAULTS apply only to XML schema-based data. If you
specify SHOW DEFAULTS and the input data is missing any optional elements or
attributes for which the XML schema defines default values, then those elements
or attributes are included in the output with their default values. If you specify
HIDE DEFAULTS, then no such elements or attributes are included in the output.
HIDE DEFAULTS is the default behavior.
See Also: Oracle XML DB Developer's Guide for more information on
this function

Examples
The following statement uses the DUAL table to illustrate the syntax of XMLSerialize:
SELECT XMLSERIALIZE(CONTENT XMLTYPE('Grandco')) AS xmlserialize_doc

Functions

5-373

XMLSERIALIZE

FROM DUAL;
XMLSERIALIZE_DOC
---------------Grandco

5-374 Oracle Database SQL Language Reference

XMLTABLE

XMLTABLE
Syntax
XMLnamespaces_clause
XMLTABLE

,

(

XQuery_string

XMLTABLE_options

)

XMLnamespaces_clause::=
,
string
XMLNAMESPACES

AS

identifier

(

)
DEFAULT

Note:

string

You can specify at most one DEFAULT string clause.

XMLTABLE_options::=
,
XML_passing_clause

COLUMNS

XML_table_column

XML_passing_clause::=
,
BY

VALUE

PASSING

AS

identifier

expr

XML_table_column::=
FOR

ORDINALITY

column

PATH

string

DEFAULT

expr

datatype

Purpose
XMLTable maps the result of an XQuery evaluation into relational rows and columns.
You can query the result returned by the function as a virtual relational table using
SQL.
■

■

■

The XMLNAMESPACES clause contains a set of XML namespace declarations. These
declarations are referenced by the XQuery expression (the evaluated XQuery_
string), which computes the row, and by the XPath expression in the PATH clause
of XML_table_column, which computes the columns for the entire XMLTable
function. If you want to use qualified names in the PATH expressions of the
COLUMNS clause, then you need to specify the XMLNAMESPACES clause.
XQuery_string is a complete XQuery expression and can include prolog
declarations.
The expr in the XML_passing_clause is an expression returning an XMLType or an
instance of a SQL scalar data type that is used as the context for evaluating the
Functions

5-375

XMLTABLE

XQuery expression. You can specify only one expr in the PASSING clause without
an identifier. The result of evaluating each expr is bound to the corresponding
identifier in the XQuery_string. If any expr that is not followed by an AS clause,
then the result of evaluating that expression is used as the context item for
evaluating the XQuery_string.
■

The optional COLUMNS clause defines the columns of the virtual table to be created
by XMLTable.
–

If you omit the COLUMNS clause, then XMLTable returns a row with a single
XMLType pseudocolumn named COLUMN_VALUE.

–

FOR ORDINALITY specifies that column is to be a column of generated row
numbers. There must be at most one FOR ORDINALITY clause. It is created as a
NUMBER column.

–

For each resulting column except the FOR ORDINALITY column, you must
specify the column datatype, which can be XMLType or any other data type.

–

The optional PATH clause specifies that the portion of the XQuery result that is
addressed by XQuery expression string is to be used as the column content. If
you omit PATH, then the XQuery expression column is assumed. For example:
XMLTable(... COLUMNS xyz

is equivalent to
XMLTable(... COLUMNS xyz PATH 'XYZ')

You can use different PATH clauses to split the XQuery result into different
virtual-table columns.
–

The optional DEFAULT clause specifies the value to use when the PATH
expression results in an empty sequence. Its expr is an XQuery expression that
is evaluated to produce the default value.
See Also: Oracle XML DB Developer's Guide for more information on
the XMLTable function, including additional examples, and on XQuery
in general

Examples
The following example converts the result of applying the XQuery '/Warehouse' to
each value in the warehouse_spec column of the warehouses table into a virtual
relational table with columns Water and Rail:
SELECT warehouse_name warehouse,
warehouse2."Water", warehouse2."Rail"
FROM warehouses,
XMLTABLE('/Warehouse'
PASSING warehouses.warehouse_spec
COLUMNS
"Water" varchar2(6) PATH '/Warehouse/WaterAccess',
"Rail" varchar2(6) PATH '/Warehouse/RailAccess')
warehouse2;
WAREHOUSE
----------------------------------Southlake, Texas
San Francisco
New Jersey
Seattle, Washington

5-376 Oracle Database SQL Language Reference

Water
-----Y
Y
N
N

Rail
-----N
N
N
Y

XMLTRANSFORM

XMLTRANSFORM
Syntax
XMLType_instance
XMLTRANSFORM

(

XMLType_instance

,

)
string

Purpose
XMLTransform takes as arguments an XMLType instance and an XSL style sheet, which is
itself a form of XMLType instance. It applies the style sheet to the instance and returns
an XMLType.
This function is useful for organizing data according to a style sheet as you are
retrieving it from the database.
See Also: Oracle XML DB Developer's Guide for more information on
this function

Examples
The XMLTransform function requires the existence of an XSL style sheet. Here is an
example of a very simple style sheet that alphabetizes elements within a node:
CREATE TABLE xsl_tab (col1 XMLTYPE);
INSERT INTO xsl_tab VALUES (
XMLTYPE.createxml(
'













 '));
1 row created.

The next example uses the xsl_tab XSL style sheet to alphabetize the elements in one
warehouse_spec of the sample table oe.warehouses:
SELECT XMLTRANSFORM(w.warehouse_spec, x.col1).GetClobVal()
FROM warehouses w, xsl_tab x
WHERE w.warehouse_name = 'San Francisco';
XMLTRANSFORM(W.WAREHOUSE_SPEC,X.COL1).GETCLOBVAL()
-------------------------------------------------------------------------------
50000
Rented

Functions

5-377

XMLTRANSFORM

Side load
1
Lot
N
12 ft
Y


5-378 Oracle Database SQL Language Reference

ROUND and TRUNC Date Functions

ROUND and TRUNC Date Functions
Table 5–14 lists the format models you can use with the ROUND and TRUNC date
functions and the units to which they round and truncate dates. The default model,
'DD', returns the date rounded or truncated to the day with a time of midnight.
Table 5–14

Date Format Models for the ROUND and TRUNC Date Functions

Format Model

Rounding or Truncating Unit

CC
SCC

One greater than the first two digits of a four-digit year

SYYYY
YYYY
YEAR
SYEAR
YYY
YY
Y

Year (rounds up on July 1)

IYYY
IYY
IY
I

Year containing the calendar week, as defined by the ISO 8601 standard

Q

Quarter (rounds up on the sixteenth day of the second month of the
quarter)

MONTH
MON
MM
RM

Month (rounds up on the sixteenth day)

WW

Same day of the week as the first day of the year

IW

Same day of the week as the first day of the calendar week as defined by
the ISO 8601 standard, which is Monday

W

Same day of the week as the first day of the month

DDD
DD
J

Day

DAY
DY
D

Starting day of the week

HH
HH12
HH24

Hour

MI

Minute

The starting day of the week used by the format models DAY, DY, and D is specified
implicitly by the initialization parameter NLS_TERRITORY.
Oracle Database Reference and Oracle Database Globalization
Support Guide for information on this parameter

See Also:

Functions

5-379

About User-Defined Functions

About User-Defined Functions
You can write user-defined functions in PL/SQL, Java, or C to provide functionality
that is not available in SQL or SQL built-in functions. User-defined functions can
appear in a SQL statement wherever an expression can occur.
For example, user-defined functions can be used in the following:
■

The select list of a SELECT statement

■

The condition of a WHERE clause

■

CONNECT BY, START WITH, ORDER BY, and GROUP BY clauses

■

The VALUES clause of an INSERT statement

■

The SET clause of an UPDATE statement
Oracle SQL does not support calling of functions with
Boolean parameters or returns. Therefore, if your user-defined
functions will be called from SQL statements, you must design
them to return numbers (0 or 1) or character strings ('TRUE' or
'FALSE').

Note:

user_defined_function::=
package
schema

.

.

function
user_defined_operator

DISTINCT
ALL

,
expr

@

dblink

.

(

)

The optional expression list must match attributes of the function, package, or
operator.
Restriction on User-defined Functions The DISTINCT and ALL keywords are valid
only with a user-defined aggregate function.
See Also:
■

■

CREATE FUNCTION on page 14-58 for information on creating
functions, including restrictions on user-defined functions
Oracle Database Advanced Application Developer's Guide for a
complete discussion of the creation and use of user functions

5-380 Oracle Database SQL Language Reference

About User-Defined Functions

Prerequisites
User-defined functions must be created as top-level functions or declared with a
package specification before they can be named within a SQL statement.
To use a user function in a SQL expression, you must own or have EXECUTE privilege
on the user function. To query a view defined with a user function, you must have
SELECT privileges on the view. No separate EXECUTE privileges are needed to select
from the view.
CREATE FUNCTION on page 14-58 for information on
creating top-level functions and CREATE PACKAGE on page 15-42 for
information on specifying packaged functions

See Also:

Functions

5-381

Name Precedence

Name Precedence
Within a SQL statement, the names of database columns take precedence over the
names of functions with no parameters. For example, if the Human Resources
manager creates the following two objects in the hr schema:
CREATE TABLE new_emps (new_sal NUMBER, ...);
CREATE FUNCTION new_sal RETURN NUMBER IS BEGIN ... END;

then in the following two statements, the reference to new_sal refers to the column
new_emps.new_sal:
SELECT new_sal FROM new_emps;
SELECT new_emps.new_sal FROM new_emps;

To access the function new_sal, you would enter:
SELECT hr.new_sal FROM new_emps;

Here are some sample calls to user functions that are allowed in SQL expressions:
circle_area (radius)
payroll.tax_rate (empno)
hr.employees.tax_rate (dependent, empno)@remote

To call the tax_rate user function from schema hr, execute it against the
ss_no and sal columns in tax_table, specify the following:
Example

SELECT hr.tax_rate (ss_no, sal)
INTO income_tax
FROM tax_table WHERE ss_no = tax_id;

The INTO clause is PL/SQL that lets you place the results into the variable income_tax.

Naming Conventions
If only one of the optional schema or package names is given, then the first identifier
can be either a schema name or a package name. For example, to determine whether
PAYROLL in the reference PAYROLL.TAX_RATE is a schema or package name, Oracle
Database proceeds as follows:
1.

Check for the PAYROLL package in the current schema.

2.

If a PAYROLL package is not found, then look for a schema name PAYROLL that
contains a top-level TAX_RATE function. If no such function is found, then return an
error.

3.

If the PAYROLL package is found in the current schema, then look for a TAX_RATE
function in the PAYROLL package. If no such function is found, then return an error.

You can also refer to a stored top-level function using any synonym that you have
defined for it.

5-382 Oracle Database SQL Language Reference

6
6

Expressions

This chapter describes how to combine values, operators, and functions into
expressions.
This chapter includes these sections:
■

About SQL Expressions

■

Simple Expressions

■

Compound Expressions

■

CASE Expressions

■

Column Expressions

■

CURSOR Expressions

■

Datetime Expressions

■

Function Expressions

■

Interval Expressions

■

Model Expressions

■

Object Access Expressions

■

Placeholder Expressions

■

Scalar Subquery Expressions

■

Type Constructor Expressions

■

Expression Lists

About SQL Expressions
An expression is a combination of one or more values, operators, and SQL functions
that evaluates to a value. An expression generally assumes the data type of its
components.

Expressions 6-1

About SQL Expressions

The combined values of the NLS_COMP and NLS_SORT settings
determine the rules by which characters are sorted and compared. If
NLS_COMP is set to LINGUISTIC for your database, then all entities in
this chapter will be interpreted according to the rules specified by the
NLS_SORT parameter. If NLS_COMP is not set to LINGUISTIC, then the
functions are interpreted without regard to the NLS_SORT setting. NLS_
SORT can be explicitly set. If it is not set explicitly, it is derived from
NLS_LANGUAGE. Refer to Oracle Database Globalization Support Guide for
more information on these settings.
Note:

This simple expression evaluates to 4 and has data type NUMBER (the same data type as
its components):
2*2

The following expression is an example of a more complex expression that uses both
functions and operators. The expression adds seven days to the current date, removes
the time component from the sum, and converts the result to CHAR data type:
TO_CHAR(TRUNC(SYSDATE+7))

You can use expressions in:
■

The select list of the SELECT statement

■

A condition of the WHERE clause and HAVING clause

■

The CONNECT BY, START WITH, and ORDER BY clauses

■

The VALUES clause of the INSERT statement

■

The SET clause of the UPDATE statement

For example, you could use an expression in place of the quoted string 'Smith' in this
UPDATE statement SET clause:
SET last_name = 'Smith';

This SET clause has the expression INITCAP(last_name) instead of the quoted string
'Smith':
SET last_name = INITCAP(last_name);

Expressions have several forms, as shown in the following syntax:

6-2 Oracle Database SQL Language Reference

Simple Expressions

expr::=
simple_expression
compound_expression
case_expression
cursor_expression
datetime_expression
function_expression
interval_expression
object_access_expression
scalar_subquery_expression
model_expression
type_constructor_expression
variable_expression

Oracle Database does not accept all forms of expressions in all parts of all SQL
statements. Refer to the individual SQL statements in Chapter 10 through Chapter 19
for information on restrictions on the expressions in that statement.
You must use appropriate expression notation whenever expr appears in conditions,
SQL functions, or SQL statements in other parts of this reference. The sections that
follow describe and provide examples of the various forms of expressions.

Simple Expressions
A simple expression specifies a column, pseudocolumn, constant, sequence number, or
null.
simple_expression::=
query_name
schema

.

table

.

view
materialized view

column
ROWID

ROWNUM
string
number
CURRVAL
sequence

.
NEXTVAL

NULL

In addition to the schema of a user, schema can also be "PUBLIC" (double quotation
marks required), in which case it must qualify a public synonym for a table, view, or
Expressions 6-3

Compound Expressions

materialized view. Qualifying a public synonym with "PUBLIC" is supported only in
data manipulation language (DML) statements, not data definition language (DDL)
statements.
You can specify ROWID only with a table, not with a view or materialized view. NCHAR
and NVARCHAR2 are not valid pseudocolumn data types.
See Also: Chapter 2, "Pseudocolumns" for more information on
pseudocolumns and subquery_factoring_clause on page 19-13 for
information on query_name

Some valid simple expressions are:
employees.last_name
'this is a text string'
10
N'this is an NCHAR string'

Compound Expressions
A compound expression specifies a combination of other expressions.
compound_expression::=
(

expr

)

+
–

expr

PRIOR
*
/
expr

+

expr

–
||

You can use any built-in function as an expression ("Function Expressions" on
page 6-10). However, in a compound expression, some combinations of functions are
inappropriate and are rejected. For example, the LENGTH function is inappropriate
within an aggregate function.
The PRIOR operator is used in CONNECT BY clauses of hierarchical queries.
See Also: "Operator Precedence" on page 4-2 and "Hierarchical
Queries" on page 9-3

Some valid compound expressions are:
('CLARK' || 'SMITH')
LENGTH('MOOSE') * 57
SQRT(144) + 72
my_fun(TO_CHAR(sysdate,'DD-MMM-YY'))

6-4 Oracle Database SQL Language Reference

CASE Expressions

CASE Expressions
CASE expressions let you use IF ... THEN ... ELSE logic in SQL statements without having
to invoke procedures. The syntax is:
else_clause

simple_case_expression
CASE

END
searched_case_expression

simple_case_expression::=
expr

WHEN

comparison_expr

THEN

return_expr

searched_case_expression::=
WHEN

condition

THEN

return_expr

else_clause::=
ELSE

else_expr

In a simple CASE expression, Oracle Database searches for the first WHEN ... THEN pair for
which expr is equal to comparison_expr and returns return_expr. If none of the WHEN
... THEN pairs meet this condition, and an ELSE clause exists, then Oracle returns else_
expr. Otherwise, Oracle returns null.
In a searched CASE expression, Oracle searches from left to right until it finds an
occurrence of condition that is true, and then returns return_expr. If no condition is
found to be true, and an ELSE clause exists, then Oracle returns else_expr. Otherwise,
Oracle returns null.
Oracle Database uses short-circuit evaluation. For a simple CASE expression, the
database evaluates each comparison_expr value only before comparing it to expr,
rather than evaluating all comparison_expr values before comparing any of them with
expr. Consequently, Oracle never evaluates a comparison_expr if a previous
comparison_expr is equal to expr. For a searched CASE expression, the database
evaluates each condition to determine whether it is true, and never evaluates a
condition if the previous condition was true.
For a simple CASE expression, the expr and all comparison_expr values must either
have the same data type (CHAR, VARCHAR2, NCHAR, or NVARCHAR2, NUMBER, BINARY_FLOAT,
or BINARY_DOUBLE) or must all have a numeric data type. If all expressions have a
numeric data type, then Oracle determines the argument with the highest numeric
precedence, implicitly converts the remaining arguments to that data type, and returns
that data type.
For both simple and searched CASE expressions, all of the return_exprs must either
have the same data type (CHAR, VARCHAR2, NCHAR, or NVARCHAR2, NUMBER, BINARY_FLOAT,
or BINARY_DOUBLE) or must all have a numeric data type. If all return expressions have
a numeric data type, then Oracle determines the argument with the highest numeric
precedence, implicitly converts the remaining arguments to that data type, and returns
that data type.
The maximum number of arguments in a CASE expression is 65535. All expressions
count toward this limit, including the initial expression of a simple CASE expression
and the optional ELSE expression. Each WHEN ... THEN pair counts as two arguments. To
Expressions 6-5

Column Expressions

avoid exceeding this limit, you can nest CASE expressions so that the return_expr itself
is a CASE expression.
See Also:
■

■

■

■

Table 3–10, " Implicit Type Conversion Matrix" on page 3-40 for
more information on implicit conversion
"Numeric Precedence" on page 3-14 for information on numeric
precedence
COALESCE on page 5-48 and NULLIF on page 5-170 for
alternative forms of CASE logic
Oracle Database Data Warehousing Guide for examples using various
forms of the CASE expression

Simple CASE Example For each customer in the sample oe.customers table, the

following statement lists the credit limit as "Low" if it equals $100, "High" if it equals
$5000, and "Medium" if it equals anything else.
SELECT cust_last_name,
CASE credit_limit WHEN 100 THEN 'Low'
WHEN 5000 THEN 'High'
ELSE 'Medium' END AS credit
FROM customers
ORDER BY cust_last_name, credit;
CUST_LAST_NAME
-------------------Adjani
Adjani
Alexander
Alexander
Altman
Altman
. . .

CREDIT
-----Medium
Medium
Medium
Medium
High
Medium

The following statement finds the average salary of the
employees in the sample table oe.employees, using $2000 as the lowest salary
possible:

Searched CASE Example

SELECT AVG(CASE WHEN e.salary > 2000 THEN e.salary
ELSE 2000 END) "Average Salary" FROM employees e;
Average Salary
-------------6461.68224

Column Expressions
A column expression, which is designated as column_expression in subsequent syntax
diagrams, is a limited form of expr. A column expression can be a simple expression,
compound expression, function expression, or expression list, but it can contain only
the following forms of expression:
■

Columns of the subject table — the table being created, altered, or indexed

■

Constants (strings or numbers)

■

Deterministic functions — either SQL built-in functions or user-defined functions

6-6 Oracle Database SQL Language Reference

CURSOR Expressions

No other expression forms described in this chapter are valid. In addition, compound
expressions using the PRIOR keyword are not supported, nor are aggregate functions.
You can use a column expression for these purposes:
■
■

To create a function-based index.
To explicitly or implicitly define a virtual column. When you define a virtual
column, the defining column_expression must refer only to columns of the subject
table that have already been defined, in the current statement or in a prior
statement.

The combined components of a column expression must be deterministic. That is, the
same set of input values must return the same set of output values.
"Simple Expressions" on page 6-3, "Compound
Expressions" on page 6-4, "Function Expressions" on page 6-10, and
"Expression Lists" on page 6-16 for information on these forms of expr
See Also:

CURSOR Expressions
A CURSOR expression returns a nested cursor. This form of expression is equivalent to
the PL/SQL REF CURSOR and can be passed as a REF CURSOR argument to a function.
CURSOR

(

subquery

)

A nested cursor is implicitly opened when the cursor expression is evaluated. For
example, if the cursor expression appears in a select list, a nested cursor will be
opened for each row fetched by the query. The nested cursor is closed only when:
■

The nested cursor is explicitly closed by the user

■

The parent cursor is reexecuted

■

The parent cursor is closed

■

The parent cursor is cancelled

■

An error arises during fetch on one of its parent cursors (it is closed as part of the
clean-up)

Restrictions on CURSOR Expressions The following restrictions apply to CURSOR
expressions:
■

■

If the enclosing statement is not a SELECT statement, then nested cursors can
appear only as REF CURSOR arguments of a procedure.
If the enclosing statement is a SELECT statement, then nested cursors can also
appear in the outermost select list of the query specification or in the outermost
select list of another nested cursor.

■

Nested cursors cannot appear in views.

■

You cannot perform BIND and EXECUTE operations on nested cursors.

Examples The following example shows the use of a CURSOR expression in the select
list of a query:
SELECT department_name, CURSOR(SELECT salary, commission_pct
FROM employees e
WHERE e.department_id = d.department_id)
FROM departments d
ORDER BY department_name;
Expressions 6-7

Datetime Expressions

The next example shows the use of a CURSOR expression as a function argument. The
example begins by creating a function in the sample OE schema that can accept the REF
CURSOR argument. (The PL/SQL function body is shown in italics.)
CREATE FUNCTION f(cur SYS_REFCURSOR, mgr_hiredate DATE)
RETURN NUMBER IS
emp_hiredate DATE;
before number :=0;
after number:=0;
begin
loop
fetch cur into emp_hiredate;
exit when cur%NOTFOUND;
if emp_hiredate > mgr_hiredate then
after:=after+1;
else
before:=before+1;
end if;
end loop;
close cur;
if before > after then
return 1;
else
return 0;
end if;
end;
/

The function accepts a cursor and a date. The function expects the cursor to be a query
returning a set of dates. The following query uses the function to find those managers
in the sample employees table, most of whose employees were hired before the
manager.
SELECT e1.last_name FROM employees e1
WHERE f(
CURSOR(SELECT e2.hire_date FROM employees e2
WHERE e1.employee_id = e2.manager_id),
e1.hire_date) = 1
ORDER BY last_name;
LAST_NAME
------------------------Cambrault
Higgins
Hunold
Kochhar
Mourgos
Zlotkey

Datetime Expressions
A datetime expression yields a value of one of the datetime data types.

6-8 Oracle Database SQL Language Reference

Datetime Expressions

datetime_expression::=
LOCAL
+
expr

–

AT

’

hh

:

mi

’

DBTIMEZONE
TIME

ZONE

SESSIONTIMEZONE
’

time_zone_name

’

expr

The initial expr is any expression, except a scalar subquery expression, that evaluates
to a value of data type TIMESTAMP, TIMESTAMP WITH TIME ZONE, or TIMESTAMP WITH LOCAL
TIME ZONE. The DATE data type is not supported. If this expr is itself a datetime_
expression, then it must be enclosed in parentheses.
Datetimes and intervals can be combined according to the rules defined in Table 3–5
on page 3-21. The three combinations that yield datetime values are valid in a datetime
expression.
If you specify AT LOCAL, then Oracle uses the current session time zone.
The settings for AT TIME ZONE are interpreted as follows:
■

■

■

■

The string '[+|-]hh:mi ' specifies a time zone as an offset from UTC. For hh,
specify the number of hours. For mi, specify the number of minutes.
DBTIMEZONE: Oracle uses the database time zone established (explicitly or by
default) during database creation.
SESSIONTIMEZONE: Oracle uses the session time zone established by default or in
the most recent ALTER SESSION statement.
time_zone_name: Oracle returns the datetime_value_expr in the time zone
indicated by time_zone_name. For a listing of valid time zone region names, query
the V$TIMEZONE_NAMES dynamic performance view.

Time zone region names are needed by the daylight saving
feature. These names are stored in two types of time zone files: one
large and one small. One of these files is the default file, depending
on your environment and the release of Oracle Database you are
using. For more information regarding time zone files and names,
see Oracle Database Globalization Support Guide.

Note:

See Also:
■

■

■

Oracle Database Globalization Support Guide for a complete listing of
the time zone region names in both files
Oracle Database Reference for information on the dynamic
performance views

expr: If expr returns a character string with a valid time zone format, then Oracle
returns the input in that time zone. Otherwise, Oracle returns an error.

Expressions 6-9

Function Expressions

Example The following example converts the datetime value of one time zone to
another time zone:
SELECT FROM_TZ(CAST(TO_DATE('1999-12-01 11:00:00',
'YYYY-MM-DD HH:MI:SS') AS TIMESTAMP), 'America/New_York')
AT TIME ZONE 'America/Los_Angeles' "West Coast Time"
FROM DUAL;
West Coast Time
-----------------------------------------------01-DEC-99 08.00.00.000000 AM AMERICA/LOS_ANGELES

Function Expressions
You can use any built-in SQL function or user-defined function as an expression. Some
valid built-in function expressions are:
LENGTH('BLAKE')
ROUND(1234.567*43)
SYSDATE

See Also: "About SQL Functions" on page 5-2' and "Aggregate
Functions" on page 5-10 for information on built-in functions

A user-defined function expression specifies a call to:
■

■

■

A function in an Oracle-supplied package (see Oracle Database PL/SQL Packages and
Types Reference)
A function in a user-defined package or type or in a standalone user-defined
function (see "About User-Defined Functions" on page 5-380)
A user-defined function or operator (see CREATE OPERATOR on page 15-35,
CREATE FUNCTION on page 14-58, and Oracle Database Data Cartridge Developer's
Guide)

Some valid user-defined function expressions are:
circle_area(radius)
payroll.tax_rate(empno)
hr.employees.comm_pct@remote(dependents, empno)
DBMS_LOB.getlength(column_name)
my_function(a_column)

In a user-defined function being used as an expression, positional, named, and mixed
notation are supported. For example, all of the following notations are correct:
CALL my_function(arg1 => 3, arg2 => 4) ...
CALL my_function(3, 4) ...
CALL my_function(3, arg2 => 4) ...

You cannot pass arguments of
object type or XMLType to remote functions and procedures.

Restriction on User-Defined Function Expressions

Interval Expressions
An interval expression yields a value of INTERVAL YEAR TO MONTH or INTERVAL DAY TO
SECOND.

6-10 Oracle Database SQL Language Reference

Model Expressions

interval_expression::=
(

expr1

–

expr2

(

)

leading_field_precision

)

(

DAY

TO
(

leading_field_precision

YEAR

fractional_second_precision

)

SECOND

)
TO

MONTH

The expressions expr1 and expr2 can be any expressions that evaluate to values of
data type DATE, TIMESTAMP, TIMESTAMP WITH TIME ZONE, or TIMESTAMP WITH LOCAL TIME
ZONE.
Datetimes and intervals can be combined according to the rules defined in Table 3–5
on page 3-21. The six combinations that yield interval values are valid in an interval
expression.
Both leading_field_precision and fractional_second_precision can be any
integer from 0 to 9. If you omit the leading_field_precision for either DAY or YEAR,
then Oracle Database uses the default value of 2. If you omit the fractional_second_
precision for second, then the database uses the default value of 6. If the value
returned by a query contains more digits that the default precision, then Oracle
Database returns an error. Therefore, it is good practice to specify a precision that you
know will be at least as large as any value returned by the query.
For example, the following statement subtracts the value of the order_date column in
the sample table orders (a datetime value) from the system timestamp (another
datetime value) to yield an interval value expression. It is not known how many days
ago the oldest order was placed, so the maximum value of 9 for the DAY leading field
precision is specified:
SELECT (SYSTIMESTAMP - order_date) DAY(9) TO SECOND FROM orders
WHERE order_id = 2458;

Model Expressions
A model expression is used only in the model_clause of a SELECT statement and then
only on the right-hand side of a model rule. It yields a value for a cell in a measure
column previously defined in the model_clause. For additional information, refer to
model_clause on page 19-28.

Expressions 6-11

Model Expressions

model_expression::=
,
condition
measure_column

[

]
expr
,
condition
expr

aggregate_function

,

[

]

single_column_for_loop
multi_column_for_loop
analytic_function

When you specify a measure column in a model expression, any conditions and
expressions you specify must resolve to single values.
When you specify an aggregate function in a model expression, the argument to the
function is a measure column that has been previously defined in the model_clause.
An aggregate function can be used only on the right-hand side of a model rule.
Specifying an analytic function on the right-hand side of the model rule lets you
express complex calculations directly in the model_clause. The following restrictions
apply when using an analytic function in a model expression:
■
■

■

■

Analytic functions can be used only in an UPDATE rule.
You cannot specify an analytic function on the right-hand side of the model rule if
the left-hand side of the rule contains a FOR loop or an ORDER BY clause.
The arguments in the OVER clause of the analytic function cannot contain an
aggregate.
The arguments before the OVER clause of the analytic function cannot contain a cell
reference.
"The MODEL clause: Examples" on page 19-45 for an
example of using an analytic function on the right-hand side of a
model rule

See Also:

When expr is itself a model expression, it is referred to as a nested cell reference. The
following restrictions apply to nested cell references:
■

Only one level of nesting is allowed.

■

A nested cell reference must be a single-cell reference.

■

When AUTOMATIC ORDER is specified in the model_rules_clause, a nested cell
reference can be used on the left-hand side of a model rule only if the measures
used in the nested cell reference remain static.

The model expressions shown below are based on the model_clause of the following
SELECT statement:
SELECT country,prod,year,s
FROM sales_view_ref
MODEL

6-12 Oracle Database SQL Language Reference

Object Access Expressions

PARTITION BY (country)
DIMENSION BY (prod, year)
MEASURES (sale s)
IGNORE NAV
UNIQUE DIMENSION
RULES UPSERT SEQUENTIAL ORDER
(
s[prod='Mouse Pad', year=2000] =
s['Mouse Pad', 1998] + s['Mouse Pad', 1999],
s['Standard Mouse', 2001] = s['Standard Mouse', 2000]
)
ORDER BY country, prod, year;

The following model expression represents a single cell reference using symbolic
notation. It represents the sales of the Mouse Pad for the year 2000.
s[prod='Mouse Pad',year=2000]

The following model expression represents a multiple cell reference using positional
notation, using the CV function. It represents the sales of the current value of the
dimension column prod for the year 2001.
s[CV(prod), 2001]

The following model expression represents an aggregate function. It represents the
sum of sales of the Mouse Pad for the years between the current value of the
dimension column year less two and the current value of the dimension column year
less one.
SUM(s)['Mouse Pad',year BETWEEN CV()-2 AND CV()-1]

See Also:

CV on page 5-73 and model_clause on page 19-28

Object Access Expressions
An object access expression specifies attribute reference and method invocation.
object_access_expression::=
,
argument

.
table_alias

.

column

object_table_alias
(

expr

)

.

.

(

)

attribute
,

.
.

method

argument
method

(

)

The column parameter can be an object or REF column. If you specify expr, then it
must resolve to an object type.
When a type's member function is invoked in the context of a SQL statement, if the
SELF argument is null, Oracle returns null and the function is not invoked.
The following example creates a table based on the sample oe.order_
item_typ object type, and then shows how you would update and select from the
object column attributes.
Examples

CREATE TABLE short_orders (
Expressions 6-13

Placeholder Expressions

sales_rep VARCHAR2(25), item order_item_typ);
UPDATE short_orders s SET sales_rep = 'Unassigned';
SELECT o.item.line_item_id, o.item.quantity FROM short_orders o;

Placeholder Expressions
A placeholder expression provides a location in a SQL statement for which a
third-generation language bind variable will provide a value. You can specify the
placeholder expression with an optional indicator variable. This form of expression
can appear only in embedded SQL statements or SQL statements processed in an
Oracle Call Interface (OCI) program.
placeholder_expression::=
INDICATOR
:
:

indicator_variable

host_variable

Some valid placeholder expressions are:
:employee_name INDICATOR :employee_name_indicator_var
:department_location

Scalar Subquery Expressions
A scalar subquery expression is a subquery that returns exactly one column value
from one row. The value of the scalar subquery expression is the value of the select list
item of the subquery. If the subquery returns 0 rows, then the value of the scalar
subquery expression is NULL. If the subquery returns more than one row, then Oracle
returns an error.
You can use a scalar subquery expression in most syntax that calls for an expression
(expr). In all cases, a scalar subquery must be enclosed in its own parentheses, even if
its syntactic location already positions it within parentheses (for example, when the
scalar subquery is used as the argument to a built-in function).
Scalar subqueries are not valid expressions in the following places:
■

As default values for columns

■

As hash expressions for clusters

■

In the RETURNING clause of DML statements

■

As the basis of a function-based index

■

In CHECK constraints

■

In GROUP BY clauses

■

In statements that are unrelated to queries, such as CREATE PROFILE

Type Constructor Expressions
A type constructor expression specifies a call to a constructor method. The argument
to the type constructor is any expression. Type constructors can be invoked anywhere
functions are invoked.

6-14 Oracle Database SQL Language Reference

Type Constructor Expressions

type_constructor_expression::=
,
NEW

schema

.

expr
type_name

(

)

The NEW keyword applies to constructors for object types but not for collection types. It
instructs Oracle to construct a new object by invoking an appropriate constructor. The
use of the NEW keyword is optional, but it is good practice to specify it.
If type_name is an object type, then the expressions must be an ordered list, where the
first argument is a value whose type matches the first attribute of the object type, the
second argument is a value whose type matches the second attribute of the object type,
and so on. The total number of arguments to the constructor must match the total
number of attributes of the object type.
If type_name is a varray or nested table type, then the expression list can contain zero
or more arguments. Zero arguments implies construction of an empty collection.
Otherwise, each argument corresponds to an element value whose type is the element
type of the collection type.
Restriction on Type Constructor Invocation In an invocation of a type constructor
method, the number of parameters (expr) specified cannot exceed 999, even if the
object type has more than 999 attributes. This limitation applies only when the
constructor is called from SQL. For calls from PL/SQL, the PL/SQL limitations apply.
See Also: Oracle Database Object-Relational Developer's Guide for
additional information on constructor methods and Oracle Database
PL/SQL Language Reference for information on PL/SQL limitations on
calls to type constructors

This example uses the cust_address_typ type in the sample oe
schema to show the use of an expression in the call to a constructor method (the
PL/SQL is shown in italics):

Expression Example

CREATE TYPE address_book_t AS TABLE OF cust_address_typ;
DECLARE
myaddr cust_address_typ := cust_address_typ(
'500 Oracle Parkway', 94065, 'Redwood Shores', 'CA','USA');
alladdr address_book_t := address_book_t();
BEGIN
INSERT INTO customers VALUES (
666999, 'Joe', 'Smith', myaddr, NULL, NULL, NULL, NULL,
NULL, NULL, NULL, NULL, NULL, NULL, NULL);
END;
/

This example uses the warehouse_typ type in the sample schema
oe to illustrate the use of a subquery in the call to the constructor method.
Subquery Example

CREATE TABLE warehouse_tab OF warehouse_typ;
INSERT INTO warehouse_tab
VALUES (warehouse_typ(101, 'new_wh', 201));
CREATE TYPE facility_typ AS OBJECT (
facility_id NUMBER,
warehouse_ref REF warehouse_typ);

Expressions 6-15

Expression Lists

CREATE TABLE buildings (b_id NUMBER, building facility_typ);
INSERT INTO buildings VALUES (10, facility_typ(102,
(SELECT REF(w) FROM warehouse_tab w
WHERE warehouse_name = 'new_wh')));
SELECT b.b_id, b.building.facility_id "FAC_ID",
DEREF(b.building.warehouse_ref) "WH" FROM buildings b;
B_ID
FAC_ID WH(WAREHOUSE_ID, WAREHOUSE_NAME, LOCATION_ID)
---------- ---------- --------------------------------------------10
102 WAREHOUSE_TYP(101, 'new_wh', 201)

Expression Lists
An expression list is a combination of other expressions.
expression_list::=
,
expr
,
expr
(

)

Expression lists can appear in comparison and membership conditions and in GROUP BY
clauses of queries and subqueries. An expression lists in a comparision or membership
condition is sometimes referred to as a row value constructor or row constructor.
Comparison and membership conditions appear in the conditions of WHERE clauses.
They can contain either one or more comma-delimited expressions or one or more sets
of expressions where each set contains one or more comma-delimited expressions. In
the latter case (multiple sets of expressions):
■

Each set is bounded by parentheses

■

Each set must contain the same number of expressions

■

The number of expressions in each set must match the number of expressions
before the operator in the comparison condition or before the IN keyword in the
membership condition.

A comma-delimited list of expressions can contain no more than 1000 expressions. A
comma-delimited list of sets of expressions can contain any number of sets, but each
set can contain no more than 1000 expressions.
The following are some valid expression lists in conditions:
(10, 20, 40)
('SCOTT', 'BLAKE', 'TAYLOR')
( ('Guy', 'Himuro', 'GHIMURO'),('Karen', 'Colmenares', 'KCOLMENA') )

In the third example, the number of expressions in each set must equal the number of
expressions in the first part of the condition. For example:
SELECT * FROM employees
WHERE (first_name, last_name, email) IN
(('Guy', 'Himuro', 'GHIMURO'),('Karen', 'Colmenares', 'KCOLMENA'))

6-16 Oracle Database SQL Language Reference

Expression Lists

"Comparison Conditions" on page 7-4 and IN Condition
conditions on page 7-23

See Also:

In a simple GROUP BY clause, you can use either the upper or lower form of expression
list:
SELECT department_id, MIN(salary) min, MAX(salary) max FROM employees
GROUP BY department_id, salary
ORDER BY department_id, min, max;
SELECT department_id, MIN(salary) min, MAX(salary) max FROM employees
GROUP BY (department_id, salary)
ORDER BY department_id, min, max;

In ROLLUP, CUBE, and GROUPING SETS clauses of GROUP BY clauses, you can combine
individual expressions with sets of expressions in the same expression list. The
following example shows several valid grouping sets expression lists in one SQL
statement:
SELECT
prod_category, prod_subcategory, country_id, cust_city, count(*)
FROM products, sales, customers
WHERE sales.prod_id = products.prod_id
AND sales.cust_id=customers.cust_id
AND sales.time_id = '01-oct-00'
AND customers.cust_year_of_birth BETWEEN 1960 and 1970
GROUP BY GROUPING SETS
(
(prod_category, prod_subcategory, country_id, cust_city),
(prod_category, prod_subcategory, country_id),
(prod_category, prod_subcategory),
country_id
)
ORDER BY prod_category, prod_subcategory, country_id, cust_city;

See Also:

SELECT on page 19-4

Expressions 6-17

Expression Lists

6-18 Oracle Database SQL Language Reference

7
7

Conditions

A condition specifies a combination of one or more expressions and logical (Boolean)
operators and returns a value of TRUE, FALSE, or UNKNOWN.
This chapter contains the following sections:
■

About SQL Conditions

■

Comparison Conditions

■

Floating-Point Conditions

■

Logical Conditions

■

Model Conditions

■

Multiset Conditions

■

Pattern-matching Conditions

■

Null Conditions

■

XML Conditions

■

Compound Conditions

■

BETWEEN Condition

■

EXISTS Condition

■

IN Condition

■

IS OF type Condition

About SQL Conditions
Conditions can have several forms, as shown in the following syntax.

Conditions 7-1

About SQL Conditions

condition::=
comparison_condition
floating_point_condition
logical_condition
model_condition
multiset_condition
pattern_matching_condition
range_condition
null_condition
XML_condition
compound_condition
exists_condition
in_condition
is_of_type_condition

If you have installed Oracle Text, then you can create conditions with the built-in
operators that are part of that product, including CONTAINS, CATSEARCH, and MATCHES.
For more information on these Oracle Text elements, refer to Oracle Text Reference.
If you are using Oracle Expression Filter, then you can create conditions with the
built-in EVALUATE operator that is part of that product. For more information, refer to
Oracle Database Rules Manager and Expression Filter Developer's Guide.
The sections that follow describe the various forms of conditions. You must use
appropriate condition syntax whenever condition appears in SQL statements.
You can use a condition in the WHERE clause of these statements:
■

DELETE

■

SELECT

■

UPDATE

You can use a condition in any of these clauses of the SELECT statement:
■

WHERE

■

START WITH

■

CONNECT BY

■

HAVING

7-2 Oracle Database SQL Language Reference

About SQL Conditions

The combined values of the NLS_COMP and NLS_SORT settings
determine the rules by which characters are sorted and compared. If
NLS_COMP is set to LINGUISTIC for your database, then all entities in
this chapter will be interpreted according to the rules specified by the
NLS_SORT parameter. If NLS_COMP is not set to LINGUISTIC, then the
functions are interpreted without regard to the NLS_SORT setting. NLS_
SORT can be explicitly set. If it is not set explicitly, it is derived from
NLS_LANGUAGE. Refer to Oracle Database Globalization Support Guide for
more information on these settings.
Note:

A condition could be said to be of a logical data type, although Oracle Database does
not formally support such a data type.
The following simple condition always evaluates to TRUE:
1 = 1

The following more complex condition adds the salary value to the commission_pct
value (substituting the value 0 for null) and determines whether the sum is greater
than the number constant 25000:
NVL(salary, 0) + NVL(salary + (salary*commission_pct, 0) > 25000)

Logical conditions can combine multiple conditions into a single condition. For
example, you can use the AND condition to combine two conditions:
(1 = 1) AND (5 < 7)

Here are some valid conditions:
name = 'SMITH'
employees.department_id = departments.department_id
hire_date > '01-JAN-08'
job_id IN ('SA_MAN', 'SA_REP')
salary BETWEEN 5000 AND 10000
commission_pct IS NULL AND salary = 2100

See Also: The description of each statement in Chapter 10 through
Chapter 19 for the restrictions on the conditions in that statement

Condition Precedence
Precedence is the order in which Oracle Database evaluates different conditions in the
same expression. When evaluating an expression containing multiple conditions,
Oracle evaluates conditions with higher precedence before evaluating those with
lower precedence. Oracle evaluates conditions with equal precedence from left to right
within an expression, with the following exceptions:
■

■

Left to right evaluation is not guaranteed for multiple conditions connected using
AND
Left to right evaluation is not guaranteed for multiple conditions connected using
OR

Table 7–1 lists the levels of precedence among SQL condition from high to low.
Conditions listed on the same line have the same precedence. As the table indicates,
Oracle evaluates operators before conditions.

Conditions 7-3

Comparison Conditions

Table 7–1

SQL Condition Precedence

Type of Condition

Purpose

SQL operators are evaluated before SQL
conditions

See "Operator Precedence" on page 4-2

=, !=, <, >, <=, >=,

comparison

IS [NOT] NULL, LIKE, [NOT] BETWEEN, [NOT]
IN, EXISTS, IS OF type

comparison

NOT

exponentiation, logical negation

AND

conjunction

OR

disjunction

Comparison Conditions
Comparison conditions compare one expression with another. The result of such a
comparison can be TRUE, FALSE, or UNKNOWN.
Large objects (LOBs) are not supported in comparison conditions. However, you can
use PL/SQL programs for comparisons on CLOB data.
When comparing numeric expressions, Oracle uses numeric precedence to determine
whether the condition compares NUMBER, BINARY_FLOAT, or BINARY_DOUBLE values.
Refer to "Numeric Precedence" on page 3-14 for information on numeric precedence.
When comparing character expressions, Oracle uses the rules described in "Data Type
Comparison Rules" on page 3-36. The rules define how the character sets of the
expressions are aligned before the comparison, the use of binary or linguistic
comparison (collation), and the use of blank-padded comparison semantics.
When character values are compared linguistically using the comparison conditions,
they are first transformed to collation keys and then compared like RAW values. The
collation keys are the same values that are returned by the function NLSSORT and are
subject to the same restrictions that are described in "NLSSORT" on page 5-164. As a
result of these restrictions, two expressions may compare as linguistically equal if they
do not differ in the prefix that was used to produce the collation key, even if they differ
in the rest of the value.
Two objects of nonscalar type are comparable if they are of the same named type and
there is a one-to-one correspondence between their elements. In addition, nested tables
of user-defined object types, even if their elements are comparable, must have MAP
methods defined on them to be used in equality or IN conditions.
See Also: Oracle Database Object-Relational Developer's Guide for
information on using MAP methods to compare objects

Table 7–2 lists comparison conditions.

7-4 Oracle Database SQL Language Reference

Comparison Conditions

Table 7–2

Comparison Conditions

Type of
Condition

Purpose

Example

=

Equality test.

SELECT *
FROM employees
WHERE salary = 2500
ORDER BY employee_id;

!=

Inequality test. Some forms of the inequality
condition may be unavailable on some platforms.

SELECT *
FROM employees
WHERE salary != 2500
ORDER BY employee_id;

Greater-than and less-than tests.

SELECT * FROM employees
WHERE salary > 2500
ORDER BY employee_id;
SELECT * FROM employees
WHERE salary < 2500
ORDER BY employee_id;

Greater-than-or-equal-to and less-than-or-equal-to
tests.

SELECT * FROM employees
WHERE salary >= 2500
ORDER BY employee_id;
SELECT * FROM employees
WHERE salary <= 2500
ORDER BY employee_id;

Compares a value to each value in a list or returned
by a query. Must be preceded by =, !=, >, <, <=, >=.
Can be followed by any expression or subquery that
returns one or more values.

SELECT * FROM employees
WHERE salary = ANY
(SELECT salary
FROM employees
WHERE department_id = 30)
ORDER BY employee_id;

^=
<>
¬=
>

<

>=

<=

ANY

SOME

Evaluates to FALSE if the query returns no rows.
ALL

Compares a value to every value in a list or returned
by a query. Must be preceded by =, !=, >, <, <=, >=.
Can be followed by any expression or subquery that
returns one or more values.

SELECT * FROM employees
WHERE salary >=
ALL (1400, 3000)
ORDER BY employee_id;

Evaluates to TRUE if the query returns no rows.

Simple Comparison Conditions
A simple comparison condition specifies a comparison with expressions or subquery
results.

Conditions 7-5

Comparison Conditions

simple_comparison_condition::=
=
!=
^=
<>
expr

expr
>
<
>=
<=
=
,
!=

(

expr

)

expression_list
(

^=

)
subquery

<>

expression_list::=
,
expr
,
expr
(

)

If you use the lower form of this condition (with multiple expressions to the left of the
operator), then you must use the lower form of expression_list, and the expressions
in the expression_list must match in number and data type the expressions to the
left of the operator. If you specify subquery, then the values returned by the subquery
must match in number and data type the expressions to the left of the operator.
See Also: "Expression Lists" on page 6-16 for more information
about combining expressions and SELECT on page 19-4 for
information about subqueries

Group Comparison Conditions
A group comparison condition specifies a comparison with any or all members in a list
or subquery.

7-6 Oracle Database SQL Language Reference

Floating-Point Conditions

group_comparison_condition::=
=
!=
^=
ANY
<>

expression_list

expr

SOME

(

)

>

subquery
ALL

<
>=
<=
=

’

,

ANY
!=

(

expr

expression_list

)

SOME
^=

(

)
subquery

ALL
<>

expression_list::=
,
expr
,
expr
(

)

If you use the upper form of this condition (with a single expression to the left of the
operator), then you must use the upper form of expression_list. If you use the lower
form of this condition (with multiple expressions to the left of the operator), then you
must use the lower form of expression_list, and the expressions in each
expression_list must match in number and data type the expressions to the left of
the operator. If you specify subquery, then the values returned by the subquery must
match in number and data type the expressions to the left of the operator.
See Also:
■

"Expression Lists" on page 6-16

■

SELECT on page 19-4

Floating-Point Conditions
The floating-point conditions let you determine whether an expression is infinite or is
the undefined result of an operation (is not a number or NaN).
floating_point_condition::=
NOT
expr

NAN

IS
INFINITE

Conditions 7-7

Logical Conditions

In both forms of floating-point condition, expr must resolve to a numeric data type or
to any data type that can be implicitly converted to a numeric data type. Table 7–3
describes the floating-point conditions.
Table 7–3
Type of
Condition

Floating-Point Conditions
Operation

Example

IS [NOT]
NAN

Returns TRUE if expr is the special
SELECT COUNT(*) FROM employees
value NaN when NOT is not specified.
WHERE commission_pct IS NOT NAN;
Returns TRUE if expr is not the
special value NaN when NOT is
specified.

IS [NOT]
INFINITE

Returns TRUE if expr is the special
value +INF or -INF when NOT is not
specified. Returns TRUE if expr is
neither +INF nor -INF when NOT is
specified.

SELECT last_name FROM employees
WHERE salary IS NOT INFINITE;

See Also:
■

■

"Floating-Point Numbers" on page 3-12 for more information on
the Oracle implementation of floating-point numbers
"Implicit Data Conversion" on page 3-40 for more information on
how Oracle converts floating-point data types

Logical Conditions
A logical condition combines the results of two component conditions to produce a
single result based on them or to invert the result of a single condition. Table 7–4 lists
logical conditions.
Table 7–4
Type of
Condition

Logical Conditions
Operation

Examples

NOT

Returns TRUE if the following
condition is FALSE. Returns FALSE
if it is TRUE. If it is UNKNOWN, then it
remains UNKNOWN.

SELECT *
FROM employees
WHERE NOT (job_id IS NULL)
ORDER BY employee_id;
SELECT *
FROM employees
WHERE NOT
(salary BETWEEN 1000 AND 2000)
ORDER BY employee_id;

AND

Returns TRUE if both component
SELECT *
conditions are TRUE. Returns
FROM employees
FALSE if either is FALSE. Otherwise
WHERE job_id = 'PU_CLERK'
returns UNKNOWN.
AND department_id = 30
ORDER BY employee_id;

OR

Returns TRUE if either component SELECT *
condition is TRUE. Returns FALSE if
FROM employees
both are FALSE. Otherwise returns
WHERE job_id = 'PU_CLERK'
UNKNOWN.
OR department_id = 10
ORDER BY employee_id;

7-8 Oracle Database SQL Language Reference

Model Conditions

Table 7–5 shows the result of applying the NOT condition to an expression.
Table 7–5

NOT Truth Table

--

TRUE

FALSE

UNKNOWN

NOT

FALSE

TRUE

UNKNOWN

Table 7–6 shows the results of combining the AND condition to two expressions.
Table 7–6

AND Truth Table

AND

TRUE

FALSE

UNKNOWN

TRUE

TRUE

FALSE

UNKNOWN

FALSE

FALSE

FALSE

FALSE

UNKNOWN

UNKNOWN

FALSE

UNKNOWN

For example, in the WHERE clause of the following SELECT statement, the AND logical
condition is used to ensure that only those hired before 2004 and earning more than
$2500 a month are returned:
SELECT * FROM employees
WHERE hire_date < TO_DATE('01-JAN-2004', 'DD-MON-YYYY')
AND salary > 2500
ORDER BY employee_id;

Table 7–7 shows the results of applying OR to two expressions.
Table 7–7

OR Truth Table

OR

TRUE

FALSE

UNKNOWN

TRUE

TRUE

TRUE

TRUE

FALSE

TRUE

FALSE

UNKNOWN

UNKNOWN

TRUE

UNKNOWN

UNKNOWN

For example, the following query returns employees who have a 40% commission rate
or a salary greater than $20,000:
SELECT employee_id FROM employees
WHERE commission_pct = .4 OR salary > 20000
ORDER BY employee_id;

Model Conditions
Model conditions can be used only in the MODEL clause of a SELECT statement.

IS ANY Condition
The IS ANY condition can be used only in the model_clause of a SELECT statement. Use
this condition to qualify all values of a dimension column, including NULL.
is_any_condition::=
dimension_column

IS
ANY

Conditions 7-9

Model Conditions

The condition always returns a Boolean value of TRUE in order to qualify all values of
the column.
model_clause on page 19-28 and "Model Expressions" on
page 6-11 for information

See Also:

Example
The following example sets sales for each product for year 2000 to 0:
SELECT country, prod, year, s
FROM sales_view_ref
MODEL
PARTITION BY (country)
DIMENSION BY (prod, year)
MEASURES (sale s)
IGNORE NAV
UNIQUE DIMENSION
RULES UPSERT SEQUENTIAL ORDER
(
s[ANY, 2000] = 0
)
ORDER BY country, prod, year;
COUNTRY
---------France
France
France
France
France
France
France
France
Germany
Germany
Germany
Germany
Germany
Germany
Germany
Germany

PROD
----------------------------------Mouse Pad
Mouse Pad
Mouse Pad
Mouse Pad
Standard Mouse
Standard Mouse
Standard Mouse
Standard Mouse
Mouse Pad
Mouse Pad
Mouse Pad
Mouse Pad
Standard Mouse
Standard Mouse
Standard Mouse
Standard Mouse

YEAR
-------1998
1999
2000
2001
1998
1999
2000
2001
1998
1999
2000
2001
1998
1999
2000
2001

S
--------2509.42
3678.69
0
3269.09
2390.83
2280.45
0
2164.54
5827.87
8346.44
0
9535.08
7116.11
6263.14
0
6456.13

16 rows selected.

The preceding example requires the view sales_view_ref. Refer to "The MODEL
clause: Examples" on page 19-45 to create this view.

IS PRESENT Condition
is_present_condition::=
The IS PRESENT condition can be used only in the model_clause of a SELECT statement.
Use this condition to test whether the cell referenced is present prior to the execution
of the model_clause.
cell_reference

IS

PRESENT

7-10 Oracle Database SQL Language Reference

Multiset Conditions

The condition returns TRUE if the cell exists prior to the execution of the model_clause
and FALSE if it does not.
model_clause on page 19-28 and "Model Expressions" on
page 6-11 for information

See Also:

Example
In the following example, if sales of the Mouse Pad for year 1999 exist, then sales of the
Mouse Pad for year 2000 is set to sales of the Mouse Pad for year 1999. Otherwise,
sales of the Mouse Pad for year 2000 is set to 0.
SELECT country, prod, year, s
FROM sales_view_ref
MODEL
PARTITION BY (country)
DIMENSION BY (prod, year)
MEASURES (sale s)
IGNORE NAV
UNIQUE DIMENSION
RULES UPSERT SEQUENTIAL ORDER
(
s['Mouse Pad', 2000] =
CASE WHEN s['Mouse Pad', 1999] IS PRESENT
THEN s['Mouse Pad', 1999]
ELSE 0
END
)
ORDER BY country, prod, year;
COUNTRY
PROD
-------------------------------------------France
Mouse Pad
France
Mouse Pad
France
Mouse Pad
France
Mouse Pad
France
Standard Mouse
France
Standard Mouse
France
Standard Mouse
France
Standard Mouse
Germany
Mouse Pad
Germany
Mouse Pad
Germany
Mouse Pad
Germany
Mouse Pad
Germany
Standard Mouse
Germany
Standard Mouse
Germany
Standard Mouse
Germany
Standard Mouse
16 rows selected.

YEAR
-------1998
1999
2000
2001
1998
1999
2000
2001
1998
1999
2000
2001
1998
1999
2000
2001

S
--------2509.42
3678.69
3678.69
3269.09
2390.83
2280.45
1274.31
2164.54
5827.87
8346.44
8346.44
9535.08
7116.11
6263.14
2637.31
6456.13

The preceding example requires the view sales_view_ref. Refer to "The MODEL
clause: Examples" on page 19-45 to create this view.

Multiset Conditions
Multiset conditions test various aspects of nested tables.

Conditions 7-11

Multiset Conditions

IS A SET Condition
Use IS A SET conditions to test whether a specified nested table is composed of unique
elements. The condition returns NULL if the nested table is NULL. Otherwise, it returns
TRUE if the nested table is a set, even if it is a nested table of length zero, and FALSE
otherwise.
is_a_set_condition::=
NOT
nested_table

IS

A

SET

Example
The following example selects from the table customers_demo those rows in which the
cust_address_ntab nested table column contains unique elements:
SELECT customer_id, cust_address_ntab
FROM customers_demo
WHERE cust_address_ntab IS A SET
ORDER BY customer_id;
CUSTOMER_ID CUST_ADDRESS_NTAB(STREET_ADDRESS, POSTAL_CODE, CITY, STATE_PROVINCE, COUNTRY_ID)
---------------------------------------------------------------------------------------------101 CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP('514 W Superior St', '46901', 'Kokomo', 'IN', 'US'))
102 CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP('2515 Bloyd Ave', '46218', 'Indianapolis', 'IN', 'US'))
103 CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP('8768 N State Rd 37', '47404', 'Bloomington', 'IN', 'US'))
104 CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP('6445 Bay Harbor Ln', '46254', 'Indianapolis', 'IN', 'US'))
105 CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP('4019 W 3Rd St', '47404', 'Bloomington', 'IN', 'US'))

The preceding example requires the table customers_demo and a nested table column
containing data. Refer to "Multiset Operators" on page 4-6 to create this table and
nested table column.

IS EMPTY Condition
Use the IS [NOT] EMPTY conditions to test whether a specified nested table is empty. A
nested table that consists of a single value, a NULL, is not considered an empty nested
table.
is_empty_condition::=
NOT
nested_table

IS

EMPTY

The condition returns a Boolean value: TRUE for an IS EMPTY condition if the collection
is empty, and TRUE for an IS NOT EMPTY condition if the collection is not empty. If you
specify NULL for the nested table or varray, then the result is NULL.

Example
The following example selects from the sample table pm.print_media those rows in
which the ad_textdocs_ntab nested table column is not empty:
SELECT product_id, TO_CHAR(ad_finaltext) AS text
FROM print_media
WHERE ad_textdocs_ntab IS NOT EMPTY
ORDER BY product_id, text;

7-12 Oracle Database SQL Language Reference

Multiset Conditions

MEMBER Condition
member_condition::=
NOT
expr

OF
MEMBER

nested_table

A member_condition is a membership condition that tests whether an element is a
member of a nested table. The return value is TRUE if expr is equal to a member of the
specified nested table or varray. The return value is NULL if expr is null or if the nested
table is empty.
■

expr must be of the same type as the element type of the nested table.

■

The OF keyword is optional and does not change the behavior of the condition.

■

■

The NOT keyword reverses the Boolean output: Oracle returns FALSE if expr is a
member of the specified nested table.
The element types of the nested table must be comparable. Refer to "Comparison
Conditions" on page 7-4 for information on the comparability of nonscalar types.

Example
The following example selects from the table customers_demo those rows in which the
cust_address_ntab nested table column contains the values specified in the WHERE
clause:
SELECT customer_id, cust_address_ntab
FROM customers_demo
WHERE cust_address_typ('8768 N State Rd 37', 47404,
'Bloomington', 'IN', 'US')
MEMBER OF cust_address_ntab
ORDER BY customer_id;
CUSTOMER_ID CUST_ADDRESS_NTAB(STREET_ADDRESS, POSTAL_CODE, CITY, STATE_PROVINCE, COUNTRY_ID)
------------ --------------------------------------------------------------------------------103 CUST_ADDRESS_TAB_TYP(CUST_ADDRESS_TYP('8768 N State Rd 37', '47404', 'Bloomington', 'IN', 'US'))

The preceding example requires the table customers_demo and a nested table column
containing data. Refer to "Multiset Operators" on page 4-6 to create this table and
nested table column.

SUBMULTISET Condition
The SUBMULTISET condition tests whether a specified nested table is a submultiset of
another specified nested table.
The operator returns a Boolean value. TRUE is returned when nested_table1 is a
submultiset of nested_table2. nested_table1 is a submultiset of nested_table2
when one of the following conditions occur:
■

■

nested_table1 is not null and contains no rows. TRUE is returned even if nested_
table2 is null since an empty multiset is a submultiset of any non-null
replacement for nested_table2.
nested_table1 and nested_table2 are not null, nested_table1 does not contain a
null element, and there is a one-to-one mapping of each element in nested_table1
to an equal element in nested_table2.

NULL is returned when one of the following conditions occurs:

Conditions 7-13

Pattern-matching Conditions

■

nested_table1 is null.

■

nested_table2 is null, and nested_table1 is not null and not empty.

■

nested_table1 is a submultiset of nested_table2 after modifying each null
element of nested_table1 and nested_table2 to some non-null value, enabling a
one-to-one mapping of each element in nested_table1 to an equal element in
nested_table2.

If none of the above conditions occur, then FALSE is returned.
submultiset_condition::=
NOT
nested_table1
■
■

■

OF
SUBMULTISET

nested_table2

The OF keyword is optional and does not change the behavior of the operator.
The NOT keyword reverses the Boolean output: Oracle returns FALSE if nested_
table1 is a subset of nested_table2.
The element types of the nested table must be comparable. Refer to "Comparison
Conditions" on page 7-4 for information on the comparability of nonscalar types.

Example
The following example selects from the customers_demo table those rows in which the
cust_address_ntab nested table is a submultiset of the cust_address2_ntab nested
table:
SELECT customer_id, cust_address_ntab
FROM customers_demo
WHERE cust_address_ntab SUBMULTISET OF cust_address2_ntab
ORDER BY customer_id;

The preceding example requires the table customers_demo and two nested table
columns containing data. Refer to "Multiset Operators" on page 4-6 to create this table
and nested table columns.

Pattern-matching Conditions
The pattern-matching conditions compare character data.

LIKE Condition
The LIKE conditions specify a test involving pattern matching. Whereas the equality
operator (=) exactly matches one character value to another, the LIKE conditions match
a portion of one character value to another by searching the first value for the pattern
specified by the second. LIKE calculates strings using characters as defined by the
input character set. LIKEC uses Unicode complete characters. LIKE2 uses UCS2 code
points. LIKE4 uses UCS4 code points.

7-14 Oracle Database SQL Language Reference

Pattern-matching Conditions

like_condition::=
LIKE
NOT

ESCAPE

LIKEC

char1

esc_char

char2
LIKE2
LIKE4

In this syntax:
■

char1 is a character expression, such as a character column, called the search
value.

■

char2 is a character expression, usually a literal, called the pattern.

■

esc_char is a character expression, usually a literal, called the escape character.

The LIKE condition is the best choice in almost all situations. Use the following
guidelines to determine whether any of the variations would be helpful in your
environment:
■

■

■

Use LIKE2 to process strings using UCS-2 semantics. LIKE2 treats a Unicode
supplementary character as two characters.
Use LIKE4 to process strings using UCS-4 semantics. LIKE4 treats a Unicode
supplementary character as one character.
Use LIKEC to process strings using Unicode complete character semantics. LIKEC
treats a composite character as one character.

If esc_char is not specified, then there is no default escape character. If any of char1,
char2, or esc_char is null, then the result is unknown. Otherwise, the escape character,
if specified, must be a character string of length 1.
All of the character expressions (char1, char2, and esc_char) can be of any of the data
types CHAR, VARCHAR2, NCHAR, or NVARCHAR2. If they differ, then Oracle converts all of
them to the data type of char1.
The pattern can contain special pattern-matching characters:
■

■

An underscore (_) in the pattern matches exactly one character (as opposed to one
byte in a multibyte character set) in the value.
A percent sign (%) in the pattern can match zero or more characters (as opposed to
bytes in a multibyte character set) in the value. The pattern '%' cannot match a
null.

You can include the actual characters % or _ in the pattern by using the ESCAPE clause,
which identifies the escape character. If the escape character precedes the character %
or _ in the pattern, then Oracle interprets this character literally in the pattern rather
than as a special pattern-matching character. You can also search for the escape
character itself by repeating it. For example, if @ is the escape character, then you can
use @@ to search for @.

Only ASCII-equivalent underscore (_) and percent (%)
characters are recognized as pattern-matching characters. Their
full-width variants, present in East Asian character sets and in
Unicode, are treated as normal characters.

Note:

Conditions 7-15

Pattern-matching Conditions

Table 7–8 describes the LIKE conditions.
Table 7–8

LIKE Condition

Type of
Condition

Operation

x [NOT]
LIKE y
[ESCAPE
'z']

Example

TRUE if x does [not] match the pattern y. SELECT last_name
Within y, the character % matches any
FROM employees
string of zero or more characters except
WHERE last_name
null. The character _ matches any single
LIKE '%A\_B%' ESCAPE '\'
character. Any character can follow
ORDER BY last_name;
ESCAPE except percent (%) and underbar
(_). A wildcard character is treated as a
literal if preceded by the escape
character.

To process the LIKE conditions, Oracle divides the pattern into subpatterns consisting
of one or two characters each. The two-character subpatterns begin with the escape
character and the other character is %, or _, or the escape character.
Let P1, P2, ..., Pn be these subpatterns. The like condition is true if there is a way to
partition the search value into substrings S1, S2, ..., Sn so that for all i between 1 and n:
■

If Pi is _, then Si is a single character.

■

If Pi is %, then Si is any string.

■

■

If Pi is two characters beginning with an escape character, then Si is the second
character of Pi.
Otherwise, Pi = Si.

With the LIKE conditions, you can compare a value to a pattern rather than to a
constant. The pattern must appear after the LIKE keyword. For example, you can issue
the following query to find the salaries of all employees with names beginning with R:
SELECT salary
FROM employees
WHERE last_name LIKE 'R%'
ORDER BY salary;

The following query uses the = operator, rather than the LIKE condition, to find the
salaries of all employees with the name 'R%':
SELECT salary
FROM employees
WHERE last_name = 'R%'
ORDER BY salary;

The following query finds the salaries of all employees with the name 'SM%'. Oracle
interprets 'SM%' as a text literal, rather than as a pattern, because it precedes the LIKE
keyword:
SELECT salary
FROM employees
WHERE 'SM%' LIKE last_name
ORDER BY salary;

Case Sensitivity
Case is significant in all conditions comparing character expressions that use the LIKE
condition and the equality (=) operators. You can perform case or accent insensitive
LIKE searches by setting the NLS_SORT and the NLS_COMP session parameters.
7-16 Oracle Database SQL Language Reference

Pattern-matching Conditions

See Also: Oracle Database Globalization Support Guide for more
information on this case- and accent-insensitive linguistic sorts

Pattern Matching on Indexed Columns
When you use LIKE to search an indexed column for a pattern, Oracle can use the
index to improve performance of a query if the leading character in the pattern is not %
or _. In this case, Oracle can scan the index by this leading character. If the first
character in the pattern is % or _, then the index cannot improve performance because
Oracle cannot scan the index.
LIKE Condition: General Examples
This condition is true for all last_name values beginning with Ma:
last_name LIKE 'Ma%'

All of these last_name values make the condition true:
Mallin, Markle, Marlow, Marvins, Mavris, Matos

Case is significant, so last_name values beginning with MA, ma, and mA make the
condition false.
Consider this condition:
last_name LIKE 'SMITH_'

This condition is true for these last_name values:
SMITHE, SMITHY, SMITHS

This condition is false for SMITH because the special underscore character (_) must
match exactly one character of the last_name value.
ESCAPE Clause Example The following example searches for employees with the
pattern A_B in their name:
SELECT last_name
FROM employees
WHERE last_name LIKE '%A\_B%' ESCAPE '\'
ORDER BY last_name;

The ESCAPE clause identifies the backslash (\) as the escape character. In the pattern,
the escape character precedes the underscore (_). This causes Oracle to interpret the
underscore literally, rather than as a special pattern matching character.
If a pattern does not contain the % character, then the
condition can be true only if both operands have the same length. Consider the
definition of this table and the values inserted into it:

Patterns Without % Example

CREATE TABLE ducks (f CHAR(6), v VARCHAR2(6));
INSERT INTO ducks VALUES ('DUCK', 'DUCK');
SELECT '*'||f||'*' "char",
'*'||v||'*' "varchar"
FROM ducks;
char
varchar
-------- -------*DUCK * *DUCK*

Conditions 7-17

Pattern-matching Conditions

Because Oracle blank-pads CHAR values, the value of f is blank-padded to 6 bytes. v is
not blank-padded and has length 4.

REGEXP_LIKE Condition
REGEXP_LIKE is similar to the LIKE condition, except REGEXP_LIKE performs regular
expression matching instead of the simple pattern matching performed by LIKE. This
condition evaluates strings using characters as defined by the input character set.
This condition complies with the POSIX regular expression standard and the Unicode
Regular Expression Guidelines. For more information, refer to Appendix D, "Oracle
Regular Expression Support".
regexp_like_condition::=
,
REGEXP_LIKE

■

■

■

(

source_char

,

match_param

pattern

)

source_char is a character expression that serves as the search value. It is
commonly a character column and can be of any of the data types CHAR, VARCHAR2,
NCHAR, NVARCHAR2, CLOB, or NCLOB.
pattern is the regular expression. It is usually a text literal and can be of any of
the data types CHAR, VARCHAR2, NCHAR, or NVARCHAR2. It can contain up to 512 bytes.
If the data type of pattern is different from the data type of source_char, Oracle
converts pattern to the data type of source_char. For a listing of the operators
you can specify in pattern, refer to Appendix D, "Oracle Regular Expression
Support".
match_parameter is a text literal that lets you change the default matching
behavior of the function. You can specify one or more of the following values for
match_parameter:
–

'i' specifies case-insensitive matching.

–

'c' specifies case-sensitive matching.

–

'n' allows the period (.), which is the match-any-character wildcard character,
to match the newline character. If you omit this parameter, then the period
does not match the newline character.

–

'm' treats the source string as multiple lines. Oracle interprets ^ and $ as the
start and end, respectively, of any line anywhere in the source string, rather
than only at the start or end of the entire source string. If you omit this
parameter, then Oracle treats the source string as a single line.

–

'x' ignores whitespace characters. By default, whitespace characters match
themselves.

If you specify multiple contradictory values, then Oracle uses the last value. For
example, if you specify 'ic', then Oracle uses case-sensitive matching. If you
specify a character other than those shown above, then Oracle returns an error.
If you omit match_parameter, then:
–

The default case sensitivity is determined by the value of the NLS_SORT
parameter.

–

A period (.) does not match the newline character.

–

The source string is treated as a single line.

7-18 Oracle Database SQL Language Reference

Null Conditions

See Also:

"LIKE Condition" on page 7-14

■

REGEXP_INSTR on page 5-218, REGEXP_REPLACE on
page 5-221, and REGEXP_SUBSTR on page 5-224 for functions
that provide regular expression support

■

Examples
The following query returns the first and last names for those employees with a first
name of Steven or Stephen (where first_name begins with Ste and ends with en and
in between is either v or ph):
SELECT first_name, last_name
FROM employees
WHERE REGEXP_LIKE (first_name, '^Ste(v|ph)en$')
ORDER BY first_name, last_name;
FIRST_NAME
-------------------Steven
Steven
Stephen

LAST_NAME
------------------------King
Markle
Stiles

The following query returns the last name for those employees with a double vowel in
their last name (where last_name contains two adjacent occurrences of either a, e, i, o,
or u, regardless of case):
SELECT last_name
FROM employees
WHERE REGEXP_LIKE (last_name, '([aeiou])\1', 'i')
ORDER BY last_name;
LAST_NAME
------------------------De Haan
Greenberg
Khoo
Gee
Greene
Lee
Bloom
Feeney

Null Conditions
A NULL condition tests for nulls. This is the only condition that you should use to test
for nulls.
null_condition::=
NOT
expr

IS

NULL

Table 7–9 lists the null conditions.

Conditions 7-19

XML Conditions

Table 7–9

Null Condition

Type of
Condition

Operation

Example

Tests for nulls.

SELECT last_name
FROM employees
WHERE commission_pct
IS NULL
ORDER BY last_name;

IS [NOT]
NULL

See Also: "Nulls" on page 3-71

XML Conditions
XML conditions determine whether a specified XML resource can be found in a
specified path.

EQUALS_PATH Condition
The EQUALS_PATH condition determines whether a resource in the Oracle XML
database can be found in the database at a specified path.
Use this condition in queries to RESOURCE_VIEW and PATH_VIEW. These public views
provide a mechanism for SQL access to data stored in the XML database repository.
RESOURCE_VIEW contains one row for each resource in the repository, and PATH_VIEW
contains one row for each unique path in the repository.
equals_path_condition::=
,
EQUALS_PATH

(

column

,

correlation_integer

path_string

)

This condition applies only to the path as specified. It is similar to but more restrictive
than UNDER_PATH.
For path_string, specify the (absolute) path name to resolve. This can contain
components that are hard or weak resource links.
The optional correlation_integer argument correlates the EQUALS_PATH condition
with its ancillary functions DEPTH and PATH.
UNDER_PATH Condition on page 7-21, DEPTH on
page 5-84, and PATH on page 5-179

See Also:

Example
The view RESOURCE_VIEW computes the paths (in the any_path column) that lead to all
XML resources (in the res column) in the database repository. The following example
queries the RESOURCE_VIEW view to find the paths to the resources in the sample
schema oe. The EQUALS_PATH condition causes the query to return only the specified
path:
SELECT ANY_PATH FROM RESOURCE_VIEW
WHERE EQUALS_PATH(res, '/sys/schemas/OE/www.example.com')=1;
ANY_PATH
----------------------------------------------/sys/schemas/OE/www.example.com

Compare this example with that for UNDER_PATH Condition on page 7-21.

7-20 Oracle Database SQL Language Reference

Compound Conditions

UNDER_PATH Condition
The UNDER_PATH condition determines whether resources specified in a column can be
found under a particular path specified by path_string in the Oracle XML database
repository. The path information is computed by the RESOURCE_VIEW view, which you
query to use this condition.
Use this condition in queries to RESOURCE_VIEW and PATH_VIEW. These public views
provide a mechanism for SQL access to data stored in the XML database repository.
RESOURCE_VIEW contains one row for each resource in the repository, and PATH_VIEW
contains one row for each unique path in the repository.
under_path_condition::=
,
UNDER_PATH

(

levels

,

column

,

correlation_integer

path_string

)

The optional levels argument indicates the number of levels down from path_string
Oracle should search. For levels, specify any nonnegative integer.
The optional correlation_integer argument correlates the UNDER_PATH condition
with its ancillary functions PATH and DEPTH.
The related condition EQUALS_PATH Condition on
page 7-20 and the ancillary functions DEPTH on page 5-84 and PATH
on page 5-179

See Also:

Example
The view RESOURCE_VIEW computes the paths (in the any_path column) that lead to all
XML resources (in the res column) in the database repository. The following example
queries the RESOURCE_VIEW view to find the paths to the resources in the sample
schema oe. The query returns the path of the XML schema that was created in
"XMLType Table Examples" on page 16-75:
SELECT ANY_PATH FROM RESOURCE_VIEW
WHERE UNDER_PATH(res, '/sys/schemas/OE/www.example.com')=1;
ANY_PATH
---------------------------------------------/sys/schemas/OE/www.example.com/xwarehouses.xsd

Compound Conditions
A compound condition specifies a combination of other conditions.
compound_condition::=
(

condition

NOT

)

condition
AND

condition

condition
OR

See Also: "Logical Conditions" on page 7-8 for more information
about NOT, AND, and OR conditions

Conditions 7-21

BETWEEN Condition

BETWEEN Condition
A BETWEEN condition determines whether the value of one expression is in an interval
defined by two other expressions.
between_condition::=
NOT
expr1

BETWEEN

expr2

AND

expr3

All three expressions must be numeric, character, or datetime expressions. In SQL, it is
possible that expr1 will be evaluated more than once. If the BETWEEN expression
appears in PL/SQL, expr1 is guaranteed to be evaluated only once. If the expressions
are not all the same data type, then Oracle Database implicitly converts the
expressions to a common data type. If it cannot do so, then it returns an error.
"Implicit Data Conversion" on page 3-40 for more
information on SQL data type conversion

See Also:

The value of
expr1 NOT BETWEEN expr2 AND expr3

is the value of the expression
NOT (expr1 BETWEEN expr2 AND expr3)

And the value of
expr1 BETWEEN expr2 AND expr3

is the value of the boolean expression:
expr2 <= expr1 AND expr1 <= expr3

If expr3 < expr2, then the interval is empty. If expr1 is NULL, then the result is NULL. If
expr1 is not NULL, then the value is FALSE in the ordinary case and TRUE when the
keyword NOT is used.
The boolean operator AND may produce unexpected results. Specifically, in the
expression x AND y, the condition x IS NULL is not sufficient to determine the value of
the expression. The second operand still must be evaluated. The result is FALSE if the
second operand has the value FALSE and NULL otherwise. See "Logical Conditions" on
page 7-8 for more information on AND.
Table 7–10

BETWEEN Condition

Type of
Condition

Operation

Example

[NOT]
BETWEEN x
AND y

[NOT] (expr2 less than or equal to
expr1 AND expr1 less than or
equal to expr3)

SELECT * FROM employees
WHERE salary
BETWEEN 2000 AND 3000
ORDER BY employee_id;

EXISTS Condition
An EXISTS condition tests for existence of rows in a subquery.

7-22 Oracle Database SQL Language Reference

IN Condition

EXISTS

(

subquery

)

Table 7–11 shows the EXISTS condition.
Table 7–11

EXISTS Condition

Type of
Condition
EXISTS

Operation

Example

TRUE if a subquery returns at
least one row.

SELECT department_id
FROM departments d
WHERE EXISTS
(SELECT * FROM employees e
WHERE d.department_id
= e.department_id)
ORDER BY department_id;

IN Condition
An in_condition is a membership condition. It tests a value for membership in a list
of values or subquery
in_condition::=
NOT

expression_list

expr

IN

(

)
subquery
,

,
(

NOT

expr

)

expression_list
IN

(

)
subquery

expression_list::=
,
expr
,
expr
(

)

If you use the upper form of the in_condition condition (with a single expression to
the left of the operator), then you must use the upper form of expression_list. If you
use the lower form of this condition (with multiple expressions to the left of the
operator), then you must use the lower form of expression_list, and the expressions
in each expression_list must match in number and data type the expressions to the
left of the operator. You can specify up to 1000 expressions in expression_list.
Oracle Database does not always evaluate the expressions in an expression_list in
the order in which they appear in the IN list. However, expressions in the select list of a
subquery are evaluated in their specified order.
See Also:

"Expression Lists" on page 6-16

Table 7–12 lists the form of IN condition.

Conditions 7-23

IN Condition

Table 7–12

IN Condition

Type of Condition

Operation

Example

IN

Equal-to-any-member-of test.
Equivalent to =ANY.

SELECT * FROM employees
WHERE job_id IN
('PU_CLERK','SH_CLERK')
ORDER BY employee_id;
SELECT * FROM employees
WHERE salary IN
(SELECT salary
FROM employees
WHERE department_id =30)
ORDER BY employee_id;

NOT IN

Equivalent to !=ALL. Evaluates
to FALSE if any member of the
set is NULL.

SELECT * FROM employees
WHERE salary NOT IN
(SELECT salary
FROM employees
WHERE department_id = 30)
ORDER BY employee_id;
SELECT * FROM employees
WHERE job_id NOT IN
('PU_CLERK', 'SH_CLERK')
ORDER BY employee_id;

If any item in the list following a NOT IN operation evaluates to null, then all rows
evaluate to FALSE or UNKNOWN, and no rows are returned. For example, the following
statement returns the string 'True' for each row:
SELECT 'True' FROM employees
WHERE department_id NOT IN (10, 20);

However, the following statement returns no rows:
SELECT 'True' FROM employees
WHERE department_id NOT IN (10, 20, NULL);

The preceding example returns no rows because the WHERE clause condition evaluates
to:
department_id != 10 AND department_id != 20 AND department_id != null

Because the third condition compares department_id with a null, it results in an
UNKNOWN, so the entire expression results in FALSE (for rows with department_id equal
to 10 or 20). This behavior can easily be overlooked, especially when the NOT IN
operator references a subquery.
Moreover, if a NOT IN condition references a subquery that returns no rows at all, then
all rows will be returned, as shown in the following example:
SELECT 'True' FROM employees
WHERE department_id NOT IN (SELECT 0 FROM DUAL WHERE 1=2);

Restriction on LEVEL in WHERE Clauses In a [NOT] IN condition in a WHERE clause, if
the right-hand side of the condition is a subquery, you cannot use LEVEL on the
left-hand side of the condition. However, you can specify LEVEL in a subquery of the
FROM clause to achieve the same result. For example, the following statement is not
valid:
SELECT employee_id, last_name FROM employees
WHERE (employee_id, LEVEL)

7-24 Oracle Database SQL Language Reference

IS OF type Condition

IN (SELECT employee_id, 2 FROM employees)
START WITH employee_id = 2
CONNECT BY PRIOR employee_id = manager_id;

But the following statement is valid because it encapsulates the query containing the
LEVEL information in the FROM clause:
SELECT v.employee_id, v.last_name, v.lev FROM
(SELECT employee_id, last_name, LEVEL lev
FROM employees v
START WITH employee_id = 100
CONNECT BY PRIOR employee_id = manager_id) v
WHERE (v.employee_id, v.lev) IN
(SELECT employee_id, 2 FROM employees);

IS OF type Condition
Use the IS OF type condition to test object instances based on their specific type
information.
is_of_type_condition::=
,
NOT
expr

IS

TYPE
OF

ONLY
(

schema

.
type

)

You must have EXECUTE privilege on all types referenced by type, and all types must
belong to the same type family.
This condition evaluates to null if expr is null. If expr is not null, then the condition
evaluates to true (or false if you specify the NOT keyword) under either of these
circumstances:
■

■

The most specific type of expr is the subtype of one of the types specified in the
type list and you have not specified ONLY for the type, or
The most specific type of expr is explicitly specified in the type list.

The expr frequently takes the form of the VALUE function with a correlation variable.
The following example uses the sample table oe.persons, which is built on a type
hierarchy in "Substitutable Table and Column Examples" on page 16-71. The example
uses the IS OF type condition to restrict the query to specific subtypes:
SELECT * FROM persons p
WHERE VALUE(p) IS OF TYPE (employee_t);
NAME
SSN
---------------------------Joe
32456
Tim
5678
SELECT * FROM persons p
WHERE VALUE(p) IS OF (ONLY part_time_emp_t);
NAME
SSN
---------------------------Tim
5678

Conditions 7-25

IS OF type Condition

7-26 Oracle Database SQL Language Reference

8
8

Common SQL DDL Clauses

This chapter describes some SQL data definition clauses that appear in multiple SQL
statements.
This chapter contains these sections:
■

allocate_extent_clause

■

constraint

■

deallocate_unused_clause

■

file_specification

■

logging_clause

■

parallel_clause

■

physical_attributes_clause

■

size_clause

■

storage_clause

Common SQL DDL Clauses

8-1

allocate_extent_clause

allocate_extent_clause
Purpose
8

Use the allocate_extent_clause clause to explicitly allocate a new extent for a
database object.
Explicitly allocating an extent with this clause does not change the values of the NEXT
and PCTINCREASE storage parameters, so does not affect the size of the next extent to be
allocated implicitly by Oracle Database. Refer to storage_clause on page 8-48 for
information about the NEXT and PCTINCREASE storage parameters.
You can allocate an extent in the following SQL statements:
■
■

■

■

■

ALTER CLUSTER (see ALTER CLUSTER on page 10-5)
ALTER INDEX: to allocate an extent to the index, an index partition, or an index
subpartition (see ALTER INDEX on page 10-78)
ALTER MATERIALIZED VIEW: to allocate an extent to the materialized view, one of its
partitions or subpartitions, or the overflow segment of an index-organized
materialized view (see ALTER MATERIALIZED VIEW on page 11-3)
ALTER MATERIALIZED VIEW LOG (see ALTER MATERIALIZED VIEW LOG on
page 11-18)
ALTER TABLE: to allocate an extent to the table, a table partition, a table
subpartition, the mapping table of an index-organized table, the overflow segment
of an index-organized table, or a LOB storage segment (see ALTER TABLE on
page 12-2)

Syntax
8

allocate_extent_clause::=
SIZE
(

size_clause

DATAFILE
INSTANCE

ALLOCATE

’

filename

’

)

integer

EXTENT

(size_clause::= on page 8-47)

Semantics
8

This section describes the parameters of the allocate_extent_clause. For additional
information, refer to the SQL statement in which you set or reset these parameters for
a particular database object.
You cannot specify the allocate_extent_clause and the deallocate_unused_clause
in the same statement.

SIZE
Specify the size of the extent in bytes. The value of integer can be 0 through
2147483647. To specify a larger extent size, use an integer within this range with K, M, G,
or T to specify the extent size in kilobytes, megabytes, gigabytes, or terabytes.

8-2 Oracle Database SQL Language Reference

allocate_extent_clause

For a table, index, materialized view, or materialized view log, if you omit SIZE, then
Oracle Database determines the size based on the values of the storage parameters of
the object. However, for a cluster, Oracle does not evaluate the cluster's storage
parameters, so you must specify SIZE if you do not want Oracle to use a default value.

DATAFILE 'filename'
Specify one of the data files in the tablespace of the table, cluster, index, materialized
view, or materialized view log to contain the new extent. If you omit DATAFILE, then
Oracle chooses the data file.

INSTANCE integer
Use this parameter only if you are using Oracle Real Application Clusters.
Specifying INSTANCE integer makes the new extent available to the freelist group
associated with the specified instance. If the instance number exceeds the maximum
number of freelist groups, then Oracle divides the specified number by the maximum
number and uses the remainder to identify the freelist group to be used. An instance is
identified by the value of its initialization parameter INSTANCE_NUMBER.
If you omit this parameter, then the space is allocated to the table, cluster, index,
materialized view, or materialized view log but is not drawn from any particular
freelist group. Instead, Oracle uses the master freelist and allocates space as needed.
If you are using automatic segment-space management, then
the INSTANCE parameter of the allocate_extent_clause may not
reserve the newly allocated space for the specified instance, because
automatic segment-space management does not maintain rigid
affinity between extents and instances.
Note:

Common SQL DDL Clauses

8-3

constraint

constraint
Purpose
8

Use a constraint to define an integrity constraint—a rule that restricts the values in a
database. Oracle Database lets you create six types of constraints and lets you declare
them in two ways.
The six types of integrity constraint are described briefly here and more fully in
"Semantics" on page 8-8:
■
■

■

■

■

■

A NOT NULL constraint prohibits a database value from being null.
A unique constraint prohibits multiple rows from having the same value in the
same column or combination of columns but allows some values to be null.
A primary key constraint combines a NOT NULL constraint and a unique constraint
in a single declaration. It prohibits multiple rows from having the same value in
the same column or combination of columns and prohibits values from being null.
A foreign key constraint requires values in one table to match values in another
table.
A check constraint requires a value in the database to comply with a specified
condition.
A REF column by definition references an object in another object type or in a
relational table. A REF constraint lets you further describe the relationship
between the REF column and the object it references.

You can define constraints syntactically in two ways:
■

■

As part of the definition of an individual column or attribute. This is called inline
specification.
As part of the table definition. This is called out-of-line specification.

NOT NULL constraints must be declared inline. All other constraints can be declared
either inline or out of line.
Constraint clauses can appear in the following statements:
■

CREATE TABLE (see CREATE TABLE on page 16-6)

■

ALTER TABLE (see ALTER TABLE on page 12-2)

■

CREATE VIEW (see CREATE VIEW on page 17-14)

■

ALTER VIEW (see ALTER VIEW on page 13-14)

View Constraints Oracle Database does not enforce view constraints. However, you
can enforce constraints on views through constraints on base tables.

You can specify only unique, primary key, and foreign key constraints on views, and
they are supported only in DISABLE NOVALIDATE mode. You cannot define view
constraints on attributes of an object column.
See Also: "View Constraints" on page 8-19 for additional
information on view constraints and "DISABLE Clause" on page 8-16
for information on DISABLE NOVALIDATE mode

8-4 Oracle Database SQL Language Reference

constraint

Prerequisites
8

You must have the privileges necessary to issue the statement in which you are
defining the constraint.
To create a foreign key constraint, in addition, the parent table or view must be in your
own schema or you must have the REFERENCES privilege on the columns of the
referenced key in the parent table or view.

Syntax
8

constraint::=
inline_constraint
out_of_line_constraint
inline_ref_constraint
out_of_line_ref_constraint

(inline_constraint::= on page 8-5, out_of_line_constraint::= on page 8-5, inline_ref_
constraint::= on page 8-6, out_of_line_ref_constraint::= on page 8-6)
inline_constraint::=
NOT
NULL
CONSTRAINT

constraint_name

UNIQUE

constraint_state

PRIMARY

KEY

references_clause
CHECK

(

condition

)

(references_clause::= on page 8-6)
out_of_line_constraint::=
,
UNIQUE

(

column

)
,

CONSTRAINT

constraint_name

PRIMARY

KEY

(

column

constraint_state

)

,
FOREIGN
CHECK

KEY
(

(

condition

column

)

references_clause

)

(references_clause::= on page 8-6, constraint_state::= on page 8-6)

Common SQL DDL Clauses

8-5

constraint

inline_ref_constraint::=
schema
SCOPE
WITH

.

IS

scope_table

ROWID

CONSTRAINT

constraint_name

constraint_state
references_clause

(references_clause::= on page 8-6, constraint_state::= on page 8-6)
out_of_line_ref_constraint::=
schema

ref_col
SCOPE

FOR

(

)

.

IS

scope_table

ref_attr
ref_col
REF

(

)

WITH

ROWID

ref_attr
,
CONSTRAINT

ref_col

constraint_name
FOREIGN

KEY

(

constraint_state

,

)

references_clause

ref_attr

(references_clause::= on page 8-6, constraint_state::= on page 8-6)
references_clause::=
,

CASCADE
ON

schema

.

REFERENCES

(

column

DELETE

)

object

constraint_state::=
NOT
DEFERRABLE
IMMEDIATE
INITIALLY
DEFERRED
RELY
NORELY
using_index_clause
ENABLE
DISABLE
VALIDATE
NOVALIDATE
exceptions_clause

(using_index_clause::= on page 8-7, exceptions_clause::= on page 8-8)

8-6 Oracle Database SQL Language Reference

SET

NULL

constraint

using_index_clause::=
schema

.
index

USING

INDEX

(

create_index_statement

)

index_properties

(create_index::= on page 14-61, index_properties::= on page 8-7)
index_properties::=
global_partitioned_index
local_partitioned_index
index_attributes
domain_index_clause
INDEXTYPE

IS
XMLIndex_clause

(global_partitioned_index::= on page 14-64, local_partitioned_index::= on page 14-65--part
of CREATE INDEX, index_attributes::= on page 8-7. The INDEXTYPE IS ... clause is not valid
when defining a constraint.)
index_attributes::=
physical_attributes_clause
logging_clause
ONLINE
tablespace
TABLESPACE
DEFAULT
key_compression
SORT
NOSORT
REVERSE
VISIBLE
INVISIBLE
parallel_clause

(physical_attributes_clause::= on page 14-3, logging_clause::= on page 8-38, key_
compression::= on page 14-63--all part of CREATE INDEX, parallel_clause: not
supported in using_index_clause)

Common SQL DDL Clauses

8-7

constraint

exceptions_clause::=
schema
EXCEPTIONS

.

INTO

table

Semantics
8

This section describes the semantics of constraint. For additional information, refer to
the SQL statement in which you define or redefine a constraint for a table or view.
Oracle Database does not support constraints on columns or attributes whose type is a
user-defined object, nested table, VARRAY, REF, or LOB, with two exceptions:
■

■

NOT NULL constraints are supported for a column or attribute whose type is
user-defined object, VARRAY, REF, or LOB.
NOT NULL, foreign key, and REF constraints are supported on a column of type REF.

Specify a name for the constraint. If you omit this
identifier, then Oracle Database generates a name with the form SYS_Cn. Oracle stores
the name and the definition of the integrity constraint in the USER_, ALL_, and DBA_
CONSTRAINTS data dictionary views (in the CONSTRAINT_NAME and SEARCH_CONDITION
columns, respectively).
CONSTRAINT constraint_name

Oracle Database Reference for information on the data
dictionary views

See Also:

NOT NULL Constraints
A NOT NULL constraint prohibits a column from containing nulls. The NULL keyword by
itself does not actually define an integrity constraint, but you can specify it to explicitly
permit a column to contain nulls. You must define NOT NULL and NULL using inline
specification. If you specify neither NOT NULL nor NULL, then the default is NULL.
NOT NULL constraints are the only constraints you can specify inline on XMLType and
VARRAY columns.
To satisfy a NOT NULL constraint, every row in the table must contain a value for the
column.
Oracle Database does not index table rows in which all key
columns are null except in the case of bitmap indexes. Therefore, if
you want an index on all rows of a table, then you must either specify
NOT NULL constraints for at least one of the index key columns or create
a bitmap index.
Note:

Restrictions on NOT NULL Constraints NOT NULL constraints are subject to the
following restrictions:
■
■

You cannot specify NULL or NOT NULL in a view constraint.
You cannot specify NULL or NOT NULL for an attribute of an object. Instead, use a
CHECK constraint with the IS [NOT] NULL condition.
See Also: "Attribute-Level Constraints Example" on page 8-25 and
"NOT NULL Example" on page 8-21

8-8 Oracle Database SQL Language Reference

constraint

Unique Constraints
A unique constraint designates a column as a unique key. A composite unique key
designates a combination of columns as the unique key. When you define a unique
constraint inline, you need only the UNIQUE keyword. When you define a unique
constraint out of line, you must also specify one or more columns. You must define a
composite unique key out of line.
To satisfy a unique constraint, no two rows in the table can have the same value for the
unique key. However, the unique key made up of a single column can contain nulls. To
satisfy a composite unique key, no two rows in the table or view can have the same
combination of values in the key columns. Any row that contains nulls in all key
columns automatically satisfies the constraint. However, two rows that contain nulls
for one or more key columns and the same combination of values for the other key
columns violate the constraint.
When you specify a unique constraint on one or more columns, Oracle implicitly
creates an index on the unique key. If you are defining uniqueness for purposes of
query performance, then Oracle recommends that you instead create the unique index
explicitly using a CREATE UNIQUE INDEX statement. You can also use the CREATE UNIQUE
INDEX statement to create a unique function-based index that defines a conditional
unique constraint. See "Using a Function-based Index to Define Conditional
Uniqueness: Example" on page 14-82 for more information.
Restrictions on Unique Constraints Unique constraints are subject to the following
restrictions:
■

■
■

■

None of the columns in the unique key can be of LOB, LONG, LONG RAW, VARRAY,
NESTED TABLE, OBJECT, REF, TIMESTAMP WITH TIME ZONE, or user-defined type.
However, the unique key can contain a column of TIMESTAMP WITH LOCAL TIME
ZONE.
A composite unique key cannot have more than 32 columns.
You cannot designate the same column or combination of columns as both a
primary key and a unique key.
You cannot specify a unique key when creating a subview in an inheritance
hierarchy. The unique key can be specified only for the top-level (root) view.
See Also: "Unique Key Example" on page 8-19 and Composite
Unique Key Example on page 8-20

Primary Key Constraints
A primary key constraint designates a column as the primary key of a table or view. A
composite primary key designates a combination of columns as the primary key.
When you define a primary key constraint inline, you need only the PRIMARY KEY
keywords. When you define a primary key constraint out of line, you must also
specify one or more columns. You must define a composite primary key out of line.
To satisfy a primary key constraint:
■

No primary key value can appear in more than one row in the table.

■

No column that is part of the primary key can contain a null.

When you create a primary key constraint:
■

Oracle Database uses an existing index if it contains a unique set of values before
enforcing the primary key constraint. The existing index can be defined as unique

Common SQL DDL Clauses

8-9

constraint

or nonunique. When a DML operation is performed, the primary key constraint is
enforced using this existing index.
■

If no existing index can be used, then Oracle Database generates a unique index.

When you drop a primary key constraint:
■

■

If the primary key was created using an existing index, then the index is not
dropped.
If the primary key was created using a system-generated index, then the index is
dropped.

Restrictions on Primary Key Constraints Primary constraints are subject to the

following restrictions:
■
■

A table or view can have only one primary key.
None of the columns in the primary key can be LOB, LONG, LONG RAW, VARRAY,
NESTED TABLE, BFILE, REF, TIMESTAMP WITH TIME ZONE, or user-defined type.
However, the primary key can contain a column of TIMESTAMP WITH LOCAL TIME
ZONE.

■

The size of the primary key cannot exceed approximately one database block.

■

A composite primary key cannot have more than 32 columns.

■

■

You cannot designate the same column or combination of columns as both a
primary key and a unique key.
You cannot specify a primary key when creating a subview in an inheritance
hierarchy. The primary key can be specified only for the top-level (root) view.
"Primary Key Example" on page 8-20 and "Composite
Primary Key Example" on page 8-21

See Also:

Foreign Key Constraints
A foreign key constraint (also called a referential integrity constraint) designates a
column as the foreign key and establishes a relationship between that foreign key and
a specified primary or unique key, called the referenced key. A composite foreign key
designates a combination of columns as the foreign key.
The table or view containing the foreign key is called the child object, and the table or
view containing the referenced key is called the parent object. The foreign key and the
referenced key can be in the same table or view. In this case, the parent and child tables
are the same. If you identify only the parent table or view and omit the column name,
then the foreign key automatically references the primary key of the parent table or
view. The corresponding column or columns of the foreign key and the referenced key
must match in order and data type.
You can define a foreign key constraint on a single key column either inline or out of
line. You must specify a composite foreign key and a foreign key on an attribute out of
line.
To satisfy a composite foreign key constraint, the composite foreign key must refer to a
composite unique key or a composite primary key in the parent table or view, or the
value of at least one of the columns of the foreign key must be null.
You can designate the same column or combination of columns as both a foreign key
and a primary or unique key. You can also designate the same column or combination
of columns as both a foreign key and a cluster key.

8-10 Oracle Database SQL Language Reference

constraint

You can define multiple foreign keys in a table or view. Also, a single column can be
part of more than one foreign key.
Restrictions on Foreign Key Constraints Foreign key constraints are subject to the
following restrictions:
■

■

■
■

■

■

■

None of the columns in the foreign key can be of LOB, LONG, LONG RAW, VARRAY,
NESTED TABLE, BFILE, REF, TIMESTAMP WITH TIME ZONE, or user-defined type.
However, the primary key can contain a column of TIMESTAMP WITH LOCAL TIME
ZONE.
The referenced unique or primary key constraint on the parent table or view must
already be defined.
A composite foreign key cannot have more than 32 columns.
The child and parent tables must be on the same database. To enable referential
integrity constraints across nodes of a distributed database, you must use database
triggers. See CREATE TRIGGER on page 16-98.
If either the child or parent object is a view, then the constraint is subject to all
restrictions on view constraints. See "View Constraints" on page 8-19.
You cannot define a foreign key constraint in a CREATE TABLE statement that
contains an AS subquery clause. Instead, you must create the table without the
constraint and then add it later with an ALTER TABLE statement.
When a table has a foreign key, and the parent of the foreign key is an
index-organized table, a session that updates a row that contains the foreign key
can hang when another session is updating a non-key column in the parent table.
See Also:
■

■

Oracle Database Advanced Application Developer's Guide for more
information on using constraints
"Foreign Key Constraint Example" on page 8-21 and "Composite
Foreign Key Constraint Example" on page 8-23

references_clause Foreign key constraints use the references_clause syntax. When

you specify a foreign key constraint inline, you need only the references_clause.
When you specify a foreign key constraint out of line, you must also specify the
FOREIGN KEY keywords and one or more columns.
ON DELETE Clause The ON DELETE clause lets you determine how Oracle Database
automatically maintains referential integrity if you remove a referenced primary or
unique key value. If you omit this clause, then Oracle does not allow you to delete
referenced key values in the parent table that have dependent rows in the child table.
■
■

Specify CASCADE if you want Oracle to remove dependent foreign key values.
Specify SET NULL if you want Oracle to convert dependent foreign key values to
NULL. You cannot specify this clause for a virtual column, because the values in a
virtual column cannot be updated directly. Rather, the values from which the
virtual column are derived must be updated.

Restriction on ON DELETE
See Also:

You cannot specify this clause for a view constraint.

"ON DELETE Example" on page 8-22

Common SQL DDL Clauses 8-11

constraint

Check Constraints
A check constraint lets you specify a condition that each row in the table must satisfy.
To satisfy the constraint, each row in the table must make the condition either TRUE or
unknown (due to a null). When Oracle evaluates a check constraint condition for a
particular row, any column names in the condition refer to the column values in that
row.
The syntax for inline and out-of-line specification of check constraints is the same.
However, inline specification can refer only to the column (or the attributes of the
column if it is an object column) currently being defined, whereas out-of-line
specification can refer to multiple columns or attributes.
Oracle does not verify that conditions of check constraints are not mutually exclusive.
Therefore, if you create multiple check constraints for a column, design them carefully
so their purposes do not conflict. Do not assume any particular order of evaluation of
the conditions.
See Also:
■
■

Chapter 7, "Conditions" for additional information and syntax
"Check Constraint Examples" on page 8-23 and "Attribute-Level
Constraints Example" on page 8-25

Restrictions on Check Constraints

Check constraints are subject to the following

restrictions:
■

■

■

You cannot specify a check constraint for a view. However, you can define the
view using the WITH CHECK OPTION clause, which is equivalent to specifying a check
constraint for the view.
The condition of a check constraint can refer to any column in the table, but it
cannot refer to columns of other tables.
Conditions of check constraints cannot contain the following constructs:
–

Subqueries and scalar subquery expressions

–

Calls to the functions that are not deterministic (CURRENT_DATE, CURRENT_
TIMESTAMP, DBTIMEZONE, LOCALTIMESTAMP, SESSIONTIMEZONE, SYSDATE,
SYSTIMESTAMP, UID, USER, and USERENV)

–

Calls to user-defined functions

–

Dereferencing of REF columns (for example, using the DEREF function)

–

Nested table columns or attributes

–

The pseudocolumns CURRVAL, NEXTVAL, LEVEL, or ROWNUM

–

Date constants that are not fully specified

REF Constraints
REF constraints let you describe the relationship between a column of type REF and the
object it references.
REF constraints use the ref_constraint syntax. You define a REF
constraint either inline or out of line. Out-of-line specification requires you to specify
the REF column or attribute you are further describing.

ref_constraint

■

For ref_column, specify the name of a REF column of an object or relational table.

8-12 Oracle Database SQL Language Reference

constraint

■

For ref_attribute, specify an embedded REF attribute within an object column of
a relational table.

Both inline and out-of-line specification let you define a scope constraint, a rowid
constraint, or a referential integrity constraint on a REF column.
If the scope table or referenced table of the REF column has a primary-key-based object
identifier, then the REF column is a user-defined REF column.
See Also:
■

■

Oracle Database Object-Relational Developer's Guide for more
information on REF data types
"Foreign Key Constraints" on page 8-10, and "REF Constraint
Examples" on page 8-25

SCOPE REF Constraints
In a table with a REF column, each REF value in the column can conceivably reference a
row in a different object table. The SCOPE clause restricts the scope of references to a
single table, scope_table. The values in the REF column or attribute point to objects in
scope_table, in which object instances of the same type as the REF column are stored.
Specify the SCOPE clause to restrict the scope of references in the REF column to a single
table. For you to specify this clause, scope_table must be in your own schema or you
must have SELECT privileges on scope_table or SELECT ANY TABLE system privileges.
You can specify only one scope table for each REF column.
Restrictions on Scope Constraints Scope constraints are subject to the following
restrictions:
■

You cannot add a scope constraint to an existing column unless the table is empty.

■

You cannot specify a scope constraint for the REF elements of a VARRAY column.

■

■

You must specify this clause if you specify AS subquery and the subquery returns
user-defined REF data types.
You cannot subsequently drop a scope constraint from a REF column.

Rowid REF Constraints
Specify WITH ROWID to store the rowid along with the REF value in ref_column or ref_
attribute. Storing the rowid with the REF value can improve the performance of
dereferencing operations, but will also use more space. Default storage of REF values is
without rowids.
See Also: The function DEREF on page 5-85 for an example of
dereferencing
Restrictions on Rowid Constraints

Rowid constraints are subject to the following

restrictions:
■

You cannot define a rowid constraint for the REF elements of a VARRAY column.

■

You cannot subsequently drop a rowid constraint from a REF column.

■

If the REF column or attribute is scoped, then this clause is ignored and the rowid
is not stored with the REF value.

Common SQL DDL Clauses 8-13

constraint

Referential Integrity Constraints on REF Columns
The references_clause of the ref_constraint syntax lets you define a foreign key
constraint on the REF column. This clause also implicitly restricts the scope of the REF
column or attribute to the referenced table. However, whereas a foreign key constraint
on a non-REF column references an actual column in the parent table, a foreign key
constraint on a REF column references the implicit object identifier column of the
parent table.
If you do not specify a constraint name, then Oracle generates a system name for the
constraint of the form SYS_Cn.
If you add a referential integrity constraint to an existing REF column that is already
scoped, then the referenced table must be the same as the scope table of the REF
column. If you later drop the referential integrity constraint, then the REF column will
remain scoped to the referenced table.
As is the case for foreign key constraints on other types of columns, you can use the
references_clause alone for inline declaration. For out-of-line declaration you must
also specify the FOREIGN KEY keywords plus one or more REF columns or attributes.
See Also: Oracle Database Object-Relational Developer's Guide for more
information on object identifiers
Restrictions on Foreign Key Constraints on REF Columns

Foreign key constraints

on REF columns have the following additional restrictions:
■

■

Oracle implicitly adds a scope constraint when you add a referential integrity
constraint to an existing unscoped REF column. Therefore, all the restrictions that
apply for scope constraints also apply in this case.
You cannot specify a column after the object name in the references_clause.

Specifying Constraint State
As part of constraint definition, you can specify how and when Oracle should enforce
the constraint.
You can use the constraint_state with both inline and out-of-line
specification. Specify the clauses of constraint_state in the order shown, from top to
bottom, and do not specify any clause more than once.
constraint_state

DEFERRABLE Clause The DEFERRABLE and NOT DEFERRABLE parameters indicate
whether or not, in subsequent transactions, constraint checking can be deferred until
the end of the transaction using the SET CONSTRAINT(S) statement. If you omit this
clause, then the default is NOT DEFERRABLE.
■

Specify NOT DEFERRABLE to indicate that in subsequent transactions you cannot use
the SET CONSTRAINT[S] clause to defer checking of this constraint until the
transaction is committed. The checking of a NOT DEFERRABLE constraint can never
be deferred to the end of the transaction.
If you declare a new constraint NOT DEFERRABLE, then it must be valid at the time
the CREATE TABLE or ALTER TABLE statement is committed or the statement will fail.

■

Specify DEFERRABLE to indicate that in subsequent transactions you can use the SET
CONSTRAINT[S] clause to defer checking of this constraint until a COMMIT statement
is submitted. If the constraint check fails, then the database returns an error and
the transaction is not committed. This setting in effect lets you disable the
constraint temporarily while making changes to the database that might violate
the constraint until all the changes are complete.

8-14 Oracle Database SQL Language Reference

constraint

The optimizer does not consider indexes on deferrable
constraints as usable.

Note:

You cannot alter the deferrability of a constraint. Whether you specify either of these
parameters, or make the constraint NOT DEFERRABLE implicitly by specifying neither of
them, you cannot specify this clause in an ALTER TABLE statement. You must drop the
constraint and re-create it.
See Also:
■

■

■

SET CONSTRAINT[S] on page 19-59 for information on setting
constraint checking for a transaction
Oracle Database Administrator's Guide and Oracle Database Concepts
for more information about deferred constraints
"DEFERRABLE Constraint Examples" on page 8-26

Restriction on [NOT] DEFERRABLE

You cannot specify either of these parameters

for a view constraint.
The INITIALLY clause establishes the default checking behavior for
constraints that are DEFERRABLE. The INITIALLY setting can be overridden by a SET
CONSTRAINT(S) statement in a subsequent transaction.
INITIALLY Clause

■

Specify INITIALLY IMMEDIATE to indicate that Oracle should check this constraint
at the end of each subsequent SQL statement. If you do not specify INITIALLY at
all, then the default is INITIALLY IMMEDIATE.
If you declare a new constraint INITIALLY IMMEDIATE, then it must be valid at the
time the CREATE TABLE or ALTER TABLE statement is committed or the statement will
fail.

■

Specify INITIALLY DEFERRED to indicate that Oracle should check this constraint at
the end of subsequent transactions.

This clause is not valid if you have declared the constraint to be NOT DEFERRABLE,
because a NOT DEFERRABLE constraint is automatically INITIALLY IMMEDIATE and cannot
ever be INITIALLY DEFERRED.
VALIDATE | NOVALIDATE The behavior of VALIDATE and NOVALIDATE depends on
whether the constraint is enabled or disabled, either explicitly or by default. Therefore,
the VALIDATE and NOVALIDATE keywords are described in the context of "ENABLE
Clause" on page 8-15 and "DISABLE Clause" on page 8-16.

When a foreign key
constraint is in NOVALIDATE mode, if existing data in the table does not comply with the
constraint and the QUERY_REWRITE_INTEGRITY parameter is not set to ENFORCED, then
the optimizer may use join elimination during queries on the table. In this case, a
query may return table rows with noncompliant foreign key values even if the query
contains a join condition that should filter out those rows.
Note on Foreign Key Constraints in NOVALIDATE Mode

ENABLE Clause

Specify ENABLE if you want the constraint to be applied to the data

in the table.
If you enable a unique or primary key constraint, and if no index exists on the key,
then Oracle Database creates a unique index. Unless you specify KEEP INDEX when

Common SQL DDL Clauses 8-15

constraint

subsequently disabling the constraint, this index is dropped and the database rebuilds
the index every time the constraint is reenabled.
You can also avoid rebuilding the index and eliminate redundant indexes by creating
new primary key and unique constraints initially disabled. Then create (or use
existing) nonunique indexes to enforce the constraint. Oracle does not drop a
nonunique index when the constraint is disabled, so subsequent ENABLE operations are
facilitated.
■

ENABLE VALIDATE specifies that all old and new data also complies with the
constraint. An enabled validated constraint guarantees that all data is and will
continue to be valid.
If any row in the table violates the integrity constraint, then the constraint remains
disabled and Oracle returns an error. If all rows comply with the constraint, then
Oracle enables the constraint. Subsequently, if new data violates the constraint,
then Oracle does not execute the statement and returns an error indicating the
integrity constraint violation.
If you place a primary key constraint in ENABLE VALIDATE mode, then the
validation process will verify that the primary key columns contain no nulls. To
avoid this overhead, mark each column in the primary key NOT NULL before
entering data into the column and before enabling the primary key constraint of
the table.

■

ENABLE NOVALIDATE ensures that all new DML operations on the constrained data
comply with the constraint. This clause does not ensure that existing data in the
table complies with the constraint.

If you specify neither VALIDATE nor NOVALIDATE, then the default is VALIDATE.
If you change the state of any single constraint from ENABLE NOVALIDATE to ENABLE
VALIDATE, then the operation can be performed in parallel, and does not block reads,
writes, or other DDL operations.
Restriction on the ENABLE Clause You cannot enable a foreign key that references a

disabled unique or primary key.
DISABLE Clause Specify DISABLE to disable the integrity constraint. Disabled
integrity constraints appear in the data dictionary along with enabled constraints. If
you do not specify this clause when creating a constraint, then Oracle automatically
enables the constraint.
■

DISABLE VALIDATE disables the constraint and drops the index on the constraint,
but keeps the constraint valid. This feature is most useful in data warehousing
situations, because it lets you load large amounts of data while also saving space
by not having an index. This setting lets you load data from a nonpartitioned table
into a partitioned table using the exchange_partition_subpart clause of the
ALTER TABLE statement or using SQL*Loader. All other modifications to the table
(inserts, updates, and deletes) by other SQL statements are disallowed.
See Also: Oracle Database Data Warehousing Guide for more
information on using this setting

■

DISABLE NOVALIDATE signifies that Oracle makes no effort to maintain the
constraint (because it is disabled) and cannot guarantee that the constraint is true
(because it is not being validated).

8-16 Oracle Database SQL Language Reference

constraint

You cannot drop a table whose primary key is being referenced by a foreign key
even if the foreign key constraint is in DISABLE NOVALIDATE state. Further, the
optimizer can use constraints in DISABLE NOVALIDATE state.
See Also: Oracle Database Performance Tuning Guide for information
on when to use this setting

If you specify neither VALIDATE nor NOVALIDATE, then the default is NOVALIDATE.
If you disable a unique or primary key constraint that is using a unique index, then
Oracle drops the unique index. Refer to the CREATE TABLE enable_disable_clause on
page 16-64 for additional notes and restrictions.
The RELY and NORELY parameters specify whether a constraint in
NOVALIDATE mode is to be taken into account for query rewrite. Specify RELY to activate
a constraint in NOVALIDATE mode for query rewrite in an unenforced query rewrite
integrity mode. The constraint is in NOVALIDATE mode, so Oracle does not enforce it.
The default is NORELY.
RELY Clause

Unenforced constraints are generally useful only with materialized views and query
rewrite. Depending on the QUERY_REWRITE_INTEGRITY mode, query rewrite can use
only constraints that are in VALIDATE mode, or that are in NOVALIDATE mode with the
RELY parameter set, to determine join information.
Restriction on the RELY Clause You cannot set a nondeferrable NOT NULL constraint
to RELY.
See Also: Oracle Database Data Warehousing Guide for more
information on materialized views and query rewrite

Using Indexes to Enforce Constraints
When defining the state of a unique or primary key constraint, you can specify an
index for Oracle to use to enforce the constraint, or you can instruct Oracle to create
the index used to enforce the constraint.
You can specify the using_index_clause only when enabling
unique or primary key constraints. You can specify the clauses of the using_index_
clause in any order, but you can specify each clause only once.

using_index_clause

■

■

■

If you specify schema.index, then Oracle attempts to enforce the constraint using
the specified index. If Oracle cannot find the index or cannot use the index to
enforce the constraint, then Oracle returns an error.
If you specify the create_index_statement, then Oracle attempts to create the
index and use it to enforce the constraint. If Oracle cannot create the index or
cannot use the index to enforce the constraint, then Oracle returns an error.
If you neither specify an existing index nor create a new index, then Oracle creates
the index. In this case:
–

The index receives the same name as the constraint.

–

If table is partitioned, then you can specify a locally or globally partitioned
index for the unique or primary key constraint.

Restrictions on the using_index_clause The following restrictions apply to the

using_index_clause:
■

You cannot specify this clause for a view constraint.

Common SQL DDL Clauses 8-17

constraint

■
■

■
■

You cannot specify this clause for a NOT NULL, foreign key, or check constraint.
You cannot specify an index (schema.index) or create an index (create_index_
statement) when enabling the primary key of an index-organized table.
You cannot specify the parallel_clause of index_attributes.
The INDEXTYPE IS ... clause of index_properties is not valid in the definition of a
constraint.
See Also:
■

■

■

CREATE INDEX on page 14-60 for a description of index_
attributes, the global_partitioned_index and local_partitioned_index
clauses, and for a description of NOSORT and the logging_clause
in relation to indexes
physical_attributes_clause on page 8-44 and PCTFREE parameters
and storage_clause on page 8-48
"Explicit Index Control Example" on page 8-26

Handling Constraint Exceptions
When defining the state of a constraint, you can specify a table into which Oracle
places the rowids of all rows violating the constraint.
exceptions_clause Use the exceptions_clause syntax to define exception handling.
If you omit schema, then Oracle assumes the exceptions table is in your own schema. If
you omit this clause altogether, then Oracle assumes that the table is named
EXCEPTIONS. The EXCEPTIONS table or the table you specify must exist on your local
database.

You can create the EXCEPTIONS table using one of these scripts:
■

■

UTLEXCPT.SQL uses physical rowids. Therefore it can accommodate rows from
conventional tables but not from index-organized tables. (See the Note that
follows.)
UTLEXPT1.SQL uses universal rowids, so it can accommodate rows from both
conventional and index-organized tables.

If you create your own exceptions table, then it must follow the format prescribed by
one of these two scripts.
If you are collecting exceptions from index-organized tables based on primary keys
(rather than universal rowids), then you must create a separate exceptions table for
each index-organized table to accommodate its primary-key storage. You create
multiple exceptions tables with different names by modifying and resubmitting the
script.
Restrictions on the exceptions_clause

The following restrictions apply to the

exceptions_clause:
■
■

You cannot specify this clause for a view constraint.
You cannot specify this clause in a CREATE TABLE statement, because no rowids
exist until after the successful completion of the statement.

8-18 Oracle Database SQL Language Reference

constraint

See Also:
■

■

The DBMS_IOT package in Oracle Database PL/SQL Packages and
Types Reference for information on the SQL scripts
Oracle Database Performance Tuning Guide for information on
eliminating migrated and chained rows

View Constraints
Data warehousing applications recognize multidimensional data in the Oracle
Database by identifying referential integrity constraints in the relational schema. These
constraints represent primary and foreign key relationships among tables. By querying
the Oracle Database data dictionary, applications can recognize such constraints and
therefore recognize the multidimensional data in the database. For schema complexity
or security reasons, you might want to define views on fact and dimension tables.
Oracle Database provides the ability to constrain these views. By allowing constraint
definitions between views, you can propagate base table constraints to the views,
thereby allowing applications to recognize multidimensional data even in the
restricted environment provided by the view.
Oracle does not enforce view constraints. However, operations on views are subject to
the integrity constraints defined on the underlying base tables. This means that you
can enforce constraints on views through constraints on base tables.
Notes on View Constraints View constraints are a subset of table constraints and are

subject to the following restrictions:
■

■

■

■

■

■

You can specify only unique, primary key, and foreign key constraints on views.
However, you can define the view using the WITH CHECK OPTION clause, which is
equivalent to specifying a check constraint for the view.
View constraints are supported only in DISABLE NOVALIDATE mode. You cannot
specify any other mode. You must specify the keyword DISABLE when you declare
the view constraint. You need not specify NOVALIDATE explicitly, as it is the default.
The RELY and NORELY parameters are optional. View constraints, because they are
unenforced, are usually specified with the RELY parameter to make them more
useful. The RELY or NORELY keyword must precede the DISABLE keyword. Refer to
"RELY Clause" on page 8-17 for more information.
Because view constraints are not enforced directly, you cannot specify INITIALLY
DEFERRED or DEFERRABLE.
You cannot specify the using_index_clause, the exceptions_clause clause, or the
ON DELETE clause of the references_clause.
You cannot define view constraints on attributes of an object column.

Examples
8

Unique Key Example The following statement is a variation of the statement that
created the sample table sh.promotions. It defines inline and implicitly enables a
unique key on the promo_id column (other constraints are not shown):
CREATE TABLE promotions_var1
( promo_id
NUMBER(6)
CONSTRAINT promo_id_u
, promo_name
VARCHAR2(20)
, promo_category
VARCHAR2(15)
, promo_cost
NUMBER(10,2)

UNIQUE

Common SQL DDL Clauses 8-19

constraint

, promo_begin_date DATE
, promo_end_date
DATE
) ;

The constraint promo_id_u identifies the promo_id column as a unique key. This
constraint ensures that no two promotions in the table have the same ID. However, the
constraint does allow promotions without identifiers.
Alternatively, you can define and enable this constraint out of line:
CREATE TABLE promotions_var2
( promo_id
NUMBER(6)
, promo_name
VARCHAR2(20)
, promo_category
VARCHAR2(15)
, promo_cost
NUMBER(10,2)
, promo_begin_date DATE
, promo_end_date
DATE
, CONSTRAINT promo_id_u UNIQUE (promo_id)
USING INDEX PCTFREE 20
TABLESPACE stocks
STORAGE (INITIAL 8M) );

The preceding statement also contains the using_index_clause, which specifies
storage characteristics for the index that Oracle creates to enable the constraint.
The following statement defines and enables a
composite unique key on the combination of the warehouse_id and warehouse_name
columns of the oe.warehouses table:

Composite Unique Key Example

ALTER TABLE warehouses
ADD CONSTRAINT wh_unq UNIQUE (warehouse_id, warehouse_name)
USING INDEX PCTFREE 5
EXCEPTIONS INTO wrong_id;

The wh_unq constraint ensures that the same combination of warehouse_id and
warehouse_name values does not appear in the table more than once.
The ADD CONSTRAINT clause also specifies other properties of the constraint:
■

■

The USING INDEX clause specifies storage characteristics for the index Oracle
creates to enable the constraint.
The EXCEPTIONS INTO clause causes Oracle to write to the wrong_id table
information about any rows currently in the warehouses table that violate the
constraint. If the wrong_id exceptions table does not already exist, then this
statement will fail.

Primary Key Example The following statement is a variation of the statement that

created the sample table hr.locations. It creates the locations_demo table and
defines and enables a primary key on the location_id column (other constraints from
the hr.locations table are omitted):
CREATE TABLE locations_demo
( location_id
NUMBER(4) CONSTRAINT loc_id_pk PRIMARY KEY
, street_address VARCHAR2(40)
, postal_code
VARCHAR2(12)
, city
VARCHAR2(30)
, state_province VARCHAR2(25)
, country_id
CHAR(2)
) ;

8-20 Oracle Database SQL Language Reference

constraint

The loc_id_pk constraint, specified inline, identifies the location_id column as the
primary key of the locations_demo table. This constraint ensures that no two locations
in the table have the same location number and that no location identifier is NULL.
Alternatively, you can define and enable this constraint out of line:
CREATE TABLE locations_demo
( location_id
NUMBER(4)
, street_address VARCHAR2(40)
, postal_code
VARCHAR2(12)
, city
VARCHAR2(30)
, state_province VARCHAR2(25)
, country_id
CHAR(2)
, CONSTRAINT loc_id_pk PRIMARY KEY (location_id));

The following statement alters the locations_demo table
(created in "Primary Key Example" on page 8-20) to define and enable a NOT NULL
constraint on the country_id column:
NOT NULL Example

ALTER TABLE locations_demo
MODIFY (country_id CONSTRAINT country_nn NOT NULL);

The constraint country_nn ensures that no location in the table has a null country_id.
Composite Primary Key Example The following statement defines a composite
primary key on the combination of the prod_id and cust_id columns of the sample
table sh.sales:
ALTER TABLE sales
ADD CONSTRAINT sales_pk PRIMARY KEY (prod_id, cust_id) DISABLE;

This constraint identifies the combination of the prod_id and cust_id columns as the
primary key of the sales table. The constraint ensures that no two rows in the table
have the same combination of values for the prod_id column and cust_id columns.
The constraint clause (PRIMARY KEY) also specifies the following properties of the
constraint:
■

■

The constraint definition does not include a constraint name, so Oracle generates a
name for the constraint.
The DISABLE clause causes Oracle to define the constraint but not enable it.

The following statement creates the dept_20 table
and defines and enables a foreign key on the department_id column that references
the primary key on the department_id column of the departments table:

Foreign Key Constraint Example

CREATE TABLE dept_20
(employee_id
NUMBER(4),
last_name
VARCHAR2(10),
job_id
VARCHAR2(9),
manager_id
NUMBER(4),
hire_date
DATE,
salary
NUMBER(7,2),
commission_pct NUMBER(7,2),
department_id
CONSTRAINT fk_deptno
REFERENCES departments(department_id) );

The constraint fk_deptno ensures that all departments given for employees in the
dept_20 table are present in the departments table. However, employees can have null
department numbers, meaning they are not assigned to any department. To ensure

Common SQL DDL Clauses 8-21

constraint

that all employees are assigned to a department, you could create a NOT NULL constraint
on the department_id column in the dept_20 table in addition to the REFERENCES
constraint.
Before you define and enable this constraint, you must define and enable a constraint
that designates the department_id column of the departments table as a primary or
unique key.
The foreign key constraint definition does not use the FOREIGN KEY clause, because the
constraint is defined inline. The data type of the department_id column is not needed,
because Oracle automatically assigns to this column the data type of the referenced
key.
The constraint definition identifies both the parent table and the columns of the
referenced key. Because the referenced key is the primary key of the parent table, the
referenced key column names are optional.
Alternatively, you can define this foreign key constraint out of line:
CREATE TABLE dept_20
(employee_id
NUMBER(4),
last_name
VARCHAR2(10),
job_id
VARCHAR2(9),
manager_id
NUMBER(4),
hire_date
DATE,
salary
NUMBER(7,2),
commission_pct NUMBER(7,2),
department_id,
CONSTRAINT fk_deptno
FOREIGN KEY (department_id)
REFERENCES departments(department_id) );

The foreign key definitions in both variations of this statement omit the ON DELETE
clause, causing Oracle to prevent the deletion of a department if any employee works
in that department.
This statement creates the dept_20 table, defines and enables
two referential integrity constraints, and uses the ON DELETE clause:

ON DELETE Example

CREATE TABLE dept_20
(employee_id
NUMBER(4) PRIMARY KEY,
last_name
VARCHAR2(10),
job_id
VARCHAR2(9),
manager_id
NUMBER(4) CONSTRAINT fk_mgr
REFERENCES employees ON DELETE SET NULL,
hire_date
DATE,
salary
NUMBER(7,2),
commission_pct NUMBER(7,2),
department_id
NUMBER(2)
CONSTRAINT fk_deptno
REFERENCES departments(department_id)
ON DELETE CASCADE );

Because of the first ON DELETE clause, if manager number 2332 is deleted from the
employees table, then Oracle sets to null the value of manager_id for all employees in
the dept_20 table who previously had manager 2332.
Because of the second ON DELETE clause, Oracle cascades any deletion of a department_
id value in the departments table to the department_id values of its dependent rows
of the dept_20 table. For example, if Department 20 is deleted from the departments
table, then Oracle deletes all of the employees in Department 20 from the dept_20
table.
8-22 Oracle Database SQL Language Reference

constraint

The following statement defines and
enables a foreign key on the combination of the employee_id and hire_date columns
of the dept_20 table:
Composite Foreign Key Constraint Example

ALTER TABLE dept_20
ADD CONSTRAINT fk_empid_hiredate
FOREIGN KEY (employee_id, hire_date)
REFERENCES hr.job_history(employee_id, start_date)
EXCEPTIONS INTO wrong_emp;

The constraint fk_empid_hiredate ensures that all the employees in the dept_20 table
have employee_id and hire_date combinations that exist in the employees table.
Before you define and enable this constraint, you must define and enable a constraint
that designates the combination of the employee_id and hire_date columns of the
employees table as a primary or unique key.
The EXCEPTIONS INTO clause causes Oracle to write information to the wrong_emp table
about any rows in the dept_20 table that violate the constraint. If the wrong_emp
exceptions table does not already exist, then this statement will fail.
Check Constraint Examples The following statement creates a divisions table and
defines a check constraint in each column of the table:
CREATE TABLE divisions
(div_no
NUMBER CONSTRAINT check_divno
CHECK (div_no BETWEEN 10 AND 99)
DISABLE,
div_name VARCHAR2(9) CONSTRAINT check_divname
CHECK (div_name = UPPER(div_name))
DISABLE,
office
VARCHAR2(10) CONSTRAINT check_office
CHECK (office IN ('DALLAS','BOSTON',
'PARIS','TOKYO'))
DISABLE);

Each constraint restricts the values of the column in which it is defined:
■

check_divno ensures that no division numbers are less than 10 or greater than 99.

■

check_divname ensures that all division names are in uppercase.

■

check_office restricts office locations to Dallas, Boston, Paris, or Tokyo.

Because each CONSTRAINT clause contains the DISABLE clause, Oracle only defines the
constraints and does not enable them.
The following statement creates the dept_20 table, defining out of line and implicitly
enabling a check constraint:
CREATE TABLE dept_20
(employee_id
NUMBER(4) PRIMARY KEY,
last_name
VARCHAR2(10),
job_id
VARCHAR2(9),
manager_id
NUMBER(4),
salary
NUMBER(7,2),
commission_pct NUMBER(7,2),
department_id
NUMBER(2),
CONSTRAINT check_sal CHECK (salary * commission_pct <= 5000));

This constraint uses an inequality condition to limit an employee's total commission,
the product of salary and commission_pct, to $5000:

Common SQL DDL Clauses 8-23

constraint

■

■

If an employee has non-null values for both salary and commission, then the
product of these values must not exceed $5000 to satisfy the constraint.
If an employee has a null salary or commission, then the result of the condition is
unknown and the employee automatically satisfies the constraint.

Because the constraint clause in this example does not supply a constraint name,
Oracle generates a name for the constraint.
The following statement defines and enables a primary key constraint, two foreign key
constraints, a NOT NULL constraint, and two check constraints:
CREATE TABLE order_detail
(CONSTRAINT pk_od PRIMARY KEY (order_id, part_no),
order_id
NUMBER
CONSTRAINT fk_oid
REFERENCES oe.orders(order_id),
part_no
NUMBER
CONSTRAINT fk_pno
REFERENCES oe.product_information(product_id),
quantity
NUMBER
CONSTRAINT nn_qty NOT NULL
CONSTRAINT check_qty CHECK (quantity > 0),
cost
NUMBER
CONSTRAINT check_cost CHECK (cost > 0) );

The constraints enable the following rules on table data:
■

■

■

■
■

■

pk_od identifies the combination of the order_id and part_no columns as the
primary key of the table. To satisfy this constraint, no two rows in the table can
contain the same combination of values in the order_id and the part_no columns,
and no row in the table can have a null in either the order_id or the part_no
column.
fk_oid identifies the order_id column as a foreign key that references the order_
id column in the orders table in the sample schema oe. All new values added to
the column order_detail.order_id must already appear in the column
oe.orders.order_id.
fk_pno identifies the product_id column as a foreign key that references the
product_id column in the product_information table owned by oe. All new
values added to the column order_detail.product_id must already appear in the
column oe.product_information.product_id.
nn_qty forbids nulls in the quantity column.
check_qty ensures that values in the quantity column are always greater than
zero.
check_cost ensures the values in the cost column are always greater than zero.

This example also illustrates the following points about constraint clauses and column
definitions:
■

■

Out-of-line constraint definition can appear before or after the column definitions.
In this example, the out-of-line definition of the pk_od constraint precedes the
column definitions.
A column definition can contain multiple inline constraint definitions. In this
example, the definition of the quantity column contains the definitions of both the
nn_qty and check_qty constraints.

8-24 Oracle Database SQL Language Reference

constraint

■

A table can have multiple CHECK constraints. Multiple CHECK constraints, each with
a simple condition enforcing a single business rule, are preferable to a single CHECK
constraint with a complicated condition enforcing multiple business rules. When a
constraint is violated, Oracle returns an error identifying the constraint. Such an
error more precisely identifies the violated business rule if the identified constraint
enables a single business rule.

The following example guarantees that a value
exists for both the first_name and last_name attributes of the name column in the
students table:
Attribute-Level Constraints Example

CREATE TYPE person_name AS OBJECT
(first_name VARCHAR2(30), last_name VARCHAR2(30));
/
CREATE TABLE students (name person_name, age INTEGER,
CHECK (name.first_name IS NOT NULL AND
name.last_name IS NOT NULL));

The following example creates a duplicate of the sample
schema object type cust_address_typ, and then creates a table containing a REF
column with a SCOPE constraint:
REF Constraint Examples

CREATE TYPE cust_address_typ_new AS OBJECT
( street_address
VARCHAR2(40)
, postal_code
VARCHAR2(10)
, city
VARCHAR2(30)
, state_province
VARCHAR2(10)
, country_id
CHAR(2)
);
/
CREATE TABLE address_table OF cust_address_typ_new;
CREATE TABLE customer_addresses (
add_id NUMBER,
address REF cust_address_typ_new
SCOPE IS address_table);

The following example creates the same table but with a referential integrity constraint
on the REF column that references the object identifier column of the parent table:
CREATE TABLE customer_addresses (
add_id NUMBER,
address REF cust_address_typ REFERENCES address_table);

The following example uses the type department_typ and the table departments_obj_
t, created in "Creating Object Tables: Examples" on page 16-80. A table with a scoped
REF is then created.
CREATE TABLE employees_obj
( e_name
VARCHAR2(100),
e_number NUMBER,
e_dept
REF department_typ SCOPE IS departments_obj_t );

The following statement creates a table with a REF column which has a referential
integrity constraint defined on it:
CREATE TABLE employees_obj
( e_name
VARCHAR2(100),
e_number NUMBER,
e_dept
REF department_typ REFERENCES departments_obj_t);
Common SQL DDL Clauses 8-25

constraint

Explicit Index Control Example The following statement shows another way to
create a unique (or primary key) constraint that gives you explicit control over the
index (or indexes) Oracle uses to enforce the constraint:
CREATE TABLE promotions_var3
( promo_id
NUMBER(6)
, promo_name
VARCHAR2(20)
, promo_category
VARCHAR2(15)
, promo_cost
NUMBER(10,2)
, promo_begin_date DATE
, promo_end_date
DATE
, CONSTRAINT promo_id_u UNIQUE (promo_id, promo_cost)
USING INDEX (CREATE UNIQUE INDEX promo_ix1
ON promotions_var3 (promo_id, promo_cost))
, CONSTRAINT promo_id_u2 UNIQUE (promo_cost, promo_id)
USING INDEX promo_ix1);

This example also shows that you can create an index for one constraint and use that
index to create and enable another constraint in the same statement.
DEFERRABLE Constraint Examples The following statement creates table games
with a NOT DEFERRABLE INITIALLY IMMEDIATE constraint check (by default) on the
scores column:
CREATE TABLE games (scores NUMBER CHECK (scores >= 0));

To define a unique constraint on a column as INITIALLY DEFERRED DEFERRABLE, issue
the following statement:
CREATE TABLE games
(scores NUMBER, CONSTRAINT unq_num UNIQUE (scores)
INITIALLY DEFERRED DEFERRABLE);

8-26 Oracle Database SQL Language Reference

deallocate_unused_clause

deallocate_unused_clause
Purpose
8

Use the deallocate_unused_clause to explicitly deallocate unused space at the end of
a database object segment and make the space available for other segments in the
tablespace.
You can deallocate unused space using the following statements:
■
■

■

■

ALTER CLUSTER (see ALTER CLUSTER on page 10-5)
ALTER INDEX: to deallocate unused space from the index, an index partition, or an
index subpartition (see ALTER INDEX on page 10-78)
ALTER MATERIALIZED VIEW: to deallocate unused space from the overflow segment
of an index-organized materialized view (see ALTER MATERIALIZED VIEW on
page 11-3)
ALTER TABLE: to deallocate unused space from the table, a table partition, a table
subpartition, the mapping table of an index-organized table, the overflow segment
of an index-organized table, or a LOB storage segment (see ALTER TABLE on
page 12-2)

Syntax
8

deallocate_unused_clause::=
KEEP
DEALLOCATE

size_clause

UNUSED

(size_clause::= on page 8-47)

Semantics
8

This section describes the semantics of the deallocate_unused_clause. For additional
information, refer to the SQL statement in which you set or reset this clause for a
particular database object.
You cannot specify both the deallocate_unused_clause and the allocate_extent_
clause in the same statement.
Oracle Database frees only unused space above the high water mark (the point beyond
which database blocks have not yet been formatted to receive data). Oracle deallocates
unused space beginning from the end of the object and moving toward the beginning
of the object to the high water mark.
If an extent is completely contained in the deallocation, then the whole extent is freed
for reuse. If an extent is partially contained in the deallocation, then the used part up
to the high water mark becomes the extent, and the remaining unused space is freed
for reuse.
Oracle credits the amount of the released space to the user quota for the tablespace in
which the deallocation occurs.
The exact amount of space freed depends on the values of the INITIAL, MINEXTENTS,
and NEXT storage parameters. Refer to the storage_clause on page 8-48 for a description
of these parameters.

Common SQL DDL Clauses 8-27

deallocate_unused_clause

KEEP integer
Specify the number of bytes above the high water mark that the segment of the
database object is to have after deallocation.
■

■

■

If you omit KEEP and the high water mark is above the size of INITIAL and
MINEXTENTS, then all unused space above the high water mark is freed. When the
high water mark is less than the size of INITIAL or MINEXTENTS, then all unused
space above MINEXTENTS is freed.
If you specify KEEP, then the specified amount of space is kept and the remaining
space is freed. When the remaining number of extents is less than MINEXTENTS,
then Oracle adjusts MINEXTENTS to the new number of extents. If the initial extent
becomes smaller than INITIAL, then Oracle adjusts INITIAL to the new size.
In either case, Oracle sets the value of the NEXT storage parameter to the size of the
last extent that was deallocated.

8-28 Oracle Database SQL Language Reference

file_specification

file_specification
Purpose
8

Use one of the file_specification forms to specify a file as a data file or temp file, or
to specify a group of one or more files as a redo log file group. If you are storing your
files in Oracle Automatic Storage Management (Oracle ASM) disk groups, then you
can further specify the file as a disk group file.
A file_specification can appear in the following statements:
■

CREATE CONTROLFILE (see CREATE CONTROLFILE on page 14-12)

■

CREATE DATABASE (see CREATE DATABASE on page 14-19)

■

ALTER DATABASE (see ALTER DATABASE on page 10-8)

■

CREATE TABLESPACE (see CREATE TABLESPACE on page 16-83)

■

ALTER TABLESPACE (see ALTER TABLESPACE on page 12-90)

■

ALTER DISKGROUP (see ALTER DISKGROUP on page 10-51)

Prerequisites
8

You must have the privileges necessary to issue the statement in which the file
specification appears.

Syntax
8

file_specification::=
datafile_tempfile_spec
redo_log_file_spec

datafile_tempfile_spec::=
filename
’

’
ASM_filename

SIZE

size_clause

REUSE

autoextend_clause

(size_clause::= on page 8-47)
redo_log_file_spec::=
filename
’

’
ASM_filename
,
filename

(

’

’

)

ASM_filename

SIZE

size_clause

BLOCKSIZE

size_clause

REUSE

Common SQL DDL Clauses 8-29

file_specification

(size_clause::= on page 8-47)
ASM_filename::=
fully_qualified_file_name
numeric_file_name
incomplete_file_name
alias_file_name

fully_qualified_file_name::=
+

diskgroup_name

/

db_name

/

file_type

/

file_type_tag

.

filenumber

.

incarnation_number

numeric_file_name::=
+

diskgroup_name

.

filenumber

.

incarnation_number

incomplete_file_name::=

+

(

template_name

)

(

template_name

)

diskgroup_name

alias_file_name::=

+

diskgroup_name

/

alias_name

autoextend_clause::=
OFF
AUTOEXTEND

NEXT

size_clause

maxsize_clause

ON

(size_clause::= on page 8-47)
maxsize_clause::=
UNLIMITED
MAXSIZE
size_clause

(size_clause::= on page 8-47)

Semantics
8

This section describes the semantics of file_specification. For additional
information, refer to the SQL statement in which you specify a data file, temp file, redo
log file, or Oracle ASM disk group or disk group file.

8-30 Oracle Database SQL Language Reference

file_specification

datafile_tempfile_spec
Use this clause to specify the attributes of data files and temp files if your database
storage is in a file system or on raw devices or in Oracle ASM disk groups.

redo_log_file_spec
Use this clause to specify the attributes of redo log files if your database storage is in a
file system or on raw devices or in Oracle ASM disk groups.
filename
Use filename for files stored in a file system or on raw devices. The filename can
specify either a new file or an existing file. For a new file:
■

■

If you are not using Oracle Managed Files, then you must specify both filename
and the SIZE clause or the statement fails. When you specify a filename without a
size, Oracle attempts to reuse an existing file and returns an error if the file does
not exist.
If you are using Oracle Managed Files, then filename is optional, as are the
remaining clauses of the specification. In this case, Oracle Database creates a
unique name for the file and saves it in the directory specified by one of the
following initialization parameters:
–

The DB_RECOVERY_FILE_DEST (for logfiles and control files)

–

The DB_CREATE_FILE_DEST initialization parameter (for any type of file)

–

The DB_CREATE_ONLINE_LOG_DEST_n initialization parameter, which takes
precedence over DB_CREATE_FILE_DEST and DB_RECOVERY_FILE_DEST for log
files.

For an existing file, specify the name of either a data file, temp file, or a redo log file
member. The filename can contain only single-byte characters from 7-bit ASCII or
EBCDIC character sets. Multibyte characters are not valid.
The filename can include a path prefix. If you do not specify such a path prefix, then
the database adds the path prefix for the default storage location, which is platform
dependent.
A redo log file group can have one or more members (copies). Each filename must be
fully specified according to the conventions for your operating system.
The way the database interprets filename also depends on whether you specify it with
the SIZE and REUSE clauses.
■

■

■

If you specify filename only, or with the REUSE clause but without the SIZE clause,
then the file must already exist.
If you specify filename with SIZE but without REUSE, then the file must be a new
file.
If you specify filename with both SIZE and REUSE, then the file can be either new
or existing. If the file exists, then it is reused with the new size. If it does not exist,
then the database ignores the REUSE keyword and creates a new file of the
specified size.
See Also: Oracle Automatic Storage Management Administrator's Guide
for more information on Oracle Managed Files, "Specifying a Data
File: Example" on page 8-36, and "Specifying a Log File: Example" on
page 8-35

Common SQL DDL Clauses 8-31

file_specification

ASM_filename
Use a form of ASM_filename for files stored in Oracle ASM disk groups. You can create
or refer to data files, temp files, and redo log files with this syntax.
All forms of ASM_filename begin with the plus sign (+) followed by the name of the
disk group. You can determine the names of all Oracle ASM disk groups by querying
the V$ASM_DISKGROUP view.
See Also: Oracle Automatic Storage Management Administrator's Guide
for information on using Oracle ASM

fully_qualified_file_name
When you create a file in an Oracle ASM disk group, the file receives a
system-generated fully qualified Oracle ASM filename. You can use this form only
when referring to an existing Oracle ASM file. Therefore, if you are using this form
during file creation, you must also specify REUSE.
■

■

■

db_name is the value of the DB_UNIQUE_NAME initialization parameter. This name is
equivalent to the name of the database on which the file resides, but the parameter
distinguishes between primary and standby databases, if both exist.
file_type and file_type_tag indicate the type of database file. Table 8–1 on
page 8-32 lists all of the file types and their corresponding Oracle ASM tags.
filenumber and incarnation_number are system-generated identifiers to
guarantee uniqueness.

You can determine the fully qualified names of Oracle ASM files by querying the
dynamic performance view appropriate for the file type (for example V$DATAFILE for
data files, V$CONTROLFILE for control files, and so on). You can also obtain the
filenumber and incarnation_number portions of the fully qualified names by
querying the V$ASM_FILE view.
Table 8–1

Oracle File Types and Oracle ASM File Type Tags

Oracle ASM file_type

Description

Oracle ASM file_type_tag Comments

CONTROLFILE

Control files and backup
control files

Current

DATAFILE

Data files and data file
copies

tsname

Tablespace into which the
file is added

ONLINELOG

Online logs

group_group#

—

ARCHIVELOG

Archive logs

thread_thread#_seq_
sequence#

—

TEMPFILE

Temp files

tsname

Tablespace into which the
file is added

BACKUPSET

Data file and archive log
hasspfile_timestamp
backup pieces; data file
incremental backup pieces

hasspfile can take one of
two values: s indicates that
the backup set includes the
spfile; n indicates that the
backup set does not
include the spfile.

PARAMETERFILE

Persistent parameter files

spfile

—

DAATAGUARDCONFIG

Data Guard configuration
file

db_unique_name

Data Guard uses the value
of the DB_UNIQUE_NAME
initialization parameter.

8-32 Oracle Database SQL Language Reference

—

Backup

file_specification

Table 8–1 (Cont.) Oracle File Types and Oracle ASM File Type Tags
Oracle ASM file_type

Description

Oracle ASM file_type_tag Comments

FLASHBACK

Flashback logs

log_log#

CHANGETRACKING

Block change tracking data ctf

Used during incremental
backups

DUMPSET

Data Pump dumpset

user_obj#_file#

Dump set files encode the
user name, the job number
that created the dump set,
and the file number as part
of the tag.

XTRANSPORT

Data file convert

tsname

—

AUTOBACKUP

Automatic backup files

hasspfile_timestamp

hasspfile can take one of
two values: s indicates that
the backup set includes the
spfile; n indicates that the
backup set does not
include the spfile.

—

numeric_file_name
A numeric Oracle ASM filename is similar to a fully qualified filename except that it
uses only the unique filenumber.incarnation_number string. You can use this form
only to refer to an existing file. Therefore, if you are using this form during file
creation, you must also specify REUSE.
incomplete_file_name
Incomplete Oracle ASM filenames are used during file creation only. If you specify the
disk group name alone, then Oracle ASM uses the appropriate default template for the
file type. For example, if you are creating a data file in a CREATE TABLESPACE statement,
Oracle ASM uses the default DATAFILE template to create an Oracle ASM data file. If
you specify the disk group name with a template, then Oracle ASM uses the specified
template to create the file. In both cases, Oracle ASM also creates a fully qualified
filename.
template_name A template is a named collection of attributes. You can create
templates and apply them to files in a disk group. You can determine the names of all
Oracle ASM template names by querying the V$ASM_TEMPLATE data dictionary view.
Refer to diskgroup_template_clauses on page 10-63 for instructions on creating Oracle
ASM templates.

You can specify template only during file creation. It appears in the incomplete and
alias name forms of the ASM_filename diagram:
■

■

If you specify template immediately after the disk group name, then Oracle ASM
uses the specified template to create the file, and gives the file a fully qualified
filename.
If you specify template after specifying an alias, then Oracle ASM uses the
specified template to create the file, gives the file a fully qualified filename, and
also creates the alias so that you can subsequently use it to refer to the file. If the
alias you specify refers to an existing file, then Oracle ASM ignores the template
specification unless you also specify REUSE.
See Also: diskgroup_template_clauses on page 10-63 for information
about the default templates

Common SQL DDL Clauses 8-33

file_specification

alias_file_name
An alias is a user-friendly name for an Oracle ASM file. You can use alias filenames
during file creation or reference. You can specify a template with an alias, but only
during file creation. To determine the alias names for Oracle ASM files, query the
V$ASM_ALIAS data dictionary view.
If you are specifying an alias during file creation, then refer to diskgroup_directory_
clauses on page 10-66 and diskgroup_alias_clauses on page 10-66 for instructions on
specifying the full alias name.
SIZE Clause
Specify the size of the file in bytes. Use K, M, G, or T to specify the size in kilobytes,
megabytes, gigabytes, or terabytes.
■

■

■

For undo tablespaces, you must specify the SIZE clause for each data file. For other
tablespaces, you can omit this parameter if the file already exists, or if you are
creating an Oracle Managed File.
If you omit this clause when creating an Oracle Managed File, then Oracle creates
a 100M file.
The size of a tablespace must be one block greater than the sum of the sizes of the
objects contained in it.
See Also: Oracle Database Administrator's Guide for information on
automatic undo management and undo tablespaces and "Adding a
Log File: Example" on page 8-36

BLOCKSIZE Clause
Specify BLOCKSIZE to override the operating system-dependent sector size. If you omit
this clause, then the database uses the operating system-dependent sector size as the
block size.
When you add a redo log file to a 512-byte sector disk or to a 4KB sector disk with
512-byte emulation, the blocksize of the new file must be the original platform base
block size or 4KB.
■

■

■

If the redo log file is being added to a 512-byte sector disk, then you must specify
512 or 1024 (or 1K) as the block size, depending on your platform.
If the redo log file is being added to a 4KB sector disk (native), then you must
specify either 4096 or 4K as the block size.
If the redo log file is being added to a 4KB sector disk with 512-byte emulation,
then you can specify either 512, 1024 (or 1K), or 4096 (or 4K) as the block size,
depending on your platform.

All logs within a log group must have the same block size. Two log groups created on
separate disks can have different block sizes. However, the mixed configuration
introduces overhead at every log switch. Oracle recommends that you create all log
files with the same block size.
This clause is useful when the 4K sector size is in use, but you want to optimize disk
space use rather than performance. In such a case you can override the operating
system sector size by specifying BLOCKSIZE 512 or, for HP-UX, BLOCKSIZE 1024.
See Also:

"Adding a Log File: Example" on page 8-36

8-34 Oracle Database SQL Language Reference

file_specification

REUSE
Specify REUSE to allow Oracle to reuse an existing file.
■

■

If the file already exists, then Oracle reuses the filename and applies the new size
(if you specify SIZE) or retains the original size.
If the file does not exist, then Oracle ignores this clause and creates the file.

Restriction on the REUSE Clause You cannot specify REUSE unless you have
specified filename.

Whenever Oracle uses an existing file, the previous contents of the file are lost.
"Adding a Data File: Example" on page 8-36 and "Adding
a Log File: Example" on page 8-36

See Also:

autoextend_clause
The autoextend_clause is valid for data files and temp files but not for redo log files.
Use this clause to enable or disable the automatic extension of a new or existing data
file or temp file. If you omit this clause, then:
■

■

For Oracle Managed Files:
–

If you specify SIZE, then Oracle Database creates a file of the specified size
with AUTOEXTEND disabled.

–

If you do not specify SIZE, then the database creates a 100M file with
AUTOEXTEND enabled. When autoextension is required, the database extends
the file by its original size or 100MB, whichever is smaller. You can override
this default behavior by specifying the NEXT clause.

For user-managed files, with or without SIZE specified, Oracle creates a file with
AUTOEXTEND disabled.

ON Specify ON to enable autoextend.
OFF Specify OFF to turn off autoextend if is turned on. When you turn off

autoextend, the values of NEXT and MAXSIZE are set to zero. If you turn autoextend back
on in a subsequent statement, then you must reset these values.
NEXT Use the NEXT clause to specify the size in bytes of the next increment of disk
space to be allocated automatically when more extents are required. The default is the
size of one data block.
MAXSIZE Use the MAXSIZE clause to specify the maximum disk space allowed for
automatic extension of the data file.
UNLIMITED Use the UNLIMITED clause if you do not want to limit the disk space that

Oracle can allocate to the data file or temp file.
You cannot specify this clause as part of the
datafile_tempfile_spec in a CREATE CONTROLFILE statement or in an ALTER DATABASE
CREATE DATAFILE clause.
Restriction on the autoextend_clause

Examples
8

Specifying a Log File: Example The following statement creates a database named

payable that has two redo log file groups, each with two members, and one data file:
Common SQL DDL Clauses 8-35

file_specification

CREATE DATABASE payable
LOGFILE GROUP 1 ('diska:log1.log', 'diskb:log1.log') SIZE 50K,
GROUP 2 ('diska:log2.log', 'diskb:log2.log') SIZE 50K
DATAFILE 'diskc:dbone.dbf' SIZE 30M;

The first file specification in the LOGFILE clause specifies a redo log file group with the
GROUP value 1. This group has members named 'diska:log1.log' and
'diskb:log1.log', each 50 kilobytes in size.
The second file specification in the LOGFILE clause specifies a redo log file group with
the GROUP value 2. This group has members named 'diska:log2.log' and
'diskb:log2.log', also 50 kilobytes in size.
The file specification in the DATAFILE clause specifies a data file named
'diskc:dbone.dbf', 30 megabytes in size.
Each file specification specifies a value for the SIZE parameter and omits the REUSE
clause, so none of these files can already exist. Oracle must create them.
Adding a Log File: Example The following statement adds another redo log file
group with two members to the payable database:
ALTER DATABASE payable
ADD LOGFILE GROUP 3 ('diska:log3.log', 'diskb:log3.log')
SIZE 50K REUSE;

The file specification in the ADD LOGFILE clause specifies a new redo log file group with
the GROUP value 3. This new group has members named 'diska:log3.log' and
'diskb:log3.log', each 50 kilobytes in size. Because the file specification specifies the
REUSE clause, each member can (but need not) already exist.
The following statement adds a logfile group 5 with member log files on migration
target disks 4k_disk_a and 4k_disk_b. After executing this statement, you can switch
existing log files on disks with 512-byte block size to logs with 4K block size using the
switch_logfile_clause on page 10-32.
ALTER DATABASE ADD LOGFILE GROUP 5
('4k_disk_a:log5.log', '4k_disk_b:log5.log')
SIZE 100M BLOCKSIZE 4096 REUSE;

Specifying a Data File: Example

The following statement creates a tablespace named

stocks that has three data files:
CREATE TABLESPACE stocks
DATAFILE 'stock1.dbf' SIZE 10M,
'stock2.dbf' SIZE 10M,
'stock3.dbf' SIZE 10M;

The file specifications for the data files specify files named 'diskc:stock1.dbf',
'diskc:stock2.dbf', and 'diskc:stock3.dbf'.
Adding a Data File: Example

The following statement alters the stocks tablespace

and adds a new data file:
ALTER TABLESPACE stocks
ADD DATAFILE 'stock4.dbf' SIZE 10M REUSE;

The file specification specifies a data file named 'stock4.dbf'. If the filename does not
exist, then Oracle simply ignores the REUSE keyword.

8-36 Oracle Database SQL Language Reference

file_specification

When using Oracle
ASM, the following syntax shows how to use the fully_qualified_file_name clause
to bring online a data file in a hypothetical database, testdb:

Using a Fully Qualified Oracle ASM Data File Name: Example

ALTER DATABASE testdb
DATAFILE '+dgroup_01/testdb/datafile/system.261.1' ONLINE;

Common SQL DDL Clauses 8-37

logging_clause

logging_clause
Purpose
8

The logging_clause lets you specify whether certain DML operations will be logged
in the redo log file (LOGGING) or not (NOLOGGING).
You can specify the logging_clause in the following statements:
■

CREATE TABLE and ALTER TABLE: for logging of the table, a table partition, a LOB
segment, or the overflow segment of an index-organized table (see CREATE
TABLE on page 16-6 and ALTER TABLE on page 12-2).
Logging specified for a LOB column can differ from logging
set at the table level. If you specify LOGGING at the table level and
NOLOGGING for a LOB column, then DML changes to the base table row
are logged, but DML changes to the LOB data are not logged.
Note:

■

■

■

■

CREATE INDEX and ALTER INDEX: for logging of the index or an index partition (see
CREATE INDEX on page 14-60 and ALTER INDEX on page 10-78).
CREATE MATERIALIZED VIEW and ALTER MATERIALIZED VIEW: for logging of the
materialized view, one of its partitions, or a LOB segment (see CREATE
MATERIALIZED VIEW on page 15-4 and ALTER MATERIALIZED VIEW on
page 11-3).
CREATE MATERIALIZED VIEW LOG and ALTER MATERIALIZED VIEW LOG: for logging of
the materialized view log or one of its partitions (see CREATE MATERIALIZED
VIEW LOG on page 15-27 and ALTER MATERIALIZED VIEW LOG on
page 11-18).
CREATE TABLESPACE and ALTER TABLESPACE: to set or modify the default logging
characteristics for all objects created in the tablespace (see CREATE TABLESPACE
on page 16-83 and ALTER TABLESPACE on page 12-90).

You can also specify LOGGING or NOLOGGING for the following operations:
■

Rebuilding an index (using CREATE INDEX ... REBUILD)

■

Moving a table (using ALTER TABLE ... MOVE)

Syntax
8

logging_clause::=
LOGGING
NOLOGGING
FILESYSTEM_LIKE_LOGGING

Semantics
8

This section describes the semantics of the logging_clause. For additional
information, refer to the SQL statement in which you set or reset logging
characteristics for a particular database object.

8-38 Oracle Database SQL Language Reference

logging_clause

■

■

■

If you specify LOGGING, then the creation of a database object, as well as
subsequent inserts into the object, will be logged in the redo log file.
If you specify NOLOGGING, then the creation of a database object, as well as
subsequent conventional inserts, will be logged in the redo log file. Direct-path
inserts will not be logged.
–

For a nonpartitioned object, the value specified for this clause is the actual
physical attribute of the segment associated with the object.

–

For partitioned objects, the value specified for this clause is the default
physical attribute of the segments associated with all partitions specified in the
CREATE statement (and in subsequent ALTER ... ADD PARTITION statements),
unless you specify the logging attribute in the PARTITION description.

–

For SecureFiles LOBs, the NOLOGGING setting is converted internally to
FILESYSTEM_LIKE_LOGGING.

–

CACHE NOLOGGING is not allowed for BasicFiles LOBs.

The FILESYSTEM_LIKE_LOGGING clause is valid only for logging of SecureFiles LOB
segments. You cannot specify this setting for BasicFiles LOBs. Specify this setting if
you want to log only metadata changes. This setting is similar to the metadata
journaling of file systems, which reduces mean time to recovery from failures. The
LOGGING setting, for SecureFiles LOBs, is similar to the data journaling of file
systems. Both the LOGGING and FILESYSTEM_LIKE_LOGGING settings provide a
complete transactional file system by way of SecureFiles.

Note: For LOB segments, with the NOLOGGING and FILESYSTEM_LIKE_
LOGGING settings it is possible for data to be changed on disk during a
backup operation, resulting in an inconsistent backup. To avoid this
situation, ensure that changes to LOB segments are saved in the redo
log file by setting LOGGING for LOB storage. Alternatively, change the
database to FORCE LOGGING mode so that changes to all LOB segments
are saved in the redo.

If the object for which you are specifying the logging attributes resides in a database or
tablespace in force logging mode, then Oracle Database ignores any NOLOGGING setting
until the database or tablespace is taken out of force logging mode.
If the database is running in ARCHIVELOG mode, then media recovery from a backup
made before the LOGGING operation re-creates the object. However, media recovery
from a backup made before the NOLOGGING operation does not re-create the object.
The size of a redo log generated for an operation in NOLOGGING mode is significantly
smaller than the log generated in LOGGING mode.
In NOLOGGING mode, data is modified with minimal logging (to mark new extents
INVALID and to record dictionary changes). When applied during media recovery, the
extent invalidation records mark a range of blocks as logically corrupt, because the
redo data is not fully logged. Therefore, if you cannot afford to lose the database
object, then you should take a backup after the NOLOGGING operation.
NOLOGGING is supported in only a subset of the locations that support LOGGING. Only
the following operations support the NOLOGGING mode:
DML:

Common SQL DDL Clauses 8-39

logging_clause

■

■

Direct-path INSERT (serial or parallel) resulting either from an INSERT or a MERGE
statement. NOLOGGING is not applicable to any UPDATE operations resulting from the
MERGE statement.
Direct Loader (SQL*Loader)

DDL:
■

■

■

CREATE TABLE ... AS SELECT (In NOLOGGING mode, the creation of the table will be
logged, but direct-path inserts will not be logged.)
ALTER TABLE ... LOB_storage_clause ... LOB_parameters ... CACHE | NOCACHE |
CACHE READS (to specify logging of newly created LOB columns)
ALTER TABLE ... modify_LOB_storage_clause ... modify_LOB_parameters ... CACHE |
NOCACHE | CACHE READS (to change logging of existing LOB columns)

■

ALTER TABLE ... MOVE

■

ALTER TABLE ... (all partition operations that involve data movement)
–

ALTER TABLE ... ADD PARTITION (hash partition only)

–

ALTER TABLE ... MERGE PARTITIONS

–

ALTER TABLE ... SPLIT PARTITION

–

ALTER TABLE ... MOVE PARTITION

–

ALTER TABLE ... MODIFY PARTITION ... ADD SUBPARTITION

–

ALTER TABLE ... MODIFY PARTITION ... COALESCE SUBPARTITION

■

CREATE INDEX

■

ALTER INDEX ... REBUILD

■

ALTER INDEX ... REBUILD [SUB]PARTITION

■

ALTER INDEX ... SPLIT PARTITION

For objects other than LOBs, if you omit this clause, then the logging attribute of the
object defaults to the logging attribute of the tablespace in which it resides.
For LOBs, if you omit this clause, then:
■

■

If you specify CACHE, then LOGGING is used (because you cannot have CACHE
NOLOGGING).
If you specify NOCACHE or CACHE READS, then the logging attribute defaults to the
logging attribute of the tablespace in which it resides.

NOLOGGING does not apply to LOBs that are stored internally (in the table with row
data). If you specify NOLOGGING for LOBs with values less than 4000 bytes and you
have not disabled STORAGE IN ROW, then Oracle ignores the NOLOGGING specification and
treats the LOB data the same as other table data.

8-40 Oracle Database SQL Language Reference

parallel_clause

parallel_clause
Purpose
8

The parallel_clause lets you parallelize the creation of a database object and set the
default degree of parallelism for subsequent queries of and DML operations on the
object.
You can specify the parallel_clause in the following statements:
■

CREATE TABLE: to set parallelism for the table (see CREATE TABLE on page 16-6).

■

ALTER TABLE (see ALTER TABLE on page 12-2):

■

–

To change parallelism for the table

–

To parallelize the operations of adding, coalescing, exchanging, merging,
splitting, truncating, dropping, or moving a table partition

CREATE CLUSTER and ALTER CLUSTER: to set or alter parallelism for a cluster (see
CREATE CLUSTER on page 14-2 and ALTER CLUSTER on page 10-5).

■

CREATE INDEX: to set parallelism for the index (see CREATE INDEX on page 14-60).

■

ALTER INDEX (see ALTER INDEX on page 10-78):

■

■

■

■

■

■

–

To change parallelism for the index

–

To parallelize the rebuilding of the index or the splitting of an index partition

CREATE MATERIALIZED VIEW: to set parallelism for the materialized view (see
CREATE MATERIALIZED VIEW on page 15-4).
ALTER MATERIALIZED VIEW (see ALTER MATERIALIZED VIEW on page 11-3):
–

To change parallelism for the materialized view

–

To parallelize the operations of adding, coalescing, exchanging, merging,
splitting, truncating, dropping, or moving a materialized view partition

–

To parallelize the operations of adding or moving materialized view
subpartitions

CREATE MATERIALIZED VIEW LOG: to set parallelism for the materialized view log
(see CREATE MATERIALIZED VIEW LOG on page 15-27).
ALTER MATERIALIZED VIEW LOG (see ALTER MATERIALIZED VIEW LOG on
page 11-18):
–

To change parallelism for the materialized view log

–

To parallelize the operations of adding, coalescing, exchanging, merging,
splitting, truncating, dropping, or moving a materialized view log partition

ALTER DATABASE ... RECOVER: to recover the database (see ALTER DATABASE on
page 10-8).
ALTER DATABASE ... standby_database_clauses: to parallelize operations on the
standby database (see ALTER DATABASE on page 10-8).
See Also: Oracle Database PL/SQL Packages and Types Reference for
information on the DBMS_PARALLEL_EXECUTE package, which provides
methods to apply table changes in chunks of rows. Changes to each
chunk are independently committed when there are no errors.

Common SQL DDL Clauses 8-41

parallel_clause

Syntax
8

parallel_clause::=
NOPARALLEL
integer
PARALLEL

Semantics
8

This section describes the semantics of the parallel_clause. For additional
information, refer to the SQL statement in which you set or reset parallelism for a
particular database object or operation.

Note: The syntax of the parallel_clause supersedes syntax
appearing in earlier releases of Oracle. Superseded syntax is still
supported for backward compatibility but may result in slightly
different behavior from that documented.

The database interprets the parallel_clause based on the setting of the PARALLEL_
DEGREE_POLICY initialization parameter. When that parameter is set to AUTO, the
parallel_clause is ignored entirely, and the optimizer determines the best degree of
parallelism for all statements. When PARALLEL_DEGREE_POLICY is set to either MANUAL
or LIMITED, the parallel_clause is interpreted as follows:
NOPARALLEL
PARALLEL
■

■

Specify NOPARALLEL for serial execution. This is the default.

Specify PARALLEL for parallel execution.

If PARALLEL_DEGREE_POLICY is set to MANUAL, then the optimizer calculates a degree
of parallelism equal to the number of CPUs available on all participating instances
times the value of the PARALLEL_THREADS_PER_CPU initialization parameter.
If PARALLEL_DEGREE_POLICY is set to LIMITED, then the optimizer determines the
best degree of parallelism.

PARALLEL integer Specification of integer indicates the degree of parallelism,
which is the number of parallel threads used in the parallel operation. Each parallel
thread may use one or two parallel execution servers.
Notes on the parallel_clause
■

■

■

■

The following notes apply to the parallel_clause:

Parallelism is disabled for DML operations on tables on which you have defined a
trigger or referential integrity constraint.
Parallelism is not supported for UPDATE or DELETE operations on index-organized
tables.
When you specify the parallel_clause during creation of a table, if the table
contains any columns of LOB or user-defined object type, then subsequent INSERT,
UPDATE, DELETE or MERGE operations that modify the LOB or object type column are
executed serially without notification. Subsequent queries, however, will be
executed in parallel.
A parallel hint overrides the effect of the parallel_clause.

8-42 Oracle Database SQL Language Reference

parallel_clause

■

■

DML statements and CREATE TABLE ... AS SELECT statements that reference remote
objects can run in parallel. However, the remote object must really be on a remote
database. The reference cannot loop back to an object on the local database, for
example, by way of a synonym on the remote database pointing back to an object
on the local database.
DML operations on tables with LOB columns can be parallelized. However,
intrapartition parallelism is not supported.
See Also: Oracle Database VLDB and Partitioning Guide for more
information on parallelized operations, and "Creating a Table:
Parallelism Examples" on page 16-72

Common SQL DDL Clauses 8-43

physical_attributes_clause

physical_attributes_clause
Purpose
8

The physical_attributes_clause lets you specify the value of the PCTFREE, PCTUSED,
and INITRANS parameters and the storage characteristics of a table, cluster, index, or
materialized view.
You can specify the physical_attributes_clause in the following statements:
■

■

■

■

■

■

■

■

CREATE CLUSTER and ALTER CLUSTER: to set or change the physical attributes of the
cluster and all tables in the cluster (see CREATE CLUSTER on page 14-2 and
ALTER CLUSTER on page 10-5).
CREATE TABLE: to set the physical attributes of the table, a table partition, the
OIDINDEX of an object table, or the overflow segment of an index-organized table
(see CREATE TABLE on page 16-6).
ALTER TABLE: to change the physical attributes of the table, the default physical
attributes of future table partitions, or the physical attributes of existing table
partitions (see ALTER TABLE on page 12-2). The following restrictions apply:
–

You cannot specify physical attributes for a temporary table.

–

You cannot specify physical attributes for a clustered table. Tables in a cluster
inherit the physical attributes of the cluster.

CREATE INDEX: to set the physical attributes of an index or index partition (see
CREATE INDEX on page 14-60).
ALTER INDEX: to change the physical attributes of the index, the default physical
attributes of future index partitions, or the physical attributes of existing index
partitions (see ALTER INDEX on page 10-78).
CREATE MATERIALIZED VIEW: to set the physical attributes of the materialized view,
one of its partitions, or the index Oracle Database generates to maintain the
materialized view (see CREATE MATERIALIZED VIEW on page 15-4).
ALTER MATERIALIZED VIEW: to change the physical attributes of the materialized
view, the default physical attributes of future partitions, the physical attributes of
an existing partition, or the index Oracle creates to maintain the materialized view
(see ALTER MATERIALIZED VIEW on page 11-3).
CREATE MATERIALIZED VIEW LOG and ALTER MATERIALIZED VIEW LOG: to set or change
the physical attributes of the materialized view log (see CREATE MATERIALIZED
VIEW LOG on page 15-27 and ALTER MATERIALIZED VIEW LOG on
page 11-18).

Syntax
8

physical_attributes_clause::=
PCTFREE

integer

PCTUSED

integer

INITRANS

integer

storage_clause

8-44 Oracle Database SQL Language Reference

physical_attributes_clause

(storage_clause::= on page 8-50)

Semantics
8

This section describes the parameters of the physical_attributes_clause. For
additional information, refer to the SQL statement in which you set or reset these
parameters for a particular database object.

PCTFREE integer
Specify a whole number representing the percentage of space in each data block of the
database object reserved for future updates to rows of the object. The value of PCTFREE
must be a value from 0 to 99. A value of 0 means that the entire block can be filled by
inserts of new rows. The default value is 10. This value reserves 10% of each block for
updates to existing rows and allows inserts of new rows to fill a maximum of 90% of
each block.
PCTFREE has the same function in the statements that create and alter tables, partitions,
clusters, indexes, materialized views, and materialized view logs. The combination of
PCTFREE and PCTUSED determines whether new rows will be inserted into existing data
blocks or into new blocks. See "How PCTFREE and PCTUSED Work Together" on
page 8-45.
Restriction on the PCTFREE Clause When altering an index, you can specify this

parameter only in the modify_index_default_attrs clause and the split_index_
partition clause.

PCTUSED integer
Specify a whole number representing the minimum percentage of used space that
Oracle maintains for each data block of the database object. PCTUSED is specified as a
positive integer from 0 to 99 and defaults to 40.
PCTUSED has the same function in the statements that create and alter tables, partitions,
clusters, materialized views, and materialized view logs.
PCTUSED is not a valid table storage characteristic for an index-organized table.
The sum of PCTFREE and PCTUSED must be equal to or less than 100. You can use
PCTFREE and PCTUSED together to utilize space within a database object more efficiently.
See "How PCTFREE and PCTUSED Work Together" on page 8-45.
Restrictions on the PCTUSED Clause The PCTUSED parameter is subject to the
following restrictions:
■

■

You cannot specify this parameter for an index or for the index segment of an
index-organized table.
This parameter is not useful and is ignored for objects with automatic
segment-space management.
See Also: Oracle Database Performance Tuning Guide for information
on the performance effects of different values of PCTUSED and PCTFREE
and CREATE TABLESPACE segment_management_clause on page 16-91 for
information on automatic segment-space management

How PCTFREE and PCTUSED Work Together
In a newly allocated data block, the space available for inserts is the block size minus
the sum of the block overhead and free space (PCTFREE). Updates to existing data can

Common SQL DDL Clauses 8-45

physical_attributes_clause

use any available space in the block. Therefore, updates can reduce the available space
of a block to less than PCTFREE.
After a data block is filled to the limit determined by PCTFREE, Oracle Database
considers the block unavailable for the insertion of new rows until the percentage of
that block falls beneath the parameter PCTUSED. Until this value is achieved, Oracle
Database uses the free space of the data block only for updates to rows already
contained in the data block. A block becomes a candidate for row insertion when its
used space falls below PCTUSED.
See Also: FREELISTS on page 8-53 for information on how PCTUSED
and PCTFREE work with freelist segment space management

INITRANS integer
Specify the initial number of concurrent transaction entries allocated within each data
block allocated to the database object. This value can range from 1 to 255 and defaults
to 1, with the following exceptions:
■

■

The default INITRANS value for a cluster is 2 or the default INITRANS value of the
tablespace in which the cluster resides, whichever is greater.
The default value for an index is 2.

In general, you should not change the INITRANS value from its default.
Each transaction that updates a block requires a transaction entry in the block. This
parameter ensures that a minimum number of concurrent transactions can update the
block and helps avoid the overhead of dynamically allocating a transaction entry.
The INITRANS parameter serves the same purpose in the statements that create and
alter tables, partitions, clusters, indexes, materialized views, and materialized view
logs.

MAXTRANS Parameter
In earlier releases, the MAXTRANS parameter determined the maximum number of
concurrent update transactions allowed for each data block in the segment. This
parameter has been deprecated. Oracle now automatically allows up to 255 concurrent
update transactions for any data block, depending on the available space in the block.
Existing objects for which a value of MAXTRANS has already been set retain that setting.
However, if you attempt to change the value for MAXTRANS, Oracle ignores the new
specification and substitutes the value 255 without returning an error.
storage_clause
The storage_clause lets you specify storage characteristics for the table, object table
OIDINDEX, partition, LOB data segment, or index-organized table overflow data
segment. This clause has performance ramifications for large tables. Storage should be
allocated to minimize dynamic allocation of additional space. Refer to the storage_
clause on page 8-48 for more information.

8-46 Oracle Database SQL Language Reference

size_clause

size_clause
Purpose
8

The size_clause lets you specify a number of bytes, kilobytes (K), megabytes (M),
gigabytes (G), terabytes (T), petabytes (P), or exabytes (E) in any statement that lets
you establish amounts of disk or memory space.

Syntax
8

size_clause::=
K
M
G
T
P
E
integer

Semantics
8

Use the size_clause to specify a number or multiple of bytes. If you do not specify
any of the multiple abbreviations, then the integer is interpreted as bytes.
Not all multiples of bytes are appropriate in all cases, and
context-sensitive limitations may apply. In the latter case, Oracle
issues an error message.

Note:

Common SQL DDL Clauses 8-47

storage_clause

storage_clause
Purpose
8

The storage_clause lets you specify how Oracle Database should store a permanent
database object. Storage parameters for temporary segments always use the default
storage parameters for the associated tablespace. Storage parameters affect both how
long it takes to access data stored in the database and how efficiently space in the
database is used.
See Also: Oracle Automatic Storage Management Administrator's Guide
for a discussion of the effects of the storage parameters

When you create a cluster, index, materialized view, materialized view log, rollback
segment, table, LOB, varray, nested table, or partition, you can specify values for the
storage parameters for the segments allocated to these objects. If you omit any storage
parameter, then Oracle uses the value of that parameter specified for the tablespace in
which the object resides. If no value was specified for the tablespace, then the database
uses default values.
The specification of storage parameters for objects in locally
managed tablespaces is supported for backward compatibility. If you
are using locally managed tablespaces, then you can omit these
storage parameter when creating objects in those tablespaces.

Note:

When you alter a cluster, index, materialized view, materialized view log, rollback
segment, table, varray, nested table, or partition, you can change the values of storage
parameters. The new values affect only future extent allocations.
The storage_clause is part of the physical_attributes_clause, so you can specify
this clause in any of the statements where you can specify the physical attributes
clause (see physical_attributes_clause on page 8-44). In addition, you can specify the
storage_clause in the following statements:
■

■

■

■

■

CREATE CLUSTER and ALTER CLUSTER: to set or change the storage characteristics of
the cluster and all tables in the cluster (see CREATE CLUSTER on page 14-2 and
ALTER CLUSTER on page 10-5).
CREATE INDEX and ALTER INDEX: to set or change the storage characteristics of an
index segment created for a table index or index partition or an index segment
created for an index used to enforce a primary key or unique constraint (see
CREATE INDEX on page 14-60 and ALTER INDEX on page 10-78).
The ENABLE ... USING INDEX clause of CREATE TABLE or ALTER TABLE: to set or change
the storage characteristics of an index created by the system to enforce a primary
key or unique constraint.
CREATE MATERIALIZED VIEW and ALTER MATERIALIZED VIEW: to set or change the
storage characteristics of a materialized view, one of its partitions, or the index
Oracle generates to maintain the materialized view (see CREATE
MATERIALIZED VIEW on page 15-4 and ALTER MATERIALIZED VIEW on
page 11-3).
CREATE MATERIALIZED VIEW LOG and ALTER MATERIALIZED VIEW LOG: to set or change
the storage characteristics of the materialized view log (see CREATE
MATERIALIZED VIEW LOG on page 15-27 and ALTER MATERIALIZED VIEW

8-48 Oracle Database SQL Language Reference

storage_clause

LOG on page 11-18).
■

■

■

■

CREATE ROLLBACK SEGMENT and ALTER ROLLBACK SEGMENT: to set or change the
storage characteristics of a rollback segment (see CREATE ROLLBACK
SEGMENT on page 15-62 and ALTER ROLLBACK SEGMENT on page 11-40).
CREATE TABLE and ALTER TABLE: to set the storage characteristics of a LOB or varray
data segment of the nonclustered table or one of its partitions or subpartitions, or
the storage table of a nested table (see CREATE TABLE on page 16-6 and ALTER
TABLE on page 12-2).
CREATE TABLESPACE and ALTER TABLESPACE: to set or change the default storage
characteristics for objects created in the tablespace (see CREATE TABLESPACE
on page 16-83 and ALTER TABLESPACE on page 12-90). Changes to tablespace
storage parameters affect only new objects created in the tablespace or new extents
allocated for a segment.
constraint: to specify storage for the index (and its partitions, if it is a
partitioned index) used to enforce the constraint (see constraint on page 8-4).

Prerequisites
8

To change the value of a STORAGE parameter, you must have the privileges necessary to
use the appropriate CREATE or ALTER statement.

Common SQL DDL Clauses 8-49

storage_clause

Syntax
8

storage_clause::=
INITIAL
NEXT

size_clause
size_clause

MINEXTENTS

integer
integer

MAXEXTENTS
UNLIMITED
maxsize_clause
PCTINCREASE

STORAGE

integer

FREELISTS

integer

FREELIST

GROUPS

(

integer

size_clause

)

NULL
OPTIMAL
KEEP
BUFFER_POOL

RECYCLE
DEFAULT
KEEP

FLASH_CACHE

NONE
DEFAULT

ENCRYPT

(size_clause::= on page 8-47
maxsize_clause::=
UNLIMITED
MAXSIZE
size_clause

(size_clause::= on page 8-47

Semantics
8

This section describes the parameters of the storage_clause. For additional
information, refer to the SQL statement in which you set or reset these storage
parameters for a particular database object.

8-50 Oracle Database SQL Language Reference

storage_clause

Note: The storage_clause is interpreted differently for locally
managed tablespaces. For locally managed tablespaces, Oracle
Database uses INITIAL, NEXT, PCTINCREASE, and MINEXTENTS to
compute how many extents are allocated when the object is first
created. After object creation, these parameters are ignored. For more
information, see CREATE TABLESPACE on page 16-83.

See Also: "Specifying Table Storage Attributes: Example" on
page 8-56

INITIAL
Specify the size of the first extent of the object. Oracle allocates space for this extent
when you create the schema object. Refer to size_clause on page 8-47 for information on
that clause.
In locally managed tablespaces, Oracle uses the value of INITIAL, in conjunction with
the type of local management—AUTOALLOCATE or UNIFORM—and the values of
MINEXTENTS, NEXT and PCTINCREASE, to determine the initial size of the segment.
■

■

With AUTOALLOCATE extent management, Oracle uses the INITIAL setting to
optimize the number of extents allocated. Extents of 64K, 1M, 8M, and 64M can be
allocated. During segment creation, the system chooses the greatest of these four
sizes that is equal to or smaller than INITIAL, and allocates as many extents of that
size as are needed to reach the INITIAL setting. For example, if you set INITIAL to
4M, then the database creates four 1M extents.
For UNIFORM extent management, the number of extents is determined from initial
segment size and the uniform extent size specified at tablespace creation time. For
example, in a uniform locally managed tablespace with 1M extents, if you specify
an INITIAL value of 5M, then Oracle creates five 1M extents.
Consider this comparison: With AUTOALLOCATE, if you set INITAL to 72K, then the
initial segment size will be 128K (greater than INITIAL). The database cannot
allocate an extent smaller than 64K, so it must allocate two 64K extents. If you set
INITIAL to 72K with a UNIFORM extent size of 24K, then the database will allocate
three 24K extents to equal 72K.

In dictionary managed tablespaces, the default initial extent size is 5 blocks, and all
subsequent extents are rounded to 5 blocks. If MINIMUM EXTENT was specified at
tablespace creation time, then the extent sizes are rounded to the value of MINIMUM
EXTENT.
Restriction on INITIAL You cannot specify INITIAL in an ALTER statement.

NEXT
Specify in bytes the size of the next extent to be allocated to the object. Refer to size_
clause on page 8-47 for information on that clause.
In locally managed tablespaces, any user-supplied value for NEXT is ignored and the
size of NEXT is determined by Oracle if the tablespace is set for autoallocate extent
management. In UNIFORM tablespaces, the size of NEXT is the uniform extent size
specified at tablespace creation time.
In dictionary-managed tablespaces, the default value is the size of 5 data blocks. The
minimum value is the size of 1 data block. The maximum value depends on your
operating system. Oracle rounds values up to the next multiple of the data block size

Common SQL DDL Clauses 8-51

storage_clause

for values less than 5 data blocks. For values greater than 5 data blocks, Oracle rounds
up to a value that minimizes fragmentation.
Oracle Database Concepts for information on how Oracle
minimizes fragmentation

See Also:

PCTINCREASE
In locally managed tablespaces, Oracle Database uses the value of PCTINCREASE during
segment creation to determine the initial segment size and ignores this parameter
during subsequent space allocation.
In dictionary-managed tablespaces, specify the percent by which the third and
subsequent extents grow over the preceding extent. The default value is 50, meaning
that each subsequent extent is 50% larger than the preceding extent. The minimum
value is 0, meaning all extents after the first are the same size. The maximum value
depends on your operating system. Oracle rounds the calculated size of each new
extent to the nearest multiple of the data block size. If you change the value of the
PCTINCREASE parameter by specifying it in an ALTER statement, then Oracle calculates
the size of the next extent using this new value and the size of the most recently
allocated extent.
Restriction on PCTINCREASE You cannot specify PCTINCREASE for rollback

segments. Rollback segments always have a PCTINCREASE value of 0.

MINEXTENTS
In locally managed tablespaces, Oracle Database uses the value of MINEXTENTS in
conjunction with PCTINCREASE, INITIAL and NEXT to determine the initial segment size.
In dictionary-managed tablespaces, specify the total number of extents to allocate
when the object is created. The default and minimum value is 1, meaning that Oracle
allocates only the initial extent, except for rollback segments, for which the default and
minimum value is 2. The maximum value depends on your operating system.
■

■

In a locally managed tablespace, MINEXTENTS is used to compute the initial amount
of space allocated, which is equal to INITIAL * MINEXTENTS. Thereafter this value is
set to 1, which is reflected in the DBA_SEGMENTS view.
In a dictionary-managed tablespace, MINEXTENTS is simply the minimum number
of extents that must be allocated to the segment.

If the MINEXTENTS value is greater than 1, then Oracle calculates the size of subsequent
extents based on the values of the INITIAL, NEXT, and PCTINCREASE storage parameters.
When changing the value of MINEXTENTS by specifying it in an ALTER statement, you
can reduce the value from its current value, but you cannot increase it. Resetting
MINEXTENTS to a smaller value might be useful, for example, before a TRUNCATE ... DROP
STORAGE statement, if you want to ensure that the segment will maintain a minimum
number of extents after the TRUNCATE operation.
Restrictions on MINEXTENTS

The MINEXTENTS storage parameter is subject to the

following restrictions:
■
■

MINEXTENTS is not applicable at the tablespace level.
You cannot change the value of MINEXTENTS in an ALTER statement or for an object
that resides in a locally managed tablespace.

8-52 Oracle Database SQL Language Reference

storage_clause

MAXEXTENTS
This storage parameter is valid only for objects in dictionary-managed tablespaces.
Specify the total number of extents, including the first, that Oracle can allocate for the
object. The minimum value is 1 except for rollback segments, which always have a
minimum of 2. The default value depends on your data block size.
Restriction on MAXEXTENTS MAXEXTENTS is ignored for objects residing in a locally
managed tablespace, unless the value of ALLOCATION_TYPE is USER for the tablespace in
the DBA_TABLESPACES data dictionary view.

Oracle Database Reference for more information on the DBA_
TABLESPACES data dictionary view

See Also:

UNLIMITED Specify UNLIMITED if you want extents to be allocated automatically as
needed. Oracle recommends this setting as a way to minimize fragmentation.

Do not use this clause for rollback segments. Doing so allows the possibility that
long-running rogue DML transactions will continue to create new extents until a disk
is full.
A rollback segment that you create without specifying the
storage_clause has the same storage parameters as the tablespace in
which the rollback segment is created. Thus, if you create a tablespace
with MAXEXTENTS UNLIMITED, then the rollback segment will have this
same default.
Caution:

MAXSIZE
The MAXSIZE clause lets you specify the maximum size of the storage element. For
LOB storage, MAXSIZE has the following effects
■

■

If you specify RETENTION MAX in LOB_parameters, then the LOB segment increases
to the specified size before any space can be reclaimed from undo space.
If you specify RETENTION AUTO, MIN, or NONE in LOB_parameters, then the specified
size is a hard limit on the LOB segment size and has no bearing on undo
retention.

UNLIMITED Use the UNLIMITED clause if you do not want to limit the disk space of

the storage element. This clause is not compatible with a specification of RETENTION
MAX in LOB_parameters. If you specify both, then the database uses RETENTION AUTO
and MAXSIZE UNLIMITED.

FREELISTS
In tablespaces with manual segment-space management, Oracle Database uses the
FREELISTS storage parameter to improve performance of space management in OLTP
systems by increasing the number of insert points in the segment. In tablespaces with
automatic segment-space management, this parameter is ignored, because the
database adapts to varying workload.
In tablespaces with manual segment-space management, for objects other than
tablespaces and rollback segments, specify the number of free lists for each of the free
list groups for the table, partition, cluster, or index. The default and minimum value
for this parameter is 1, meaning that each free list group contains one free list. The
maximum value of this parameter depends on the data block size. If you specify a

Common SQL DDL Clauses 8-53

storage_clause

FREELISTS value that is too large, then Oracle returns an error indicating the maximum
value.
This clause is not valid or useful if you have specified the SECUREFILE parameter of
LOB_parameters on page 16-44. If you specify both the SECUREFILE parameter and
FREELISTS, then the database silently ignores the FREELISTS specification.
Restriction on FREELISTS You can specify FREELISTS in the storage_clause of any
statement except when creating or altering a tablespace or rollback segment.

FREELIST GROUPS
In tablespaces with manual segment-space management, Oracle Database uses the
value of this storage parameter to statically partition the segment free space in an
Oracle Real Application Clusters environment. This partitioning improves the
performance of space allocation and deallocation by avoiding inter instance transfer of
segment metadata. In tablespaces with automatic segment-space management, this
parameter is ignored, because Oracle dynamically adapts to inter instance workload.
In tablespaces with manual segment-space management, specify the number of groups
of free lists for the database object you are creating. The default and minimum value
for this parameter is 1. Oracle uses the instance number of Oracle Real Application
Clusters (Oracle RAC) instances to map each instance to one free list group.
Each free list group uses one database block. Therefore:
■

■

If you do not specify a large enough value for INITIAL to cover the minimum
value plus one data block for each free list group, then Oracle increases the value
of INITIAL the necessary amount.
If you are creating an object in a uniform locally managed tablespace, and the
extent size is not large enough to accommodate the number of freelist groups, then
the create operation will fail.

This clause is not valid or useful if you have specified the SECUREFILE parameter of
LOB_parameters on page 16-44. If you specify both the SECUREFILE parameter and
FREELIST GROUPS, then the database silently ignores the FREELIST GROUPS specification.
Restriction on FREELIST GROUPS You can specify the FREELIST GROUPS parameter
only in CREATE TABLE, CREATE CLUSTER, CREATE MATERIALIZED VIEW, CREATE
MATERIALIZED VIEW LOG, and CREATE INDEX statements.

OPTIMAL
The OPTIMAL keyword is relevant only to rollback segments. It specifies an optimal size
in bytes for a rollback segment. Refer to size_clause on page 8-47 for information on
that clause.
Oracle tries to maintain this size for the rollback segment by dynamically deallocating
extents when their data is no longer needed for active transactions. Oracle deallocates
as many extents as possible without reducing the total size of the rollback segment
below the OPTIMAL value.
The value of OPTIMAL cannot be less than the space initially allocated by the
MINEXTENTS, INITIAL, NEXT, and PCTINCREASE parameters. The maximum value
depends on your operating system. Oracle rounds values up to the next multiple of
the data block size.
Specify NULL for no optimal size for the rollback segment, meaning that Oracle
never deallocates the extents of the rollback segment. This is the default behavior.

NULL

8-54 Oracle Database SQL Language Reference

storage_clause

BUFFER_POOL
The BUFFER_POOL clause lets you specify a default buffer pool or cache for a schema
object. All blocks for the object are stored in the specified cache.
■

■

If you define a buffer pool for a partitioned table or index, then the partitions
inherit the buffer pool from the table or index definition unless overridden by a
partition-level definition.
For an index-organized table, you can specify a buffer pool separately for the
index segment and the overflow segment.

Restrictions on the BUFFER_POOL Parameter BUFFER_POOL is subject to the
following restrictions:
■

■

You cannot specify this clause for a cluster table. However, you can specify it for a
cluster.
You cannot specify this clause for a tablespace or a rollback segment.

KEEP Specify KEEP to put blocks from the segment into the KEEP buffer pool.
Maintaining an appropriately sized KEEP buffer pool lets Oracle retain the schema
object in memory to avoid I/O operations. KEEP takes precedence over any NOCACHE
clause you specify for a table, cluster, materialized view, or materialized view log.

Specify RECYCLE to put blocks from the segment into the RECYCLE pool. An
appropriately sized RECYCLE pool reduces the number of objects whose default pool is
the RECYCLE pool from taking up unnecessary cache space.
RECYCLE

DEFAULT Specify DEFAULT to indicate the default buffer pool. This is the default for

objects not assigned to KEEP or RECYCLE.
See Also: Oracle Database Performance Tuning Guide for more
information about using multiple buffer pools

FLASH_CACHE
The FLASH_CACHE clause lets you override the automatic buffer cache policy and
specify how specific schema objects are cached in flash memory. To use this clause,
Database Smart Flash Cache (flash cache) must be configured on your system. The
flash cache is an extension of the database buffer cache that is stored on a flash disk, a
storage device that uses flash memory. Because flash memory is faster than magnetic
disks, the database can improve performance by caching buffers in the flash cache
instead of reading from magnetic disk.
Specify KEEP if you want the schema object buffers to remain cached in the
flash cache as long as the flash cache is large enough.

KEEP

NONE Specify NONE to ensure that the schema object buffers are never cached in the
flash cache. This allows you to reserve the flash cache space for more frequently
accessed objects.
DEFAULT Specify DEFAULT if you want the schema object buffers to be written to the

flash cache when they are aged out of main memory, and then be aged out of the flash
cache with the standard buffer cache replacement algorithm. This is the default if flash
cache is configured and you do not specify KEEP or NONE.

Common SQL DDL Clauses 8-55

storage_clause

Database Smart Flash Cache is available only in Solaris and
Oracle Linux.

Note:

See Also:
■

■

Oracle Database Concepts for more information about Database
Smart Flash Cache
Oracle Database Administrator's Guide to learn how to configure
Database Smart Flash Cache

ENCRYPT
This clause is valid only when you are creating a tablespace. Specify ENCRYPT to
encrypt the entire tablespace. You must also specify the ENCRYPTION clause in the
CREATE TABLESPACE statement.
See Also: The CREATE TABLESPACE "ENCRYPTION Clause" on
page 16-89

Example
8

Specifying Table Storage Attributes: Example The following statement creates a

table and provides storage parameter values:
CREATE TABLE divisions
(div_no
NUMBER(2),
div_name
VARCHAR2(14),
location
VARCHAR2(13) )
STORAGE ( INITIAL 8M MAXSIZE 1G );

Oracle allocates space for the table based on the STORAGE parameter values as follows:
■

The INITIAL value is 8M, so the size of the first extent is 8 megabytes.

■

The MAXSIZE value is 1G, so the maximum size of the storage element is 1 gigabyte.

8-56 Oracle Database SQL Language Reference

9
9

SQL Queries and Subqueries

This chapter describes SQL queries and subqueries.
This chapter contains these sections:
■

About Queries and Subqueries

■

Creating Simple Queries

■

Hierarchical Queries

■

The UNION [ALL], INTERSECT, MINUS Operators

■

Sorting Query Results

■

Joins

■

Using Subqueries

■

Unnesting of Nested Subqueries

■

Selecting from the DUAL Table

■

Distributed Queries

About Queries and Subqueries
A query is an operation that retrieves data from one or more tables or views. In this
reference, a top-level SELECT statement is called a query, and a query nested within
another SQL statement is called a subquery.
This section describes some types of queries and subqueries and how to use them. The
top level of the syntax is shown in this chapter. Refer to SELECT on page 19-4 for the
full syntax of all the clauses and the semantics of this statement.
select::=
for_update_clause
subquery

;

SQL Queries and Subqueries

9-1

Creating Simple Queries

subquery::=
query_block
ALL
order_by_clause

UNION
subquery

INTERSECT

subquery

MINUS
(

subquery

)

query_block::=
DISTINCT
UNIQUE
subquery_factoring_clause

hint
SELECT

ALL
select_list

,
table_reference
FROM

where_clause

hierarchical_query_clause

group_by_clause

join_clause
(

join_clause

)

model_clause

Creating Simple Queries
The list of expressions that appears after the SELECT keyword and before the FROM
clause is called the select list. Within the select list, you specify one or more columns
in the set of rows you want Oracle Database to return from one or more tables, views,
or materialized views. The number of columns, as well as their data type and length,
are determined by the elements of the select list.
If two or more tables have some column names in common, then you must qualify
column names with names of tables. Otherwise, fully qualified column names are
optional. However, it is always a good idea to qualify table and column references
explicitly. Oracle often does less work with fully qualified table and column names.
You can use a column alias, c_alias, to label the immediately preceding expression in
the select list so that the column is displayed with a new heading. The alias effectively
renames the select list item for the duration of the query. The alias can be used in the
ORDER BY clause, but not other clauses in the query.
You can use comments in a SELECT statement to pass instructions, or hints, to the
Oracle Database optimizer. The optimizer uses hints to choose an execution plan for
the statement. Refer to "Hints" on page 3-74 for more information on hints.

9-2 Oracle Database SQL Language Reference

Hierarchical Queries

Hierarchical Queries
If a table contains hierarchical data, then you can select rows in a hierarchical order
using the hierarchical query clause:
hierarchical_query_clause::=
NOCYCLE
CONNECT

BY

START

WITH

condition

condition
NOCYCLE

START

WITH

condition

CONNECT

BY

condition

condition can be any condition as described in Chapter 7, "Conditions."
START WITH specifies the root row(s) of the hierarchy.
CONNECT BY specifies the relationship between parent rows and child rows of the
hierarchy.
■

■

The NOCYCLE parameter instructs Oracle Database to return rows from a query
even if a CONNECT BY loop exists in the data. Use this parameter along with the
CONNECT_BY_ISCYCLE pseudocolumn to see which rows contain the loop. Refer to
CONNECT_BY_ISCYCLE Pseudocolumn on page 2-1 for more information.
In a hierarchical query, one expression in condition must be qualified with the
PRIOR operator to refer to the parent row. For example,
... PRIOR expr = expr
or
... expr = PRIOR expr

If the CONNECT BY condition is compound, then only one condition requires the
PRIOR operator, although you can have multiple PRIOR conditions. For example:
CONNECT BY last_name != 'King' AND PRIOR employee_id = manager_id ...
CONNECT BY PRIOR employee_id = manager_id and
PRIOR account_mgr_id = customer_id ...

PRIOR is a unary operator and has the same precedence as the unary + and arithmetic operators. It evaluates the immediately following expression for the
parent row of the current row in a hierarchical query.
PRIOR is most commonly used when comparing column values with the equality
operator. (The PRIOR keyword can be on either side of the operator.) PRIOR causes
Oracle to use the value of the parent row in the column. Operators other than the
equal sign (=) are theoretically possible in CONNECT BY clauses. However, the
conditions created by these other operators can result in an infinite loop through
the possible combinations. In this case Oracle detects the loop at run time and
returns an error.
Both the CONNECT BY condition and the PRIOR expression can take the form of an
uncorrelated subquery. However, CURRVAL and NEXTVAL are not valid PRIOR
expressions, so the PRIOR expression cannot refer to a sequence.
You can further refine a hierarchical query by using the CONNECT_BY_ROOT operator to
qualify a column in the select list. This operator extends the functionality of the
CONNECT BY [PRIOR] condition of hierarchical queries by returning not only the
immediate parent row but all ancestor rows in the hierarchy.

SQL Queries and Subqueries

9-3

Hierarchical Queries

See Also: CONNECT_BY_ROOT on page 4-5 for more information
about this operator and "Hierarchical Query Examples" on page 9-5

Oracle processes hierarchical queries as follows:
A join, if present, is evaluated first, whether the join is specified in the FROM clause
or with WHERE clause predicates.

■

■

The CONNECT BY condition is evaluated.

■

Any remaining WHERE clause predicates are evaluated.

Oracle then uses the information from these evaluations to form the hierarchy using
the following steps:
1.

Oracle selects the root row(s) of the hierarchy—those rows that satisfy the START
WITH condition.

2.

Oracle selects the child rows of each root row. Each child row must satisfy the
condition of the CONNECT BY condition with respect to one of the root rows.

3.

Oracle selects successive generations of child rows. Oracle first selects the children
of the rows returned in step 2, and then the children of those children, and so on.
Oracle always selects children by evaluating the CONNECT BY condition with respect
to a current parent row.

4.

If the query contains a WHERE clause without a join, then Oracle eliminates all rows
from the hierarchy that do not satisfy the condition of the WHERE clause. Oracle
evaluates this condition for each row individually, rather than removing all the
children of a row that does not satisfy the condition.

5.

Oracle returns the rows in the order shown in Figure 9–1. In the diagram, children
appear below their parents. For an explanation of hierarchical trees, see Figure 2–1,
"Hierarchical Tree" on page 2-3.

Figure 9–1 Hierarchical Queries
ROOT
1

7

2

4

3

5

8

6

9

10

12

11

To find the children of a parent row, Oracle evaluates the PRIOR expression of the
CONNECT BY condition for the parent row and the other expression for each row in the
table. Rows for which the condition is true are the children of the parent. The CONNECT
BY condition can contain other conditions to further filter the rows selected by the
query.
If the CONNECT BY condition results in a loop in the hierarchy, then Oracle returns an
error. A loop occurs if one row is both the parent (or grandparent or direct ancestor)
and a child (or a grandchild or a direct descendent) of another row.

9-4 Oracle Database SQL Language Reference

Hierarchical Queries

In a hierarchical query, do not specify either ORDER BY or GROUP
BY, as they will override the hierarchical order of the CONNECT BY
results. If you want to order rows of siblings of the same parent, then
use the ORDER SIBLINGS BY clause. See order_by_clause on page 19-33.
Note:

Hierarchical Query Examples
CONNECT BY Example The following hierarchical query uses the CONNECT BY clause
to define the relationship between employees and managers:
SELECT employee_id, last_name, manager_id
FROM employees
CONNECT BY PRIOR employee_id = manager_id;
EMPLOYEE_ID
----------101
108
109
110
111
112
113
200
203
204
. . .

LAST_NAME
MANAGER_ID
------------------------- ---------Kochhar
100
Greenberg
101
Faviet
108
Chen
108
Sciarra
108
Urman
108
Popp
108
Whalen
101
Mavris
101
Baer
101

The next example is similar to the preceding example, but uses the
LEVEL pseudocolumn to show parent and child rows:
LEVEL Example

SELECT employee_id, last_name, manager_id, LEVEL
FROM employees
CONNECT BY PRIOR employee_id = manager_id;
EMPLOYEE_ID
----------101
108
109
110
111
112
113
200
203
204
205
206
102
...

LAST_NAME
MANAGER_ID
LEVEL
------------------------- ---------- ---------Kochhar
100
1
Greenberg
101
2
Faviet
108
3
Chen
108
3
Sciarra
108
3
Urman
108
3
Popp
108
3
Whalen
101
2
Mavris
101
2
Baer
101
2
Higgins
101
2
Gietz
205
3
De Haan
100
1

The next example adds a START WITH clause to specify a root
row for the hierarchy and an ORDER BY clause using the SIBLINGS keyword to preserve
ordering within the hierarchy:
START WITH Examples

SELECT last_name, employee_id, manager_id, LEVEL
FROM employees

SQL Queries and Subqueries

9-5

Hierarchical Queries

START WITH employee_id = 100
CONNECT BY PRIOR employee_id = manager_id
ORDER SIBLINGS BY last_name;
LAST_NAME
EMPLOYEE_ID MANAGER_ID
LEVEL
------------------------- ----------- ---------- ---------King
100
1
Cambrault
148
100
2
Bates
172
148
3
Bloom
169
148
3
Fox
170
148
3
Kumar
173
148
3
Ozer
168
148
3
Smith
171
148
3
De Haan
102
100
2
Hunold
103
102
3
Austin
105
103
4
Ernst
104
103
4
Lorentz
107
103
4
Pataballa
106
103
4
Errazuriz
147
100
2
Ande
166
147
3
Banda
167
147
3
...

In the hr.employees table, the employee Steven King is the head of the company and
has no manager. Among his employees is John Russell, who is the manager of
department 80. If you update the employees table to set Russell as King's manager, you
create a loop in the data:
UPDATE employees SET manager_id = 145
WHERE employee_id = 100;
SELECT last_name "Employee",
LEVEL, SYS_CONNECT_BY_PATH(last_name, '/') "Path"
FROM employees
WHERE level <= 3 AND department_id = 80
START WITH last_name = 'King'
CONNECT BY PRIOR employee_id = manager_id AND LEVEL <= 4;
ERROR:
ORA-01436: CONNECT BY loop in user data

The NOCYCLE parameter in the CONNECT BY condition causes Oracle to return the rows in
spite of the loop. The CONNECT_BY_ISCYCLE pseudocolumn shows you which rows
contain the cycle:
SELECT last_name "Employee", CONNECT_BY_ISCYCLE "Cycle",
LEVEL, SYS_CONNECT_BY_PATH(last_name, '/') "Path"
FROM employees
WHERE level <= 3 AND department_id = 80
START WITH last_name = 'King'
CONNECT BY NOCYCLE PRIOR employee_id = manager_id AND LEVEL <= 4
ORDER BY "Employee", "Cycle", LEVEL, "Path";
Employee
Cycle
LEVEL Path
------------------------- ---------- ---------- ------------------------Abel
0
3 /King/Zlotkey/Abel
Ande
0
3 /King/Errazuriz/Ande
Banda
0
3 /King/Errazuriz/Banda
Bates
0
3 /King/Cambrault/Bates
9-6 Oracle Database SQL Language Reference

Hierarchical Queries

Bernstein
Bloom
Cambrault
Cambrault
Doran
Errazuriz
Fox
...

0
0
0
0
0
0
0

3
3
2
3
3
2
3

/King/Russell/Bernstein
/King/Cambrault/Bloom
/King/Cambrault
/King/Russell/Cambrault
/King/Partners/Doran
/King/Errazuriz
/King/Cambrault/Fox

CONNECT_BY_ISLEAF Example The following statement shows how you can use a
hierarchical query to turn the values in a column into a comma-delimited list:
SELECT LTRIM(SYS_CONNECT_BY_PATH (warehouse_id,','),',') FROM
(SELECT ROWNUM r, warehouse_id FROM warehouses)
WHERE CONNECT_BY_ISLEAF = 1
START WITH r = 1
CONNECT BY r = PRIOR r + 1
ORDER BY warehouse_id;
LTRIM(SYS_CONNECT_BY_PATH(WAREHOUSE_ID,','),',')
-------------------------------------------------------------------------------1,2,3,4,5,6,7,8,9

CONNECT_BY_ROOT Examples The following example returns the last name of

each employee in department 110, each manager at the highest level above that
employee in the hierarchy, the number of levels between manager and employee, and
the path between the two:
SELECT last_name "Employee", CONNECT_BY_ROOT last_name "Manager",
LEVEL-1 "Pathlen", SYS_CONNECT_BY_PATH(last_name, '/') "Path"
FROM employees
WHERE LEVEL > 1 and department_id = 110
CONNECT BY PRIOR employee_id = manager_id
ORDER BY "Employee", "Manager", "Pathlen", "Path";
Employee
--------------Gietz
Gietz
Gietz
Higgins
Higgins

Manager
Pathlen Path
--------------- ---------- -----------------------------Higgins
1 /Higgins/Gietz
King
3 /King/Kochhar/Higgins/Gietz
Kochhar
2 /Kochhar/Higgins/Gietz
King
2 /King/Kochhar/Higgins
Kochhar
1 /Kochhar/Higgins

The following example uses a GROUP BY clause to return the total salary of each
employee in department 110 and all employees above that employee in the hierarchy:
SELECT name, SUM(salary) "Total_Salary" FROM (
SELECT CONNECT_BY_ROOT last_name as name, Salary
FROM employees
WHERE department_id = 110
CONNECT BY PRIOR employee_id = manager_id)
GROUP BY name
ORDER BY name, "Total_Salary";
NAME
Total_Salary
------------------------- -----------Gietz
8300
Higgins
20300
King
20300
Kochhar
20300

SQL Queries and Subqueries

9-7

The UNION [ALL], INTERSECT, MINUS Operators

See Also:
■

■

■

■

LEVEL Pseudocolumn on page 2-2 and CONNECT_BY_ISCYCLE
Pseudocolumn on page 2-1 for a discussion of how these
pseudocolumns operate in a hierarchical query
SYS_CONNECT_BY_PATH on page 5-278 for information on
retrieving the path of column values from root to node
order_by_clause on page 19-33 for more information on the
SIBLINGS keyword of ORDER BY clauses
subquery_factoring_clause on page 19-13, which supports recursive
subquery factoring (recursive WITH) and lets you query
hierarchical data. This feature is more powerful than CONNECT BY
in that it provides depth-first search and breadth-first search, and
supports multiple recursive branches.

The UNION [ALL], INTERSECT, MINUS Operators
You can combine multiple queries using the set operators UNION, UNION ALL, INTERSECT,
and MINUS. All set operators have equal precedence. If a SQL statement contains
multiple set operators, then Oracle Database evaluates them from the left to right
unless parentheses explicitly specify another order.
The corresponding expressions in the select lists of the component queries of a
compound query must match in number and must be in the same data type group
(such as numeric or character).
If component queries select character data, then the data type of the return values are
determined as follows:
■

■

If both queries select values of data type CHAR of equal length, then the returned
values have data type CHAR of that length. If the queries select values of CHAR with
different lengths, then the returned value is VARCHAR2 with the length of the larger
CHAR value.
If either or both of the queries select values of data type VARCHAR2, then the
returned values have data type VARCHAR2.

If component queries select numeric data, then the data type of the return values is
determined by numeric precedence:
■

■

■

If any query selects values of type BINARY_DOUBLE, then the returned values have
data type BINARY_DOUBLE.
If no query selects values of type BINARY_DOUBLE but any query selects values of
type BINARY_FLOAT, then the returned values have data type BINARY_FLOAT.
If all queries select values of type NUMBER, then the returned values have data type
NUMBER.

In queries using set operators, Oracle does not perform implicit conversion across data
type groups. Therefore, if the corresponding expressions of component queries resolve
to both character data and numeric data, Oracle returns an error.
See Also: Table 3–10, " Implicit Type Conversion Matrix" on
page 3-40 for more information on implicit conversion and "Numeric
Precedence" on page 3-14 for information on numeric precedence
Examples

The following query is valid:

9-8 Oracle Database SQL Language Reference

The UNION [ALL], INTERSECT, MINUS Operators

SELECT 3 FROM DUAL
INTERSECT
SELECT 3f FROM DUAL;

This is implicitly converted to the following compound query:
SELECT TO_BINARY_FLOAT(3) FROM DUAL
INTERSECT
SELECT 3f FROM DUAL;

The following query returns an error:
SELECT '3' FROM DUAL
INTERSECT
SELECT 3f FROM DUAL;

Restrictions on the Set Operators

The set operators are subject to the following

restrictions:
■

■
■

The set operators are not valid on columns of type BLOB, CLOB, BFILE, VARRAY, or
nested table.
The UNION, INTERSECT, and MINUS operators are not valid on LONG columns.
If the select list preceding the set operator contains an expression, then you must
provide a column alias for the expression in order to refer to it in the order_by_
clause.

■

You cannot also specify the for_update_clause with the set operators.

■

You cannot specify the order_by_clause in the subquery of these operators.

■

You cannot use these operators in SELECT statements containing TABLE collection
expressions.
To comply with emerging SQL standards, a future release of
Oracle will give the INTERSECT operator greater precedence than the
other set operators. Therefore, you should use parentheses to specify
order of evaluation in queries that use the INTERSECT operator with
other set operators.
Note:

UNION Example The following statement combines the results of two queries with

the UNION operator, which eliminates duplicate selected rows. This statement shows
that you must match data type (using the TO_CHAR function) when columns do not
exist in one or the other table:
SELECT location_id, department_name "Department",
TO_CHAR(NULL) "Warehouse" FROM departments
UNION
SELECT location_id, TO_CHAR(NULL) "Department", warehouse_name
FROM warehouses;
LOCATION_ID
----------1400
1400
1500
1500
1600
1700
1700

Department
Warehouse
------------------------------ --------------------------IT
Southlake, Texas
Shipping
San Francisco
New Jersey
Accounting
Administration
SQL Queries and Subqueries

9-9

Sorting Query Results

1700
1700
1700
1700

Benefits
Construction
Contracting
Control And Credit

...

UNION ALL Example The UNION operator returns only distinct rows that appear in
either result, while the UNION ALL operator returns all rows. The UNION ALL operator
does not eliminate duplicate selected rows:
SELECT product_id FROM order_items
UNION
SELECT product_id FROM inventories
ORDER BY product_id;
SELECT location_id FROM locations
UNION ALL
SELECT location_id FROM departments
ORDER BY location_id;

A location_id value that appears multiple times in either or both queries (such as
'1700') is returned only once by the UNION operator, but multiple times by the UNION
ALL operator.
The following statement combines the results with the
INTERSECT operator, which returns only those unique rows returned by both queries:
INTERSECT Example

SELECT product_id FROM inventories
INTERSECT
SELECT product_id FROM order_items
ORDER BY product_id;

MINUS Example The following statement combines results with the MINUS operator,
which returns only unique rows returned by the first query but not by the second:
SELECT product_id FROM inventories
MINUS
SELECT product_id FROM order_items
ORDER BY product_id;

Sorting Query Results
Use the ORDER BY clause to order the rows selected by a query. Sorting by position is
useful in the following cases:
■

■

To order by a lengthy select list expression, you can specify its position in the
ORDER BY clause rather than duplicate the entire expression.
For compound queries containing set operators UNION, INTERSECT, MINUS, or UNION
ALL, the ORDER BY clause must specify positions or aliases rather than explicit
expressions. Also, the ORDER BY clause can appear only in the last component
query. The ORDER BY clause orders all rows returned by the entire compound query.

The mechanism by which Oracle Database sorts character values for the ORDER BY
clause, also known as the collation, is specified by the NLS_SORT session parameter. If
this parameter is not set, then its default is derived from the NLS_LANGUAGE session
parameter. You can change the collation dynamically using the ALTER SESSION SET
NLS_SORT statement. You can also apply a specific collation by including the character
expressions to be sorted as arguments to the NLSSORT function, with the collation
specified in the second parameter.
9-10 Oracle Database SQL Language Reference

Joins

When character values are compared linguistically for the ORDER BY clause, they are
first transformed to collation keys and then compared like RAW values. The collation
keys are generated either explicitly as specified in NLSSORT or implicitly using the same
method that NLSSORT uses. Both explicitly and implicitly generated collation keys are
subject to the same restrictions that are described in "NLSSORT" on page 5-164. As a
result of these restrictions, two values may compare as linguistically equal if they do
not differ in the prefix that was used to produce the collation key, even if they differ in
the rest of the value.
See Also: NLSSORT on page 5-164 and Oracle Database Globalization
Support Guide for information on the NLS parameters

Joins
A join is a query that combines rows from two or more tables, views, or materialized
views. Oracle Database performs a join whenever multiple tables appear in the FROM
clause of the query. The select list of the query can select any columns from any of
these tables. If any two of these tables have a column name in common, then you must
qualify all references to these columns throughout the query with table names to avoid
ambiguity.

Join Conditions
Most join queries contain at least one join condition, either in the FROM clause or in the
WHERE clause. The join condition compares two columns, each from a different table. To
execute a join, Oracle Database combines pairs of rows, each containing one row from
each table, for which the join condition evaluates to TRUE. The columns in the join
conditions need not also appear in the select list.
To execute a join of three or more tables, Oracle first joins two of the tables based on
the join conditions comparing their columns and then joins the result to another table
based on join conditions containing columns of the joined tables and the new table.
Oracle continues this process until all tables are joined into the result. The optimizer
determines the order in which Oracle joins tables based on the join conditions, indexes
on the tables, and, any available statistics for the tables.
A WHERE clause that contains a join condition can also contain other conditions that
refer to columns of only one table. These conditions can further restrict the rows
returned by the join query.
You cannot specify LOB columns in the WHERE clause if the
WHERE clause contains the join condition. The use of LOBs in WHERE
clauses is also subject to other restrictions. See Oracle Database
SecureFiles and Large Objects Developer's Guide for more information.
Note:

Equijoins
An equijoin is a join with a join condition containing an equality operator. An equijoin
combines rows that have equivalent values for the specified columns. Depending on
the internal algorithm the optimizer chooses to execute the join, the total size of the
columns in the equijoin condition in a single table may be limited to the size of a data
block minus some overhead. The size of a data block is specified by the initialization
parameter DB_BLOCK_SIZE.
See Also:

"Using Join Queries: Examples" on page 19-49

SQL Queries and Subqueries 9-11

Joins

Self Joins
A self join is a join of a table to itself. This table appears twice in the FROM clause and is
followed by table aliases that qualify column names in the join condition. To perform a
self join, Oracle Database combines and returns rows of the table that satisfy the join
condition.
See Also:

"Using Self Joins: Example" on page 19-50

Cartesian Products
If two tables in a join query have no join condition, then Oracle Database returns their
Cartesian product. Oracle combines each row of one table with each row of the other.
A Cartesian product always generates many rows and is rarely useful. For example,
the Cartesian product of two tables, each with 100 rows, has 10,000 rows. Always
include a join condition unless you specifically need a Cartesian product. If a query
joins three or more tables and you do not specify a join condition for a specific pair,
then the optimizer may choose a join order that avoids producing an intermediate
Cartesian product.

Inner Joins
An inner join (sometimes called a simple join) is a join of two or more tables that
returns only those rows that satisfy the join condition.

Outer Joins
An outer join extends the result of a simple join. An outer join returns all rows that
satisfy the join condition and also returns some or all of those rows from one table for
which no rows from the other satisfy the join condition.
■

■

■

To write a query that performs an outer join of tables A and B and returns all rows
from A (a left outer join), use the LEFT [OUTER] JOIN syntax in the FROM clause, or
apply the outer join operator (+) to all columns of B in the join condition in the
WHERE clause. For all rows in A that have no matching rows in B, Oracle Database
returns null for any select list expressions containing columns of B.
To write a query that performs an outer join of tables A and B and returns all rows
from B (a right outer join), use the RIGHT [OUTER] JOIN syntax in the FROM clause, or
apply the outer join operator (+) to all columns of A in the join condition in the
WHERE clause. For all rows in B that have no matching rows in A, Oracle returns
null for any select list expressions containing columns of A.
To write a query that performs an outer join and returns all rows from A and B,
extended with nulls if they do not satisfy the join condition (a full outer join), use
the FULL [OUTER] JOIN syntax in the FROM clause.

You cannot compare a column with a subquery in the WHERE clause of any outer join,
regardless which form you specify.
You can use outer joins to fill gaps in sparse data. Such a join is called a partitioned
outer join and is formed using the query_partition_clause of the join_clause
syntax. Sparse data is data that does not have rows for all possible values of a
dimension such as time or department. For example, tables of sales data typically do
not have rows for products that had no sales on a given date. Filling data gaps is
useful in situations where data sparsity complicates analytic computation or where
some data might be missed if the sparse data is queried directly.

9-12 Oracle Database SQL Language Reference

Joins

See Also:
■

■

join_clause on page 19-23 for more information about using outer
joins to fill gaps in sparse data
Oracle Database Data Warehousing Guide for a complete discussion
of group outer joins and filling gaps in sparse data

Oracle recommends that you use the FROM clause OUTER JOIN syntax rather than the
Oracle join operator. Outer join queries that use the Oracle join operator (+) are subject
to the following rules and restrictions, which do not apply to the FROM clause OUTER
JOIN syntax:
■

■

■

■

■

You cannot specify the (+) operator in a query block that also contains FROM clause
join syntax.
The (+) operator can appear only in the WHERE clause or, in the context of
left-correlation (when specifying the TABLE clause) in the FROM clause, and can be
applied only to a column of a table or view.
If A and B are joined by multiple join conditions, then you must use the (+)
operator in all of these conditions. If you do not, then Oracle Database will return
only the rows resulting from a simple join, but without a warning or error to
advise you that you do not have the results of an outer join.
The (+) operator does not produce an outer join if you specify one table in the
outer query and the other table in an inner query.
You cannot use the (+) operator to outer-join a table to itself, although self joins are
valid. For example, the following statement is not valid:
-- The following statement is not valid:
SELECT employee_id, manager_id
FROM employees
WHERE employees.manager_id(+) = employees.employee_id;

However, the following self join is valid:
SELECT e1.employee_id, e1.manager_id, e2.employee_id
FROM employees e1, employees e2
WHERE e1.manager_id(+) = e2.employee_id
ORDER BY e1.employee_id, e1.manager_id, e2.employee_id;
■

■

■

The (+) operator can be applied only to a column, not to an arbitrary expression.
However, an arbitrary expression can contain one or more columns marked with
the (+) operator.
A WHERE condition containing the (+) operator cannot be combined with another
condition using the OR logical operator.
A WHERE condition cannot use the IN comparison condition to compare a column
marked with the (+) operator with an expression.

If the WHERE clause contains a condition that compares a column from table B with a
constant, then the (+) operator must be applied to the column so that Oracle returns
the rows from table A for which it has generated nulls for this column. Otherwise
Oracle returns only the results of a simple join.
In a query that performs outer joins of more than two pairs of tables, a single table can
be the null-generated table for only one other table. For this reason, you cannot apply
the (+) operator to columns of B in the join condition for A and B and the join
condition for B and C. Refer to SELECT on page 19-4 for the syntax for an outer join.

SQL Queries and Subqueries 9-13

Using Subqueries

Antijoins
An antijoin returns rows from the left side of the predicate for which there are no
corresponding rows on the right side of the predicate. It returns rows that fail to match
(NOT IN) the subquery on the right side.
See Also:

"Using Antijoins: Example" on page 19-53

Semijoins
A semijoin returns rows that match an EXISTS subquery without duplicating rows
from the left side of the predicate when multiple rows on the right side satisfy the
criteria of the subquery.
Semijoin and antijoin transformation cannot be done if the subquery is on an OR
branch of the WHERE clause.
See Also:

"Using Semijoins: Example" on page 19-53

Using Subqueries
A subquery answers multiple-part questions. For example, to determine who works in
Taylor's department, you can first use a subquery to determine the department in
which Taylor works. You can then answer the original question with the parent SELECT
statement. A subquery in the FROM clause of a SELECT statement is also called an inline
view. you can nest any number of subqueries in an inline view. A subquery in the
WHERE clause of a SELECT statement is also called a nested subquery. You can nest up
to 255 levels of subqueries in the a nested subquery.
A subquery can contain another subquery. Oracle Database imposes no limit on the
number of subquery levels in the FROM clause of the top-level query. You can nest up to
255 levels of subqueries in the WHERE clause.
If columns in a subquery have the same name as columns in the containing statement,
then you must prefix any reference to the column of the table from the containing
statement with the table name or alias. To make your statements easier to read, always
qualify the columns in a subquery with the name or alias of the table, view, or
materialized view.
Oracle performs a correlated subquery when a nested subquery references a column
from a table referred to a parent statement one level above the subquery. The parent
statement can be a SELECT, UPDATE, or DELETE statement in which the subquery is
nested. A correlated subquery conceptually is evaluated once for each row processed
by the parent statement. However, the optimizer may choose to rewrite the query as a
join or use some other technique to formulate a query that is semantically equivalent.
Oracle resolves unqualified columns in the subquery by looking in the tables named in
the subquery and then in the tables named in the parent statement.
A correlated subquery answers a multiple-part question whose answer depends on the
value in each row processed by the parent statement. For example, you can use a
correlated subquery to determine which employees earn more than the average
salaries for their departments. In this case, the correlated subquery specifically
computes the average salary for each department.
See Also:

"Using Correlated Subqueries: Examples" on page 19-57

Use subqueries for the following purposes:

9-14 Oracle Database SQL Language Reference

Unnesting of Nested Subqueries

■

■

■

■

■

To define the set of rows to be inserted into the target table of an INSERT or CREATE
TABLE statement
To define the set of rows to be included in a view or materialized view in a CREATE
VIEW or CREATE MATERIALIZED VIEW statement
To define one or more values to be assigned to existing rows in an UPDATE
statement
To provide values for conditions in a WHERE clause, HAVING clause, or START WITH
clause of SELECT, UPDATE, and DELETE statements
To define a table to be operated on by a containing query
You do this by placing the subquery in the FROM clause of the containing query as
you would a table name. You may use subqueries in place of tables in this way as
well in INSERT, UPDATE, and DELETE statements.
Subqueries so used can employ correlation variables, but only those defined
within the subquery itself, not outer references. Refer to table_collection_expression
on page 19-20 for more information.
Scalar subqueries, which return a single column value from a single row, are a
valid form of expression. You can use scalar subquery expressions in most of the
places where expr is called for in syntax. Refer to "Scalar Subquery Expressions"
on page 6-14 for more information.

Unnesting of Nested Subqueries
Subqueries are nested when they appear in the WHERE clause of the parent statement.
When Oracle Database evaluates a statement with a nested subquery, it must evaluate
the subquery portion multiple times and may overlook some efficient access paths or
joins.
Subquery unnesting unnests and merges the body of the subquery into the body of
the statement that contains it, allowing the optimizer to consider them together when
evaluating access paths and joins. The optimizer can unnest most subqueries, with
some exceptions. Those exceptions include hierarchical subqueries and subqueries that
contain a ROWNUM pseudocolumn, one of the set operators, a nested aggregate function,
or a correlated reference to a query block that is not the immediate outer query block
of the subquery.
Assuming no restrictions exist, the optimizer automatically unnests some (but not all)
of the following nested subqueries:
■
■

Uncorrelated IN subqueries
IN and EXISTS correlated subqueries, as long as they do not contain aggregate
functions or a GROUP BY clause

You can enable extended subquery unnesting by instructing the optimizer to unnest
additional types of subqueries:
■

■

You can unnest an uncorrelated NOT IN subquery by specifying the HASH_AJ or
MERGE_AJ hint in the subquery.
You can unnest other subqueries by specifying the UNNEST hint in the subquery.
See Also:

"Hints" on page 3-74 for information on hints

SQL Queries and Subqueries 9-15

Selecting from the DUAL Table

Selecting from the DUAL Table
DUAL is a table automatically created by Oracle Database along with the data
dictionary. DUAL is in the schema of the user SYS but is accessible by the name DUAL to
all users. It has one column, DUMMY, defined to be VARCHAR2(1), and contains one row
with a value X. Selecting from the DUAL table is useful for computing a constant
expression with the SELECT statement. Because DUAL has only one row, the constant is
returned only once. Alternatively, you can select a constant, pseudocolumn, or
expression from any table, but the value will be returned as many times as there are
rows in the table. Refer to "About SQL Functions" on page 5-2 for many examples of
selecting a constant value from DUAL.

Beginning with Oracle Database 10g Release 1, logical I/O is
not performed on the DUAL table when computing an expression that
does not include the DUMMY column. This optimization is listed as FAST
DUAL in the execution plan. If you SELECT the DUMMY column from DUAL,
then this optimization does not take place and logical I/O occurs.
Note:

Distributed Queries
The Oracle distributed database management system architecture lets you access data
in remote databases using Oracle Net and an Oracle Database server. You can identify
a remote table, view, or materialized view by appending @dblink to the end of its
name. The dblink must be a complete or partial name for a database link to the
database containing the remote table, view, or materialized view.
See Also:
■

■

"References to Objects in Remote Databases" on page 3-117 for
more information on referring to database links
Oracle Database Net Services Administrator's Guide for information
on accessing remote databases

Restrictions on Distributed Queries Distributed queries are currently subject to the
restriction that all tables locked by a FOR UPDATE clause and all tables with LONG
columns selected by the query must be located on the same database. In addition,
Oracle Database currently does not support distributed queries that select
user-defined types or object REF data types on remote tables.

9-16 Oracle Database SQL Language Reference

10
10

SQL Statements: ALTER CLUSTER to ALTER
JAVA
This chapter lists the various types of SQL statements and then describes the first set
(in alphabetical order) of SQL statements. The remaining SQL statements appear in
alphabetical order in Chapter 11 through Chapter 19.
This chapter contains the following sections:
■

Types of SQL Statements

■

How the SQL Statement Chapters are Organized

■

ALTER CLUSTER

■

ALTER DATABASE

■

ALTER DATABASE LINK

■

ALTER DIMENSION

■

ALTER DISKGROUP

■

ALTER FLASHBACK ARCHIVE

■

ALTER FUNCTION

■

ALTER INDEX

■

ALTER INDEXTYPE

■

ALTER JAVA

Types of SQL Statements
The lists in the following sections provide a functional summary of SQL statements
and are divided into these categories:
■

Data Definition Language (DDL) Statements

■

Data Manipulation Language (DML) Statements

■

Transaction Control Statements

■

Session Control Statements

■

System Control Statement

■

Embedded SQL Statements

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-1

Types of SQL Statements

Data Definition Language (DDL) Statements
Data definition language (DDL) statements let you to perform these tasks:
■

Create, alter, and drop schema objects

■

Grant and revoke privileges and roles

■

Analyze information on a table, index, or cluster

■

Establish auditing options

■

Add comments to the data dictionary

The CREATE, ALTER, and DROP commands require exclusive access to the specified
object. For example, an ALTER TABLE statement fails if another user has an open
transaction on the specified table.
The GRANT, REVOKE, ANALYZE, AUDIT, and COMMENT commands do not require exclusive
access to the specified object. For example, you can analyze a table while other users
are updating the table.
Oracle Database implicitly commits the current transaction before and after every DDL
statement.
Many DDL statements may cause Oracle Database to recompile or reauthorize schema
objects. For information on how Oracle Database recompiles and reauthorizes schema
objects and the circumstances under which a DDL statement would cause this, see
Oracle Database Concepts.
DDL statements are supported by PL/SQL with the use of the DBMS_SQL package.
See Also: Oracle Database PL/SQL Packages and Types Reference for
more information about this package

The DDL statements are:
ALTER ... (All statements beginning with ALTER, except ALTER SESSION and ALTER
SYSTEM—see "Session Control Statements" on page 10-3 and "System Control
Statement" on page 10-3)
ANALYZE
ASSOCIATE STATISTICS
AUDIT
COMMENT
CREATE ... (All statements beginning with CREATE)
DISASSOCIATE STATISTICS
DROP ... (All statements beginning with DROP)
FLASHBACK ... (All statements beginning with FLASHBACK)
GRANT
NOAUDIT
PURGE
RENAME
REVOKE
TRUNCATE

Data Manipulation Language (DML) Statements
Data manipulation language (DML) statements access and manipulate data in existing
schema objects. These statements do not implicitly commit the current transaction. The
data manipulation language statements are:
CALL
10-2 Oracle Database SQL Language Reference

Types of SQL Statements

DELETE
EXPLAIN PLAN
INSERT
LOCK TABLE
MERGE
SELECT
UPDATE
The SELECT statement is a limited form of DML statement in that it can only access
data in the database. It cannot manipulate data stored in the database, although it can
manipulate the accessed data before returning the results of the query.
The SELECT statement is supported in PL/SQL only when executed dynamically.
However, you can use the similar PL/SQL statement SELECT INTO in PL/SQL code,
and you do not have to execute it dynamically. The CALL and EXPLAIN PLAN statements
are supported in PL/SQL only when executed dynamically. All other DML statements
are fully supported in PL/SQL.

Transaction Control Statements
Transaction control statements manage changes made by DML statements. The
transaction control statements are:
COMMIT
ROLLBACK
SAVEPOINT
SET TRANSACTION
SET CONSTRAINT
All transaction control statements, except certain forms of the COMMIT and ROLLBACK
commands, are supported in PL/SQL. For information on the restrictions, see
COMMIT on page 13-49 and ROLLBACK on page 18-96.

Session Control Statements
Session control statements dynamically manage the properties of a user session. These
statements do not implicitly commit the current transaction.
PL/SQL does not support session control statements. The session control statements
are:
ALTER SESSION
SET ROLE

System Control Statement
The single system control statement, ALTER SYSTEM, dynamically manages the
properties of an Oracle Database instance. This statement does not implicitly commit
the current transaction and is not supported in PL/SQL.

Embedded SQL Statements
Embedded SQL statements place DDL, DML, and transaction control statements
within a procedural language program. Embedded SQL is supported by the Oracle
precompilers and is documented in the following books:
■

Pro*COBOL Programmer's Guide

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-3

How the SQL Statement Chapters are Organized

■

Pro*C/C++ Programmer's Guide

■

Oracle SQL*Module for Ada Programmer's Guide

How the SQL Statement Chapters are Organized
All SQL statements in this chapter, as well as in Chapter 11 through Chapter 19, are
organized into the following sections:
Syntax The syntax diagrams show the keywords and parameters that make up the

statement.
Caution: Not all keywords and parameters are valid in all
circumstances. Be sure to refer to the "Semantics" section of each
statement and clause to learn about any restrictions on the syntax.
Purpose

The "Purpose" section describes the basic uses of the statement.

Prerequisites The "Prerequisites" section lists privileges you must have and steps
that you must take before using the statement. In addition to the prerequisites listed,
most statements also require that the database be opened by your instance, unless
otherwise noted.
Semantics The "Semantics" section describes the purpose of the keywords,
parameters, and clauses that make up the syntax, as well as restrictions and other
usage notes that may apply to them. (The conventions for keywords and parameters
used in this chapter are explained in the "Preface" of this reference.)

The "Examples" section shows how to use the various clauses and
parameters of the statement.

Examples

10-4 Oracle Database SQL Language Reference

ALTER CLUSTER

ALTER CLUSTER
Purpose
10

Use the ALTER CLUSTER statement to redefine storage and parallelism characteristics of
a cluster.
You cannot use this statement to change the number or the
name of columns in the cluster key, and you cannot change the
tablespace in which the cluster is stored.

Note:

See Also: CREATE CLUSTER on page 14-2 for information on
creating a cluster, DROP CLUSTER on page 17-36 and DROP TABLE
on page 18-5 for information on removing tables from a cluster, and
CREATE TABLE ... physical_properties on page 16-32 for information on
adding a table to a cluster

Prerequisites
10

The cluster must be in your own schema or you must have the ALTER ANY CLUSTER
system privilege.

Syntax
10

alter_cluster::=
physical_attributes_clause
SIZE
schema
ALTER

CLUSTER

.
cluster

size_clause

allocate_extent_clause

parallel_clause
;

deallocate_unused_clause
CACHE
NOCACHE

(physical_attributes_clause::= on page 10-5, size_clause::= on page 8-47, allocate_extent_
clause::= on page 10-6, deallocate_unused_clause::= on page 10-6, parallel_clause::= on
page 10-6)
physical_attributes_clause::=
PCTFREE

integer

PCTUSED

integer

INITRANS

integer

storage_clause

(storage_clause::= on page 8-50)

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-5

ALTER CLUSTER

allocate_extent_clause::=
SIZE
(

size_clause

DATAFILE
INSTANCE

ALLOCATE

’

filename

’

)

integer

EXTENT

(size_clause::= on page 8-47)
deallocate_unused_clause::=
KEEP
DEALLOCATE

size_clause

UNUSED

(size_clause::= on page 8-47)
parallel_clause::=
NOPARALLEL
integer
PARALLEL

Semantics
10

schema
Specify the schema containing the cluster. If you omit schema, then Oracle Database
assumes the cluster is in your own schema.

cluster
Specify the name of the cluster to be altered.

physical_attributes_clause
Use this clause to change the values of the PCTUSED, PCTFREE, and INITRANS parameters
of the cluster.
Use the STORAGE clause to change the storage characteristics of the cluster.
See Also:
■

■

physical_attributes_clause on page 8-44 for information on the
parameters
storage_clause on page 8-46 for a full description of that clause

Restriction on Physical Attributes You cannot change the values of the storage
parameters INITIAL and MINEXTENTS for a cluster.

SIZE integer
Use the SIZE clause to specify the number of cluster keys that will be stored in data
blocks allocated to the cluster.

10-6 Oracle Database SQL Language Reference

ALTER CLUSTER

Restriction on SIZE You can change the SIZE parameter only for an indexed cluster,
not for a hash cluster.

CREATE CLUSTER on page 14-2 for a description of the
SIZE parameter and "Modifying a Cluster: Example" on page 10-7
See Also:

allocate_extent_clause
Specify the allocate_extent_clause to explicitly allocate a new extent for the cluster.
When you explicitly allocate an extent with this clause, Oracle Database does not
evaluate the storage parameters of the cluster and determine a new size for the next
extent to be allocated (as it does when you create a table). Therefore, specify SIZE if
you do not want Oracle Database to use a default value.
See Also: allocate_extent_clause on page 8-2 for a full description of
this clause

deallocate_unused_clause
Use the deallocate_unused_clause to explicitly deallocate unused space at the end of
the cluster and make the freed space available for other segments.
See Also: deallocate_unused_clause on page 8-27 for a full description
of this clause and "Deallocating Unused Space: Example" on page 10-7

CACHE | NOCACHE
This clause has the same behavior in CREATE CLUSTER and ALTER CLUSTER statements.
See Also: "CACHE | NOCACHE" on page 14-7 for information on
this clause.

parallel_clause
Specify the parallel_clause to change the default degree of parallelism for queries on
the cluster.
parallel_clause on page 16-63 in the documentation on
CREATE TABLE for complete information on this clause
See Also:

Examples
10

The following examples modify the clusters that were created in the CREATE CLUSTER
"Examples" on page 14-7.
Modifying a Cluster: Example

The next statement alters the personnel cluster:

ALTER CLUSTER personnel
SIZE 1024 CACHE;

Oracle Database allocates 1024 bytes for each cluster key value and enables the cache
attribute. Assuming a data block size of 2 kilobytes, future data blocks within this
cluster contain 2 cluster keys in each data block, or 2 kilobytes divided by 1024 bytes.
Deallocating Unused Space: Example The following statement deallocates unused
space from the language cluster, keeping 30 kilobytes of unused space for future use:
ALTER CLUSTER language
DEALLOCATE UNUSED KEEP 30 K;
SQL Statements: ALTER CLUSTER to ALTER JAVA

10-7

ALTER DATABASE

ALTER DATABASE
Purpose
10

Use the ALTER DATABASE statement to modify, maintain, or recover an existing
database.
In earlier versions of Oracle Database, you could use the ALTER
DATABASE for two conversion operations:
Note:
■

■

The RESET COMPATIBILITY clause lets you reset the database to an
earlier version at the next instance startup.
The CONVERT clause lets you upgrade an Oracle7 data dictionary to
an Oracle8i or Oracle9i data dictionary.

These clauses are no longer supported. Refer to Oracle Database
Upgrade Guide for more information on migration and interoperability
issues.

See Also:
■

■

■

Oracle Database Backup and Recovery User's Guide for examples of
performing media recovery
Oracle Data Guard Concepts and Administration for additional
information on using the ALTER DATABASE statement to maintain
standby databases
CREATE DATABASE on page 14-19 for information on creating a
database

Prerequisites
10

You must have the ALTER DATABASE system privilege.
To specify the startup_clauses, you must also be connected AS SYSDBA or AS SYSOPER.
To specify the general_recovery clause, you must also have the SYSDBA system
privilege.

10-8 Oracle Database SQL Language Reference

ALTER DATABASE

Syntax
10

alter_database::=
startup_clauses
recovery_clauses
database_file_clauses
database
ALTER

DATABASE

logfile_clauses
controlfile_clauses

;

standby_database_clauses
default_settings_clauses
instance_clauses
security_clause

Groups of ALTER DATABASE syntax:
■
startup_clauses::= on page 10-9
■

recovery_clauses::= on page 10-9

■

database_file_clauses::= on page 10-11

■

logfile_clauses::= on page 10-13

■

controlfile_clauses::= on page 10-14

■

standby_database_clauses::= on page 10-15

■

default_settings_clauses::= on page 10-17

■

instance_clauses::= on page 10-17

■

security_clause::= on page 10-17

startup_clauses::=
STANDBY
DATABASE
CLONE
MOUNT

READ
OPEN

READ

WRITE

RESETLOGS

UPGRADE

NORESETLOGS

DOWNGRADE

ONLY

recovery_clauses::=
general_recovery
managed_standby_recovery
BEGIN
BACKUP
END

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-9

ALTER DATABASE

(general_recovery::= on page 10-10, managed_standby_recovery::= on page 10-11)
general_recovery::=
AUTOMATIC

FROM

’

location

’

RECOVER

TEST
ALLOW
full_database_recovery

integer

CORRUPTION

parallel_clause

partial_database_recovery
LOGFILE

’

filename

’

DEFAULT
CONTINUE
CANCEL

(full_database_recovery::= on page 10-10, partial_database_recovery::= on page 10-10,
parallel_clause::= on page 10-10)
full_database_recovery::=
CANCEL
TIME

date

UNTIL
CHANGE

integer

CONSISTENT
STANDBY

USING

BACKUP

DATABASE

partial_database_recovery::=
,
TABLESPACE

tablespace
,
’

filename

’

DATAFILE
filenumber

parallel_clause::=
NOPARALLEL
integer
PARALLEL

10-10 Oracle Database SQL Language Reference

CONTROLFILE

ALTER DATABASE

managed_standby_recovery::=
USING

CURRENT

LOGFILE
FROM

SESSION

DISCONNECT
NODELAY
UNTIL

CHANGE

UNTIL

CONSISTENT

integer

parallel_clause
FINISH
CANCEL
MANAGED

STANDBY

DATABASE

RECOVER

db_name
TO

LOGICAL

STANDBY
KEEP

IDENTITY

(parallel_clause::= on page 10-10)
Several subclauses of managed_standby_recovery are no
longer needed and have been deprecated. These clauses no longer
appear in the syntax diagrams. Refer to the semantics of managed_
standby_recovery on page 10-22.

Note:

database_file_clauses::=
,
RENAME

FILE

’

filename

’

TO

’

filename

’

create_datafile_clause
alter_datafile_clause
alter_tempfile_clause

(create_datafile_clause::= on page 10-11, alter_datafile_clause::= on page 10-12, alter_
tempfile_clause::= on page 10-12)
create_datafile_clause::=
,
file_specification

,
’
CREATE

filename

AS
’

NEW

DATAFILE
filenumber

(file_specification::= on page 8-29)

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-11

ALTER DATABASE

alter_datafile_clause::=
ONLINE
,
’

filename

FOR
’

DATAFILE

RESIZE

filenumber

DROP

OFFLINE
size_clause

autoextend_clause
END

BACKUP

(autoextend_clause::= on page 10-12, size_clause::= on page 8-47)
alter_tempfile_clause::=
RESIZE
,
’

filename

size_clause

autoextend_clause
’

TEMPFILE

INCLUDING

DATAFILES

DROP

filenumber

ONLINE
OFFLINE

(autoextend_clause::= on page 10-12, size_clause::= on page 8-47)
autoextend_clause::=
OFF
AUTOEXTEND

NEXT

size_clause

ON

maxsize_clause::=
UNLIMITED
MAXSIZE
size_clause

(size_clause::= on page 8-47)

10-12 Oracle Database SQL Language Reference

maxsize_clause

ALTER DATABASE

logfile_clauses::=
MANUAL
ARCHIVELOG
NOARCHIVELOG
NO
FORCE

LOGGING
,

RENAME

FILE

’

filename

’

TO

’

filename

’

,

UNARCHIVED
CLEAR

LOGFILE

UNRECOVERABLE

DATAFILE

logfile_descriptor

add_logfile_clauses
drop_logfile_clauses
switch_logfile_clause
supplemental_db_logging

(logfile_descriptor::= on page 10-14, add_logfile_clauses::= on page 10-13, drop_logfile_
clauses::= on page 10-13, switch_logfile_clause::= on page 10-14, supplemental_db_
logging::= on page 10-14)
add_logfile_clauses::=
STANDBY
ADD

LOGFILE

INSTANCE
THREAD

’

instance_name

’

,

integer

GROUP

integer
redo_logfile_spec

,
,

REUSE
MEMBER

’

filename

’

TO

logfile_descriptor

(redo_log_file_spec::= on page 8-29, logfile_descriptor::= on page 10-14)
drop_logfile_clauses::=
,
logfile_descriptor

STANDBY
DROP

LOGFILE

,
MEMBER

’

filename

’

(logfile_descriptor::= on page 10-14)

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-13

ALTER DATABASE

switch_logfile_clause::=
SWITCH

ALL

LOGFILES

TO

BLOCKSIZE

integer

supplemental_db_logging::=
DATA
ADD
SUPPLEMENTAL

LOG

supplemental_id_key_clause

DROP
supplemental_plsql_clause

(supplemental_id_key_clause::= on page 10-14)
supplemental_id_key_clause::=
,
ALL
PRIMARY
DATA

KEY

(

)

COLUMNS

UNIQUE
FOREIGN

KEY

supplemental_plsql_clause::=
DATA

FOR

PROCEDURAL

REPLICATION

logfile_descriptor::=
GROUP

integer
,

(

’

filename

’

filename

’

’

)

controlfile_clauses::=
LOGICAL
PHYSICAL

REUSE

CREATE

STANDBY

CONTROLFILE

AS

REUSE
’
BACKUP

CONTROLFILE

filename

’

TO
trace_file_clause

(trace_file_clause::= on page 10-15)

10-14 Oracle Database SQL Language Reference

’

filename

’

ALTER DATABASE

trace_file_clause::=
RESETLOGS

REUSE
AS

’

filename

’

NORESETLOGS

TRACE

standby_database_clauses::=
activate_standby_db_clause
maximize_standby_db_clause
register_logfile_clause

parallel_clause

commit_switchover_clause
start_standby_clause
stop_standby_clause
convert_database_clause

(activate_standby_db_clause::= on page 10-15, maximize_standby_db_clause::= on
page 10-15, register_logfile_clause::= on page 10-15, commit_switchover_clause::= on
page 10-16, start_standby_clause::= on page 10-16, stop_standby_clause::= on page 10-16,
convert_database_clause::= on page 10-16, parallel_clause::= on page 10-10)
activate_standby_db_clause::=
PHYSICAL
LOGICAL

FINISH

ACTIVATE

STANDBY

APPLY

DATABASE

maximize_standby_db_clause::=
PROTECTION
SET

STANDBY

DATABASE

TO

MAXIMIZE

AVAILABILITY
PERFORMANCE

register_logfile_clause::=
PHYSICAL
OR

REPLACE

LOGICAL

REGISTER

FOR

,
file_specification
LOGFILE

logminer_session_name

(file_specification::= on page 8-29)

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-15

ALTER DATABASE

commit_switchover_clause::=
PREPARE
TO

SWITCHOVER

COMMIT

PHYSICAL
WITH

LOGICAL
PRIMARY

WAIT
SESSION

WITHOUT

PHYSICAL
STANDBY

TO
LOGICAL

STANDBY

CANCEL

start_standby_clause::=
IMMEDIATE
START

LOGICAL

NEW

STANDBY

PRIMARY

APPLY

dblink

scn_value
INITIAL
SKIP

FAILED

TRANSACTION

FINISH

stop_standby_clause::=
STOP
LOGICAL

STANDBY

APPLY

ABORT

convert_database_clause::=
PHYSICAL
CONVERT

TO

STANDBY
SNAPSHOT

10-16 Oracle Database SQL Language Reference

NODELAY

SHUTDOWN
NOWAIT

ALTER DATABASE

default_settings_clauses::=
DEFAULT

EDITION

=

edition_name

BIGFILE
SET

DEFAULT

TABLESPACE
SMALLFILE

DEFAULT

TABLESPACE

tablespace

DEFAULT

TEMPORARY

TABLESPACE

tablespace
tablespace_group_name
RENAME

GLOBAL_NAME

TO

database

.

domain
REUSE
USING

ENABLE

BLOCK

CHANGE

TRACKING

DISABLE

BLOCK

CHANGE

TRACKING

FILE

’

filename

’

flashback_mode_clause
set_time_zone_clause

(flashback_mode_clause::= on page 10-17, set_time_zone_clause::= on page 10-17)
set_time_zone_clause::=
+
hh
SET

TIME_ZONE

=

:

–

’

mi
’

time_zone_region

flashback_mode_clause::=
ON
FLASHBACK
OFF

instance_clauses::=
ENABLE
INSTANCE

’

instance_name

’

DISABLE

security_clause::=
ALL
GUARD

STANDBY
NONE

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-17

ALTER DATABASE

Semantics
10

database
Specify the name of the database to be altered. The database name can contain only
ASCII characters. If you omit database, then Oracle Database alters the database
identified by the value of the initialization parameter DB_NAME. You can alter only the
database whose control files are specified by the initialization parameter CONTROL_
FILES. The database identifier is not related to the Oracle Net database specification.

startup_clauses
The startup_clauses let you mount and open the database so that it is accessible to
users.
MOUNT Clause
Use the MOUNT clause to mount the database. Do not use this clause when the database
is already mounted.
MOUNT STANDBY DATABASE You can specify MOUNT STANDBY DATABASE to mount a

physical standby database. The keywords STANDBY DATABASE are optional, because
Oracle Database determines automatically whether the database to be mounted is a
primary or standby database. As soon as this statement executes, the standby instance
can receive redo data from the primary instance.
See Also: Oracle Data Guard Concepts and Administration for more
information on standby databases
MOUNT CLONE DATABASE Specify MOUNT CLONE DATABASE to mount the clone
database.

OPEN Clause
Use the OPEN clause to make the database available for normal use. You must mount
the database before you can open it.
If you specify only OPEN without any other keywords, then the default is OPEN READ
WRITE NORESETLOGS on a primary database, logical standby database, or snapshot
standby database and OPEN READ ONLY on a physical standby database.
OPEN READ WRITE Specify OPEN READ WRITE to open the database in read/write
mode, allowing users to generate redo logs. This is the default if you are opening a
primary database. You cannot specify this clause for a physical standby database.
See Also:

"READ ONLY / READ WRITE: Example" on page 10-42

This clause determines whether Oracle Database
resets the current log sequence number to 1, archives any unarchived logs (including
the current log), and discards any redo information that was not applied during
recovery, ensuring that it will never be applied. Oracle Database uses NORESETLOGS
automatically except in the following specific situations, which require a setting for
this clause:

RESETLOGS | NORESETLOGS

■

You must specify RESETLOGS:
–

After performing incomplete media recovery or media recovery using a
backup control file

–

After a previous OPEN RESETLOGS operation that did not complete

10-18 Oracle Database SQL Language Reference

ALTER DATABASE

–
■

After a FLASHBACK DATABASE operation

If a created control file is mounted, then you must specify RESETLOGS if the online
logs are lost, or you must specify NORESETLOGS if they are not lost.

UPGRADE | DOWNGRADE Use these OPEN clause parameters only if you are
upgrading or downgrading a database. This clause instructs Oracle Database to
modify system parameters dynamically as required for upgrade and downgrade,
respectively. You can achieve the same result using the SQL*Plus STARTUP UPGRADE or
STARTUP DOWNGRADE command.
See Also:
■

■

Oracle Database Upgrade Guide for information on the steps
required to upgrade or downgrade a database from one release to
another
SQL*Plus User's Guide and Reference for information on the
SQL*Plus STARTUP command

OPEN READ ONLY Specify OPEN READ ONLY to restrict users to read-only transactions,
preventing them from generating redo logs. This setting is the default when you are
opening a physical standby database, so that the physical standby database is
available for queries even while archive logs are being copied from the primary
database site.
Restrictions on Opening a Database The following restrictions apply to opening a
database:
■

■
■

You cannot open a database in READ ONLY mode if it is currently opened in READ
WRITE mode by another instance.
You cannot open a database in READ ONLY mode if it requires recovery.
You cannot take tablespaces offline while the database is open in READ ONLY mode.
However, you can take data files offline and online, and you can recover offline
data files and tablespaces while the database is open in READ ONLY mode.
See Also: Oracle Data Guard Concepts and Administration for
additional information about opening a physical standby database

recovery_clauses
The recovery_clauses include post-backup operations. For all of these clauses, Oracle
Database recovers the database using any incarnations of data files and log files that
are known to the current control file.
See Also: Oracle Database Backup and Recovery User's Guide for
information on backing up the database and "Database Recovery:
Examples" on page 10-44

general_recovery
The general_recovery clause lets you control media recovery for the database or
standby database or for specified tablespaces or files. You can use this clause when
your instance has the database mounted, open or closed, and the files involved are not
in use.

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-19

ALTER DATABASE

Parallelism is enabled by default during full or partial
database recovery and logfile recovery. The database computes the
degree of parallelism. You can disable parallelism of these operations
by specifying NOPARALLEL, or specify a degree of parallelism with
PARALLEL integer, as shown in the respective syntax diagrams.
Note:

Restrictions on General Database Recovery General recovery is subject to the
following restrictions:
■

You can recover the entire database only when the database is closed.

■

Your instance must have the database mounted in exclusive mode.

■

■

You can recover tablespaces or data files when the database is open or closed, if
the tablespaces or data files to be recovered are offline.
You cannot perform media recovery if you are connected to Oracle Database
through the shared server architecture.
See Also:
■

■

Oracle Database Backup and Recovery User's Guide for more
information on RMAN media recovery and user-defined media
recovery
SQL*Plus User's Guide and Reference for information on the
SQL*Plus RECOVER command

AUTOMATIC
Specify AUTOMATIC if you want Oracle Database to automatically generate the name of
the next archived redo log file needed to continue the recovery operation. If the LOG_
ARCHIVE_DEST_n parameters are defined, then Oracle Database scans those that are
valid and enabled for the first local destination. It uses that destination in conjunction
with LOG_ARCHIVE_FORMAT to generate the target redo log filename. If the LOG_
ARCHIVE_DEST_n parameters are not defined, then Oracle Database uses the value of
the LOG_ARCHIVE_DEST parameter instead.
If the resulting file is found, then Oracle Database applies the redo contained in that
file. If the file is not found, then Oracle Database prompts you for a filename,
displaying the generated filename as a suggestion.
If you specify neither AUTOMATIC nor LOGFILE, then Oracle Database prompts you for a
filename, displaying the generated filename as a suggestion. You can then accept the
generated filename or replace it with a fully qualified filename. If you know that the
archived filename differs from what Oracle Database would generate, then you can
save time by using the LOGFILE clause.
FROM 'location'
Specify FROM 'location' to indicate the location from which the archived redo log file
group is read. The value of location must be a fully specified file location following
the conventions of your operating system. If you omit this parameter, then Oracle
Database assumes that the archived redo log file group is in the location specified by
the initialization parameter LOG_ARCHIVE_DEST or LOG_ARCHIVE_DEST_1.

10-20 Oracle Database SQL Language Reference

ALTER DATABASE

full_database_recovery
The full_database_recovery clause lets you recover an entire database.
DATABASE Specify the DATABASE clause to recover the entire database. This is the

default. You can use this clause only when the database is closed.
STANDBY DATABASE Specify the STANDBY DATABASE clause to manually recover a
physical standby database using the control file and archived redo log files copied
from the primary database. The standby database must be mounted but not open.

This clause recovers only online data files.
■

■

Use the UNTIL clause to specify the duration of the recovery operation.
–

CANCEL indicates cancel-based recovery. This clause recovers the database until
you issue the ALTER DATABASE statement with the RECOVER CANCEL clause.

–

TIME indicates time-based recovery. This parameter recovers the database to
the time specified by the date. The date must be a character literal in the
format 'YYYY-MM-DD:HH24:MI:SS'.

–

CHANGE indicates change-based recovery. This parameter recovers the database
to a transaction-consistent state immediately before the system change
number specified by integer.

–

CONSISTENT recovers the database until all online files are brought to a
consistent SCN point so that the database can be open in read only mode. This
clauses requires the controlfile to be a backup controlfile.

Specify USING BACKUP CONTROLFILE if you want to use a backup control file instead
of the current control file.

partial_database_recovery
The partial_database_recovery clause lets you recover individual tablespaces and
data files.
Specify the TABLESPACE clause to recover only the specified
tablespaces. You can use this clause if the database is open or closed, provided the
tablespaces to be recovered are offline.
TABLESPACE

See Also: "Using Parallel Recovery Processes: Example" on
page 10-42
DATAFILE Specify the DATAFILE clause to recover the specified data files. You can use

this clause when the database is open or closed, provided the data files to be recovered
are offline.
You can identify the data file by name or by number. If you identify it by number, then
filenumber is an integer representing the number found in the FILE# column of the
V$DATAFILE dynamic performance view or in the FILE_ID column of the DBA_DATA_
FILES data dictionary view.
In earlier releases, you could specify STANDBY
TABLESPACE or STANDBY DATAFILE to recover older backups of a specific tablespace or a
specific datafile on the standby to be consistent with the rest of the standby database.
These two clauses are now obsolete. Instead, to recover the standby database to a
consistent point, but no further, use the statement ALTER DATABASE RECOVER MANAGED
STANDBY DATABASE UNTIL CONSISTENT.
STANDBY {TABLESPACE | DATAFILE}

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-21

ALTER DATABASE

LOGFILE
Specify the LOGFILE 'filename' to continue media recovery by applying the specified
redo log file.
TEST
Use the TEST clause to conduct a trial recovery. A trial recovery is useful if a normal
recovery procedure has encountered some problem. It lets you look ahead into the
redo stream to detect possible additional problems. The trial recovery applies redo in a
way similar to normal recovery, but it does not write changes to disk, and it rolls back
its changes at the end of the trial recovery.
You can use this clause only if you have restored a backup taken since the last
RESETLOGS operation. Otherwise, Oracle Database returns an error.
ALLOW ... CORRUPTION
The ALLOW integer CORRUPTION clause lets you specify, in the event of logfile
corruption, the number of corrupt blocks that can be tolerated while allowing recovery
to proceed.
See Also:
■

■

Oracle Database Backup and Recovery User's Guide for information
on database recovery in general
Oracle Data Guard Concepts and Administration for information on
managed recovery of standby databases

CONTINUE
Specify CONTINUE to continue multi-instance recovery after it has been interrupted to
disable a thread.
Specify CONTINUE DEFAULT to continue recovery using the redo log file that Oracle
Database would automatically generate if no other logfile were specified. This clause is
equivalent to specifying AUTOMATIC, except that Oracle Database does not prompt for a
filename.
CANCEL
Specify CANCEL to terminate cancel-based recovery.
managed_standby_recovery
Use the managed_standby_recovery clause to start and stop Redo Apply on a physical
standby database. Redo Apply keeps the standby database transactionally consistent
with the primary database by continuously applying redo received from the primary
database.
A primary database transmits its redo data to standby sites. As the redo data is written
to redo log files at the physical standby site, the log files become available for use by
Redo Apply. You can use the managed_standby_recovery clause when your standby
instance has the database mounted or is opened read-only.

10-22 Oracle Database SQL Language Reference

ALTER DATABASE

Parallelism is enabled by default during Redo Apply. The
database computes the degree of parallelism. You can disable
parallelism of these operations by specifying NOPARALLEL, or specify a
degree of parallelism with PARALLEL integer, as shown in the
respective syntax diagrams.
Note:

The same restrictions listed under
general_recovery on page 10-19 apply to this clause.

Restrictions on Managed Standby Recovery

See Also: Oracle Data Guard Concepts and Administration for more
information on the use of this clause

Specify USING CURRENT LOGFILE to invoke
real-time apply, which recovers redo from the standby redo log files as soon as they
are written, without requiring them to be archived first at the physical standby
database.

USING CURRENT LOGFILE Clause

See Also: Oracle Data Guard Concepts and Administration for more
information on real-time apply
DISCONNECT Specify DISCONNECT to indicate that Redo Apply should be performed
in the background, leaving the current session available for other tasks. The FROM
SESSION keywords are optional and are provided for semantic clarity.

The NODELAY clause overrides the DELAY attribute on the LOG_ARCHIVE_
DEST_n parameter on the primary database. If you do not specify the NODELAY clause,
then application of the archived redo log file is delayed according to the DELAY
attribute of the LOG_ARCHIVE_DEST_n setting (if any). If the DELAY attribute was not
specified on that parameter, then the archived redo log file is applied immediately to
the standby database.
NODELAY

If you specify real-time apply with the USING CURRENT LOGFILE clause, then any DELAY
value specified for the LOG_ARCHIVE_DEST_n parameter at the primary for this standby
is ignored, and NODELAY is the default.
UNTIL CHANGE Clause Use this clause to instruct Redo Apply to recover redo data

up to, but not including, the specified system change number.
Use this clause to recover the standby database to a consistent
SCN point so that the standby database can be opened in read only mode.

UNTIL CONSISTENT

FINISH Specify FINISH to complete applying all available redo data in preparation
for a failover.

Use the FINISH clause only in the event of the failure of the primary database. This
clause overrides any specified delay intervals and applies all available redo
immediately. After the FINISH command completes, this database can no longer run in
the standby database role, and it must be converted to a primary database by issuing
the ALTER DATABASE COMMIT TO SWITCHOVER TO PRIMARY statement.
Specify CANCEL to stop Redo Apply immediately. Control is returned as
soon as Redo Apply stops.

CANCEL

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-23

ALTER DATABASE

Use this clause to convert a physical standby
database into a logical standby database.

TO LOGICAL STANDBY Clause

Specify a database name to identify the new logical standby database. If
you are using a server parameter file (spfile) at the time you issue this statement, then
the database will update the file with appropriate information about the new logical
standby database. If you are not using an spfile, then the database issues a message
reminding you to set the name of the DB_NAME parameter after shutting down the
database. In addition, you must invoke the DBMS_LOGSTDBY.BUILD PL/SQL procedure
on the primary database before using this clause on the standby database.
db_name

See Also: Oracle Database PL/SQL Packages and Types Reference for
information about the DBMS_LOGSTDBY.BUILD procedure

Use this clause if you want to use the rolling upgrade feature
provided by a logical standby and also revert to the original configuration of a
primary database and a physical standby. A logical standby database created using
this clause provides only limited support for switchover and failover. Therefore, do
not use this clause create a general-purpose logical standby database.

KEEP IDENTITY

See Also: Oracle Data Guard Concepts and Administration for more
information on rolling upgrade

Deprecated Managed Standby Recovery Clauses
The following clauses appeared in the syntax of earlier releases. They have been
deprecated and are no longer needed. Oracle recommends that you do not use these
clauses.
FINISH FORCE, FINISH WAIT, FINISH NOWAIT These optional forms of the FINISH
clause are deprecated. Their semantics are presented here for backward compatibility:
■

■
■

FORCE terminates inactive redo transport sessions that would otherwise prevent
FINISH processing from beginning.
NOWAIT returns control to the foreground process before the recovery completes
WAIT (the default) returns control to the foreground process after recovery
completes

When specified, these clauses are ignored. Terminal recovery now runs in the
foreground and always terminates all redo transport sessions. Therefore control is not
returned to the user until recovery completes.
These optional forms of the
CANCEL clause are deprecated. Their semantics are presented here for backward
compatibility:
CANCEL IMMEDIATE, CANCEL WAIT, CANCEL NOWAIT

■

■

Include the IMMEDIATE keyword to stop Redo Apply before completely applying the
current redo log file. Session control returns when Redo Apply actually stops.
Include the NOWAIT keyword to return session control without waiting for the
CANCEL operation to complete.

When specified, these clauses are ignored. Redo Apply is now always cancelled
immediately and control returns to the session only after the operation completes.

10-24 Oracle Database SQL Language Reference

ALTER DATABASE

BACKUP Clauses
Use these clauses to move all the data files in the database into or out of online backup
mode (also called hot backup mode).
See Also: ALTER TABLESPACE on page 12-90 for information on
moving all data files in an individual tablespace into and out of online
backup mode

BEGIN BACKUP Clause
Specify BEGIN BACKUP to move all data files in the database into online backup mode.
The database must be mounted and open, and media recovery must be enabled (the
database must be in ARCHIVELOG mode).
While the database is in online backup mode, you cannot shut down the instance
normally, begin backup of an individual tablespace, or take any tablespace offline or
make it read only.
This clause has no effect on data files that are in offline or on read-only tablespaces.
END BACKUP Clause
Specify END BACKUP to take out of online backup mode any data files in the database
currently in online backup mode. The database must be mounted (either open or
closed) when you perform this operation.
After a system failure, instance failure, or SHUTDOWN ABORT operation, Oracle Database
does not know whether the files in online backup mode match the files at the time the
system crashed. If you know the files are consistent, then you can take either
individual data files or all data files out of online backup mode. Doing so avoids
media recovery of the files upon startup.
■

■

To take an individual data file out of online backup mode, use the ALTER DATABASE
DATAFILE ... END BACKUP statement. See database_file_clauses on page 10-25.
To take all data files in a tablespace out of online backup mode, use an ALTER
TABLESPACE ... END BACKUP statement.

database_file_clauses
The database_file_clauses let you modify data files and temp files. You can use any
of the following clauses when your instance has the database mounted, open or closed,
and the files involved are not in use.
RENAME FILE Clause
Use the RENAME FILE clause to rename data files, temp files, or redo log file members.
You must create each filename using the conventions for filenames on your operating
system before specifying this clause.
■

■
■

To use this clause for a data file or temp file, the database must be mounted. The
database can also be open, but the data file or temp file being renamed must be
offline. In addition, you must first rename the file on the file system to the new
name.
To use this clause for logfiles, the database must be mounted but not open.
If you have enabled block change tracking, then you can use this clause to rename
the block change tracking file. The database must be mounted but not open when
you rename the block change tracking file.

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-25

ALTER DATABASE

This clause renames only files in the control file. It does not actually rename them on
your operating system. The operating system files continue to exist, but Oracle
Database no longer uses them.
See Also:
■

■

Oracle Database Backup and Recovery User's Guide for information
on recovery of data files and temp files
"Renaming a Log File Member: Example" on page 10-43 and
"Manipulating Temp Files: Example" on page 10-44

create_datafile_clause
Use the CREATE DATAFILE clause to create a new empty data file in place of an old one.
You can use this clause to re-create a data file that was lost with no backup. The
filename or filenumber must identify a file that is or was once part of the database. If
you identify the file by number, then filenumber is an integer representing the
number found in the FILE# column of the V$DATAFILE dynamic performance view or
in the FILE_ID column of the DBA_DATA_FILES data dictionary view.
■

■

Specify AS NEW to create an Oracle-managed data file with a system-generated
filename, the same size as the file being replaced, in the default file system location
for data files.
Specify AS file_specification to assign a file name (and optional size) to the
new data file. Use the datafile_tempfile_spec form of file_specification (see
file_specification on page 8-29) to list regular data files and temp files in an
operating system file system or to list Oracle Automatic Storage Management
(Oracle ASM) disk group files.

If the original file (filename or filenumber) is an existing Oracle-managed data file,
then Oracle Database attempts to delete the original file after creating the new file. If
the original file is an existing user-managed data file, then Oracle Database does not
attempt to delete the original file.
If you omit the AS clause entirely, then Oracle Database creates the new file with the
same name and size as the file specified by filename or filenumber.
During recovery, all archived redo logs written to since the original data file was
created must be applied to the new, empty version of the lost data file.
Oracle Database creates the new file in the same state as the old file when it was
created. You must perform media recovery on the new file to return it to the state of
the old file at the time it was lost.
Restrictions on Creating New Data Files The creation of new data files is subject to
the following restrictions:
■
■

You cannot create a new file based on the first data file of the SYSTEM tablespace.
You cannot specify the autoextend_clause of datafile_tempfile_spec in this
CREATE DATAFILE clause.

10-26 Oracle Database SQL Language Reference

ALTER DATABASE

See Also:
■

■

"DATAFILE Clause" on page 14-15 of CREATE DATABASE for
information on the result of this clause if you do not specify a
name for the new data file
file_specification on page 8-29 for a full description of the file
specification (datafile_tempfile_spec) and "Creating a New
Data File: Example" on page 10-43

alter_datafile_clause
The DATAFILE clause lets you manipulate a file that you identify by name or by
number. If you identify it by number, then filenumber is an integer representing the
number found in the FILE# column of the V$DATAFILE dynamic performance view or
in the FILE_ID column of the DBA_DATA_FILES data dictionary view. The DATAFILE
clauses affect your database files as follows:
ONLINE

Specify ONLINE to bring the data file online.

OFFLINE Specify OFFLINE to take the data file offline. If the database is open, then
you must perform media recovery on the data file before bringing it back online,
because a checkpoint is not performed on the data file before it is taken offline.
FOR DROP If the database is in NOARCHIVELOG mode, then you must specify FOR DROP
clause to take a data file offline. However, this clause does not remove the data file
from the database. To do that, you must use an operating system command or drop
the tablespace in which the data file resides. Until you do so, the data file remains in
the data dictionary with the status RECOVER or OFFLINE.

If the database is in ARCHIVELOG mode, then Oracle Database ignores the FOR DROP
clause.
RESIZE Specify RESIZE if you want Oracle Database to attempt to increase or

decrease the size of the data file to the specified absolute size in bytes. There is no
default, so you must specify a size.
If sufficient disk space is not available for the increased size, or if the file contains data
beyond the specified decreased size, then Oracle Database returns an error.
See Also:

"Resizing a Data File: Example" on page 10-44

END BACKUP Specify END BACKUP to take the data file out of online backup mode.
The END BACKUP clause is described more fully at the top level of the syntax of ALTER
DATABASE. See "END BACKUP Clause" on page 10-25.

alter_tempfile_clause
Use the TEMPFILE clause to resize your temporary data file or specify the autoextend_
clause, with the same effect as for a permanent data file. The database must be open.
You can identify the temp file by name or by number. If you identify it by number,
then filenumber is an integer representing the number found in the FILE# column of
the V$TEMPFILE dynamic performance view.

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-27

ALTER DATABASE

On some operating systems, Oracle does not allocate space
for a temp file until the temp file blocks are actually accessed. This
delay in space allocation results in faster creation and resizing of
temp files, but it requires that sufficient disk space is available
when the temp files are later used. To avoid potential problems,
before you create or resize a temp file, ensure that the available disk
space exceeds the size of the new temp file or the increased size of a
resized temp file. The excess space should allow for anticipated
increases in disk space use by unrelated operations as well. Then
proceed with the creation or resizing operation.

Note:

DROP

Specify DROP to drop tempfile from the database. The tablespace remains.

If you specify INCLUDING DATAFILES, then Oracle Database also deletes the associated
operating system files and writes a message to the alert log for each such deleted file.
You can achieve the same result using an ALTER TABLESPACE ... DROP TEMPFILE
statement. Refer to the ALTER TABLESPACE DROP Clause on page 12-96 for more
information.
autoextend_clause
Use the autoextend_clause to enable or disable the automatic extension of a new or
existing data file or temp file. Refer to file_specification on page 8-29 for information
about this clause.

logfile_clauses
The logfile clauses let you add, drop, or modify log files.
ARCHIVELOG
Specify ARCHIVELOG if you want the contents of a redo log file group to be archived
before the group can be reused. This mode prepares for the possibility of media
recovery. Use this clause only after shutting down your instance normally, or
immediately with no errors, and then restarting it and mounting the database.
MANUAL Specify MANUAL to indicate that Oracle Database should create redo log files,
but the archiving of the redo log files is controlled entirely by the user. This clause is
provided for backward compatibility, for example for users who archive directly to
tape. If you specify MANUAL, then:
■

■

Oracle Database does not archive redo log files when a log switch occurs. You
must handle this manually.
You cannot have specified a standby database as an archivelog destinations. As a
result, the database cannot be in MAXIMUM PROTECTION or MAXIMUM AVAILABILITY
standby protection mode.

If you omit this clause, then Oracle Database automatically archives the redo log files
to the destination specified in the LOG_ARCHIVE_DEST_n initialization parameters.
NOARCHIVELOG
Specify NOARCHIVELOG if you do not want the contents of a redo log file group to be
archived so that the group can be reused. This mode does not prepare for recovery
after media failure. Use this clause only if your instance has the database mounted but
not open.

10-28 Oracle Database SQL Language Reference

ALTER DATABASE

[NO] FORCE LOGGING
Use this clause to put the database into or take the database out of FORCE LOGGING
mode. The database must be mounted or open.
In FORCE LOGGING mode, Oracle Database logs all changes in the database except
changes in temporary tablespaces and temporary segments. This setting takes
precedence over and is independent of any NOLOGGING or FORCE LOGGING settings you
specify for individual tablespaces and any NOLOGGING settings you specify for
individual database objects.
If you specify FORCE LOGGING, then Oracle Database waits for all ongoing unlogged
operations to finish.
See Also: Oracle Database Administrator's Guide for information on
when to use FORCE LOGGING mode

RENAME FILE Clause
This clause has the same function for logfiles that it has for data files and temp files.
See "RENAME FILE Clause" on page 10-25.
CLEAR LOGFILE Clause
Use the CLEAR LOGFILE clause to reinitialize an online redo log, optionally without
archiving the redo log. CLEAR LOGFILE is similar to adding and dropping a redo log,
except that the statement may be issued even if there are only two logs for the thread
and may be issued for the current redo log of a closed thread.
For a standby database, if the STANDBY_FILE_MANAGEMENT initialization parameter is set
to AUTO, and if any of the log files are Oracle Managed Files, Oracle Database will
create as many Oracle-managed log files as are in the control file. The log file members
will reside in the current default log file destination.
■

You must specify UNARCHIVED if you want to reuse a redo log that was not
archived.
Caution: Specifying UNARCHIVED makes backups unusable if the redo
log is needed for recovery.

■

You must specify UNRECOVERABLE DATAFILE if you have taken the data file offline
with the database in ARCHIVELOG mode (that is, you specified ALTER DATABASE ...
DATAFILE OFFLINE without the DROP keyword), and if the unarchived log to be
cleared is needed to recover the data file before bringing it back online. In this
case, you must drop the data file and the entire tablespace once the CLEAR LOGFILE
statement completes.
Do not use CLEAR LOGFILE to clear a log needed for media recovery. If it is
necessary to clear a log containing redo after the database checkpoint, then you
must first perform incomplete media recovery. The current redo log of an open
thread can be cleared. The current log of a closed thread can be cleared by
switching logs in the closed thread.
If the CLEAR LOGFILE statement is interrupted by a system or instance failure, then
the database may hang. In this case, reissue the statement after the database is
restarted. If the failure occurred because of I/O errors accessing one member of a
log group, then that member can be dropped and other members added.
See Also:

"Clearing a Log File: Example" on page 10-44

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-29

ALTER DATABASE

add_logfile_clauses
Use these clauses to add redo log file groups to the database and to add new members
to existing redo log file groups.
Use the ADD LOGFILE clause to add one or more redo log file
groups to the online redo log or standby redo log.

ADD LOGFILE Clause

See Also:
■

■
■

"LOGFILE Clause" on page 14-24 of CREATE DATABASE for
information on the result of this clause for Oracle Managed Files if
you do not specify a name for the new log file group
"Adding Redo Log File Groups: Examples" on page 10-42
Oracle Data Guard Concepts and Administration for more
information on standby redo logs

STANDBY Use the STANDBY clause to add a redo log file group to the standby redo
log. If you do not specify this clause, then a log file group is added to the online redo
log.
INSTANCE The INSTANCE clause is applicable only for Oracle Real Application
Clusters (Oracle RAC) or Oracle RAC One Node databases. Specify the name of the
instance for which you want to add a redo log file group. The instance name is a string
of up to 80 characters. Oracle Database automatically uses the thread that is mapped to
the specified instance. If no thread is mapped to the specified instance, then Oracle
Database automatically acquires an available unmapped thread and assigns it to that
instance. If you do not specify this clause, then Oracle Database executes the command
as if you had specified the current instance. If the specified instance has no current
thread mapping and there are no available unmapped threads, then Oracle Database
returns an error.
THREAD When adding a redo log file group to the standby redo log, use the THREAD
clause to assign the log file group to a specific primary database redo thread. Query
the V$INSTANCE view on the primary database to determine which redo threads have
been opened, and specify one of these thread numbers.

You can also use the THREAD clause to assign a log file group to a specific redo thread
when adding the log file group to the online redo log. This usage has been deprecated.
The INSTANCE clause achieves the same purpose and is easier to use.
GROUP The GROUP clause uniquely identifies the redo log file group among all
groups in all threads and can range from 1 to the value specified for MAXLOGFILES in
the CREATE DATABASE statement. You cannot add multiple redo log file groups having
the same GROUP value. If you omit this parameter, then Oracle Database generates its
value automatically. You can examine the GROUP value for a redo log file group through
the dynamic performance view V$LOG.
redo_log_file_spec Each redo_log_file_spec specifies a redo log file group
containing one or more members (copies). If you do not specify a filename for the new
log file, then Oracle Database creates Oracle Managed Files according to the rules
described in the "LOGFILE Clause" on page 14-24 of CREATE DATABASE.

10-30 Oracle Database SQL Language Reference

ALTER DATABASE

See Also:
■
■

file_specification on page 8-29
Oracle Database Reference for information on dynamic performance
views

Use the ADD LOGFILE MEMBER clause to add new
members to existing redo log file groups. Each new member is specified by
'filename'. If the file already exists, then it must be the same size as the other group
members and you must specify REUSE. If the file does not exist, then Oracle Database
creates a file of the correct size. You cannot add a member to a group if all of the
members of the group have been lost through media failure.
ADD LOGFILE MEMBER Clause

STANDBY You must specify STANDBY when adding a member to a standby redo log

file group. Otherwise, Oracle Database returns an error.
You can use the logfile_descriptor clause to specify an existing redo log file group
in one of two ways:
GROUP integer Specify the value of the GROUP parameter that identifies the redo log
file group.

List all members of the redo log file group. You must fully specify each
filename according to the conventions of your operating system.

filename(s)

See Also:
■

■

"LOGFILE Clause" on page 14-24 of CREATE DATABASE for
information on the result of this clause for Oracle Managed Files if
you do not specify a name for the new log file group
"Adding Redo Log File Group Members: Example" on page 10-43

drop_logfile_clauses
Use these clauses to drop redo log file groups or redo log file members.
DROP LOGFILE Clause Use the DROP LOGFILE clause to drop all members of a redo
log file group. If you use this clause to drop Oracle Managed Files, then Oracle
Database also removes all log file members from disk. Specify a redo log file group as
indicated for the ADD LOGFILE MEMBER clause.
■

■
■

To drop the current log file group, you must first issue an ALTER SYSTEM SWITCH
LOGFILE statement.
You cannot drop a redo log file group if it needs archiving.
You cannot drop a redo log file group if doing so would cause the redo thread to
contain less than two redo log file groups.
ALTER SYSTEM on page 11-58 and "Dropping Log File
Members: Example" on page 10-43

See Also:

Use the DROP LOGFILE MEMBER clause to drop one or
more redo log file members. Each 'filename' must fully specify a member using the
conventions for filenames on your operating system.
DROP LOGFILE MEMBER Clause

■

To drop a log file in the current log, you must first issue an ALTER SYSTEM SWITCH
LOGFILE statement. Refer to ALTER SYSTEM on page 11-58 for more information.

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-31

ALTER DATABASE

■

You cannot use this clause to drop all members of a redo log file group that
contains valid data. To perform that operation, use the DROP LOGFILE clause.
See Also:

"Dropping Log File Members: Example" on page 10-43

switch_logfile_clause
This clause is useful when you are migrating the database to disks with a different
block size that the block size of the current database. Use this clause to switch logfiles
to a different block size for all externally enabled threads, including both open and
closed threads. If you are migrating the database to use 4KB sector disks, then you
must specify 4096 for integer. If you are unmigrating the database back to using 512B
sector disks, then you must specify 512 for integer.
This clause is an extension of the existing ALTER SYSTEM SWITCH LOGFILE statement.
That statement switches logs for a single thread. This clause switches logfiles for all
externally enabled threads, including both open and closed threads.
Before using this clause, you must already have created at least two redo log groups
with the same target block size on the migration target disk.
See Also: Oracle Database Administrator's Guide for more information
on migrating the database to disks with a different block size, and
"Adding a Log File: Example" on page 8-36

supplemental_db_logging
Use these clauses to instruct Oracle Database to add or stop adding supplemental data
into the log stream.
Specify ADD SUPPLEMENTAL LOG DATA to enable
minimal supplemental logging. Specify ADD SUPPLEMENTAL LOG supplemental_id_
key_clause to enable column data logging in addition to minimal supplemental
logging. Specify ADD SUPPLEMENTAL LOG supplemental_plsql_clause to enable
supplemental logging of PL/SQL calls. Oracle Database does not enable either
minimal supplemental logging or supplemental logging by default.
ADD SUPPLEMENTAL LOG Clause

Minimal supplemental logging ensures that LogMiner (and any products building on
LogMiner technology) will have sufficient information to support chained rows and
various storage arrangements such as cluster tables.
If the redo generated on one database is to be the source of changes (to be mined and
applied) at another database, as is the case with logical standby, then the affected rows
need to be identified using column data (as opposed to rowids). In this case, you
should specify the supplemental_id_key_clause.
You can query the appropriate columns in the V$DATABASE view to determine whether
any supplemental logging has already been enabled.
You can issue this statement when the database is open. However, Oracle Database
will invalidate all DML cursors in the cursor cache, which will have an effect on
performance until the cache is repopulated.
For a full discussion of the supplemental_id_clause, refer to supplemental_id_key_
clause on page 16-31 in the documentation on CREATE TABLE.

10-32 Oracle Database SQL Language Reference

ALTER DATABASE

See Also:

Oracle Data Guard Concepts and Administration for information on
supplemental logging on the primary database to support a logical
standby database
Oracle Database Utilities for examples using the supplemental_db_
logging clause syntax
DROP SUPPLEMENTAL LOG Clause
Use this clause to stop supplemental logging.
■

■

■

Specify DROP SUPPLEMENTAL LOG DATA to instruct Oracle Database to stop placing
minimal additional log information into the redo log stream whenever an update
operation occurs. If Oracle Database is doing column data supplemental logging
specified with the supplemental_id_key_clause, then you must first stop the
column data supplemental logging with the DROP SUPPLEMENTAL LOG
supplemental_id_key_clause and then specify this clause.
Specify DROP SUPPLEMENTAL LOG supplemental_id_key_clause to drop some or all
of the system-generated supplemental log groups. You must specify the
supplemental_id_key_clause if the supplemental log groups you want to drop
were added using that clause.
Specify DROP SUPPLEMENTAL LOG supplemental_plsql_clause disable
supplemental logging of PL/SQL calls.
See Also: Oracle Data Guard Concepts and Administration for
information on supplemental logging

controlfile_clauses
The controlfile_clauses let you create or back up a control file.
CREATE STANDBY CONTROLFILE Clause
The CREATE STANDBY CONTROLFILE clause lets you create a control file to be used to
maintain a physical or logical standby database. If you do not specify LOGICAL or
PHYSICAL, then the default is PHYSICAL. If the file already exists, then you must specify
REUSE. In an Oracle RAC environment, the control file must be on shared storage.
See Also:

Oracle Data Guard Concepts and Administration

BACKUP CONTROLFILE Clause
Use the BACKUP CONTROLFILE clause to back up the current control file. The database
must be open or mounted when you specify this clause.
Use this clause to specify a binary backup of the control file. You must
fully specify the filename using the conventions for your operating system. If the
specified file already exists, then you must specify REUSE. In an Oracle RAC
environment, filename must be on shared storage.
TO 'filename'

A binary backup contains information that is not captured if you specify TO TRACE,
such as the archived log history, offline range for read-only and offline tablespaces,
and backup sets and copies (if you use RMAN). If the COMPATIBLE initialization
parameter is 10.2 or higher, binary control file backups include temp file entries.
Specify TO TRACE if you want Oracle Database to write SQL statements to
a trace file rather than making a physical backup of the control file. You can use SQL

TO TRACE

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-33

ALTER DATABASE

statements written to the trace file to start up the database, re-create the control file,
and recover and open the database appropriately, based on the created control file. If
you issue an ALTER DATABASE BACKUP CONTROLFILE TO TRACE statement while block
change tracking is enabled, then the resulting trace file will contain a command to
reenable block change tracking.
The trace file will also include ALTER DATABASE REGISTER LOGFILE statements for
existing logfiles that reside in the current archivelog destinations. This will implicitly
create database incarnation records for the branches of redo to which the logfiles
apply.
You can copy the statements from the trace file into a script file, edit the statements as
necessary, and use the script if all copies of the control file are lost (or to change the
size of the control file).
■

■

■

■

Specify AS filename if you want Oracle Database to place the trace output into a
file called filename rather than into the standard trace file.
Specify REUSE to allow Oracle Database to overwrite any existing file called
filename.
RESETLOGS indicates that the SQL statement written to the trace file for starting the
database is ALTER DATABASE OPEN RESETLOGS. This setting is valid only if the online
logs are unavailable.
NORESETLOGS indicates that the SQL statement written to the trace file for starting
the database is ALTER DATABASE OPEN NORESETLOGS. This setting is valid only if all
the online logs are available.

If you cannot predict the future state of the online logs, then specify neither RESETLOGS
nor NORESETLOGS. In this case, Oracle Database puts both versions of the script into the
trace file, and you can choose which version is appropriate when the script becomes
necessary.
The trace files are stored in a subdirectory determined by the DIAGNOSTIC_DEST
initialization parameter. You can find the name and location of the trace file to which
the CREATE CONTROLFILE statements were written by looking in the alert log. You can
also find the directory for trace files by querying the NAME and VALUE columns of the
V$DIAG_INFO dynamic performance view.
See Also: Oracle Database Administrator's Guide for information on
viewing the alert log

standby_database_clauses
Use these clauses to activate the standby database or to specify whether it is in
protected or unprotected mode.
See Also: Oracle Data Guard Concepts and Administration for
descriptions of the physical and logical standby database and for
information on maintaining and using standby databases

activate_standby_db_clause
Use the ACTIVATE STANDBY DATABASE clause to convert a standby database into a
primary database.
Before using this command, refer to Oracle Data Guard
Concepts and Administration for important usage information.

Caution:

10-34 Oracle Database SQL Language Reference

ALTER DATABASE

PHYSICAL

Specify PHYSICAL to activate a physical standby database. This is the

default.
Specify LOGICAL to activate a logical standby database. If you have more
than one logical standby database, then you should first ensure that the same log data
is available on all the standby systems.

LOGICAL

FINISH APPLY This clause applies only to logical standby databases. Use it to initiate
terminal apply, which is the application of any remaining redo to bring the logical
standby database to the same state as the primary database. When terminal apply is
complete, the database completes the switchover from logical standby to primary
database.

If you require immediate restoration of the database in spite of data loss, then omit this
clause. The database will execute the switchover from logical standby to primary
database immediately without terminal apply.
maximize_standby_db_clause
Use this clause to specify the level of protection for the data in your database
environment. You specify this clause from the primary database.
The PROTECTED and UNPROTECTED keywords have been replaced
for clarity but are still supported. PROTECTED is equivalent to TO
MAXIMIZE PROTECTION. UNPROTECTED is equivalent to TO MAXIMIZE
PERFORMANCE.
Note:

This setting establishes maximum protection mode and
offers the highest level of data protection. A transaction does not commit until all data
needed to recover that transaction has been written to at least one physical standby
database that is configured to use the SYNC log transport mode. If the primary database
is unable to write the redo records to at least one such standby database, then the
primary database is shut down. This mode guarantees zero data loss, but it has the
greatest potential impact on the performance and availability of the primary database.

TO MAXIMIZE PROTECTION

Restriction on Establishing Maximum Protection Mode You can specify TO MAXIMIZE

PROTECTION on an open database only if the current data protection mode is MAXIMUM
AVAILABILITY and there is at least one synchronized standby database.
TO MAXIMIZE AVAILABILITY This setting establishes maximum availability mode
and offers the next highest level of data protection. A transaction does not commit
until all data needed to recover that transaction has been written to at least one
physical or logical standby database that is configured to use the SYNC log transport
mode. Unlike maximum protection mode, the primary database does not shut down if
it is unable to write the redo records to at least one such standby database. Instead, the
protection is lowered to maximum performance mode until the fault has been
corrected and the standby database has caught up with the primary database. This
mode guarantees zero data loss unless the primary database fails while in maximum
performance mode. Maximum availability mode provides the highest level of data
protection that is possible without affecting the availability of the primary database.
TO MAXIMIZE PERFORMANCE This setting establishes maximum performance
mode and is the default setting. A transaction commits before the data needed to
recover that transaction has been written to a standby database. Therefore, some
transactions may be lost if the primary database fails and you are unable to recover the

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-35

ALTER DATABASE

redo records from the primary database. This mode provides the highest level of data
protection that is possible without affecting the performance of the primary database.
To determine the current mode of the database, query the PROTECTION_MODE column of
the V$DATABASE dynamic performance view.
See Also: Oracle Data Guard Concepts and Administration for full
information on using these standby database settings

register_logfile_clause
Specify the REGISTER LOGFILE clause from the standby database to manually register
log files from the failed primary. Use the redo_log_file_spec form of file_
specification (see file_specification on page 8-29) to list regular redo log files in an
operating system file system or to list Oracle ASM disk group redo log files.
When a log file is from an unknown incarnation, the REGISTER LOGFILE clause causes
an incarnation record to be added to the V$DATABASE_INCARNATION view. If the newly
registered log file belongs to an incarnation having a higher RESETLOGS_TIME than the
current RECOVERY_TARGET_INCARNATION#, then the REGISTER LOGFILE clause also causes
RECOVERY_TARGET_INCARNATION# to be changed to correspond to the newly added
incarnation record.
OR REPLACE Specify OR REPLACE to allow an existing archivelog entry in the standby
database to be updated, for example, when its location or file specification changes.
The system change numbers of the entries must match exactly, and the original entry
must have been created by the managed standby log transmittal mechanism.
FOR logminer_session_name This clause is useful in a Streams environment. It lets
you register the log file with one specified LogMiner session.

commit_switchover_clause
Use this clause to perform database role transitions in a Data Guard configuration.
Before using this command, refer to Oracle Data Guard
Concepts and Administration for complete usage information.

Caution:

PREPARE TO SWITCHOVER This clause prepares a primary database to become a
logical standby database or a logical standby database to become a primary database.
■
■

Specify PREPARE TO SWITCHOVER TO LOGICAL STANDBY on a primary database.
Specify PREPARE TO SWITCHOVER TO PRIMARY DATABASE on a logical standby
database.

This clause switches a primary database to a standby
database role or switches a standby database to the primary database role.

COMMIT TO SWITCHOVER
■

■

Specify COMMIT TO SWITCHOVER TO PHYSICAL STANDBY or COMMIT TO SWITCHOVER TO
LOGICAL STANDBY on a primary database.
Specify COMMIT TO SWITCHOVER TO PRIMARY DATABASE on a standby database.

This clause is always optional. Use of this clause with the COMMIT TO
SWITCHOVER TO PRIMARY clause has been deprecated.
PHYSICAL

This clause is specified with the PREPARE TO SWITCHOVER or COMMIT TO
SWITCHOVER clauses when switching a primary database to the logical standby
LOGICAL

10-36 Oracle Database SQL Language Reference

ALTER DATABASE

database role. Use of this clause with the COMMIT TO SWITCHOVER TO PRIMARY clause has
been deprecated.
WITH SESSION SHUTDOWN This clause causes all database sessions to be closed and
uncommitted transactions to be rolled back before performing a database role
transition.
WITHOUT SESSION SHUTDOWN This clause prevents a requested role transition

from occurring if there are any database sessions. This is the default.
Specify this clause to wait for a role transition to complete before returning
control to the user.

WAIT

Specify this clause to return control to the user without waiting for a role
transition to complete. This is the default.

NOWAIT

Specify this clause to reverse the effect of a previously specified PREPARE TO
SWITCHOVER statement.
CANCEL

See Also: Oracle Data Guard Concepts and Administration for full
information on switchover between primary and standby databases

start_standby_clause
Specify the START LOGICAL STANDBY APPLY clause to begin applying redo logs to a
logical standby database. This clause enables primary key, unique index, and unique
constraint supplemental logging as well as PL/SQL call logging.
■
■

Specify IMMEDIATE to apply redo data from the current standby redo log file.
Specify NODELAY if you want Oracle Database to ignore a delay for this apply. This
is useful if the primary database is no longer present, which would otherwise
require a PL/SQL call to be made.

■

Specify INITIAL the first time you apply the logs to the standby database.

■

The NEW PRIMARY clause is needed in two situations:
■

■

■

■

On a failover to a logical standby, specify this clause on a logical standby not
participating in the failover operation, and on the old primary database after it
has been reinstated as a logical standby database.
During a rolling upgrade using a logical standby database (which uses an
unprepared switchover operation), specify this clause after the original
primary database has been upgraded to the new database software.

Specify SKIP FAILED [TRANSACTION] to skip the last transaction in the events table
and restart the apply.
Specify FINISH to force the standby redo logfile information into archived logs. If
the primary database becomes disabled, then you can then apply the data in the
redo log files.

stop_standby_clause
Use this clause to stop the log apply services. This clause applies only to logical
standby databases, not to physical standby databases. Use the STOP clause to stop the
apply in an orderly fashion.

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-37

ALTER DATABASE

convert_database_clause
Use this clause to convert a database from one form to another.
■

Specify CONVERT TO PHYSICAL STANDBY to convert a primary database, a logical
standby database, or a snapshot standby database into a physical standby
database.
Perform these steps before specifying this clause:
–

On an Oracle Real Application Clusters (Oracle RAC) database, shut down all
but one instance.

–

Ensure that the database is mounted, but not open.

The database is dismounted after conversion and must be restarted.
■

Specify CONVERT TO SNAPSHOT STANDBY to convert a physical standby database into
a snapshot standby database.
Ensure that redo apply is stopped before specifying this clause.
A snapshot standby database must be opened at least once in
read/write mode before it can be converted into a physical standby
database.

Note:

See Also: Oracle Data Guard Concepts and Administration for more
information about standby databases

default_settings_clauses
Use these clauses to modify the default settings of the database.
DEFAULT EDITION Clause
Use this clause to designate the specified edition as the default edition for the
database. The specified edition must already have been created and must be USABLE.
The change takes place immediately and is visible to all nodes in an Oracle RAC
environment. New database sessions automatically start out in the specified edition.
The new setting persists across database shutdown and startup.
When you designate an edition as the database default edition, all object privileges on
the edition itself are revoked from all users except SYS, and the USE object privilege is
granted on the specified edition to the role PUBLIC.
You can determine the current default edition of the database with the following
query:
SELECT PROPERTY_VALUE FROM DATABASE_PROPERTIES
WHERE PROPERTY_NAME = 'DEFAULT_EDITION';

See Also: CREATE EDITION on page 14-51 for more information on
editions and Oracle Database PL/SQL Language Reference for
information on how editions are designated as USABLE

CHARACTER SET, NATIONAL CHARACTER SET
You can no longer change the database character set or the national character set using
the ALTER DATABASE statement. Refer to Oracle Database Globalization Support Guide for
information on database character set migration.

10-38 Oracle Database SQL Language Reference

ALTER DATABASE

SET DEFAULT TABLESPACE Clause
Use this clause to specify or change the default type of subsequently created
tablespaces. Specify BIGFILE or SMALLFILE to indicate whether the tablespaces should
be bigfile or smallfile tablespaces.
■

■

A bigfile tablespace contains only one data file or temp file, which can contain up
to approximately 4 billion (232) blocks. The maximum size of the single data file or
temp file is 128 terabytes (TB) for a tablespace with 32K blocks and 32TB for a
tablespace with 8K blocks.
A smallfile tablespace is a traditional Oracle tablespace, which can contain 1022
data files or temp files, each of which can contain up to approximately 4 million
(222) blocks.
See Also:
■

■

Oracle Database Administrator's Guide for more information about
bigfile tablespaces
"Setting the Default Type of Tablespaces: Example" on page 10-43

DEFAULT TABLESPACE Clause
Specify this clause to establish or change the default permanent tablespace of the
database. The tablespace you specify must already have been created. After this
operation completes, Oracle Database automatically reassigns to the new default
tablespace all non-SYSTEM users. All objects subsequently created by those users will
by default be stored in the new default tablespace. If you are replacing a previously
specified default tablespace, then you can move the previously created objects from
the old to the new default tablespace, and then drop the old default tablespace if you
want to.
DEFAULT TEMPORARY TABLESPACE Clause
Specify this clause to change the default temporary tablespace of the database to a new
tablespace or tablespace group.
■

■

Specify tablespace to indicate the new default temporary tablespace of the
database. After this operation completes, Oracle Database automatically reassigns
to the new default temporary tablespace all users who had been assigned to the
old default temporary tablespace. You can then drop the old default temporary
tablespace if you want to.
Specify tablespace_group_name to indicate that all tablespaces in the tablespace
group specified by tablespace_group_name are now default temporary tablespace
for the database. After this operation completes, users who have not been
explicitly assigned a default temporary tablespace can create temporary segments
in any of the tablespaces that are part of tablespace_group_name. You cannot drop
the old default temporary tablespace if it is part of the default temporary
tablespace group.

To learn the name of the current default temporary tablespace or default temporary
tablespace group, query the TEMPORARY_TABLESPACE column of the ALL_, DBA-, or USER_
USERS data dictionary views.
Restrictions on Default Temporary Tablespaces

Default temporary tablespaces are

subject to the following restrictions:
■

The tablespace you assign or reassign as the default temporary tablespace must
have a standard block size.

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-39

ALTER DATABASE

■

If the SYSTEM tablespace is locally managed, then the tablespace you specify as the
default temporary tablespace must also be locally managed.
See Also:
■

■

Oracle Database Administrator's Guide for information on tablespace
groups
"Changing the Default Temporary Tablespace: Examples" on
page 10-43

instance_clauses
In an Oracle Real Application Clusters environment, specify ENABLE INSTANCE to
enable the thread that is mapped to the specified database instance. The thread must
have at least two redo log file groups, and the database must be open.
Specify DISABLE INSTANCE to disable the thread that is mapped to the specified
database instance. The name of the instance is a string of up to 80 characters. If no
thread is currently mapped to the specified instance, then Oracle Database returns an
error. The database must be open, but you cannot disable a thread if an instance using
it has the database mounted.
See Also: Oracle Real Application Clusters Administration and
Deployment Guide for more information on enabling and disabling
instances

RENAME GLOBAL_NAME Clause
Specify RENAME GLOBAL_NAME to change the global name of the database. The database
must be open. The database is the new database name and can be as long as eight
bytes. The optional domain specifies where the database is effectively located in the
network hierarchy. If you specify a domain name, then the components of the domain
name must be legal identifiers. See "Database Object Naming Rules" on page 3-111 for
information on valid identifiers.
Renaming your database does not change global references to
your database from existing database links, synonyms, and stored
procedures and functions on remote databases. Changing such
references is the responsibility of the administrator of the remote
databases.

Note:

See Also:

"Changing the Global Database Name: Example" on

page 10-44
BLOCK CHANGE TRACKING Clauses
The block change tracking feature causes Oracle Database to keep track of the
physical locations of all database updates on both the primary database and any
physical standby database. You must enable block change tracking on each database
for which you want tracking to be performed. The tracking information is maintained
in a separate file called the block change tracking file. If you are using Oracle Managed
Files, then Oracle Database automatically creates the block change tracking file in the
location specified by DB_CREATE_FILE_DEST. If you are not using Oracle Managed Files,
then you must specify the change tracking filename. Oracle Database uses change
tracking data for some internal tasks, such as increasing the performance of
incremental backups. You can enable or disable block change tracking with the
database either open or mounted, in either archivelog or NOARCHIVELOG mode.
10-40 Oracle Database SQL Language Reference

ALTER DATABASE

ENABLE BLOCK CHANGE TRACKING This clause enables block change tracking
and causes Oracle Database to create a block change tracking file.
■

■

Specify USING FILE 'filename' if you want to name the block change tracking file
instead of letting Oracle Database generate a name for it. You must specify this
clause if you are not using Oracle Managed Files.
Specify REUSE to allow Oracle Database to overwrite an existing block change
tracking file of the same name.

DISABLE BLOCK CHANGE TRACKING Specify this clause if you want Oracle
Database to stop tracking changes and delete the existing block change tracking file.
See Also: Oracle Database Backup and Recovery User's Guide for
information on setting up block change tracking and "Enabling and
Disabling Block Change Tracking: Examples" on page 10-44

flashback_mode_clause
Use this clause to put the database in or take the database out of FLASHBACK mode. You
can specify this clause only if the database is in ARCHIVELOG mode and you have
already prepared a fast recovery area for the database. You can specify this clause
when the database is mounted or open. This clause cannot be specified on a physical
standby database if redo apply is active.
See Also: Oracle Database Backup and Recovery User's Guide for
information on preparing the fast recovery area for Flashback
operations

Use this clause to put the database in FLASHBACK mode. When the
database is in FLASHBACK mode, Oracle Database automatically creates and manages
Flashback Database logs in the fast recovery area. Users with SYSDBA system privilege
can then issue a FLASHBACK DATABASE statement.

FLASHBACK ON

FLASHBACK OFF Use this clause to take the database out of FLASHBACK mode.
Oracle Database stops logging Flashback data and deletes all existing Flashback
Database logs. Any attempt to issue a FLASHBACK DATABASE will fail with an error.

set_time_zone_clause
This clause has the same semantics in CREATE DATABASE and ALTER DATABASE
statements. When used in with ALTER DATABASE, this clause resets the time zone of the
database. To determine the time zone of the database, query the built-in function
DBTIMEZONE on page 5-76. After setting or changing the time zone with this clause,
you must restart the database for the new time zone to take effect.
Oracle Database normalizes all new TIMESTAMP WITH LOCAL TIME ZONE data to the time
zone of the database when the data is stored on disk.Oracle Database does not
automatically update existing data in the database to the new time zone. Therefore,
you cannot reset the database time zone if there is any TIMESTAMP WITH LOCAL TIME
ZONE data in the database. You must first delete or export the TIMESTAMP WITH LOCAL
TIME ZONE data and then reset the database time zone. For this reason, Oracle does not
encourage you to change the time zone of a database that contains data.
For a full description of this clause, refer to set_time_zone_clause on page 14-30 in the
documentation on CREATE DATABASE.

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-41

ALTER DATABASE

security_clause
Use the security_clause (GUARD) to protect data in the database from being changed.
You can override this setting for a current session using the ALTER SESSION DISABLE
GUARD statement. Refer to ALTER SESSION on page 11-45 for more information.
Specify ALL to prevent all users other than SYS from making any changes to the
database.

ALL

STANDBY Specify STANDBY to prevent all users other than SYS from making changes
to any database object being maintained by logical standby. This setting is useful if you
want report operations to be able to modify data as long as it is not being replicated by
logical standby.
See Also: Oracle Data Guard Concepts and Administration for
information on logical standby
NONE

Specify NONE if you want normal security for all data in the database.
Oracle strongly recommends that you not use this setting
on a logical standby database.

Caution:

Examples
10

READ ONLY / READ WRITE: Example

The following statement opens the database in

read-only mode:
ALTER DATABASE OPEN READ ONLY;

The following statement opens the database in read/write mode and clears the online
redo logs:
ALTER DATABASE OPEN READ WRITE RESETLOGS;

Using Parallel Recovery Processes: Example The following statement performs
tablespace recovery using parallel recovery processes:
ALTER DATABASE
RECOVER TABLESPACE tbs_03
PARALLEL;

The following statement adds a redo log
file group with two members and identifies it with a GROUP parameter value of 3:

Adding Redo Log File Groups: Examples
ALTER DATABASE
ADD LOGFILE GROUP 3
('diska:log3.log' ,
'diskb:log3.log') SIZE 50K;

The following statement adds a redo log file group containing two members to thread
5 (in a Real Application Clusters environment) and assigns it a GROUP parameter value
of 4:
ALTER DATABASE
ADD LOGFILE THREAD 5 GROUP 4
('diska:log4.log',
'diskb:log4:log');

10-42 Oracle Database SQL Language Reference

ALTER DATABASE

The following statement adds a
member to the redo log file group added in the previous example:

Adding Redo Log File Group Members: Example
ALTER DATABASE
ADD LOGFILE MEMBER 'diskc:log3.log'
TO GROUP 3;

Dropping Log File Members: Example The following statement drops one redo log
file member added in the previous example:
ALTER DATABASE
DROP LOGFILE MEMBER 'diskb:log3.log';

The following statement drops all members of the redo log file group 3:
ALTER DATABASE DROP LOGFILE GROUP 3;

Renaming a Log File Member: Example The following statement renames a redo log
file member:
ALTER DATABASE
RENAME FILE 'diskc:log3.log' TO 'diskb:log3.log';

The preceding statement only changes the member of the redo log group from one file
to another. The statement does not actually change the name of the file
diskc:log3.log to diskb:log3.log. Before issuing this statement, you must change
the name of the file through your operating system.
Setting the Default Type of Tablespaces: Example The following statement specifies
that subsequently created tablespaces be created as bigfile tablespaces by default:
ALTER DATABASE
SET DEFAULT BIGFILE TABLESPACE;

Changing the Default Temporary Tablespace: Examples The following statement

makes the tbs_05 tablespace (created in "Creating a Temporary Tablespace: Example"
on page 16-95) the default temporary tablespace of the database. This statement either
establishes a default temporary tablespace if none was specified at create time, or
replaces an existing default temporary tablespace with tbs_05:
ALTER DATABASE
DEFAULT TEMPORARY TABLESPACE tbs_05;

Alternatively, a group of tablespaces can be defined as the default temporary
tablespace by using a tablespace group. The following statement makes the
tablespaces in the tablespace group tbs_group_01 (created in "Adding a Temporary
Tablespace to a Tablespace Group: Example" on page 16-95) the default temporary
tablespaces of the database:
ALTER DATABASE
DEFAULT TEMPORARY TABLESPACE tbs_grp_01;

The following statement creates a new data file
tbs_f04.dbf based on the file tbs_f03.dbf. Before creating the new data file, you
must take the existing data file (or the tablespace in which it resides) offline.
Creating a New Data File: Example

ALTER DATABASE
CREATE DATAFILE 'tbs_f03.dbf'
AS 'tbs_f04.dbf';

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-43

ALTER DATABASE

Manipulating Temp Files: Example The following takes offline the temp file
temp02.dbf created in Adding and Dropping Data Files and Temp Files: Examples on
page 12-101 and then renames the temp file:
ALTER DATABASE TEMPFILE 'temp02.dbf' OFFLINE;
ALTER DATABASE RENAME FILE 'temp02.dbf' TO 'temp03.dbf';

The statement renaming the temp file requires that you first create the file temp03.dbf
on the operating system.
The following statement changes the
global name of the database and includes both the database name and domain:

Changing the Global Database Name: Example

ALTER DATABASE
RENAME GLOBAL_NAME TO demo.world.example.com;

Enabling and Disabling Block Change Tracking: Examples The following statement

enables block change tracking and causes Oracle Database to create a block change
tracking file named tracking_file and overwrite the file if it already exists:
ALTER DATABASE
ENABLE BLOCK CHANGE TRACKING
USING FILE 'tracking_file' REUSE;

The following statement disables block change tracking and deletes the existing block
change tracking file:
ALTER DATABASE
DISABLE BLOCK CHANGE TRACKING;

Resizing a Data File: Example

The following statement attempts to change the size

of data file diskb:tbs_f5.dbf:
ALTER DATABASE
DATAFILE 'diskb:tbs_f5.dbf' RESIZE 10 M;

Clearing a Log File: Example

The following statement clears a log file:

ALTER DATABASE
CLEAR LOGFILE 'diskc:log3.log';

The following statement performs complete
recovery of the entire database, letting Oracle Database generate the name of the next
archived redo log file needed:

Database Recovery: Examples

ALTER DATABASE
RECOVER AUTOMATIC DATABASE;

The following statement explicitly names a redo log file for Oracle Database to apply:
ALTER DATABASE
RECOVER LOGFILE 'diskc:log3.log';

The following statement performs time-based recovery of the database:
ALTER DATABASE
RECOVER AUTOMATIC UNTIL TIME '2001-10-27:14:00:00';

Oracle Database recovers the database until 2:00 p.m. on October 27, 2001.

10-44 Oracle Database SQL Language Reference

ALTER DATABASE

For an example of recovering a tablespace, see "Using Parallel Recovery Processes:
Example" on page 10-42.

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-45

ALTER DATABASE LINK

ALTER DATABASE LINK
Purpose
10

Use the ALTER DATABASE LINK statement to modify a fixed-user database link when the
password of the connection or authentication user changes.
Notes:
■

■

■

You cannot use this statement to change the connection or
authentication user associated with the database link. To change
user, you must re-create the database link.
You cannot use this statement to change the password of a
connection or authentication user. You must use the ALTER USER
on page 13-6 statement for this purpose, and then alter the
database link with the ALTER DATABASE LINK statement.
This statement is valid only for fixed-user database links, not for
connected-user or current user database links. See CREATE
DATABASE LINK on page 14-31 for more information on these
two types of database links.

Prerequisites
10

To alter a private database link, you must have the ALTER DATABASE LINK system
privilege. To alter a public database link, you must have the ALTER PUBLIC DATABASE
LINK system privilege.

Syntax
10

alter_database_link::=
SHARED

PUBLIC

ALTER

DATABASE

LINK

dblink_name

dblink_authentication
CONNECT

TO

user

IDENTIFIED

BY

password
;

dblink_authentication

dblink_authentication::=
AUTHENTICATED

BY

user

IDENTIFIED

BY

password

Semantics
10

The ALTER DATABASE LINK statement is intended only to update fixed-user database
links with the current passwords of connection and authentication users. Therefore,
any clauses valid in a CREATE DATABASE LINK statement that do not appear in the syntax
diagram above are not valid in an ALTER DATABASE LINK statement. The semantics of all
of the clauses permitted in this statement are the same as the semantics for those
clauses in CREATE DATABASE LINK. Refer to CREATE DATABASE LINK on page 14-31
for this information.

10-46 Oracle Database SQL Language Reference

ALTER DATABASE LINK

Examples
10

The following statements show the valid variations of the ALTER DATABASE LINK
statement:
ALTER DATABASE LINK private_link
CONNECT TO hr IDENTIFIED BY hr_new_password;
ALTER PUBLIC DATABASE LINK public_link
CONNECT TO scott IDENTIFIED BY scott_new_password;
ALTER SHARED PUBLIC DATABASE LINK shared_pub_link
CONNECT TO scott IDENTIFIED BY scott_new_password
AUTHENTICATED BY hr IDENTIFIED BY hr_new_password;
ALTER SHARED DATABASE LINK shared_pub_link
CONNECT TO scott IDENTIFIED BY scott_new_password;

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-47

ALTER DIMENSION

ALTER DIMENSION
Purpose
10

Use the ALTER DIMENSION statement to change the hierarchical relationships or
dimension attributes of a dimension.
See Also: CREATE DIMENSION on page 14-36 and DROP
DIMENSION on page 17-41

Prerequisites
10

The dimension must be in your schema or you must have the ALTER ANY DIMENSION
system privilege to use this statement.
A dimension is always altered under the rights of the owner.

Syntax
10

alter_dimension::=
schema
ALTER

.

DIMENSION

dimension

level_clause
hierarchy_clause
ADD
attribute_clause
extended_attribute_clause
RESTRICT
CASCADE
LEVEL

level
;

HIERARCHY

hierarchy
,

DROP
COLUMN
LEVEL
ATTRIBUTE

column

level

attribute

COMPILE

(level_clause::= on page 10-48, hierarchy_clause::= on page 10-49, attribute_clause::= on
page 10-49, extended_attribute_clause::= on page 10-49)
level_clause::=
level_table
LEVEL

level

.

level_column

IS

SKIP

,
(

level_table

.

level_column

10-48 Oracle Database SQL Language Reference

)

WHEN

NULL

ALTER DIMENSION

hierarchy_clause::=
dimension_join_clause
HIERARCHY

hierarchy

(

child_level

CHILD

OF

parent_level

)

(dimension_join_clause::= on page 10-49)
dimension_join_clause::=
child_key_column
JOIN

KEY

,
(

REFERENCES

child_key_column

parent_level

)

attribute_clause::=
dependent_column
ATTRIBUTE

level

DETERMINES

,
(

dependent_column

)

extended_attribute_clause::=
dependent_column
ATTRIBUTE

attribute

LEVEL

level

DETERMINES

,
(

dependent_column

)

Semantics
10

The following keywords, parameters, and clauses have meaning unique to ALTER
DIMENSION. Keywords, parameters, and clauses that do not appear here have the same
functionality that they have in the CREATE DIMENSION statement. Refer to CREATE
DIMENSION on page 14-36 for more information.

schema
Specify the schema of the dimension you want to modify. If you do not specify schema,
then Oracle Database assumes the dimension is in your own schema.

dimension
Specify the name of the dimension. This dimension must already exist.

ADD
The ADD clauses let you add a level, hierarchy, or attribute to the dimension. Adding
one of these elements does not invalidate any existing materialized view.
Oracle Database processes ADD LEVEL clauses prior to any other ADD clauses.

DROP
The DROP clauses let you drop a level, hierarchy, or attribute from the dimension. Any
level, hierarchy, or attribute you specify must already exist.

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-49

ALTER DIMENSION

Within one attribute, you can drop one or more level-to-column relationships
associated with one level.
Restriction on DROP If any attributes or hierarchies reference a level, then you

cannot drop the level until you either drop all the referencing attributes and
hierarchies or specify CASCADE.
Specify CASCADE if you want Oracle Database to drop any attributes or
hierarchies that reference the level, along with the level itself.

CASCADE

Specify RESTRICT if you want to prevent Oracle Database from dropping a
level that is referenced by any attributes or hierarchies. This is the default.

RESTRICT

COMPILE
Specify COMPILE to explicitly recompile an invalidated dimension. Oracle Database
automatically compiles a dimension when you issue an ADD clause or DROP clause.
However, if you alter an object referenced by the dimension (for example, if you drop
and then re-create a table referenced in the dimension), Oracle Database invalidates,
and you must recompile it explicitly.

Examples
10

The following examples modify the customers_
dim dimension in the sample schema sh:

Modifying a Dimension: Examples
ALTER DIMENSION customers_dim
DROP ATTRIBUTE country;

ALTER DIMENSION customers_dim
ADD LEVEL zone IS customers.cust_postal_code
ADD ATTRIBUTE zone DETERMINES (cust_city);

10-50 Oracle Database SQL Language Reference

ALTER DISKGROUP

ALTER DISKGROUP
This SQL statement is valid only if you are using Oracle ASM
and you have started an Oracle ASM instance. You must issue this
statement from within the Oracle ASM instance, not from a normal
database instance. For information on starting an Oracle ASM
instance, refer to Oracle Automatic Storage Management Administrator's
Guide.

Note:

Purpose
10

The ALTER DISKGROUP statement lets you perform a number of operations on a disk
group or on the disks in a disk group.
See Also:
■

■

CREATE DISKGROUP on page 14-43 for information on creating
disk groups
Oracle Automatic Storage Management Administrator's Guide for
information on Oracle ASM and using disk groups to simplify
database administration

Prerequisites
10

You must have an Oracle ASM instance started from which you issue this statement.
The disk group to be modified must be mounted.
You can issue all ALTER DISKGROUP clauses if you have the SYSASM system privilege. You
can issue specific clauses as follows:
■

■

The SYSOPER privilege permits the following subset of the ALTER DISKGROUP
operations: diskgroup_availability, rebalance_diskgroup_clause, check_
diskgroup_clause (without the REPAIR option).
If you are connected as SYSDBA, you have limited privileges to use this statement.
The following operations are always granted to users connected as SYSDBA:
■

ALTER DISKGROUP ... ADD DIRECTORY

■

ALTER DISKGROUP ... ADD/ALTER/DROP TEMPLATE (nonsystem templates only)

■

ALTER DISKGROUP ... ADD USERGROUP

■

SELECT

■

SHOW PARAMETER

Table 10–1 shows additional privileges granted to users connected as SYSDBA under the
conditions shown:
Table 10–1

Conditional Diskgroup Privileges for SYSDBA

ALTER DISKGROUP
Operation

Condition

DROP FILE

User must have read-write permission on the file.

ADD ALIAS

User must have read-write permission on the related file.

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-51

ALTER DISKGROUP

Table 10–1 (Cont.) Conditional Diskgroup Privileges for SYSDBA
ALTER DISKGROUP
Operation

Condition

RENAME ALIAS

User must have read-write permission on the related file.

DROP ALIAS

User must have read-write permission on the related file.

RENAME DIRECTORY

Directory must contain only aliases and no files. User must have
DROP ALIAS permissions on all aliases under the directory.

DROP DIRECTORY

Directory must contain only aliases and no files. User must have
DROP ALIAS permissions on all aliases under the directory.

DROP USERGROUP

User must be the owner of the user group.

MODIFY FILE

User must be the owner of the file.

MODIFY USERGROUP ADD
MEMBER

User must be the owner of the user group.

MODIFY USERGROUP DROP
MEMBER

User must be the owner of the user group.

SET PERMISSION

User must be the owner of the file.

SET OWNER GROUP

User must be the owner of the file and a member of the user
group.

10-52 Oracle Database SQL Language Reference

ALTER DISKGROUP

Syntax
10

alter_diskgroup::=
add_disk_clause
drop_disk_clause

rebalance_diskgroup_clause

resize_disk_clause
disk_online_clause
disk_offline_clause
rebalance_diskgroup_clause
diskgroup_name

check_diskgroup_clause
diskgroup_template_clauses
diskgroup_directory_clauses
diskgroup_alias_clauses
diskgroup_volume_clauses
diskgroup_attributes
modify_diskgroup_file

ALTER

DISKGROUP

;
drop_diskgroup_file_clause
usergroup_clauses
user_clauses
file_permissions_clause
file_owner_clause
,
undrop_disk_clause
diskgroup_name
diskgroup_availability
ALL
enable_disable_volume

(add_disk_clause::= on page 10-54, drop_disk_clause::= on page 10-54, resize_disk_clause::=
on page 10-54, disk_online_clause::= on page 10-55, disk_offline_clause::= on page 10-55,
rebalance_diskgroup_clause::= on page 10-55, check_diskgroup_clause::= on page 10-55,
diskgroup_template_clauses::= on page 10-55, diskgroup_directory_clauses::= on page 10-56,
diskgroup_alias_clauses::= on page 10-56, diskgroup_volume_clauses::= on page 10-57,
diskgroup_attributes::= on page 10-57, modify_diskgroup_file::= on page 10-58, drop_
diskgroup_file_clause::= on page 10-58, usergroup_clauses::= on page 10-58, user_clauses::=
on page 10-58, file_permissions_clause::= on page 10-58, file_owner_clause::= on
page 10-58, undrop_disk_clause::= on page 10-58, diskgroup_availability::= on page 10-59,
enable_disable_volume::= on page 10-59.)

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-53

ALTER DISKGROUP

add_disk_clause::=
QUORUM
REGULAR

FAILGROUP

,

failgroup_name

ADD

DISK

qualified_disk_clause

(qualified_disk_clause::= on page 10-54)
qualified_disk_clause::=
FORCE
NAME

disk_name

SIZE

size_clause

NOFORCE

search_string

(size_clause::= on page 8-47)
drop_disk_clause::=
,
QUORUM

FORCE

REGULAR

NOFORCE
DISK

disk_name
,

DROP

QUORUM

FORCE

REGULAR
DISKS

NOFORCE

IN

FAILGROUP

failgroup_name

resize_disk_clause::=
SIZE

size_clause

ALL
QUORUM

,

REGULAR
RESIZE

SIZE
DISK

size_clause

disk_name

QUORUM

,

REGULAR
DISKS

IN

(size_clause::= on page 8-47)

10-54 Oracle Database SQL Language Reference

SIZE
FAILGROUP

failgroup_name

size_clause

ALTER DISKGROUP

disk_online_clause::=
QUORUM
,

REGULAR
DISK

disk_name
WAIT

QUORUM
,

REGULAR
ONLINE

DISKS

IN

FAILGROUP

NOWAIT

failgroup_name

ALL

disk_offline_clause::=
QUORUM
,

REGULAR
DISK

disk_name
timeout_clause

QUORUM

OFFLINE

,

REGULAR
DISKS

IN

FAILGROUP

failgroup_name

timeout_clause::=
M
DROP

AFTER

integer
H

rebalance_diskgroup_clause::=
WAIT
POWER

integer

NOWAIT

REBALANCE

check_diskgroup_clause::=
REPAIR
NOREPAIR
CHECK

diskgroup_template_clauses::=
,
ADD
TEMPLATE

template_name

qualified_template_clause

MODIFY
,
DROP

TEMPLATE

template_name

(qualified_template_clause::= on page 10-56)

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-55

ALTER DISKGROUP

qualified_template_clause::=
ATTRIBUTE

(

redundancy_clause

striping_clause

disk_region_clause

)

redundancy_clause::=
MIRROR
HIGH
UNPROTECTED

striping_clause::=
FINE
COARSE

disk_region_clause::=
HOT

MIRRORHOT

COLD

MIRRORCOLD

diskgroup_directory_clauses::=
,
ADD

DIRECTORY

’

filename

’
,
FORCE
NOFORCE

DROP

DIRECTORY

’

filename

’
,

RENAME

DIRECTORY

’

old_dir_name

’

TO

’

new_dir_name

’

diskgroup_alias_clauses::=
,
ADD

ALIAS

’

alias_name

’

FOR

’

filename

’

,
DROP

ALIAS

’

alias_name

’
,

RENAME

ALIAS

’

old_alias_name

’

TO

10-56 Oracle Database SQL Language Reference

’

new_alias_name

’

ALTER DISKGROUP

diskgroup_volume_clauses::=
add_volume_clause
modify_volume_clause
RESIZE
DROP

VOLUME

asm_volume

VOLUME

SIZE

size_clause

asm_volume

(add_volume_clause::= on page 10-57, modify_volume_clause::= on page 10-57
add_volume_clause::=
K
STRIPE_WIDTH

integer

redundancy_clause
ADD

VOLUME

asm_volume

STRIPE_COLUMNS

SIZE

integer

M

size_clause

ATTRIBUTE

(

disk_region_clause

)

(size_clause::= on page 10-57, redundancy_clause::= on page 10-56, disk_region_clause::=
on page 10-56)
size_clause::=
K
M
G
T
P
E
integer

modify_volume_clause::=
ATTRIBUTE
MODIFY

VOLUME

MOUNTPATH

(

disk_region_clause

)

asm_volume

’

mountpath_name

’

USAGE

’

usage_name

’

(disk_region_clause::= on page 10-56)
diskgroup_attributes::=
SET

ATTRIBUTE

’

attribute_name

’

=

’

attribute_value

’

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-57

ALTER DISKGROUP

modify_diskgroup_file::=
,
MODIFY

FILE

’

filename

’

ATTRIBUTE

(

disk_region_clause

)

drop_diskgroup_file_clause::=
,
DROP

FILE

’

filename

’

usergroup_clauses::=
,
ADD

USERGROUP

usergroup

WITH

MEMBER

user
,

ADD
MODIFY

USERGROUP

usergroup

MEMBER

user

DROP
DROP

USERGROUP

usergroup

user_clauses::=
,
ADD

USER

user
,

DROP

USER

CASCADE

user

file_permissions_clause::=
,
OWNER
SET

PERMISSION

NONE

GROUP

=

OTHER

,

READ

ONLY

READ

WRITE

FOR

FILE

’

filename

file_owner_clause::=
,
,
SET

OWNER

=

user

GROUP

=

usergroup

OWNERSHIP

FOR

undrop_disk_clause::=
UNDROP

DISKS

10-58 Oracle Database SQL Language Reference

FILE

’

filename

’

’

ALTER DISKGROUP

diskgroup_availability::=
RESTRICTED

FORCE

NORMAL

NOFORCE

MOUNT
FORCE
NOFORCE
DISMOUNT

enable_disable_volume::=
,
ENABLE

asm_volume
VOLUME

DISABLE

ALL

Semantics
10

diskgroup_name
Specify the name of the disk group you want to modify. To determine the names of
existing disk groups, query the V$ASM_DISKGROUP dynamic performance view.
add_disk_clause
Use this clause to add one or more disks to the disk group and specify attributes for
the newly added disk. Oracle ASM automatically rebalances the disk group as part of
this operation.
You cannot use this clause to change the failure group of a disk. Instead you must drop
the disk from the disk group and then add the disk back into the disk group as part of
the new failure group.
To determine the names of the disks already in this disk group, query the V$ASM_DISK
dynamic performance view.
QUORUM | REGULAR The semantics of these keyword are the same as the semantics
in a CREATE DISKGROUP statement. See QUORUM | REGULAR on page 14-45 for more
information on these keywords.

You cannot change this qualifier for an existing disk or disk group. Therefore, you
cannot specify in this clause a keyword different from the keyword that was specified
when the disk group was created.
See Also: Oracle Automatic Storage Management Administrator's Guide
for more information about the use of these keywords

Use this clause to assign the newly added disk to a failure
group. If you omit this clause and you are adding the disk to a normal or high
redundancy disk group, then Oracle Database automatically adds the newly added
disk to its own failure group. The implicit name of the failure group is the same as the
operating system independent disk name (see "NAME Clause" on page 14-46).

FAILGROUP Clause

You cannot specify this clause if you are creating an external redundancy disk group.

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-59

ALTER DISKGROUP

qualified_disk_clause
This clause has the same semantics in CREATE DISKGROUP and ALTER DISKGROUP
statements. For complete information on this clause, refer to qualified_disk_clause on
page 14-45 in the documentation on CREATE DISKGROUP.
drop_disk_clause
Use this clause to drop one or more disks from the disk group.
DROP DISK The DROP DISK clause lets you drop one or more disks from the disk
group and automatically rebalance the disk group. When you drop a disk, Oracle ASM
relocates all the data from the disk and clears the disk header so that it no longer is
part of the disk group. The disk header is not cleared if you specify the FORCE
keyword.

Specify disk_name as shown in the NAME column of the V$ASM_DISK dynamic
performance view.
If a disk to be dropped is a quorum disk or belongs to a quorum failure group, then
you must specify QUORUM in order to drop the disk. See QUORUM | REGULAR on
page 10-59.
DROP DISKS IN FAILGROUP The DROP DISKS IN FAILGROUP clause lets you drop all
the disks in the specified failure group. The behavior is otherwise the same as that for
the DROP DISK clause.

If the specified failure group is a quorum failure group, then you must specify the
QUORUM keyword in order to drop the disks. See QUORUM | REGULAR on page 10-59.
FORCE | NOFORCE These keywords let you specify when the disk is considered to
be no longer part of the disk group. The default and recommended setting is NOFORCE.
■

When you specify NOFORCE, Oracle ASM reallocates all of the extents of the disk to
other disks and then expels the disk from the disk group and rebalances the disk
group.
DROP DISK ... NOFORCE returns control to the user before the
disk can be safely reused or removed from the system. To ensure that
the drop disk operation has completed, query the V$ASM_DISK view to
verify that HEADER_STATUS has the value FORMER. Do not attempt to
remove or reuse a disk if STATE has the value DROPPING. Query the
V$ASM_OPERATION view for approximate information on how long it
will take to complete the rebalance resulting from dropping the disk.
Caution:

If you also specify REBALANCE ... WAIT (see rebalance_diskgroup_clause on
page 10-62), then the statement will not return until the rebalance
operation is complete and the disk has been cleared. However, you
should always verify that the HEADER_STATUS column of V$ASM_DISK is
FORMER, because of the unlikely event the rebalance operations fails.
■

When you specify FORCE, Oracle Database expels the disk from the disk group
immediately. It then reconstructs the data from the redundant copies on other
disks, reallocates the data to other disks, and rebalances the disk group.
The FORCE clause can be useful, for example, if Oracle ASM can no longer read the
disk to be dropped. However, it is more time consuming than a NOFORCE drop, and
it can leave portions of a file with reduced protection. You cannot specify FORCE for

10-60 Oracle Database SQL Language Reference

ALTER DISKGROUP

an external redundancy disk group at all, because in the absence of redundant
data on the disk, Oracle ASM must read the data from the disk before it can be
dropped.
The rebalance operation invoked when a disk is dropped is time consuming, whether
or not you specify FORCE or NOFORCE. You can monitor the progress by querying the
V$ASM_OPERATION dynamic performance view. Refer to rebalance_diskgroup_clause on
page 10-62 for more information on rebalance operations.
resize_disk_clause
Use this clause to specify a new size for one or more disks in the disk group. This
clause lets you override the size being returned by the operating system or the size
you specified previously for the disks. The QUORUM and REGULAR keywords have the
same semantics here as they have when adding a disk to a disk group. See QUORUM
| REGULAR on page 10-59.
RESIZE ALL

Specify this clause to perform a resize operation on every disk in the

disk group.
RESIZE DISK Specify this clause to resize only the specified disk.

Specify disk_name as shown in the NAME column of the V$ASM_DISK dynamic
performance view.
If a disk to be resized is a quorum group, then you must specify QUORUM in order to
drop the disk group. See QUORUM | REGULAR on page 10-59.
RESIZE DISKS IN FAILGROUP Specify this clause to resize every disk in the specified
failure group. If you omit both of the optional keywords REGULAR and QUORUM, then
REGULAR is the default. If the failure group you are resizing is a QUORUM failure group,
then you must specify the QUORUM keyword.
SIZE Specify the size of the disk in kilobytes, megabytes, gigabytes, or terabytes. You
cannot specify a size greater than the capacity of the disk. If you specify a size smaller
than the disk capacity, then you limit the amount of disk space Oracle ASM will use. If
you omit this clause, then Oracle ASM uses the size being returned by the operating
system.

disk_offline_clause
Use the disk_offline_clause to take one or more disks offline. This clause fails if the
redundancy level of the disk group would be violated by taking the specified disks
offline.
Specify disk_name as shown in the NAME column of the V$ASM_DISK dynamic
performance view.
The QUORUM and REGULAR keywords have the same semantics here as they have when
adding a disk to a disk group. See QUORUM | REGULAR on page 10-59.
By default, Oracle ASM drops a disk shortly after it is taken offline. You can delay this
operation by specifying the timeout_clause, which gives you the opportunity to
repair the disk and bring it back online. You can specify the timeout value in units of
minute or hour. If you omit the unit, then the default is hour.
You can change the timeout period by specifying this clause multiple times. Each time
you specify it, Oracle ASM measures the time from the most recent previous disk_
offline_clause while the disk group is mounted. To learn how much time remains

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-61

ALTER DISKGROUP

before Oracle ASM will drop an offline disk, query the repair_timer column of
V$ASM_DISK.
This clause overrides any previous setting of the disk_repair_time attribute. Refer to
Table 14–1, " Disk Group Attributes" for more information about disk group attributes.
disk_online_clause
Use the disk_online_clause to bring one or more disks online and rebalance the disk
group. Specify disk_name as shown in the NAME column of the V$ASM_DISK dynamic
performance view. The QUORUM and REGULAR keywords have the same semantics here as
they have when adding a disk to a disk group. See QUORUM | REGULAR on
page 10-59. The WAIT and NOWAIT keywords have the same semantics here as they have
for a manual rebalancing of a disk group. See WAIT | NOWAIT on page 10-62 for
more information.
See Also: Oracle Automatic Storage Management Administrator's Guide
for more information about taking Oracle ASM disks online and
offline

rebalance_diskgroup_clause
Use this clause to manually rebalance the disk group. Oracle ASM redistributes data
files evenly across all drives. This clause is rarely necessary, because Oracle ASM
allocates files evenly and automatically rebalances disk groups when the storage
configuration changes. However, it is useful if you want to use the POWER clause to
control the speed of what would otherwise be an automatic rebalance operation.
In the POWER clause, specify a value from 0 to 11, where 0 stops the rebalance
operation and 11 permits Oracle ASM to execute the rebalance as fast as possible. The
value you specify in the POWER clause defaults to the value of the ASM_POWER_LIMIT
initialization parameter.
POWER

If you omit the POWER clause, then Oracle ASM executes both automatic and specified
rebalance operations at the power determined by the value of the ASM_POWER_LIMIT
initialization parameter.

Beginning with Oracle Database 11g Release 2 (11.2.0.2), if the
COMPATIBLE.ASM disk group attribute is set to 11.2.0.2 or higher, then
you can specify a value from 0 to 1024 in the POWER clause.
Note:

Use this clause to specify when in the course of the rebalance
operation control should be returned to the user.

WAIT | NOWAIT
■

■

Specify WAIT to allow a script that adds or removes disks to wait for the disk group
to be rebalanced before returning control to the user. You can explicitly terminate a
rebalance operation running in WAIT mode, although doing so does not undo any
completed disk add or drop operation in the same statement.
Specify NOWAIT if you want control returned to the user immediately after the
statement is issued. This is the default.

You can monitor the progress of the rebalance operation by querying the V$ASM_
OPERATION dynamic performance view.

10-62 Oracle Database SQL Language Reference

ALTER DISKGROUP

ASM_POWER_LIMIT and Oracle Automatic Storage
Management Administrator's Guide for more information on rebalancing
disk groups and "Rebalancing a Disk Group: Example" on page 10-72

See Also:

check_diskgroup_clause
The check_diskgroup_clause lets you verify the internal consistency of Oracle ASM
disk group metadata. The disk group must be mounted. Oracle ASM displays
summary errors and writes the details of the detected errors in the alert log.
The CHECK keyword performs the following operations:
■

Checks the consistency of the disk.

■

Cross checks all the file extent maps and allocation tables for consistently.

■

Checks that the alias metadata directory and file directory are linked correctly.

■

Checks that the alias directory tree is linked correctly.

■

Checks that Oracle ASM metadata directories do not have unreachable allocated
blocks.

This clause lets you instruct Oracle ASM whether or not to
attempt to repair any errors found during the consistency check. The default is
NOREPAIR. The NOREPAIR setting is useful if you want to be alerted to any
inconsistencies but do not want Oracle ASM to take any automatic action to resolve
them.
REPAIR | NOREPAIR

In earlier releases, you could specify CHECK for ALL, DISK, DISKS
IN FAILGROUP, or FILE. Those clauses have been deprecated as they are no longer
needed. If you specify them, then their behavior is the same as in earlier releases and a
message is added to the alert log. However, Oracle recommends that you do not
introduce these clauses into your new code, as they are scheduled for desupport. The
deprecated clauses are these:
Deprecated Clauses

■

ALL checks all disks and files in the disk group.

■

DISK checks one or more specified disks in the disk group.

■

DISKS IN FAILGROUP checks all disks in a specified failure group.

■

FILE checks one or more specified files in the disk group. You must use one of the
reference forms of the filename. Refer to ASM_filename on page 8-32 for
information on the reference forms of Oracle ASM filenames.

diskgroup_template_clauses
A template is a named collection of attributes. When you create a disk group, Oracle
ASM associates a set of initial system default templates with that disk group. The
attributes defined by the template are applied to all files in the disk group. Table 10–2
lists the system default templates and the attributes they apply to the various file
types. The diskgroup_template_clauses described following the table let you change
the template attributes and create new templates.
You cannot use this clause to change the attributes of a disk group file after it has been
created. Instead, you must use Recovery Manager (RMAN) to copy the file into a new
file with the new attributes.

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-63

ALTER DISKGROUP

Table 10–2

Oracle Automatic Storage Management System Default File Group Templates
Mirroring
Level in
Normal
Redundancy
Disk Groups

Mirroring
Level in
High
Redundancy
Disk Groups Striped

Template Name

File Type

Mirroring
Level in
External
Redundancy
Disk Groups

CONTROLFILE

Control files

Unprotected

3-way mirror 3-way mirror FINE

COLD

DATAFILE

Data Files and
copies

Unprotected

2-way mirror 3-way mirror COARSE

COLD

ONLINELOG

Online logs

Unprotected

2-way mirror 3-way mirror COARSE

COLD

ARCHIVELOG

Archive logs

Unprotected

2-way mirror 3-way mirror COARSE

COLD

TEMPFILE

Temp files

Unprotected

2-way mirror 3-way mirror COARSE

COLD

BACKUPSET

Data File backup
pieces, data file
incremental backup
pieces, and archive
log backup pieces

Unprotected

2-way mirror 3-way mirror COARSE

COLD

PARAMETERFILE

SPFILE

Unprotected

2-way mirror 3-way mirror COARSE

COLD

DATAGUARDCONFIG

Disaster recovery
Unprotected
configurations (used
in standby
databases)

2-way mirror 3-way mirror COARSE

COLD

FLASHBACK

Flashback logs

Unprotected

2-way mirror 3-way mirror COARSE

COLD

CHANGETRACKING

Block change
tracking data (used
during incremental
backups)

Unprotected

2-way mirror 3-way mirror COARSE

COLD

DUMPSET

Data Pump dumpset Unprotected

2-way mirror 3-way mirror COARSE

COLD

XTRANSPORT

Cross-platform
converted data file

Unprotected

2-way mirror 3-way mirror COARSE

COLD

AUTOBACKUP

Automatic backup
files

Unprotected

2-way mirror 3-way mirror COARSE

COLD

ASMPARAMETERFILE

SPFILE

Unprotected

2-way mirror 3-way mirror COARSE

COLD

OCRFILE

Oracle Cluster
Registry file

Unprotected

2-way mirror 3-way mirror COARSE

COLD

Region

Use this clause to add one or more named templates to a disk
group. To determine the names of existing templates, query the V$ASM_TEMPLATE
dynamic performance view.

ADD TEMPLATE

MODIFY TEMPLATE Use this clause to modify the attributes of a system default or
user-defined disk group template. Only the specified attributes are altered.
Unspecified properties retain their current values.
Note: In earlier releases, the keywords ALTER TEMPLATE were used
instead of MODIFY TEMPLATE. The ALTER keyword is still supported for
backward compatibility, but it replaced with MODIFY for consistency
with other Oracle SQL.

10-64 Oracle Database SQL Language Reference

ALTER DISKGROUP

Specify the name of the template to be added or modified. Template
names are subject to the same naming conventions and restrictions as database schema
objects. Refer to "Database Object Naming Rules" on page 3-111 for information on
database object names.

template_name

redundancy_clause

Specify the redundancy level of the newly added or modified

template:
■

■

■

MIRROR: Files to which this template are applied are protected by mirroring their
data blocks. In normal redundancy disk groups, each primary extent has one
mirror extent (2-way mirroring). For high redundancy disk groups, each primary
extent has two mirror extents (3-way mirroring). You cannot specify MIRROR for
templates in external redundancy disk groups.
HIGH: Files to which this template are applied are protected by mirroring their data
blocks. Each primary extent has two mirror extents (3-way mirroring) for both
normal redundancy and high redundancy disk groups. You cannot specify HIGH
for templates in external redundancy disk groups.
UNPROTECTED: Files to which this template are applied are not protected by
Automated Storage Management from media failures. Disks taken offline, either
through system action or by user command, can cause loss of unprotected files.
UNPROTECTED is the only valid setting for external redundancy disk groups.
UNPROTECTED may not be specified for templates in high redundancy disk groups.
Oracle discourages the use of unprotected files in high and normal redundancy
disk groups.

If you omit this clause, then the value defaults to MIRROR for a normal redundancy disk
group, HIGH for a high redundancy disk group, and UNPROTECTED for an external
redundancy disk group.
striping_clause

Specify how the files to which this template are applied will be

striped:
■

■

FINE: Files to which this template are applied are striped every 128KB. This
striping mode is not valid for an Oracle ASM spfile.
COARSE: Files to which this template are applied are striped every 1MB. This is the
default value.

disk_region_clause This clause lets you determine the Intelligent Data Placement
attribute of the disk group file. Specify the region of the disk in which you want Oracle
ASM to allocate extents for the file:
■

■
■

HOT: Extents are allocated in the region of the disk furthest away from the spindle.
These outer tracks on the disk are longer than inner tracks, and so have more
sectors and increased throughput.
COLD: Extents are allocated in the region of the disk closest to the spindle.
MIRRORHOT and MIRRORCOLD: Specify the region desired for the mirrored datablocks
of the file.

If no space is available in the desired disk region, then Oracle ASM allocates extents in
the other region but initiates a rebalance to adjust the size of the region.
See Also: Oracle Automatic Storage Management Administrator's Guide
for more information on Intelligent Data Placement

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-65

ALTER DISKGROUP

DROP TEMPLATE Use this clause to drop one or more templates from the disk
group. You can use this clause to drop only user-defined templates, not system default
templates.

diskgroup_directory_clauses
Before you can create alias names for Oracle ASM filenames (see diskgroup_alias_clauses
on page 10-66), you must specify the full directory structure in which the alias name
will reside. The diskgroup_directory_clauses let you create and manipulate such a
directory structure.
Use this clause to create a new directory path for hierarchically
named aliases. Use a slash (/) to separate components of the directory. Each directory
component can be up to 48 bytes in length and must not contain the slash character.
You cannot use a space for the first or last character of any component. The total length
of the directory path cannot exceed 256 bytes minus the length of any alias name you
intend to create in this directory (see diskgroup_alias_clauses on page 10-66).

ADD DIRECTORY

Use this clause to drop a directory for hierarchically named
aliases. Oracle ASM will not drop the directory if it contains any alias definitions
unless you also specify FORCE. This clause is not valid for dropping directories created
as part of a system alias. Such directories are labeled with the value Y in the SYSTEM_
CREATED column of the V$ASM_ALIAS dynamic performance view.
DROP DIRECTORY

RENAME DIRECTORY Use this clause to change the name of a directory for
hierarchically named aliases. This clause is not valid for renaming directories created
as part of a system alias. Such directories are labeled with the value Y in the SYSTEM_
CREATED column of the V$ASM_ALIAS dynamic performance view.

diskgroup_alias_clauses
When an Oracle ASM file is created, either implicitly or by user specification, Oracle
ASM assigns to the file a fully qualified name ending in a dotted pair of numbers (see
file_specification on page 8-29). The diskgroup_alias_clauses let you create more
user-friendly alias names for the Oracle ASM filenames. You cannot specify an alias
name that ends in a dotted pair of numbers, as this format is indistinguishable from an
Oracle ASM filename.
Before specifying this clause, you must first create the directory structure appropriate
for your naming conventions (see diskgroup_directory_clauses on page 10-66). The total
length of the alias name, including the directory prefix, is limited to 256 bytes. Alias
names are case insensitive but case retentive.
ADD ALIAS Use this clause to create an alias name for an Oracle ASM filename. The

alias_name consists of the full directory path and the alias itself. To determine the
names of existing Oracle ASM aliases, query the V$ASM_ALIAS dynamic performance
view. Refer to ASM_filename on page 8-32 for information on Oracle ASM filenames.
DROP ALIAS Use this clause to remove an alias name from the disk group directory.

Each alias name consists of the full directory path and the alias itself. The underlying
file to which the alias refers remains unchanged.
Use this clause to change the name of an existing alias. The alias_
name consists of the full directory path and the alias itself.

RENAME ALIAS

10-66 Oracle Database SQL Language Reference

ALTER DISKGROUP

Restriction on Dropping and Renaming Aliases You cannot drop or rename a
system-generated alias. To determine whether an alias was system generated, query
the SYSTEM_CREATED column of the V$ASM_ALIAS dynamic performance view.

diskgroup_volume_clauses
Use these clauses to manipulate logical Oracle ASM Dynamic Volume Manager
(Oracle ADVM) volumes corresponding to physical volume devices. To use these
clauses, Oracle ASM must be started and the disk group being modified must be
mounted.
See Also: Oracle Automatic Storage Management Administrator's Guide
for more information about disk group volumes, including examples
add_volume_clause

Use this clause to add a volume to the disk group.

For asm_volume, specify the name of the volume. The name can contain only
alphanumeric characters and the first character must be alphabetic. The maximum
length of the name is platform dependant. Refer to Oracle Automatic Storage
Management Administrator's Guide for more information.
For size_clause, specify the size of the Oracle ADVM volume. The Oracle ASM
instance determines whether sufficient space exists to create the volume. If sufficient
space does not exist, then the Oracle ASM instance returns an error. If sufficient space
does exist, then all nodes in the cluster with an Oracle ASM instance running and the
disk group mounted are notified of the addition. Oracle ASM creates and enables on
those nodes a volume device that can be used to create and mount file systems.
The following optional settings are also available:
■

In the redundancy_clause, specify the redundancy level of the Oracle ADVM
volume. You can specify this clause only when creating a volume in a normal
redundancy disk group. You can specify the following volume redundancy levels:
–

MIRROR: 2-way mirroring of the volume. This is the default.

–

HIGH: 3-way mirroring of the volume.

–

UNPROTECTED: No mirroring of the volume.

You cannot specify the redundancy_clause when creating a volume in a high
redundancy disk group or an external redundancy disk group. If you do so, then
an error will result. In high redundancy disk groups, Oracle Database
automatically sets the volume redundancy to HIGH (3-way mirroring). In external
redundancy disk groups, Oracle Database automatically sets the volume
redundancy to UNPROTECTED (no mirroring).
■

■

■

In the STRIPE_WIDTH clause, specify a stripe width for the Oracle ADVM Volume.
The valid range is from 4KB to 1MB, at intervals of the power of 2. The default
value is 128K.
In the STRIPE_COLUMNS clause, specify the number of stripes in a stripe set of the
Oracle ADVM volume. The valid range is 1 to 8. The default is 4. If STRIPE_
COLUMNS is set to 1, then striping becomes disabled. In this case, the stripe width is
the extent size of the volume. This volume extent size is 64 times the allocation
unit (AU) size of the disk group
In the disk_region_clause clause, specify the Intelligent Data Placement attribute
of both the primary and nonprimary mirror of the disk group volume. The default
for both is COLD. See disk_region_clause on page 10-65 for details on this clause.

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-67

ALTER DISKGROUP

Use this clause to modify the characteristics of an existing
Oracle ADVM volume. You must specify at least one of the following clauses:

modify_volume_clause
■

■

■

In the disk_region_clause clause, specify the Intelligent Data Placement attribute
of both the primary and nonprimary mirror of the disk group volume. The default
for the primary mirror is COLD. The default for mirror and high redundancy is HOT.
See disk_region_clause on page 10-65 for details on this clause.
In the MOUNTPATH clause, specify the mountpath name associated with the volume.
The mountpath_name can be up to 1024 characters.
In the USAGE clause, specify the usage name associated with the volume. The
usage_name can be up to 30 characters.

RESIZE VOLUME Clause Use this clause to change the size of an existing Oracle
ADVM volume. In an Oracle ASM cluster, the new size is propagated to all nodes. If
an Oracle Automatic Storage Management File System (ACFS) exists on the volume,
then you must use the acfsutil size command instead of the ALTER DISKGROUP
statement.
See Also: Oracle Automatic Storage Management Administrator's Guide
for more information about the acfsutil size command

Use this clause to remove the Oracle ASM file that is the
storage container for an existing Oracle ADVM volume. In an Oracle ASM cluster, all
nodes with an Oracle ASM instance running and with this disk group open are
notified of the drop operation, which results in removal of the volume device. If the
volume file is open, then this clause returns an error.

DROP VOLUME Clause

diskgroup_attributes
Use this clause to specify attributes for the disk group. Table 14–1, " Disk Group
Attributes" on page 14-47 lists the attributes you can set with this clause. Refer to the
CREATE DISKGROUP "ATTRIBUTE Clause" on page 14-46 for information on the behavior
of this clause.
modify_diskgroup_file
Use this clause to modify the Intelligent Data Placement attributes of an existing disk
group file. When you modify the Intelligent Data Placement for a file, this action will
apply to new extensions of the file, but existing file contents are not affected until a
rebalance operation. To apply the new Intelligent Data Placement policy for existing
file contents, you can manually initiate a rebalance. A rebalance operation uses the last
specified policy for the file extents.
See Also:
■

■

disk_region_clause on page 10-65 for more information on this
clause
Oracle Automatic Storage Management Administrator's Guide for
more information on manually rebalancing disk groups

drop_diskgroup_file_clause
Use this clause to drop a file from the disk group. Oracle ASM also drops all aliases
associated with the file being dropped. You must use one of the reference forms of the
filename. Most Oracle ASM files do not need to be manually deleted because, as
Oracle Managed Files, they are removed automatically when they are no longer

10-68 Oracle Database SQL Language Reference

ALTER DISKGROUP

needed. Refer to ASM_filename on page 8-32 for information on the reference forms of
Oracle ASM filenames.
You cannot drop a disk group file it if is the spfile that was used to start up the current
instance or any instance in the Oracle ASM cluster.
usergroup_clauses
Use these clauses to add a user group to the disk group, remove a user group from the
disk group, or to add a member to or drop a member from an existing user group.
See Also: Oracle Automatic Storage Management Administrator's Guide
for detailed information about user groups and members, including
examples
ADD USERGROUP Use this clause to add a user group to the disk group. You must
have SYSASM or SYSDBA privilege to create a user group. The user group name can be
up to 30 characters long. If you specify the user name, then it must be in the OS
password file.

Use these clauses to add a member to or drop a member
from an existing user group. You must be an Oracle ASM administrator (with SYSASM
privilege) or the creator (with SYSDBA privilege) of the user group to use these
clauses. The user name must be an existing user in the OS password file.

MODIFY USERGROUP

Use this clause to drop an existing user group from the disk
group. You must be an Oracle ASM administrator (with SYSASM privilege) or the
creator (with SYSDBA privilege) of the user group to use this clause. Dropping a user
group may leave a disk group file without a valid user group. In this case, you can
update the disk group file manually to add a new, valid group using the file_
permissions_clause on page 10-70.
DROP USERGROUP

user_clauses
Use these clauses to add a user to or drop a user from a disk group.
ADD USER Use this clause to add one or more operating system (OS) users to an
Oracle ASM disk group and give those users access privileges on the disk group. The
user names must be existing OS users, and their user names must have a
corresponding OS user ID or system ID. If a specified user already exists in the disk
group, as shown by V$ASM_USER, then the command records an error and continues to
add other users, if any have been specified. This command is seldom needed, because
the OS user running the database instance is added to a disk group automatically
when the instance accesses the disk group. However, this clause is useful when adding
users that are not associated with a particular database instance.

Use this clause to drop one or more users from the disk group. If a
specified user is not in the disk group, then this clause records an error and continues
to drop other users, if any are specified. If the user owns any files, then you must
specify the CASCADE keyword, which drops the user and all the user's files. If any files
owned by the user are open, then DROP USER CASCADE fails with an error.

DROP USER

To delete a user without deleting the files owned by that user, change the owner of
each of these files to another user and then issue an ALTER DISKGROUP ... DROP USER
statement on the user.

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-69

ALTER DISKGROUP

file_permissions_clause
Use this clause to change the permission settings of a disk group file. The three classes
of permissions are owner, user group, and other. You must be the file owner or the
Oracle ASM administrator to use this clause. You cannot use this clause on an open
file.
file_owner_clause
Use this clause to set the owner or user group for a specified file. You must be the
Oracle ASM administrator to change the owner of the file. You must be the owner of
the file or the Oracle ASM administrator to change the user group of a file. In addition,
to change the associated user group of a file, the specified user group must already
exist in the disk group, and the owner of the file must be a member of that user group.
You cannot use this clause on an open file.
undrop_disk_clause
Use this clause to cancel the drop of disks from the disk group. You can cancel the
pending drop of all the disks in one or more disk groups (by specifying diskgroup_
name) or of all the disks in all disk groups (by specifying ALL).
This clause is not relevant for disks that have already been completely dropped from
the disk group or for disk groups that have been completely dropped. This clause
results in a long-running operation. You can see the status of the operation by
querying the V$ASM_OPERATION dynamic performance view.
V$ASM_OPERATION for more information on the details of
long-running Oracle ASM operations

See Also:

diskgroup_availability
Use this clause to make one or more disk groups available or unavailable to the
database instances running on the same node as the Oracle ASM instance. This clause
does not affect the status of the disk group on other nodes in a cluster.
MOUNT Specify MOUNT to mount the disk groups in the local Oracle ASM instance.
Specify ALL MOUNT to mount all disk groups specified in the ASM_DISKGROUPS
initialization parameter. File operations can only be performed when a disk group is
mounted. If Oracle ASM is running in a cluster or a standalone server managed by
Oracle Restart, then the MOUNT clause automatically brings the corresponding resource
online.
RESTRICTED | NORMAL Use these clauses to determine the manner in which the
disk groups are mounted.
■

■

In the RESTRICTED mode, the disk group is mounted in single-instance exclusive
mode. No other Oracle ASM instance in the same cluster can mount that disk
group. In this mode the disk group is not usable by any Oracle ASM client.
In the NORMAL mode, the disk group is mounted in shared mode, so that other
Oracle ASM instances and clients can access the disk group. This is the default.

FORCE | NOFORCE Use these clauses to determine the circumstances under which
the disk groups are mounted.
■

In the FORCE mode, Oracle ASM attempts to mount the disk group even if it cannot
discover all of the devices that belong to the disk group. This setting is useful if
some of the disks in a normal or high redundancy disk group became unavailable

10-70 Oracle Database SQL Language Reference

ALTER DISKGROUP

while the disk group was dismounted. When MOUNT FORCE succeeds, Oracle ASM
takes the missing disks offline.
If Oracle ASM discovers all of the disks in the disk group, then MOUNT FORCE fails.
Therefore, use the MOUNT FORCE setting only if some disks are unavailable.
Otherwise, use NOFORCE.
In normal- and high-redundancy disk groups, disks from one failure group can be
unavailable and MOUNT FORCE will succeed. Also in high-redundancy disk groups,
two disks in two different failure groups can be unavailable and MOUNT FORCE will
succeed. Any other combination of unavailable disks causes the operation to fail,
because Oracle ASM cannot guarantee that a valid copy of all user data or
metadata exists on the available disks.
■

In the NOFORCE mode, Oracle ASM does not attempt to mount the disk group
unless it can discover all the member disks. This is the default.
See Also: ASM_DISKGROUPS for more information about adding disk
group names to the initialization parameter file

DISMOUNT Specify DISMOUNT to dismount the specified disk groups. Oracle ASM
returns an error if any file in the disk group is open unless you also specify FORCE.
Specify ALL DISMOUNT to dismount all currently mounted disk groups. File operations
can only be performed when a disk group is mounted. If Oracle ASM is running in a
cluster or a standalone server managed by Oracle Restart, then the DISMOUNT clause
automatically takes the corresponding resource offline.
FORCE Specify FORCE if you want Oracle ASM to dismount the disk groups even if
some files in the disk group are open.

enable_disable_volume
Use this clause to enable or disable one or more volumes in the disk group.
■

■

For each volume you enable, Oracle ASM creates a volume device file on the local
node that can be used to create or mount the file system.
For each volume you disable, Oracle ASM deletes the device file on the local node.
If the volume file is open on the local node, then the DISABLE clause returns an
error.

Use the ALL keyword to enable or disable all volumes in the disk group. If you specify
ALTER DISKGROUP ALL ..., then you must use the ALL keyword in this clause as well.
See Also: Oracle Automatic Storage Management Administrator's Guide
for more information about disk group volumes

Examples
10

The following examples require a disk group called dgroup_01. They assume that ASM_
DISKSTRING is set to /devices/disks/*. In addition, they assume the Oracle user has
read/write permission to /devices/disks/d100. Refer to "Creating a Diskgroup:
Example" on page 14-50 to create dgroup_01.
Adding a Disk to a Disk Group: Example To add a disk, d100, to a disk group,

dgroup_01, issue the following statement:
ALTER DISKGROUP dgroup_01
ADD DISK '/devices/disks/d100';

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-71

ALTER DISKGROUP

To drop a disk, dgroup_01_0000, from
a disk group, dgroup_01, issue the following statement:

Dropping a Disk from a Disk Group: Example
ALTER DISKGROUP dgroup_01
DROP DISK dgroup_01_0000;

Undropping a Disk from a Disk Group: Example To cancel the drop of disks from a
disk group, dgroup_01, issue the following statement:
ALTER DISKGROUP dgroup_01
UNDROP DISKS;

Resizing a Disk Group: Example To resize every disk in a disk group, dgroup_01,

issue the following statement:
ALTER DISKGROUP dgroup_01
RESIZE ALL
SIZE 36G;

To manually rebalance a disk group, dgroup_
01, and permit Oracle ASM to execute the rebalance as fast as possible, issue the
following statement:

Rebalancing a Disk Group: Example

ALTER DISKGROUP dgroup_01
REBALANCE POWER 11 WAIT;

The WAIT keyword causes the database to wait for the disk group to be rebalanced
before returning control to the user.
Verifying the Internal Consistency of Disk Group Metadata: Example To verify the
internal consistency of Oracle ASM disk group metadata and instruct Oracle ASM to
repair any errors found, issue the following statement:
ALTER DISKGROUP dgroup_01
CHECK ALL
REPAIR;

To add a named template,
template_01 to a disk group, dgroup_01, issue the following statement:
Adding a Named Template to a Disk Group: Example
ALTER DISKGROUP dgroup_01
ADD TEMPLATE template_01
ATTRIBUTES (UNPROTECTED COARSE);

Changing the Attributes of a Disk Group Template: Example To modify the
attributes of a system default or user-defined disk group template, template_01, issue
the following statement:
ALTER DISKGROUP dgroup_01
MODIFY TEMPLATE template_01
ATTRIBUTES (FINE);

Dropping a User-Defined Template from a Disk Group: Example To drop a
user-defined template, template_01, from a disk group, dgroup_01, issue the following
statement:
ALTER DISKGROUP dgroup_01
DROP TEMPLATE template_01;

To specify the
directory structure in which alias names will reside, issue the following statement:

Creating a Directory Path for Hierarchically Named Aliases: Example

10-72 Oracle Database SQL Language Reference

ALTER DISKGROUP

ALTER DISKGROUP dgroup_01
ADD DIRECTORY '+dgroup_01/alias_dir';

To create a user alias
by specifying the numeric Oracle ASM filename, issue the following statement:

Creating an Alias Name for an Oracle ASM Filename: Example
ALTER DISKGROUP dgroup_01
ADD ALIAS '+dgroup_01/alias_dir/datafile.dbf'
FOR '+dgroup_01.261.1';

To dismount a disk group, dgroup_01, issue the
following statement. This statement dismounts the disk group even if one or more files
are active:

Dismounting a Disk Group: Example

ALTER DISKGROUP dgroup_01
DISMOUNT FORCE;

Mounting a Disk Group: Example

To mount a disk group, dgroup_01, issue the

following statement:
ALTER DISKGROUP dgroup_01
MOUNT;

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-73

ALTER FLASHBACK ARCHIVE

ALTER FLASHBACK ARCHIVE
Purpose
10

Use the ALTER FLASHBACK ARCHIVE statement for these operations:
Designate a flashback data archive as the default flashback data archive for the
system

■

■

Add a tablespace for use by the flashback data archive

■

Change the quota of a tablespace used by the flashback data archive

■

Remove a tablespace from use by the flashback data archive

■

Change the retention period of the flashback data archive

■

Purge the flashback data archive of old data that is no longer needed
See Also: Oracle Database Advanced Application Developer's Guide and
CREATE FLASHBACK ARCHIVE on page 14-55 for more information
on using flashback data archives

Prerequisites
10

You must have the FLASHBACK ARCHIVE ADMINISTER system privilege to alter a
flashback data archive in any way. You must also have appropriate privileges on the
affected tablespaces to add, modify, or remove a flashback data archive tablespace.

Syntax
10

alter_flashback_archive::=
ALTER

FLASHBACK

SET

ARCHIVE

flashback_archive

DEFAULT
flashback_archive_quota

ADD
TABLESPACE

tablespace

MODIFY
REMOVE
MODIFY

TABLESPACE
RETENTION

tablespace_name
flashback_archive_retention

ALL
PURGE

SCN

expr

BEFORE
TIMESTAMP

expr

NO
OPTIMIZE

DATA

10-74 Oracle Database SQL Language Reference

;

ALTER FLASHBACK ARCHIVE

flashback_archive_quota::=
M
G
QUOTA

integer

T
P
E

flashback_archive_retention::=
YEAR
RETENTION

integer

MONTH
DAY

Semantics
10

flashback_archive
Specify the name of an existing flashback data archive.

SET DEFAULT
You must be logged in as SYSDBA to specify this clause. Use this clause to designate this
flashback data archive as the default flashback data archive for the system. When a
CREATE TABLE or ALTER TABLE statement specifies the flashback_archive_clause
without specifying a flashback data archive name, the database uses the default
flashback data archive to store data from that table.
This statement overrides any previous designation of a different flashback data archive
as the default.
See Also: The CREATE TABLE flashback_archive_clause on page 16-66
for more information

ADD TABLESPACE
Use this clause to add a tablespace to the flashback data archive. You can use the
flashback_archive_quota clause to specify the amount of space that can be used by
the flashback data archive in the new tablespace. If you omit that clause, then the
flashback data archive has unlimited space in the newly added tablespace.

MODIFY TABLESPACE
Use this clause to change the tablespace quota of a tablespace already used by the
flashback data archive.

REMOVE TABLESPACE
Use this clause to remove a tablespace from use by the flashback data archive. You
cannot remove the last remaining tablespace used by the flashback data archive.
If the tablespace to be removed contains any data within the retention period of the
flashback archive, then that data will be dropped as well. Therefore, you should move
your data to another tablespace before removing the tablespace with this clause.

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-75

ALTER FLASHBACK ARCHIVE

MODIFY RETENTION
Use this clause to change the retention period of the flashback data archive.

PURGE
Use this clause to purge data from the flashback data archive.
■

■

■

Specify PURGE ALL to remove all data from the flashback data archive. This
historical information can be retrieved using a flashback query only if the SCN or
timestamp specified in the flashback query is within the undo retention duration.
Specify PURGE BEFORE SCN to remove all data from the flashback data archive before
the specified system change number.
Specify PURGE BEFORE TIMESTAMP to remove all data from the flashback data
archive before the specified timestamp.

[NO] OPTIMIZE DATA
This clause has the same semantics as the [NO] OPTIMIZE DATA clause of CREATE
FLASHBACK ARCHIVE on page 14-57.
See Also: CREATE FLASHBACK ARCHIVE on page 14-55 for
information on creating flashback data archives and for some simple
examples of using flashback data archives

10-76 Oracle Database SQL Language Reference

ALTER FUNCTION

ALTER FUNCTION
Purpose
10

Functions are defined using PL/SQL. Therefore, this section provides some general
information but refers to Oracle Database PL/SQL Language Reference for complete
syntax, semantics, and examples.
Use the ALTER FUNCTION statement to recompile an invalid standalone stored function.
Explicit recompilation eliminates the need for implicit run-time recompilation and
prevents associated run-time compilation errors and performance overhead.
This statement does not change the declaration or definition of an existing function. To
redeclare or redefine a function, use the CREATE FUNCTION statement with the OR
REPLACE clause. See CREATE FUNCTION on page 14-58.

Prerequisites
10

The function must be in your own schema or you must have ALTER ANY PROCEDURE
system privilege.

Syntax
10

alter_function::=
schema
ALTER

FUNCTION

.
function

function_compile_clause

(function_compile_clause: See Oracle Database PL/SQL Language Reference for the
syntax of this clause.)

Semantics
10

schema
Specify the schema containing the function. If you omit schema, then Oracle Database
assumes the function is in your own schema.

function
Specify the name of the function to be recompiled.

function_compile_clause
See Oracle Database PL/SQL Language Reference for the syntax and semantics of this
clause and for complete information on creating and compiling functions.

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-77

ALTER INDEX

ALTER INDEX
Purpose
10

Use the ALTER INDEX statement to change or rebuild an existing index.
See Also: CREATE INDEX on page 14-60 for information on
creating an index

Prerequisites
10

The index must be in your own schema or you must have the ALTER ANY INDEX system
privilege.
To rebuild an online index in another user's schema, you must have the CREATE ANY
INDEX and CREATE ANY TABLE system privileges.
To execute the MONITORING USAGE clause, the index must be in your own schema.
To modify a domain index, you must have EXECUTE object privilege on the indextype
of the index.
Object privileges are granted on the parent index, not on individual index partitions or
subpartitions.
You must have tablespace quota to modify, rebuild, or split an index partition or to
modify or rebuild an index subpartition.
See Also: CREATE INDEX on page 14-60 and Oracle Database Data
Cartridge Developer's Guide for information on domain indexes

10-78 Oracle Database SQL Language Reference

ALTER INDEX

Syntax
10

alter_index::=
deallocate_unused_clause
allocate_extent_clause
shrink_clause
parallel_clause
physical_attributes_clause
logging_clause
rebuild_clause
PARAMETERS

(

’

ODCI_parameters

’

)

COMPILE
ENABLE
schema
ALTER

DISABLE

.

INDEX

index

;

UNUSABLE
VISIBLE
INVISIBLE
RENAME

TO

new_name

COALESCE
MONITORING
USAGE
NOMONITORING
UPDATE

BLOCK

REFERENCES

alter_index_partitioning

(deallocate_unused_clause::= on page 10-79, allocate_extent_clause::= on page 10-80,
shrink_clause::= on page 10-80, parallel_clause::= on page 10-80, physical_attributes_
clause::= on page 10-80, logging_clause::= on page 8-38, rebuild_clause::= on page 10-81,
alter_index_partitioning::= on page 10-81)
(The ODCI_parameters are documented in Oracle Database Data Cartridge Developer's
Guide.)
deallocate_unused_clause::=
KEEP
DEALLOCATE

size_clause

UNUSED

(size_clause::= on page 8-47)

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-79

ALTER INDEX

allocate_extent_clause::=
SIZE
(

size_clause

DATAFILE
INSTANCE

ALLOCATE

’

filename
integer

EXTENT

(size_clause::= on page 8-47)
shrink_clause::=
COMPACT
SHRINK

CASCADE

SPACE

parallel_clause::=
NOPARALLEL
integer
PARALLEL

physical_attributes_clause::=
PCTFREE

integer

PCTUSED

integer

INITRANS

integer

storage_clause

(storage_clause::= on page 8-50)
logging_clause::=
NOLOGGING
FILESYSTEM_LIKE_

10-80 Oracle Database SQL Language Reference

’

)

ALTER INDEX

rebuild_clause::=
PARTITION
SUBPARTITION

partition
subpartition

REVERSE
NOREVERSE
REBUILD

parallel_clause
TABLESPACE

tablespace

PARAMETERS

(

’

ODCI_parameters

’

)

XMLIndex_parameters_clause
ONLINE
physical_attributes_clause
key_compression
logging_clause

(parallel_clause::= on page 10-80, physical_attributes_clause::= on page 10-80, key_
compression::= on page 10-81, logging_clause::= on page 8-38)
(The ODCI_parameters are documented in Oracle Database Data Cartridge Developer's
Guide. The XMLIndex_parameters_clause is documented in Oracle XML DB
Developer's Guide.)
key_compression::=
integer
COMPRESS
NOCOMPRESS

alter_index_partitioning::=
modify_index_default_attrs
add_hash_index_partition
modify_index_partition
rename_index_partition
drop_index_partition
split_index_partition
coalesce_index_partition
modify_index_subpartition

(modify_index_default_attrs::= on page 10-82, add_hash_index_partition::= on page 10-82,
modify_index_partition::= on page 10-82, rename_index_partition::= on page 10-82, drop_
index_partition::= on page 10-83, split_index_partition::= on page 10-83, coalesce_index_
SQL Statements: ALTER CLUSTER to ALTER JAVA

10-81

ALTER INDEX

partition::= on page 10-82, modify_index_subpartition::= on page 10-83)
modify_index_default_attrs::=
physical_attributes_clause
FOR
MODIFY

DEFAULT

PARTITION

partition

tablespace

ATTRIBUTES

TABLESPACE
DEFAULT
logging_clause

(physical_attributes_clause::= on page 10-80, logging_clause::= on page 8-38)
add_hash_index_partition::=
partition
ADD

TABLESPACE

tablespace

key_compression

parallel_clause

PARTITION

(parallel_clause::= on page 10-80)
coalesce_index_partition::=
parallel_clause
COALESCE

PARTITION

(parallel_clause::= on page 10-80)
modify_index_partition::=
deallocate_unused_clause
allocate_extent_clause
physical_attributes_clause
logging_clause
key_compression
MODIFY

PARTITION

partition

PARAMETERS

(

’

ODCI_parameters

’

)

COALESCE
UPDATE

BLOCK

REFERENCES

UNUSABLE

(deallocate_unused_clause::= on page 10-79, allocate_extent_clause::= on page 10-80,
physical_attributes_clause::= on page 10-80, logging_clause::= on page 8-38, key_
compression::= on page 10-81)
rename_index_partition::=
PARTITION

partition

RENAME

TO
SUBPARTITION

subpartition

10-82 Oracle Database SQL Language Reference

new_name

ALTER INDEX

drop_index_partition::=
DROP

PARTITION

partition_name

split_index_partition::=
,
SPLIT

PARTITION

INTO

(

partition_name_old

index_partition_description

AT

,

(

literal

)

index_partition_description

)

parallel_clause

(parallel_clause::= on page 10-80)
index_partition_description::=
segment_attributes_clause
key_compression
PARAMETERS

(

’

ODCI_parameters

’

)

UNUSABLE

partition
PARTITION

(segment_attributes_clause::= on page 10-83, key_compression::= on page 10-81)
segment_attributes_clause::=
physical_attributes_clause
TABLESPACE

tablespace

logging_clause

(physical_attributes_clause::= on page 10-80, logging_clause::= on page 8-38)
modify_index_subpartition::=
UNUSABLE
MODIFY

SUBPARTITION

subpartition

allocate_extent_clause
deallocate_unused_clause

(allocate_extent_clause::= on page 10-80, deallocate_unused_clause::= on page 10-79)

Semantics
10

schema
Specify the schema containing the index. If you omit schema, then Oracle Database
assumes the index is in your own schema.

index
Specify the name of the index to be altered.
SQL Statements: ALTER CLUSTER to ALTER JAVA

10-83

ALTER INDEX

Restrictions on Modifying Indexes The modification of indexes is subject to the
following restrictions:
■

■

If index is a domain index, then you can specify only the PARAMETERS clause, the
RENAME clause, the rebuild_clause (with or without the PARAMETERS clause), the
parallel_clause, or the UNUSABLE clause. No other clauses are valid.
You cannot alter or rename a domain index that is marked LOADING or FAILED. If an
index is marked FAILED, then the only clause you can specify is REBUILD.
See Also: Oracle Database Data Cartridge Developer's Guide for
information on the LOADING and FAILED states of domain indexes

deallocate_unused_clause
Use the deallocate_unused_clause to explicitly deallocate unused space at the end of
the index and make the freed space available for other segments in the tablespace.
If index is range-partitioned or hash-partitioned, then Oracle Database deallocates
unused space from each index partition. If index is a local index on a
composite-partitioned table, then Oracle Database deallocates unused space from each
index subpartition.
Restrictions on Deallocating Space Deallocation of space is subject to the following
restrictions:
■

You cannot specify this clause for an index on a temporary table.

■

You cannot specify this clause and also specify the rebuild_clause.

Refer to deallocate_unused_clause on page 8-27 for a full description of this clause.
The KEEP clause lets you specify the number of bytes above the high
water mark that the index will have after deallocation. If the number of remaining
extents is less than MINEXTENTS, then MINEXTENTS is set to the current number of
extents. If the initial extent becomes smaller than INITIAL, then INITIAL is set to the
value of the current initial extent. If you omit KEEP, then all unused space is freed.
KEEP integer

Refer to ALTER TABLE on page 12-2 for a complete description of this clause.

allocate_extent_clause
The allocate_extent_clause lets you explicitly allocate a new extent for the index.
For a local index on a hash-partitioned table, Oracle Database allocates a new extent
for each partition of the index.
Restriction on Allocating Extents You cannot specify this clause for an index on a
temporary table or for a range-partitioned or composite-partitioned index.

Refer to allocate_extent_clause on page 8-2 for a full description of this clause.

shrink_clause
Use this clause to compact the index segments. Specifying ALTER INDEX ... SHRINK SPACE
COMPACT is equivalent to specifying ALTER INDEX ... COALESCE.
For complete information on this clause, refer to shrink_clause on page 12-36 in the
documentation on CREATE TABLE.
Restriction on Shrinking Index Segments You cannot specify this clause for for a

bitmap join index or for a function-based index.

10-84 Oracle Database SQL Language Reference

ALTER INDEX

parallel_clause
Use the PARALLEL clause to change the default degree of parallelism for queries and
DML on the index.
Restriction on Parallelizing Indexes You cannot specify this clause for an index on a
temporary table.

For complete information on this clause, refer to parallel_clause on page 16-63 in the
documentation on CREATE TABLE.
See Also:

"Enabling Parallel Queries: Example" on page 10-95

physical_attributes_clause
Use the physical_attributes_clause to change the values of parameters for a
nonpartitioned index, all partitions and subpartitions of a partitioned index, a
specified partition, or all subpartitions of a specified partition.
See Also:
■

■

the physical attributes parameters in CREATE TABLE on
page 16-6
"Modifying Real Index Attributes: Example" on page 10-95 and
"Changing MAXEXTENTS: Example" on page 10-95

Restrictions on Index Physical Attributes Index physical attributes are subject to the

following restrictions:
■

You cannot specify this clause for an index on a temporary table.

■

You cannot specify the PCTUSED parameter at all when altering an index.

■

You can specify the PCTFREE parameter only as part of the rebuild_clause, the
modify_index_default_attrs clause, or the split_index_partition clause.

storage_clause
Use the storage_clause to change the storage parameters for a nonpartitioned index,
index partition, or all partitions of a partitioned index, or default values of these
parameters for a partitioned index. Refer to storage_clause on page 8-46 for complete
information on this clause.
logging_clause
Use the logging_clause to change the logging attribute of the index. If you also
specify the REBUILD clause, then this new setting affects the rebuild operation. If you
specify a different value for logging in the REBUILD clause, then Oracle Database uses
the last logging value specified as the logging attribute of the index and of the rebuild
operation.
An index segment can have logging attributes different from those of the base table
and different from those of other index segments for the same base table.
Restriction on Index Logging You cannot specify this clause for an index on a
temporary table.

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-85

ALTER INDEX

See Also:
■
■

logging_clause on page 8-38 for a full description of this clause
Oracle Database VLDB and Partitioning Guide for more information
about parallel DML

RECOVERABLE | UNRECOVERABLE
These keywords are deprecated and have been replaced with LOGGING and NOLOGGING,
respectively. Although RECOVERABLE and UNRECOVERABLE are supported for backward
compatibility, Oracle strongly recommends that you use the LOGGING and NOLOGGING
keywords.
RECOVERABLE is not a valid keyword for creating partitioned tables or LOB storage
characteristics. UNRECOVERABLE is not a valid keyword for creating partitioned or
index-organized tables. Also, it can be specified only with the AS subquery clause of
CREATE INDEX.

rebuild_clause
Use the rebuild_clause to re-create an existing index or one of its partitions or
subpartitions. If index is marked UNUSABLE, then a successful rebuild will mark it
USABLE. For a function-based index, this clause also enables the index. If the function
on which the index is based does not exist, then the rebuild statement will fail.
When you rebuild the secondary index of an index-organized
table, Oracle Database preserves the primary key columns contained
in the logical rowid when the index was created. Therefore, if the
index was created with the COMPATIBLE initialization parameter set to
less than 10.0.0, the rebuilt index will contain the index key and any of
the primary key columns of the table that are not also in the index key.
If the index was created with the COMPATIBLE initialization parameter
set to 10.0.0 or greater, then the rebuilt index will contain the index
key and all the primary key columns of the table, including those also
in the index key.

Note:

Restrictions on Rebuilding Indexes The rebuilding of indexes is subject to the
following restrictions:
■
■

■

■

■

■

You cannot rebuild an index on a temporary table.
You cannot rebuild a bitmap index that is marked INVALID. Instead, you must drop
and then re-create it.
You cannot rebuild an entire partitioned index. You must rebuild each partition or
subpartition, as described for the PARTITION clause.
You cannot specify the deallocate_unused_clause in the same statement as the
rebuild_clause.
You cannot change the value of the PCTFREE parameter for the index as a whole
(ALTER INDEX) or for a partition (ALTER INDEX ... MODIFY PARTITION). You can specify
PCTFREE in all other forms of the ALTER INDEX statement.
For a domain index:
–

You can specify only the PARAMETERS clause (either for the index or for a
partition of the index) or the parallel_clause. No other rebuild clauses are
valid.

10-86 Oracle Database SQL Language Reference

ALTER INDEX

■

■

■

–

You can rebuild an index only if the index is not marked IN_PROGRESS.

–

You can rebuild an index partition only if the index is not marked IN_
PROGRESS or FAILED and the partition is not marked IN_PROGRESS.

You cannot rebuild a local index, but you can rebuild a partition of a local index
(ALTER INDEX ... REBUILD PARTITION).
For a local index on a hash partition or subpartition, the only parameter you can
specify is TABLESPACE.
You cannot rebuild an online index that is used to enforce a deferrable unique
constraint.

PARTITION Clause
Use the PARTITION clause to rebuild one partition of an index. You can also use this
clause to move an index partition to another tablespace or to change a create-time
physical attribute.
The storage of partitioned database entities in tablespaces of different block sizes is
subject to several restrictions. Refer to Oracle Database VLDB and Partitioning Guide for
a discussion of these restrictions.
Restriction on Rebuilding Partitions You cannot specify this clause for a local index

on a composite-partitioned table. Instead, use the REBUILD SUBPARTITION clause.
See Also: Oracle Database VLDB and Partitioning Guide for more
information about partition maintenance operations and "Rebuilding
Unusable Index Partitions: Example" on page 10-95

SUBPARTITION Clause
Use the SUBPARTITION clause to rebuild one subpartition of an index. You can also use
this clause to move an index subpartition to another tablespace. If you do not specify
TABLESPACE, then the subpartition is rebuilt in the same tablespace.
The storage of partitioned database entities in tablespaces of different block sizes is
subject to several restrictions. Refer to Oracle Database VLDB and Partitioning Guide for
a discussion of these restrictions.
Restriction on Modifying Index Subpartitions The only parameters you can specify
for a subpartition are TABLESPACE, ONLINE, and the parallel_clause.

REVERSE | NOREVERSE
Indicate whether the bytes of the index block are stored in reverse order:
■

■

REVERSE stores the bytes of the index block in reverse order and excludes the rowid
when the index is rebuilt.
NOREVERSE stores the bytes of the index block without reversing the order when
the index is rebuilt. Rebuilding a REVERSE index without the NOREVERSE keyword
produces a rebuilt, reverse-keyed index.

Restrictions on Reverse Indexes Reverse indexes are subject to the following
restrictions:
■

You cannot reverse a bitmap index or an index-organized table.

■

You cannot specify REVERSE or NOREVERSE for a partition or subpartition.

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-87

ALTER INDEX

See Also: "Storing Index Blocks in Reverse Order: Example" on
page 10-95

parallel_clause
Use the parallel_clause to parallelize the rebuilding of the index and to change the
degree of parallelism for the index itself. All subsequent operations on the index will
be executed with the degree of parallelism specified by this clause, unless overridden
by a subsequent data definition language (DDL) statement with the parallel_clause.
The following exceptions apply:
■

■

If ALTER SESSION DISABLE PARALLEL DDL was specified before rebuilding the index,
then the index will be rebuilt serially and the degree of parallelism for the index
will be changed to 1.
If ALTER SESSION FORCE PARALLEL DDL was specified before rebuilding the index,
then the index will be rebuilt in parallel and the degree of parallelism for the index
will be changed to the value that was specified in the ALTER SESSION statement, or
DEFAULT if no value was specified.
See Also:

"Rebuilding an Index in Parallel: Example" on page 10-95

TABLESPACE Clause
Specify the tablespace where the rebuilt index, index partition, or index subpartition
will be stored. The default is the default tablespace where the index or partition
resided before you rebuilt it.
key_compression
Specify COMPRESS to enable key compression, which eliminates repeated occurrence of
key column values. Use integer to specify the prefix length (number of prefix
columns to compress).
■

■

For unique indexes, the range of valid prefix length values is from 1 to the number
of key columns minus 1. The default prefix length is the number of key columns
minus 1.
For nonunique indexes, the range of valid prefix length values is from 1 to the
number of key columns. The default prefix length is number of key columns.

Oracle Database compresses indexes that are nonunique or unique indexes of at least
two columns. If you want to use compression for a partitioned index, then the index
must have compression enabled at the index level.
Specify NOCOMPRESS to disable key compression. This is the default.
Restriction on Key Compression You cannot specify COMPRESS for a bitmap index.

ONLINE Clause
Specify ONLINE to allow DML operations on the table or partition during rebuilding of
the index.
Restrictions on Online Indexes Online indexes are subject to the following

restrictions:
■

■

Parallel DML is not supported during online index building. If you specify ONLINE
and subsequently issue parallel DML statements, then Oracle Database returns an
error.
You cannot specify ONLINE for a bitmap join index or a cluster index.

10-88 Oracle Database SQL Language Reference

ALTER INDEX

■

For a nonunique secondary index on an index-organized table, the number of
index key columns plus the number of primary key columns that are included in
the logical rowid in the index-organized table cannot exceed 32. The logical rowid
excludes columns that are part of the index key.

logging_clause
Specify whether the ALTER INDEX ... REBUILD operation will be logged.
Refer to the logging_clause on page 8-38 for a full description of this clause.

PARAMETERS Clause
This clause is valid only for domain indexes in a top-level ALTER INDEX statement and
in the rebuild_clause. This clause specifies the parameter string that is passed
uninterpreted to the appropriate ODCI indextype routine.
The maximum length of the parameter string is 1000 characters.
If you are altering or rebuilding an entire index, then the string must refer to
index-level parameters. If you are rebuilding a partition of the index, then the string
must refer to partition-level parameters.
If index is marked UNUSABLE, then modifying the parameters alone does not make it
USABLE. You must also rebuild the UNUSABLE index to make it usable.
If you have installed Oracle Text, then you can rebuild your Oracle Text domain
indexes using parameters specific to that product. For more information on those
parameters, refer to Oracle Text Reference.
You can modify index partitions only if
index is not marked IN_PROGRESS or FAILED, no index partitions are marked IN_
PROGRESS, and the partition being modified is not marked FAILED.
Restriction on the PARAMETERS Clause

See Also:
■

■

Oracle Database Data Cartridge Developer's Guide for more
information on indextype routines for domain indexes
CREATE INDEX on page 14-60 for more information on domain
indexes

XMLIndex_parameters_clause
This clause is valid only for XMLIndex indexes. This clause specifies the parameter
string that defines the XMLIndex implementation.
The maximum length of the parameter string is 1000 characters.
If you are altering or rebuilding an entire index, then the string must refer to
index-level parameters. If you are rebuilding a partition of the index, then the string
must refer to partition-level parameters.
If index is marked UNUSABLE, then modifying the parameters alone does not make it
USABLE. You must also rebuild the UNUSABLE index to make it usable.
Oracle XML DB Developer's Guide for more information on
XMLIndex, including the syntax and semantics of the XMLIndex_
parameters_clause
See Also:

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-89

ALTER INDEX

You can modify index partitions
only if index is not marked IN_PROGRESS or FAILED, no index partitions are marked
IN_PROGRESS, and the partition being modified is not marked FAILED.
Restriction on the XMLIndex_paramaters_clause

COMPILE Clause
This clause is valid only for domain indexes. Use this clause to recompile an invalid
domain index explicitly. This clause is useful primarily when the underlying indextype
has been altered to support system-managed domain indexes, so that the existing
domain index has been marked INVALID. In this situation, this ALTER INDEX statement
migrates the domain index from a user-managed domain index to a system-managed
domain index.
The CREATE INDEXTYPE storage_table_clause on page 14-89
and Oracle Database Data Cartridge Developer's Guide for information on
creating system-managed domain indexes

See Also:

ENABLE Clause
ENABLE applies only to a function-based index that has been disabled, either by an
ALTER INDEX ... DISABLE statement, or because a user-defined function used by the
index was dropped or replaced. This clause enables such an index if these conditions
are true:
■
■

■

The function is currently valid.
The signature of the current function matches the signature of the function when
the index was created.
The function is currently marked as DETERMINISTIC.

Restrictions on Enabling Function-based Indexes The ENABLE clause is subject to the
following restrictions:
■

■

You cannot specify any other clauses of ALTER INDEX in the same statement with
ENABLE.
You cannot specify this clause for an index on a temporary table. Instead, you
must drop and recreate the index. You can retrieve the creation DDL for the index
using the DBMS_METADATA package.

DISABLE Clause
DISABLE applies only to a function-based index. This clause lets you disable the use of
a function-based index. You might want to do so, for example, while working on the
body of the function. Afterward you can either rebuild the index or specify another
ALTER INDEX statement with the ENABLE keyword.

UNUSABLE
Specify UNUSABLE to mark the index or index partition(s) or index subpartition(s)
UNUSABLE. The space allocated for an index or index partition or subpartition is freed
immediately when the object is marked UNUSABLE. An unusable index must be rebuilt,
or dropped and re-created, before it can be used. While one partition is marked
UNUSABLE, the other partitions of the index are still valid. You can execute statements
that require the index if the statements do not access the unusable partition. You can
also split or rename the unusable partition before rebuilding it. Refer to CREATE INDEX
... UNUSABLE on page 14-79 for more information.

10-90 Oracle Database SQL Language Reference

ALTER INDEX

Restriction on Marking Indexes Unusable You cannot specify this clause for an index

on a temporary table.

VISIBLE | INVISIBLE
Use this clause to specify whether the index is visible or invisible to the optimizer.
Refer to "VISIBLE | INVISIBLE" in CREATE INDEX on page 14-73 for a full description of
this clause.

RENAME Clause
Use this clause to rename an index. The new_index_name is a single identifier and does
not include the schema name.
Restriction on Renaming Indexes For a domain index, neither index nor any
partitions of index can be marked IN_PROGRESS or FAILED.
See Also:

"Renaming an Index: Example" on page 10-95

COALESCE Clause
Specify COALESCE to instruct Oracle Database to merge the contents of index blocks
where possible to free blocks for reuse.
Restrictions on Coalescing Index Blocks Coalescing of index blocks is subject to the
following restrictions:
■
■

You cannot specify this clause for an index on a temporary table.
Do not specify this clause for the primary key index of an index-organized table.
Instead use the COALESCE clause of ALTER TABLE.
See Also:
■

■

■

Oracle Database Administrator's Guide for more information on
space management and coalescing indexes
COALESCE Clause on page 12-41 for information on coalescing
the space of an index-organized table
shrink_clause on page 12-36 for an alternative method of
compacting index segments

MONITORING USAGE | NOMONITORING USAGE
Use this clause to determine whether Oracle Database should monitor index use.
■

■

Specify MONITORING USAGE to begin monitoring the index. Oracle Database first
clears existing information on index use, and then monitors the index for use until
a subsequent ALTER INDEX ... NOMONITORING USAGE statement is executed.
To terminate monitoring of the index, specify NOMONITORING USAGE.

To see whether the index has been used since this ALTER INDEX ... NOMONITORING USAGE
statement was issued, query the USED column of the V$OBJECT_USAGE dynamic
performance view.
Oracle Database Reference for information on the data
dictionary and dynamic performance views

See Also:

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-91

ALTER INDEX

UPDATE BLOCK REFERENCES Clause
The UPDATE BLOCK REFERENCES clause is valid only for normal and domain indexes on
index-organized tables. Specify this clause to update all the stale guess data block
addresses stored as part of the index row with the correct database address for the
corresponding block identified by the primary key.
For a domain index, Oracle Database executes the ODCIIndexAlter routine with the
alter_option parameter set to AlterIndexUpdBlockRefs. This routine enables the
cartridge code to update the stale guess data block addresses in the index.
Restriction on UPDATE BLOCK REFERENCES You cannot combine this clause with
any other clause of ALTER INDEX.

alter_index_partitioning
The partitioning clauses of the ALTER INDEX statement are valid only for partitioned
indexes.
The storage of partitioned database entities in tablespaces of different block sizes is
subject to several restrictions. Refer to Oracle Database VLDB and Partitioning Guide for
a discussion of these restrictions.
Restrictions on Modifying Index Partitions Modifying index partitions is subject to
the following restrictions:
■
■

You cannot specify any of these clauses for an index on a temporary table.
You can combine several operations on the base index into one ALTER INDEX
statement (except RENAME and REBUILD), but you cannot combine partition
operations with other partition operations or with operations on the base index.

modify_index_default_attrs
Specify new values for the default attributes of a partitioned index.
The only attribute you can
specify for a hash-partitioned global index or for an index on a hash-partitioned table
is TABLESPACE.

Restriction on Modifying Partition Default Attributes

TABLESPACE Specify the default tablespace for new partitions of an index or
subpartitions of an index partition.
logging_clause Specify the default logging attribute of a partitioned index or an

index partition.
Refer to logging_clause on page 8-38 for a full description of this clause.
FOR PARTITION Use the FOR PARTITION clause to specify the default attributes for the
subpartitions of a partition of a local index on a composite-partitioned table.
Restriction on FOR PARTITION You cannot specify FOR PARTITION for a list partition.
See Also:

"Modifying Default Attributes: Example" on page 10-96

add_hash_index_partition
Use this clause to add a partition to a global hash-partitioned index. Oracle Database
adds hash partitions and populates them with index entries rehashed from an existing
hash partition of the index, as determined by the hash function. If you omit the
partition name, then Oracle Database assigns a name of the form SYS_Pn. If you omit
10-92 Oracle Database SQL Language Reference

ALTER INDEX

the TABLESPACE clause, then Oracle Database places the partition in the tablespace
specified for the index. If no tablespace is specified for the index, then Oracle Database
places the partition in the default tablespace of the user, if one has been specified, or in
the system default tablespace.
modify_index_partition
Use the modify_index_partition clause to modify the real physical attributes, logging
attribute, or storage characteristics of index partition partition or its subpartitions.
For a hash-partitioned global index, the only subclause of this clause you can specify is
UNUSABLE.
COALESCE Specify this clause to merge the contents of index partition blocks where
possible to free blocks for reuse.

The UPDATE BLOCK REFERENCES clause is valid only
for normal indexes on index-organized tables. Use this clause to update all stale guess
data block addresses stored in the secondary index partition.

UPDATE BLOCK REFERENCES

Restrictions on UPDATE BLOCK REFERENCES

This clause is subject to the

following restrictions:
■

■

You cannot specify the physical_attributes_clause for an index on a
hash-partitioned table.
You cannot specify UPDATE BLOCK REFERENCES with any other clause in ALTER
INDEX.
If the index is a local index on a composite-partitioned table,
then the changes you specify here will override any attributes
specified earlier for the subpartitions of index, as well as establish
default values of attributes for future subpartitions of that partition.
To change the default attributes of the partition without overriding the
attributes of subpartitions, use ALTER TABLE ... MODIFY DEFAULT
ATTRIBUTES FOR PARTITION.
Note:

See Also:

"Marking an Index Unusable: Examples" on page 10-95

This clause has the same function for index partitions that it has
for the index as a whole. Refer to "UNUSABLE" on page 10-90.

UNUSABLE Clause

This clause is relevant for composite-partitioned indexes. Use this
clause to change the compression attribute for the partition and every subpartition in
that partition. Oracle Database marks each index subpartition in the partition
UNUSABLE and you must then rebuild these subpartitions. Key compression must
already have been specified for the table before you can specify key compression for a
partition. You can specify this clause only at the partition level. You cannot change the
compression attribute for an individual subpartition.
key_compression

You can use this clause for noncomposite index partitions. However, it is more efficient
to use the rebuild_clause for noncomposite partitions, which lets you rebuild and set
the compression attribute in one step.
rename_index_partition
Use the rename_index_partition clauses to rename index partition or subpartition
to new_name.
SQL Statements: ALTER CLUSTER to ALTER JAVA

10-93

ALTER INDEX

Restrictions on Renaming Index Partitions

Renaming index partitions is subject to

the following restrictions:
■
■

You cannot rename the subpartition of a list partition.
For a partition of a domain index, index cannot be marked IN_PROGRESS or FAILED,
none of the partitions can be marked IN_PROGRESS, and the partition you are
renaming cannot be marked FAILED.
See Also:

"Renaming an Index Partition: Example" on page 10-96

drop_index_partition
Use the drop_index_partition clause to remove a partition and the data in it from a
partitioned global index. When you drop a partition of a global index, Oracle Database
marks the next index partition UNUSABLE. You cannot drop the highest partition of a
global index.
See Also:

"Dropping an Index Partition: Example" on page 10-96

split_index_partition
Use the split_index_partition clause to split a partition of a global
range-partitioned index into two partitions, adding a new partition to the index. This
clause is not valid for hash-partitioned global indexes. Instead, use the add_hash_
index_partition clause.
Splitting a partition marked UNUSABLE results in two partitions, both marked UNUSABLE.
You must rebuild the partitions before you can use them.
Splitting a usable partition results in two partitions populated with index data. Both
new partitions are usable.
Specify the new noninclusive upper bound for split_partition_1. The
value_list must evaluate to less than the presplit partition bound for partition_
name_old and greater than the partition bound for the next lowest partition (if there is
one).
AT Clause

Specify (optionally) the name and physical attributes of each of the two
partitions resulting from the split.

INTO Clause

See Also:

"Splitting a Partition: Example" on page 10-96

coalesce_index_partition
This clause is valid only for hash-partitioned global indexes. Oracle Database reduces
by one the number of index partitions. Oracle Database selects the partition to coalesce
based on the requirements of the hash function. Use this clause if you want to
distribute index entries of a selected partition into one of the remaining partitions and
then remove the selected partition.
modify_index_subpartition
Use the modify_index_subpartition clause to mark UNUSABLE or allocate or deallocate
storage for a subpartition of a local index on a composite-partitioned table. All other
attributes of such a subpartition are inherited from partition-level default attributes.

10-94 Oracle Database SQL Language Reference

ALTER INDEX

Examples
10

Storing Index Blocks in Reverse Order: Example The following statement rebuilds
index ord_customer_ix (created in "Creating an Index: Example" on page 14-80) so
that the bytes of the index block are stored in reverse order:
ALTER INDEX ord_customer_ix REBUILD REVERSE;

Rebuilding an Index in Parallel: Example The following statement causes the index
to be rebuilt from the existing index by using parallel execution processes to scan the
old and to build the new index:
ALTER INDEX ord_customer_ix REBUILD PARALLEL;

The following statement alters the
oe.cust_lname_ix index so that future data blocks within this index use 5 initial
transaction entries:
Modifying Real Index Attributes: Example

ALTER INDEX oe.cust_lname_ix
INITRANS 5;

If the oe.cust_lname_ix index were partitioned, then this statement would also alter
the default attributes of future partitions of the index. Partitions added in the future
would then use 5 initial transaction entries and an incremental extent of 100K.
The following statement sets the parallel
attributes for index upper_ix (created in "Creating a Function-Based Index: Example"
on page 14-81) so that scans on the index will be parallelized:
Enabling Parallel Queries: Example

ALTER INDEX upper_ix PARALLEL;

Renaming an Index: Example

The following statement renames an index:

ALTER INDEX upper_ix RENAME TO upper_name_ix;

Marking an Index Unusable: Examples The following statements use the cost_ix
index, which was created in "Creating a Range-Partitioned Global Index: Example" on
page 14-83. Partition p1 of that index was dropped in "Dropping an Index Partition:
Example" on page 10-96. The first statement marks index partition p2 as UNUSABLE:
ALTER INDEX cost_ix
MODIFY PARTITION p2 UNUSABLE;

The next statement marks the entire index cost_ix as UNUSABLE:
ALTER INDEX cost_ix UNUSABLE;

Rebuilding Unusable Index Partitions: Example The following statements rebuild

partitions p2 and p3 of the cost_ix index, making the index once more usable: The
rebuilding of partition p3 will not be logged:
ALTER INDEX cost_ix
REBUILD PARTITION p2;
ALTER INDEX cost_ix
REBUILD PARTITION p3 NOLOGGING;

Changing MAXEXTENTS: Example The following statement changes the maximum

number of extents for partition p3 and changes the logging attribute:
/* This example will fail if the tablespace in which partition p3
resides is locally managed.

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-95

ALTER INDEX

*/
ALTER INDEX cost_ix MODIFY PARTITION p3
STORAGE(MAXEXTENTS 30) LOGGING;

Renaming an Index Partition: Example The following statement renames an index
partition of the cost_ix index (created in "Creating a Range-Partitioned Global Index:
Example" on page 14-83):
ALTER INDEX cost_ix
RENAME PARTITION p3 TO p3_Q3;

The following statement splits partition p2 of index
cost_ix (created in "Creating a Range-Partitioned Global Index: Example" on
page 14-83) into p2a and p2b:
Splitting a Partition: Example

ALTER INDEX cost_ix
SPLIT PARTITION p2 AT (1500)
INTO ( PARTITION p2a TABLESPACE tbs_01 LOGGING,
PARTITION p2b TABLESPACE tbs_02);

Dropping an Index Partition: Example

The following statement drops index partition

p1 from the cost_ix index:
ALTER INDEX cost_ix
DROP PARTITION p1;

The following statement alters the default
attributes of local partitioned index prod_idx, which was created in "Creating an Index
on a Hash-Partitioned Table: Example" on page 14-83 on page 14-80. Partitions added
in the future will use 5 initial transaction entries:

Modifying Default Attributes: Example

ALTER INDEX prod_idx
MODIFY DEFAULT ATTRIBUTES INITRANS 5;

10-96 Oracle Database SQL Language Reference

ALTER INDEXTYPE

ALTER INDEXTYPE
Purpose
10

Use the ALTER INDEXTYPE statement to add or drop an operator of the indextype or to
modify the implementation type or change the properties of the indextype.

Prerequisites
10

The indextype must be in your own schema or you must have the ALTER ANY
INDEXTYPE system privilege.
To add a new operator, you must have the EXECUTE object privilege on the operator.
To change the implementation type, you must have the EXECUTE object privilege on the
new implementation type.

Syntax
10

alter_indextype::=
schema
ALTER

.

INDEXTYPE

indextype

,
schema

ADD

.

using_type_clause
operator

(

parameter_types

)

DROP
COMPILE

RANGE
WITH

LOCAL

PARTITION

storage_table_clause
;

(using_type_clause::= on page 10-97, storage_table_clause on page 10-99)
using_type_clause::=
schema
USING

.

array_DML_clause
implementation_type

(array_DML_clause::= on page 10-98)

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-97

ALTER INDEXTYPE

array_DML_clause::=
WITH
WITHOUT
ARRAY

DML

,
schema
schema

.

(

,

.
varray_type

type

)

storage_table_clause::=
SYSTEM
WITH

MANAGED

STORAGE

TABLES

USER

Semantics
10

schema
Specify the name of the schema in which the indextype resides. If you omit schema,
then Oracle Database assumes the indextype is in your own schema.

indextype
Specify the name of the indextype to be modified.

ADD | DROP
Use the ADD or DROP clause to add or drop an operator.
No special privilege needed to drop.
■

■

For schema, specify the schema containing the operator. If you omit schema, then
Oracle assumes the operator is in your own schema.
For operator, specify the name of the operator supported by the indextype.
All the operators listed in this clause must be valid operators.

■

For parameter_type, list the types of parameters to the operator.

using_type_clause
The USING clause lets you specify a new type to provide the implementation for the
indextype.

array_DML_clause
Use this clause to modify the indextype to support the array interface for the
ODCIIndexInsert method.
type and varray_type If the data type of the column to be indexed is a user-defined
object type, then you must specify this clause to identify the varray varray_type that
Oracle should use to hold column values of type. If the indextype supports a list of
types, then you can specify a corresponding list of varray types. If you omit schema for
either type or varray_type, then Oracle assumes the type is in your own schema.

10-98 Oracle Database SQL Language Reference

ALTER INDEXTYPE

If the data type of the column to be indexed is a built-in system type, then any varray
type specified for the indextype takes precedence over the ODCI types defined by the
system.

COMPILE
Use this clause to recompile the indextype explicitly. This clause is required only after
some upgrade operations, because Oracle Database normally recompiles the indextype
automatically.

storage_table_clause
This clause has the same behavior when altering an indextype that it has when you are
creating an indextype. Refer to the CREATE INDEXTYPE storage_table_clause on page 14-89
for more information.

WITH LOCAL PARTITION
This clause has the same behavior when altering an indextype that it has when you
create an indextype. Refer to the CREATE INDEXTYPE clause WITH LOCAL PARTITION
on page 14-89 for more information.

Examples
10

The following example compiles the position_
indextype indextype created in "Creating an Indextype: Example" on page 14-90.

Altering an Indextype: Example

ALTER INDEXTYPE position_indextype COMPILE;

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-99

ALTER JAVA

ALTER JAVA
Purpose
10

Use the ALTER JAVA statement to force the resolution of a Java class schema object or
compilation of a Java source schema object. (You cannot call the methods of a Java
class before all its external references to Java names are associated with other classes.)
Oracle Database Java Developer's Guide for more information
on resolving Java classes and compiling Java sources

See Also:

Prerequisites
10

The Java source or class must be in your own schema, or you must have the ALTER ANY
PROCEDURE system privilege. You must also have the EXECUTE object privilege on Java
classes.

Syntax
10

alter_java::=
schema

SOURCE
ALTER

.

JAVA

object_name
CLASS

,
RESOLVER

(

(

schema_name

match_string

)

)

–

COMPILE
RESOLVE

;

invoker_rights_clause

(invoker_rights_clause::= on page 10-100)
invoker_rights_clause::=
CURRENT_USER
AUTHID
DEFINER

Semantics
10

JAVA SOURCE
Use ALTER JAVA SOURCE to compile a Java source schema object.

JAVA CLASS
Use ALTER JAVA CLASS to resolve a Java class schema object.

10-100 Oracle Database SQL Language Reference

ALTER JAVA

object_name
Specify a previously created Java class or source schema object. Use double quotation
marks to preserve lower- or mixed-case names.

RESOLVER
The RESOLVER clause lets you specify how schemas are searched for referenced fully
specified Java names, using the mapping pairs specified when the Java class or source
was created.
See Also: CREATE JAVA on page 14-91 and "Resolving a Java Class:
Example" on page 10-101

RESOLVE | COMPILE
RESOLVE and COMPILE are synonymous keywords. They let you specify that Oracle
Database should attempt to resolve the primary Java class schema object.
■

■

When applied to a class, resolution of referenced names to other class schema
objects occurs.
When applied to a source, source compilation occurs.

invoker_rights_clause
The invoker_rights_clause lets you specify whether the methods of the class execute
with the privileges and in the schema of the user who defined it or with the privileges
and in the schema of CURRENT_USER.
This clause also determines how Oracle Database resolves external names in queries,
DML operations, and dynamic SQL statements in the member functions and
procedures of the type.
Specify CURRENT_USER if you want the methods of the class
to execute with the privileges of CURRENT_USER. This clause is the default and creates
an invoker-rights class.

AUTHID CURRENT_USER

This clause also specifies that external names in queries, DML operations, and
dynamic SQL statements resolve in the schema of CURRENT_USER. External names in all
other statements resolve in the schema in which the methods reside.
AUTHID DEFINER Specify DEFINER if you want the methods of the class to execute
with the privileges of the user who defined the class.

This clause also specifies that external names resolve in the schema where the methods
reside.
Oracle Database PL/SQL Language Reference for information
on how CURRENT_USER is determined

See Also:

Examples
10

Resolving a Java Class: Example The following statement forces the resolution of a
Java class:
ALTER JAVA CLASS "Agent"
RESOLVER (("/usr/bin/bfile_dir/*" pm)(* public))
RESOLVE;

SQL Statements: ALTER CLUSTER to ALTER JAVA

10-101

ALTER JAVA

10-102 Oracle Database SQL Language Reference

11

11
SQL Statements: ALTER LIBRARY to ALTER
SYSTEM
This chapter contains the following SQL statements:
■

ALTER LIBRARY

■

ALTER MATERIALIZED VIEW

■

ALTER MATERIALIZED VIEW LOG

■

ALTER OPERATOR

■

ALTER OUTLINE

■

ALTER PACKAGE

■

ALTER PROCEDURE

■

ALTER PROFILE

■

ALTER RESOURCE COST

■

ALTER ROLE

■

ALTER ROLLBACK SEGMENT

■

ALTER SEQUENCE

■

ALTER SESSION

■

ALTER SYSTEM

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-1

ALTER LIBRARY

ALTER LIBRARY
Purpose
11

The ALTER LIBRARY statement explicitly recompiles a library. Explicit recompilation
eliminates the need for implicit run-time recompilation and prevents associated
run-time compilation errors and performance overhead.
This statement does not change the declaration or definition of
an existing library. To redeclare or redefine a library, use the "CREATE
LIBRARY" on page 15-2 with the OR REPLACE clause.

Note:

Prerequisites
11

If the library is in the SYS schema, you must be connected as SYSDBA. Otherwise, the
library must be in your own schema or you must have the ALTER ANY LIBRARY system
privilege.

Syntax
11

alter_library::=
schema
ALTER

LIBRARY

.
library_name

library_compile_clause

(library_compile_clause: See Oracle Database PL/SQL Language Reference for the
syntax of this clause.)

Semantics
11

schema
Specify the schema containing the library. If you omit schema, then Oracle Database
assumes the procedure is in your own schema.

library_name
Specify the name of the library to be recompiled.

library_compile_clause
See Oracle Database PL/SQL Language Reference for the syntax and semantics of this
clause and for complete information on creating and compiling libraries.

11-2 Oracle Database SQL Language Reference

ALTER MATERIALIZED VIEW

ALTER MATERIALIZED VIEW
Purpose
11

A materialized view is a database object that contains the results of a query. The FROM
clause of the query can name tables, views, and other materialized views. Collectively
these source objects are called master tables (a replication term) or detail tables (a
data warehousing term). This reference uses the term master tables for consistency.
The databases containing the master tables are called the master databases.
Use the ALTER MATERIALIZED VIEW statement to modify an existing materialized view
in one or more of the following ways:
■

To change its storage characteristics

■

To change its refresh method, mode, or time

■

To alter its structure so that it is a different type of materialized view

■

To enable or disable query rewrite
The keyword SNAPSHOT is supported in place of MATERIALIZED
VIEW for backward compatibility.
Note:

See Also:
■

■

■

CREATE MATERIALIZED VIEW on page 15-4 for more
information on creating materialized views
Oracle Database Advanced Replication for information on
materialized views in a replication environment
Oracle Database Data Warehousing Guide for information on
materialized views in a data warehousing environment

Prerequisites
11

The privileges required to alter a materialized view should be granted directly, as
follows:
The materialized view must be in your own schema, or you must have the ALTER ANY
MATERIALIZED VIEW system privilege.
To enable a materialized view for query rewrite:
■

■

■

If all of the master tables in the materialized view are in your schema, then you
must have the QUERY REWRITE privilege.
If any of the master tables are in another schema, then you must have the GLOBAL
QUERY REWRITE privilege.
If the materialized view is in another user's schema, then both you and the owner
of that schema must have the appropriate QUERY REWRITE privilege, as described in
the preceding two items. In addition, the owner of the materialized view must
have SELECT access to any master tables that the materialized view owner does not
own.

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-3

ALTER MATERIALIZED VIEW

Syntax
11

alter_materialized_view::=
schema
ALTER

MATERIALIZED

.

VIEW

materialized_view

physical_attributes_clause
modify_mv_column_clause
table_compression
,
LOB_storage_clause
,
modify_LOB_storage_clause
alter_table_partitioning
parallel_clause
logging_clause
allocate_extent_clause
deallocate_unused_clause
shrink_clause
CACHE
NOCACHE

alter_iot_clauses

MODIFY
USING

INDEX

physical_attributes_clause

scoped_table_ref_constraint

alter_mv_refresh

ENABLE
QUERY

REWRITE

DISABLE
COMPILE
CONSIDER

FRESH
;

(physical_attributes_clause::= on page 11-5, modify_mv_column_clause::= on page 11-5,
table_compression::= on page 11-5, LOB_storage_clause::= on page 11-5, modify_LOB_
storage_clause::= on page 11-6, alter_table_partitioning::= on page 12-19 (part of ALTER
TABLE), parallel_clause::= on page 11-7, logging_clause::= on page 11-7, allocate_extent_
clause::= on page 11-7, deallocate_unused_clause::= on page 11-8, shrink_clause::= on
page 11-8, alter_iot_clauses::= on page 11-8, scoped_table_ref_constraint::= on page 11-9,
alter_mv_refresh::= on page 11-9)

11-4 Oracle Database SQL Language Reference

ALTER MATERIALIZED VIEW

physical_attributes_clause::=
PCTFREE

integer

PCTUSED

integer

INITRANS

integer

storage_clause

(storage_clause::= on page 8-50)
modify_mv_column_clause::=
ENCRYPT

encryption_spec

DECRYPT
MODIFY

(

column

)

table_compression::=
BASIC
OLTP
LOW

FOR

HIGH

QUERY
ARCHIVE
COMPRESS
NOCOMPRESS

LOB_storage_clause::=
SECUREFILE

,
(

LOB_item

)

STORE

BASICFILE

AS
(

LOB

LOB_storage_parameters

)

SECUREFILE
BASICFILE
(

LOB_item

)

STORE

AS

LOB_segname
(

LOB_storage_parameters

)

(LOB_storage_parameters::= on page 11-6)

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-5

ALTER MATERIALIZED VIEW

LOB_storage_parameters::=
TABLESPACE

tablespace
storage_clause

LOB_parameters
storage_clause

(LOB_parameters::= on page 11-6, storage_clause::= on page 8-50)
LOB_parameters::=
ENABLE
STORAGE

IN

ROW

DISABLE
CHUNK

integer

PCTVERSION

integer

FREEPOOLS

integer

LOB_retention_clause
LOB_deduplicate_clause
LOB_compression_clause
ENCRYPT

encryption_spec

DECRYPT
CACHE

logging_clause

NOCACHE
CACHE

READS

(storage_clause::= on page 8-50, logging_clause::= on page 8-38)
modify_LOB_storage_clause::=
MODIFY

LOB

(

LOB_item

)

(

modify_LOB_parameters

(modify_LOB_parameters::= on page 11-7)

11-6 Oracle Database SQL Language Reference

)

ALTER MATERIALIZED VIEW

modify_LOB_parameters::=
storage_clause
PCTVERSION
integer
FREEPOOLS
REBUILD

FREEPOOLS

LOB_retention_clause
LOB_deduplicate_clause
LOB_compression_clause
ENCRYPT

encryption_spec

DECRYPT
CACHE
logging_clause

NOCACHE
CACHE

READS

allocate_extent_clause
shrink_clause
deallocate_unused_clause

(storage_clause::= on page 8-50, LOB_retention_clause::= on page 16-14, LOB_
compression_clause::= on page 16-14, logging_clause::= on page 8-38, allocate_extent_
clause::= on page 11-7, shrink_clause::= on page 11-8, deallocate_unused_clause::= on
page 11-8)
parallel_clause::=
NOPARALLEL
integer
PARALLEL

logging_clause::=
LOGGING
NOLOGGING
FILESYSTEM_LIKE_LOGGING

allocate_extent_clause::=
SIZE
(

size_clause

DATAFILE
INSTANCE

ALLOCATE

’

filename

’

)

integer

EXTENT

(size_clause::= on page 8-47)
SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-7

ALTER MATERIALIZED VIEW

deallocate_unused_clause::=
KEEP
DEALLOCATE

size_clause

UNUSED

(size_clause::= on page 8-47)
shrink_clause::=
COMPACT
SHRINK

CASCADE

SPACE

alter_iot_clauses::=
index_org_table_clause
alter_overflow_clause
alter_mapping_table_clauses
COALESCE

(index_org_table_clause::= on page 11-8, alter_overflow_clause::= on page 11-8, alter_
mapping_table_clauses: not supported with materialized views)
index_org_table_clause::=
mapping_table_clause
PCTTHRESHOLD

integer

key_compression

index_org_overflow_clause

(mapping_table_clause: not supported with materialized views, key_compression:
not supported with materialized views, index_org_overflow_clause::= on page 11-8)
index_org_overflow_clause::=
INCLUDING

column_name

segment_attributes_clause
OVERFLOW

(segment_attributes_clause::= on page 12-8—part of ALTER TABLE)
alter_overflow_clause::=
add_overflow_clause
segment_attributes_clause
allocate_extent_clause
OVERFLOW
shrink_clause
deallocate_unused_clause

(allocate_extent_clause::= on page 11-7, shrink_clause::= on page 11-8, deallocate_unused_
clause::= on page 11-8)
11-8 Oracle Database SQL Language Reference

ALTER MATERIALIZED VIEW

add_overflow_clause::=
,
segment_attributes_clause
segment_attributes_clause
ADD

(

PARTITION

)

OVERFLOW

(segment_attributes_clause::= on page 12-8--part of ALTER TABLE)
scoped_table_ref_constraint::=
schema

ref_column
SCOPE

FOR

(

)

.

scope_table_name

IS

ref_attribute

c_alias

alter_mv_refresh::=
FAST
COMPLETE
FORCE
DEMAND
ON
COMMIT
START
REFRESH

WITH
date

NEXT
WITH

PRIMARY

KEY

DEFAULT

MASTER

MASTER

ROLLBACK

ROLLBACK

SEGMENT

USING
SEGMENT

rollback_segment

ENFORCED
USING

CONSTRAINTS
TRUSTED

Semantics
11

schema
Specify the schema containing the materialized view. If you omit schema, then Oracle
Database assumes the materialized view is in your own schema.

materialized_view
Specify the name of the materialized view to be altered.

physical_attributes_clause
Specify new values for the PCTFREE, PCTUSED, and INITRANS parameters (or, when used
in the USING INDEX clause, for the INITRANS parameter only) and the storage
characteristics for the materialized view. Refer to ALTER TABLE on page 12-2 for
information on the PCTFREE, PCTUSED, and INITRANS parameters and to storage_clause
on page 8-46 for information about storage characteristics.

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-9

ALTER MATERIALIZED VIEW

modify_mv_column_clause
Use this clause to encrypt or decrypt this column of the materialized view. Refer to the
CREATE TABLE clause encryption_spec on page 16-27 for information on this clause.

table_compression
Use the table_compression clause to instruct Oracle Database whether to compress
data segments to reduce disk and memory use. Refer to the table_compression clause of
CREATE TABLE on page 16-34 for the full semantics of this clause.

LOB_storage_clause
The LOB_storage_clause lets you specify the storage characteristics of a new LOB.
LOB storage behaves for materialized views exactly as it does for tables. Refer to the
LOB_storage_clause on page 16-42 (in CREATE TABLE) for information on the LOB storage
parameters.

modify_LOB_storage_clause
The modify_LOB_storage_clause lets you modify the physical attributes of the LOB
attribute LOB_item or the LOB object attribute. Modification of LOB storage behaves
for materialized views exactly as it does for tables.
The modify_LOB_storage_clause of ALTER TABLE on
page 12-53 for information on the LOB storage parameters that can be
modified

See Also:

alter_table_partitioning
The syntax and general functioning of the partitioning clauses for materialized views
is the same as for partitioned tables. Refer to alter_table_partitioning on page 12-57 in
the documentation for ALTER TABLE.
Restriction on Altering Materialized View Partitions You cannot specify the LOB_

storage_clause or modify_LOB_storage_clause within any of the partitioning_
clauses.
If you want to keep the contents of the materialized view
synchronized with those of the master table, then Oracle recommends
that you manually perform a complete refresh of all materialized
views dependent on the table after dropping or truncating a table
partition.

Note:

MODIFY PARTITION UNUSABLE LOCAL INDEXES Use this clause to mark UNUSABLE
all the local index partitions associated with partition.
MODIFY PARTITION REBUILD UNUSABLE LOCAL INDEXES Use this clause to
rebuild the unusable local index partitions associated with partition.

parallel_clause
The parallel_clause lets you change the default degree of parallelism for the
materialized view.
For complete information on this clause, refer to parallel_clause on page 16-63 in the
documentation on CREATE TABLE.

11-10 Oracle Database SQL Language Reference

ALTER MATERIALIZED VIEW

logging_clause
Specify or change the logging characteristics of the materialized view. Refer to the
logging_clause on page 8-38 for a full description of this clause.

allocate_extent_clause
The allocate_extent_clause lets you explicitly allocate a new extent for the
materialized view. Refer to the allocate_extent_clause on page 8-2 for a full description
of this clause.

deallocate_unused_clause
Use the deallocate_unused_clause to explicitly deallocate unused space at the end of
the materialized view and make the freed space available for other segments. Refer to
the deallocate_unused_clause on page 8-27 for a full description of this clause.

shrink_clause
Use this clause to compact the materialized view segments. For complete information
on this clause, refer to shrink_clause on page 12-36 in the documentation on CREATE
TABLE.

CACHE | NOCACHE
For data that will be accessed frequently, CACHE specifies that the blocks retrieved for
this table are placed at the most recently used end of the LRU list in the buffer cache
when a full table scan is performed. This attribute is useful for small lookup tables.
NOCACHE specifies that the blocks are placed at the least recently used end of the LRU
list. Refer to "CACHE | NOCACHE | CACHE READS" on page 16-62 in the
documentation on CREATE TABLE for more information about this clause.

alter_iot_clauses
Use the alter_iot_clauses to change the characteristics of an index-organized
materialized view. The keywords and parameters of the components of the alter_
iot_clauses have the same semantics as in ALTER TABLE, with the restrictions that
follow.
You cannot specify the
mapping_table_clause or the key_compression clause of the index_org_table_
clause.
Restrictions on Altering Index-Organized Materialized Views

index_org_table_clause on page 15-14 of CREATE
MATERIALIZED VIEW for information on creating an index-organized
materialized view
See Also:

USING INDEX Clause
Use this clause to change the value of INITRANS and STORAGE parameters for the index
Oracle Database uses to maintain the materialized view data.
Restriction on the USING INDEX clause You cannot specify the PCTUSED or PCTFREE
parameters in this clause.

MODIFY scoped_table_ref_constraint
Use the MODIFY scoped_table_ref_constraint clause to rescope a REF column or
attribute to a new table or to an alias for a new column.

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-11

ALTER MATERIALIZED VIEW

You can rescope only one REF column or
attribute in each ALTER MATERIALIZED VIEW statement, and this must be the only clause
in this statement.

Restrictions on Rescoping REF Columns

alter_mv_refresh
Use the alter_mv_refresh clause to change the default method and mode and the
default times for automatic refreshes. If the contents of the master tables of a
materialized view are modified, then the data in the materialized view must be
updated to make the materialized view accurately reflect the data currently in its
master table(s). This clause lets you schedule the times and specify the method and
mode for Oracle Database to refresh the materialized view.
Note: This clause only sets the default refresh options. For
instructions on actually implementing the refresh, refer to Oracle
Database Advanced Replication and Oracle Database Data Warehousing
Guide.

FAST Clause
Specify FAST for the incremental refresh method, which performs the refresh according
to the changes that have occurred to the master tables. The changes are stored either in
the materialized view log associated with the master table (for conventional DML
changes) or in the direct loader log (for direct-path INSERT operations).
For both conventional DML changes and for direct-path INSERT operations, other
conditions may restrict the eligibility of a materialized view for fast refresh.
See Also:
■

■

■

Oracle Database Advanced Replication for restrictions on fast refresh
in replication environments
Oracle Database Data Warehousing Guide for restrictions on fast
refresh in data warehouse environments
"Automatic Refresh: Examples" on page 11-16

Restrictions on FAST Refresh
■

■

■

FAST refresh is subject to the following restrictions:

When you specify FAST refresh at create time, Oracle Database verifies that the
materialized view you are creating is eligible for fast refresh. When you change the
refresh method to FAST in an ALTER MATERIALIZED VIEW statement, Oracle Database
does not perform this verification. If the materialized view is not eligible for fast
refresh, then Oracle Database returns an error when you attempt to refresh this
view.
Materialized views are not eligible for fast refresh if the defining query contains an
analytic function or the XMLTable function.
You cannot fast refresh a materialized view if any of its columns is encrypted.
See Also:

"Analytic Functions" on page 5-11

COMPLETE Clause
Specify COMPLETE for the complete refresh method, which is implemented by executing
the defining query of the materialized view. If you specify a complete refresh, then
Oracle Database performs a complete refresh even if a fast refresh is possible.

11-12 Oracle Database SQL Language Reference

ALTER MATERIALIZED VIEW

See Also:

"Complete Refresh: Example" on page 11-16

FORCE Clause
Specify FORCE if, when a refresh occurs, you want Oracle Database to perform a fast
refresh if one is possible or a complete refresh otherwise.
ON COMMIT Clause
Specify ON COMMIT if you want a refresh to occur whenever Oracle Database commits a
transaction that operates on a master table of the materialized view.
You cannot specify both ON COMMIT and ON DEMAND. If you specify ON COMMIT, then you
cannot also specify START WITH or NEXT.
Restriction on ON COMMIT This clause is supported only for materialized join views

and single-table materialized aggregate views.
Oracle Database Advanced Replication and Oracle Database
Data Warehousing Guide

See Also:

ON DEMAND Clause
Specify ON DEMAND if you want the materialized view to be refreshed on demand by
calling one of the three DBMS_MVIEW refresh procedures. If you omit both ON COMMIT and
ON DEMAND, then ON DEMAND is the default.
You cannot specify both ON COMMIT and ON DEMAND. START WITH and NEXT take
precedence over ON DEMAND. Therefore, in most circumstances it is not meaningful to
specify ON DEMAND when you have specified START WITH or NEXT.
See Also:
■

■

Oracle Database PL/SQL Packages and Types Reference for
information on these procedures
Oracle Database Data Warehousing Guide on the types of
materialized views you can create by specifying REFRESH ON
DEMAND

START WITH Clause
Specify START WITH date to indicate a date for the first automatic refresh time.
NEXT Clause
Specify NEXT to indicate a date expression for calculating the interval between
automatic refreshes.
Both the START WITH and NEXT values must evaluate to a time in the future. If you omit
the START WITH value, then Oracle Database determines the first automatic refresh time
by evaluating the NEXT expression with respect to the creation time of the materialized
view. If you specify a START WITH value but omit the NEXT value, then Oracle Database
refreshes the materialized view only once. If you omit both the START WITH and NEXT
values, or if you omit the alter_mv_refresh entirely, then Oracle Database does not
automatically refresh the materialized view.
WITH PRIMARY KEY Clause
Specify WITH PRIMARY KEY to change a rowid materialized view to a primary key
materialized view. Primary key materialized views allow materialized view master

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-13

ALTER MATERIALIZED VIEW

tables to be reorganized without affecting the ability of the materialized view to
continue to fast refresh.
For you to specify this clause, the master table must contain an enabled primary key
constraint and must have defined on it a materialized view log that logs primary key
information.
See Also:
■

■

Oracle Database Advanced Replication for detailed information about
primary key materialized views
"Primary Key Materialized View: Example" on page 11-17

USING ROLLBACK SEGMENT Clause
This clause is not valid if your database is in automatic undo mode, because in that
mode Oracle Database uses undo tablespaces instead of rollback segments. Oracle
strongly recommends that you use automatic undo mode. This clause is supported for
backward compatibility with replication environments containing older versions of
Oracle Database that still use rollback segments.
For complete information on this clause, refer to CREATE MATERIALIZED VIEW ... "USING
ROLLBACK SEGMENT Clause" on page 15-19.

USING ... CONSTRAINTS Clause
This clause has the same semantics in CREATE MATERIALIZED VIEW and ALTER
MATERIALIZED VIEW statements. For complete information, refer to "USING ...
CONSTRAINTS Clause" on page 15-20 in the documentation on CREATE MATERIALIZED
VIEW.

QUERY REWRITE Clause
Use this clause to determine whether the materialized view is eligible to be used for
query rewrite.
ENABLE Clause
Specify ENABLE to enable the materialized view for query rewrite.
See Also:

"Enabling Query Rewrite: Example" on page 11-17

Restrictions on Enabling Materialized Views Enabling materialized views is subject
to the following restrictions:
■

■

■

If the materialized view is in an invalid or unusable state, then it is not eligible for
query rewrite in spite of the ENABLE mode.
You cannot enable query rewrite if the materialized view was created totally or in
part from a view.
You can enable query rewrite only if all user-defined functions in the materialized
view are DETERMINISTIC.
See Also:

■

CREATE FUNCTION on page 14-58

You can enable query rewrite only if expressions in the statement are repeatable.
For example, you cannot include CURRENT_TIME or USER.

11-14 Oracle Database SQL Language Reference

ALTER MATERIALIZED VIEW

See Also: Oracle Database Data Warehousing Guide for more
information on query rewrite

DISABLE Clause
Specify DISABLE if you do not want the materialized view to be eligible for use by
query rewrite. If a materialized view is in the invalid state, then it is not eligible for use
by query rewrite, whether or not it is disabled. However, a disabled materialized view
can be refreshed.

COMPILE
Specify COMPILE to explicitly revalidate a materialized view. If an object upon which
the materialized view depends is dropped or altered, then the materialized view
remains accessible, but it is invalid for query rewrite. You can use this clause to
explicitly revalidate the materialized view to make it eligible for query rewrite.
If the materialized view fails to revalidate, then it cannot be refreshed or used for
query rewrite.
See Also:

"Compiling a Materialized View: Example" on page 11-17

CONSIDER FRESH
This clause lets you manage the staleness state of a materialized view after changes
have been made to its master tables. CONSIDER FRESH directs Oracle Database to
consider the materialized view fresh and therefore eligible for query rewrite in the
TRUSTED or STALE_TOLERATED modes.

The CONSIDER FRESH clause also directs Oracle Database to
no longer apply any rows in a materialized view log or Partition
Change Tracking changes to the materialized view prior to the
issuance of the CONSIDER FRESH clause. In other words, the pending
changes will be ignored and deleted, not applied to the materialized
view. This may result in the materialized view containing more or less
data than the base table.
Caution:

Because Oracle Database cannot guarantee the freshness of the materialized view,
query rewrite in ENFORCED mode is not supported. This clause also sets the staleness
state of the materialized view to UNKNOWN. The staleness state is displayed in the
STALENESS column of the ALL_MVIEWS, DBA_MVIEWS, and USER_MVIEWS data dictionary
views.
A materialized view is stale if changes have been made to the contents of any of its
master tables. This clause directs Oracle Database to assume that the materialized view
is fresh and that no such changes have been made. Therefore, actual updates to those
tables pending refresh are purged with respect to the materialized view.
See Also:
■

■

Oracle Database Data Warehousing Guide for more information on
query rewrite and the implications of performing partition
maintenance operations on master tables
"CONSIDER FRESH: Example" on page 11-16

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-15

ALTER MATERIALIZED VIEW

Examples
11

The following statement changes the default refresh
method for the sales_by_month_by_state materialized view (created in "Creating
Materialized Aggregate Views: Example" on page 15-23) to FAST:
Automatic Refresh: Examples

ALTER MATERIALIZED VIEW sales_by_month_by_state
REFRESH FAST;

The next automatic refresh of the materialized view will be a fast refresh provided it is
a simple materialized view and its master table has a materialized view log that was
created before the materialized view was created or last refreshed.
Because the REFRESH clause does not specify START WITH or NEXT values, Oracle
Database will use the refresh intervals established by the REFRESH clause when the
sales_by_month_by_state materialized view was created or last altered.
The following statement establishes a new interval between automatic refreshes for the
sales_by_month_by_state materialized view:
ALTER MATERIALIZED VIEW sales_by_month_by_state
REFRESH NEXT SYSDATE+7;

Because the REFRESH clause does not specify a START WITH value, the next automatic
refresh occurs at the time established by the START WITH and NEXT values specified
when the sales_by_month_by_state materialized view was created or last altered.
At the time of the next automatic refresh, Oracle Database refreshes the materialized
view, evaluates the NEXT expression SYSDATE+7 to determine the next automatic refresh
time, and continues to refresh the materialized view automatically once a week.
Because the REFRESH clause does not explicitly specify a refresh method, Oracle
Database continues to use the refresh method specified by the REFRESH clause of the
CREATE MATERIALIZED VIEW or most recent ALTER MATERIALIZED VIEW statement.
The following statement instructs Oracle Database that
materialized view sales_by_month_by_state should be considered fresh. This
statement allows sales_by_month_by_state to be eligible for query rewrite in TRUSTED
mode even after you have performed partition maintenance operations on the master
tables of sales_by_month_by_state:
CONSIDER FRESH: Example

ALTER MATERIALIZED VIEW sales_by_month_by_state CONSIDER FRESH;

As a result of the preceding statement, any partition maintenance operations that were
done to the base table since the last refresh of the materialized view will not be applied
to the materialized view. For example, the add, drop, or change of data in a partition in
the base table will not be reflected in the materialized view if CONSIDER FRESH is used
before the next refresh of the materialized view. Refer CONSIDER FRESH on
page 11-15 for more information.
See Also: "Splitting Table Partitions: Examples" on page 12-82 for a
partitioning maintenance example that would require this ALTER
MATERIALIZED VIEW example

The following statement specifies a new refresh
method, a new NEXT refresh time, and a new interval between automatic refreshes of
the emp_data materialized view (created in "Periodic Refresh of Materialized Views:
Example" on page 15-24):
Complete Refresh: Example

ALTER MATERIALIZED VIEW emp_data

11-16 Oracle Database SQL Language Reference

ALTER MATERIALIZED VIEW

REFRESH COMPLETE
START WITH TRUNC(SYSDATE+1) + 9/24
NEXT SYSDATE+7;

The START WITH value establishes the next automatic refresh for the materialized view
to be 9:00 a.m. tomorrow. At that point, Oracle Database performs a complete refresh
of the materialized view, evaluates the NEXT expression, and subsequently refreshes the
materialized view every week.
Enabling Query Rewrite: Example The following statement enables query rewrite on
the materialized view emp_data and implicitly revalidates it:
ALTER MATERIALIZED VIEW emp_data
ENABLE QUERY REWRITE;

The following statement changes the rowid
materialized view order_data (created in "Creating Rowid Materialized Views:
Example" on page 15-24) to a primary key materialized view. This example requires
that you have already defined a materialized view log with a primary key on order_
data.

Primary Key Materialized View: Example

ALTER MATERIALIZED VIEW order_data
REFRESH WITH PRIMARY KEY;

Compiling a Materialized View: Example

The following statement revalidates the

materialized view store_mv:
ALTER MATERIALIZED VIEW order_data COMPILE;

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-17

ALTER MATERIALIZED VIEW LOG

ALTER MATERIALIZED VIEW LOG
Purpose
11

A materialized view log is a table associated with the master table of a materialized
view. Use the ALTER MATERIALIZED VIEW LOG statement to alter the storage
characteristics or type of an existing materialized view log.
The keyword SNAPSHOT is supported in place of MATERIALIZED
VIEW for backward compatibility.
Note:

See Also:
■

■

■

CREATE MATERIALIZED VIEW LOG on page 15-27 for
information on creating a materialized view log
ALTER MATERIALIZED VIEW on page 11-3 for more information
on materialized views, including refreshing them
CREATE MATERIALIZED VIEW on page 15-4 for a description of
the various types of materialized views

Prerequisites
11

You must be the owner of the master table or you must have the SELECT privilege on
the master table and the ALTER privilege on the materialized view log.
Oracle Database Advanced Replication for detailed
information about the prerequisites for ALTER MATERIALIZED VIEW LOG

See Also:

Syntax
11

alter_materialized_view_log::=
FORCE
ALTER

MATERIALIZED

VIEW

LOG

schema

.

ON

table

physical_attributes_clause
add_mv_log_column_clause
alter_table_partitioning
parallel_clause
logging_clause
allocate_extent_clause
shrink_clause
move_mv_log_clause
CACHE
NOCACHE

mv_log_augmentation

mv_log_purge_clause
;

11-18 Oracle Database SQL Language Reference

ALTER MATERIALIZED VIEW LOG

(physical_attributes_clause::= on page 11-19, add_mv_log_column_clause::= on page 11-19,
alter_table_partitioning::= on page 12-19 (in ALTER TABLE), parallel_clause::= on
page 11-19, logging_clause::= on page 8-38, allocate_extent_clause::= on page 11-19,
shrink_clause::= on page 11-19, move_mv_log_clause::= on page 11-19, mv_log_
augmentation::= on page 11-20,mv_log_purge_clause::= on page 11-20.)
physical_attributes_clause::=
PCTFREE

integer

PCTUSED

integer

INITRANS

integer

storage_clause

storage_clause::= on page 8-50
add_mv_log_column_clause::=
ADD

(

column

)

allocate_extent_clause::=
SIZE
(

size_clause

DATAFILE
INSTANCE

ALLOCATE

’

filename

’

)

integer

EXTENT

(size_clause::= on page 8-47)
shrink_clause::=
COMPACT
SHRINK

CASCADE

SPACE

move_mv_log_clause::=
parallel_clause
MOVE

segment_attributes_clause

parallel_clause::=
NOPARALLEL
integer
PARALLEL

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-19

ALTER MATERIALIZED VIEW LOG

mv_log_augmentation::=
,
OBJECT

,

ID

PRIMARY

(

KEY

column

)

ROWID
new_values_clause

SEQUENCE

ADD

,
(

column

)

(new_values_clause::= on page 11-20.
new_values_clause::=
INCLUDING
NEW

VALUES

EXCLUDING

mv_log_purge_clause::=
SYNCHRONOUS
ASYNCHRONOUS
IMMEDIATE
NEXT

datetime_expr

REPEAT

PURGE
START

WITH

START

WITH

INTERVAL

interval_expr

datetime_expr
datetime_expr

NEXT
REPEAT

datetime_expr
INTERVAL

interval_expr

Semantics
11

FORCE
If you specify FORCE and any items specified with the ADD clause have already been
specified for the materialized view log, then Oracle Database does not return an error,
but silently ignores the existing elements and adds to the materialized view log any
items that do not already exist in the log. Likewise, if you specify INCLUDING NEW
VALUES and that attribute has already been specified for the materialized view log,
Oracle Database ignores the redundancy and does not return an error.

schema
Specify the schema containing the master table. If you omit schema, then Oracle
Database assumes the materialized view log is in your own schema.

table
Specify the name of the master table associated with the materialized view log to be
altered.

11-20 Oracle Database SQL Language Reference

ALTER MATERIALIZED VIEW LOG

physical_attributes_clause
The physical_attributes_clause lets you change the value of the PCTFREE, PCTUSED,
and INITRANS parameters and the storage characteristics for the materialized view log,
the partition, the overflow data segment, or the default characteristics of a partitioned
materialized view log.
Restriction on Materialized View Log Physical Attributes You cannot use the

storage_clause to modify extent parameters if the materialized view log resides in a
locally managed tablespace. Refer to CREATE TABLE on page 16-6 for a description of
these parameters.

add_mv_log_column_clause
When you add a column to the master table of the materialized view log, the database
does not automatically add a column to the materialized view log. Therefore, use this
clause to add a column to the materialized view log. Oracle Database will encrypt the
newly added column if the corresponding column of the master table is encrypted.

alter_table_partitioning
The syntax and general functioning of the partitioning clauses is the same as described
for the ALTER TABLE statement. Refer to alter_table_partitioning on page 12-57 in the
documentation for ALTER TABLE.
Restrictions on Altering Materialized View Log Partitions Altering materialized view
log partitions is subject to the following restrictions:
■

■

You cannot use the LOB_storage_clause or modify_LOB_storage_clause when
modifying partitions of a materialized view log.
If you attempt to drop, truncate, or exchange a materialized view log partition,
then Oracle Database raises an error.

parallel_clause
The parallel_clause lets you specify whether parallel operations will be supported
for the materialized view log.
For complete information on this clause, refer to parallel_clause on page 16-63 in the
documentation on CREATE TABLE.

logging_clause
Specify the logging attribute of the materialized view log. Refer to the logging_clause
on page 8-38 for a full description of this clause.

allocate_extent_clause
Use the allocate_extent_clause to explicitly allocate a new extent for the
materialized view log. Refer to allocate_extent_clause on page 8-2 for a full description
of this clause.

shrink_clause
Use this clause to compact the materialized view log segments. For complete
information on this clause, refer to shrink_clause on page 12-36 in the documentation
on CREATE TABLE.

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-21

ALTER MATERIALIZED VIEW LOG

move_mv_log_clause
Use the MOVE clause to move the materialized view log table to a different tablespace,
to change other segment or storage attributes of the materialized view log, or to
change the parallelism of the materialized view log.
Restriction on Moving Materialized View Logs The ENCRYPT clause of the storage_

clause of segment_attributes is not valid for materialized view logs.

CACHE | NOCACHE Clause
For data that will be accessed frequently, CACHE specifies that the blocks retrieved for
this log are placed at the most recently used end of the LRU list in the buffer cache
when a full table scan is performed. This attribute is useful for small lookup tables.
NOCACHE specifies that the blocks are placed at the least recently used end of the LRU
list. Refer to "CACHE | NOCACHE | CACHE READS" on page 16-62 in the
documentation on CREATE TABLE for more information about this clause.

mv_log_augmentation
Use the ADD clause to augment the materialized view log so that it records the primary
key values, rowid values, object ID values, or a sequence when rows in the
materialized view master table are changed. This clause can also be used to record
additional columns.
To stop recording any of this information, you must first drop the materialized view
log and then re-create it. Dropping the materialized view log and then re-creating it
forces a complete refresh for each of the existing materialized views that depend on
the master table on its next refresh.
You can specify only one
PRIMARY KEY, one ROWID, one OBJECT ID, one SEQUENCE, and each column in the column
list once for each materialized view log. You can specify only a single occurrence of
PRIMARY KEY, ROWID, OBJECT ID, SEQUENCE, and column list within this ALTER statement.
Also, if any of these values was specified at create time (either implicitly or explicitly),
you cannot specify that value in this ALTER statement unless you use the FORCE option.
Restriction on Augmenting Materialized View Logs

OBJECT ID Specify OBJECT ID if you want the appropriate object identifier of all rows
that are changed to be recorded in the materialized view log.
Restriction on the OBJECT ID clause You can specify OBJECT ID only for logs on

object tables, and you cannot specify it for storage tables.
PRIMARY KEY Specify PRIMARY KEY if you want the primary key values of all rows
that are changed to be recorded in the materialized view log.
ROWID Specify ROWID if you want the rowid values of all rows that are changed to be
recorded in the materialized view log.
SEQUENCE Specify SEQUENCE to indicate that a sequence value providing additional
ordering information should be recorded in the materialized view log.
column Specify the additional columns whose values you want to be recorded in the
materialized view log for all rows that are changed. Typically these columns are filter
columns (non-primary-key columns referenced by subquery materialized views) and
join columns (non-primary-key columns that define a join in the WHERE clause of the
subquery).

11-22 Oracle Database SQL Language Reference

ALTER MATERIALIZED VIEW LOG

See Also:
■

■

■

CREATE MATERIALIZED VIEW on page 15-4 for details on
explicit and implicit inclusion of materialized view log values
Oracle Database Advanced Replication for more information about
filter columns and join columns
"Rowid Materialized View Log: Example" on page 11-23

NEW VALUES Clause
The NEW VALUES clause lets you specify whether Oracle Database saves both old and
new values for update DML operations in the materialized view log. The value you set
in this clause applies to all columns in the log, not only to columns you may have
added in this ALTER MATERIALIZED VIEW LOG statement.
INCLUDING Specify INCLUDING to save both new and old values in the log. If this log
is for a table on which you have a single-table materialized aggregate view, and if you
want the materialized view to be eligible for fast refresh, then you must specify
INCLUDING.
EXCLUDING Specify EXCLUDING to disable the recording of new values in the log. You
can use this clause to avoid the overhead of recording new values.

If you have a fast-refreshable single-table materialized aggregate view defined on this
table, then do not specify EXCLUDING NEW VALUES unless you first change the refresh
mode of the materialized view to something other than FAST.
See Also: "Materialized View Log EXCLUDING NEW VALUES:
Example" on page 11-24

mv_log_purge_clause
Use this clause alter the purge attributes of the materialized view log in the following
ways:
■

Change the purge from IMMEDIATE SYNCHRONOUS to IMMEDIATE ASYNCHRONOUS or
from IMMEDIATE ASYNCHRONOUS to IMMEDIATE SYNCHRONOUS

■

Change the purge from IMMEDIATE to scheduled or from scheduled to IMMEDIATE

■

Specify a new start time and a new next time and interval

If you are altering purge from scheduled to IMMEDIATE, then the scheduled purged job
associated with that materialized view log is dropped. If you are altering purge from
IMMEDIATE to scheduled, then a purge job is created with the attributes provided. If
you are altering scheduled purge attributes, then only those attributes specified will be
changed in the scheduler purge job.
You must specify FORCE if you are altering log purge to its current state (that is, you are
not making any change), unless you are changing scheduled purge attributes.
To learn whether the purge time or interval has already been set for this materialized
view log, query the *_MVIEW_LOGS data dictionary views. See the CREATE MATERIALIZED
VIEW LOG clause mv_log_purge_clause on page 15-33 for the full semantics of this clause.

Examples
11

Rowid Materialized View Log: Example The following statement alters an existing
primary key materialized view log to also record rowid information:

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-23

ALTER MATERIALIZED VIEW LOG

ALTER MATERIALIZED VIEW LOG ON order_items ADD ROWID;

Materialized View Log EXCLUDING NEW VALUES: Example The following
statement alters the materialized view log on hr.employees by adding a filter column
and excluding new values. Any materialized aggregate views that use this log will no
longer be fast refreshable. However, if fast refresh is no longer needed, this action
avoids the overhead of recording new values:
ALTER MATERIALIZED VIEW LOG ON employees
ADD (commission_pct)
EXCLUDING NEW VALUES;

11-24 Oracle Database SQL Language Reference

ALTER OPERATOR

ALTER OPERATOR
Purpose
11

Use the ALTER OPERATOR statement to add bindings to, drop bindings from, or compile
an existing operator.
See Also:

CREATE OPERATOR on page 15-35

Prerequisites
11

The operator must already have been created by a previous CREATE OPERATOR
statement. The operator must be in your own schema or you must have the ALTER ANY
OPERATOR system privilege. You must have the EXECUTE object privilege on the
operators and functions referenced in the ALTER OPERATOR statement.

Syntax
11

alter_operator::=
schema
ALTER

add binding_clause

.

OPERATOR

operator

drop_binding_clause

;

COMPILE

(add_binding_clause::= on page 11-25, drop_binding_clause::= on page 11-26)
add_binding_clause::=
,
ADD

BINDING

(

parameter_type

)

RETURN

(

return_type

)

implementation_clause
using_function_clause

(implementation_clause::= on page 11-25, using_function_clause::= on page 11-26)
implementation_clause::=
,
,
ANCILLARY

TO

primary_operator

(

parameter_type

)

context_clause

(context_clause::= on page 11-26)

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-25

ALTER OPERATOR

context_clause::=
COMPUTE
WITH

INDEX

CONTEXT

WITH

COLUMN

,

SCAN

CONTEXT

ANCILLARY

DATA

implementation_type

CONTEXT

using_function_clause::=
package
schema

.

type

.
.

USING

function_name

drop_binding_clause::=
,
DROP

BINDING

(

parameter_type

FORCE
)

Semantics
11

schema
Specify the schema containing the operator. If you omit this clause, then Oracle
Database assumes the operator is in your own schema.

operator
Specify the name of the operator to be altered.

add_binding_clause
Use this clause to add an operator binding and specify its parameter data types and
return type. The signature must be different from the signature of any existing binding
for this operator.
If a binding of an operator is associated with an indextype and you add another
binding to the operator, then Oracle Database does not automatically associate the new
binding with the indextype. If you want to make such an association, then you must
issue an explicit ALTER INDEXTYPE ... ADD OPERATOR statement.
implementation_clause
This clause has the same semantics in CREATE OPERATOR and ALTER OPERATOR
statements. For full information, refer to implementation_clause on page 15-36 in the
documentation on CREATE OPERATOR.
context_clause
This clause has the same semantics in CREATE OPERATOR and ALTER OPERATOR
statements. For full information, refer to context_clause on page 15-37 in the
documentation on CREATE OPERATOR.

11-26 Oracle Database SQL Language Reference

ALTER OPERATOR

using_function_clause
This clause has the same semantics in CREATE OPERATOR and ALTER OPERATOR
statements. For full information, refer to using_function_clause on page 15-37 in the
documentation on CREATE OPERATOR.

drop_binding_clause
Use this clause to specify the list of parameter data types of the binding you want to
drop from the operator. You must specify FORCE if the binding has any dependent
objects, such as an indextype or an ancillary operator binding. If you specify FORCE,
then Oracle Database marks INVALID all objects that are dependent on the binding. The
dependent objects are revalidated the next time they are referenced in a DDL or DML
statement or a query.
You cannot use this clause to drop the only binding associated with this operator.
Instead you must use the DROP OPERATOR statement. Refer to DROP OPERATOR on
page 17-59 for more information.

COMPILE
Specify COMPILE to cause Oracle Database to recompile the operator.

Examples
11

Compiling a User-defined Operator: Example The following example compiles the
operator eq_op (which was created in "Creating User-Defined Operators: Example" on
page 15-37):
ALTER OPERATOR eq_op COMPILE;

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-27

ALTER OUTLINE

ALTER OUTLINE
Purpose
11

Oracle strongly recommends that you use SQL plan
management for new applications. SQL plan management creates SQL
plan baselines, which offer superior SQL performance stability
compared with stored outlines.

Note:

You can migrate existing stored outlines to SQL plan baselines by
using the MIGRATE_STORED_OUTLINE function of the DBMS_SPM package
or Enterprise Manager DB Control. When the migration is complete,
the stored outlines are marked as migrated and can be removed. You
can drop all migrated stored outlines on your system by using the
DROP_MIGRATED_STORED_OUTLINE function of the DBMS_SPM package.
See Also: Oracle Database Performance Tuning Guide for more
information about SQL plan management and Oracle Database PL/SQL
Packages and Types Reference for information about the DBMS_SPM
package
Use the ALTER OUTLINE statement to rename a stored outline, reassign it to a different
category, or regenerate it by compiling the outline's SQL statement and replacing the
old outline data with the outline created under current conditions.
CREATE OUTLINE on page 15-38 and Oracle Database
Performance Tuning Guide for more information on outlines

See Also:

Prerequisites
11

To modify an outline, you must have the ALTER ANY OUTLINE system privilege.

Syntax
11

alter_outline::=

ALTER

PUBLIC

REBUILD

PRIVATE

RENAME

TO

CHANGE

CATEGORY

OUTLINE

outline

new_outline_name
TO

new_category_name

ENABLE
DISABLE

Semantics
11

PUBLIC | PRIVATE
Specify PUBLIC if you want to modify the public version of this outline. This is the
default.

11-28 Oracle Database SQL Language Reference

;

ALTER OUTLINE

Specify PRIVATE if you want to modify an outline that is private to the current session
and whose data is stored in the current parsing schema.

outline
Specify the name of the outline to be modified.

REBUILD
Specify REBUILD to regenerate the execution plan for outline using current conditions.
See Also:

"Rebuilding an Outline: Example" on page 11-29

RENAME TO Clause
Use the RENAME TO clause to specify an outline name to replace outline.

CHANGE CATEGORY TO Clause
Use the CHANGE CATEGORY TO clause to specify the name of the category into which the
outline will be moved.

ENABLE | DISABLE
Use this clause to selectively enable or disable this outline. Outlines are enabled by
default. The DISABLE keyword lets you disable one outline without affecting the use of
other outlines.

Examples
11

Rebuilding an Outline: Example The following statement regenerates a stored

outline called salaries by compiling the text of the outline and replacing the old
outline data with the outline created under current conditions.
ALTER OUTLINE salaries REBUILD;

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-29

ALTER PACKAGE

ALTER PACKAGE
Purpose
11

Packages are defined using PL/SQL. Therefore, this section provides some general
information but refers to Oracle Database PL/SQL Language Reference for details of
syntax and semantics.
Use the ALTER PACKAGE statement to explicitly recompile a package specification, body,
or both. Explicit recompilation eliminates the need for implicit run-time recompilation
and prevents associated run-time compilation errors and performance overhead.
Because all objects in a package are stored as a unit, the ALTER PACKAGE statement
recompiles all package objects together. You cannot use the ALTER PROCEDURE statement
or ALTER FUNCTION statement to recompile individually a procedure or function that is
part of a package.
This statement does not change the declaration or definition of
an existing package. To redeclare or redefine a package, use the
CREATE PACKAGE or the CREATE PACKAGE BODY on page 15-42
statement with the OR REPLACE clause.

Note:

Prerequisites
11

For you to modify a package, the package must be in your own schema or you must
have ALTER ANY PROCEDURE system privilege.

Syntax
11

alter_package::=
schema
ALTER

PACKAGE

.
package

package_compile_clause

(package_compile_clause: See Oracle Database PL/SQL Language Reference for the
syntax of this clause.)

Semantics
11

schema
Specify the schema containing the package. If you omit schema, then Oracle Database
assumes the package is in your own schema.

package
Specify the name of the package to be recompiled.

package_compile_clause
See Oracle Database PL/SQL Language Reference for the syntax and semantics of this
clause and for complete information on creating and compiling packages.

11-30 Oracle Database SQL Language Reference

ALTER PROCEDURE

ALTER PROCEDURE
Purpose
11

Packages are defined using PL/SQL. Therefore, this section provides some general
information but refers to Oracle Database PL/SQL Language Reference for details of
syntax and semantics.
Use the ALTER PROCEDURE statement to explicitly recompile a standalone stored
procedure. Explicit recompilation eliminates the need for implicit run-time
recompilation and prevents associated run-time compilation errors and performance
overhead.
To recompile a procedure that is part of a package, recompile the entire package using
the ALTER PACKAGE statement (see ALTER PACKAGE on page 11-30).
Note: This statement does not change the declaration or definition of
an existing procedure. To redeclare or redefine a procedure, use the
CREATE PROCEDURE statement with the OR REPLACE clause (see CREATE
PROCEDURE on page 15-48).

The ALTER PROCEDURE statement is quite similar to the ALTER FUNCTION statement. Refer
to ALTER FUNCTION on page 10-77 for more information.

Prerequisites
11

The procedure must be in your own schema or you must have ALTER ANY PROCEDURE
system privilege.

Syntax
11

alter_procedure::=
schema
ALTER

PROCEDURE

.
procedure

procedure_compile_clause

(procedure_compile_clause: See Oracle Database PL/SQL Language Reference for the
syntax of this clause.)

Semantics
11

schema
Specify the schema containing the procedure. If you omit schema, then Oracle
Database assumes the procedure is in your own schema.

procedure
Specify the name of the procedure to be recompiled.

procedure_compile_clause
See Oracle Database PL/SQL Language Reference for the syntax and semantics of this
clause and for complete information on creating and compiling procedures.

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-31

ALTER PROFILE

ALTER PROFILE
Purpose
11

Use the ALTER PROFILE statement to add, modify, or remove a resource limit or
password management parameter in a profile.
Changes made to a profile with an ALTER PROFILE statement affect users only in their
subsequent sessions, not in their current sessions.
CREATE PROFILE on page 15-50 for information on
creating a profile

See Also:

Prerequisites
11

You must have the ALTER PROFILE system privilege.

Syntax
11

alter_profile::=
resource_parameters
ALTER

PROFILE

profile

LIMIT

;
password_parameters

(resource_parameters::= on page 11-32, password_parameters::= on page 11-33)
resource_parameters::=
SESSIONS_PER_USER
CPU_PER_SESSION
CPU_PER_CALL
integer
CONNECT_TIME
UNLIMITED
IDLE_TIME
DEFAULT
LOGICAL_READS_PER_SESSION
LOGICAL_READS_PER_CALL
COMPOSITE_LIMIT
size_clause
PRIVATE_SGA

UNLIMITED
DEFAULT

(size_clause::= on page 8-47)

11-32 Oracle Database SQL Language Reference

ALTER PROFILE

password_parameters::=
FAILED_LOGIN_ATTEMPTS
PASSWORD_LIFE_TIME
expr
PASSWORD_REUSE_TIME
UNLIMITED
PASSWORD_REUSE_MAX
DEFAULT
PASSWORD_LOCK_TIME
PASSWORD_GRACE_TIME
function
PASSWORD_VERIFY_FUNCTION

NULL
DEFAULT

Semantics
11

The keywords, parameters, and clauses in the ALTER PROFILE statement all have the
same meaning as in the CREATE PROFILE statement.
You cannot remove a limit from the DEFAULT profile.

11

Refer to CREATE PROFILE on page 15-50 and to the examples in the next section for
more information.

Examples
Making a Password Unavailable: Example The following statement makes the

password of the new_profile profile (created in "Creating a Profile: Example" on
page 15-54) unavailable for reuse for 90 days:
ALTER PROFILE new_profile
LIMIT PASSWORD_REUSE_TIME 90
PASSWORD_REUSE_MAX UNLIMITED;

Setting Default Password Values: Example The following statement defaults the

PASSWORD_REUSE_TIME value of the app_user profile (created in "Setting Profile
Resource Limits: Example" on page 15-54) to its defined value in the DEFAULT profile:
ALTER PROFILE app_user
LIMIT PASSWORD_REUSE_TIME DEFAULT
PASSWORD_REUSE_MAX UNLIMITED;

Limiting Login Attempts and Password Lock Time: Example The following
statement alters profile app_user with FAILED_LOGIN_ATTEMPTS set to 5 and PASSWORD_
LOCK_TIME set to 1:
ALTER PROFILE app_user LIMIT
FAILED_LOGIN_ATTEMPTS 5
PASSWORD_LOCK_TIME 1;

This statement causes any user account to which the app_user profile is assigned to
become locked for one day after five consecutive unsuccessful login attempts.
Changing Password Lifetime and Grace Period: Example The following statement
modifies the profile app_user2 PASSWORD_LIFE_TIME to 90 days and PASSWORD_GRACE_
TIME to 5 days:
SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-33

ALTER PROFILE

ALTER PROFILE app_user2 LIMIT
PASSWORD_LIFE_TIME 90
PASSWORD_GRACE_TIME 5;

This statement defines a new limit of 5
concurrent sessions for the app_user profile:

Limiting Concurrent Sessions: Example

ALTER PROFILE app_user LIMIT SESSIONS_PER_USER 5;

If the app_user profile does not currently define a limit for SESSIONS_PER_USER, then
the preceding statement adds the limit of 5 to the profile. If the profile already defines
a limit, then the preceding statement redefines it to 5. Any user assigned the app_user
profile is subsequently limited to 5 concurrent sessions.
Removing Profile Limits: Example This statement removes the IDLE_TIME limit from
the app_user profile:
ALTER PROFILE app_user LIMIT IDLE_TIME DEFAULT;

Any user assigned the app_user profile is subject in their subsequent sessions to the
IDLE_TIME limit defined in the DEFAULT profile.
Limiting Profile Idle Time: Example This statement defines a limit of 2 minutes of
idle time for the DEFAULT profile:
ALTER PROFILE default LIMIT IDLE_TIME

2;

This IDLE_TIME limit applies to these users:
■

Users who are not explicitly assigned any profile

■

Users who are explicitly assigned a profile that does not define an IDLE_TIME limit

This statement defines unlimited idle time for the app_user2 profile:
ALTER PROFILE app_user2 LIMIT IDLE_TIME UNLIMITED;

Any user assigned the app_user2 profile is subsequently permitted unlimited idle
time.

11-34 Oracle Database SQL Language Reference

ALTER RESOURCE COST

ALTER RESOURCE COST
Purpose
11

Use the ALTER RESOURCE COST statement to specify or change the formula by which
Oracle Database calculates the total resource cost used in a session.
Although Oracle Database monitors the use of other resources, only the four resources
shown in the syntax can contribute to the total resource cost for a session.
This statement lets you apply weights to the four resources. Oracle Database then
applies the weights to the value of these resources that were specified for a profile to
establish a formula for calculating total resource cost. You can limit this cost for a
session with the COMPOSITE_LIMIT parameter of the CREATE PROFILE statement. If the
resource cost of a session exceeds the limit, then Oracle Database aborts the session
and returns an error. If you use the ALTER RESOURCE COST statement to change the
weight assigned to each resource, then Oracle Database uses these new weights to
calculate the total resource cost for all current and subsequent sessions.
CREATE PROFILE on page 15-50 for information on all
resources and on establishing resource limits

See Also:

Prerequisites
11

You must have the ALTER RESOURCE COST system privilege.

Syntax
11

alter_resource_cost::=
CPU_PER_SESSION
CONNECT_TIME
ALTER

RESOURCE

COST

integer

;

LOGICAL_READS_PER_SESSION
PRIVATE_SGA

Semantics
11

Oracle Database calculates the total resource cost by first multiplying the amount of
each resource used in the session by the weight of the resource, and then summing the
products for all four resources. For any session, this cost is limited by the value of the
COMPOSITE_LIMIT parameter in the user's profile. Both the products and the total cost
are expressed in units called service units.

CPU_PER_SESSION
Use this keyword to apply a weight to the CPU_PER_SESSION resource.

CONNECT_TIME
Use this keyword to apply a weight to the CONNECT_TIME resource.

LOGICAL_READS_PER_SESSION
Use this clause to apply a weight to the LOGICAL_READS_PER_SESSION resource. Logical
reads include blocks read from both memory and disk.
SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-35

ALTER RESOURCE COST

PRIVATE_SGA
Use this clause to apply a weight to the PRIVATE_SGA resource. This limit applies only
if you are using shared server architecture and allocating private space in the SGA for
your session.

integer
Specify the weight of each resource. The weight that you assign to each resource
determines how much the use of that resource contributes to the total resource cost. If
you do not assign a weight to a resource, then the weight defaults to 0, and use of the
resource subsequently does not contribute to the cost. The weights you assign apply to
all subsequent sessions in the database.

Examples
11

Altering Resource Costs: Examples The following statement assigns weights to the

resources CPU_PER_SESSION and CONNECT_TIME:
ALTER RESOURCE COST
CPU_PER_SESSION 100
CONNECT_TIME
1;

The weights establish this cost formula for a session:
cost = (100 * CPU_PER_SESSION) + (1 * CONNECT_TIME)

In this example, the values of CPU_PER_SESSION and CONNECT_TIME are either values in
the DEFAULT profile or in the profile of the user of the session.
Because the preceding statement assigns no weight to the resources LOGICAL_READS_
PER_SESSION and PRIVATE_SGA, these resources do not appear in the formula.
If a user is assigned a profile with a COMPOSITE_LIMIT value of 500, then a session
exceeds this limit whenever cost exceeds 500. For example, a session using 0.04
seconds of CPU time and 101 minutes of elapsed time exceeds the limit. A session
using 0.0301 seconds of CPU time and 200 minutes of elapsed time also exceeds the
limit.
You can subsequently change the weights with another ALTER RESOURCE statement:
ALTER RESOURCE COST
LOGICAL_READS_PER_SESSION 2
CONNECT_TIME 0;

These new weights establish a new cost formula:
cost = (100 * CPU_PER_SESSION) + (2 * LOGICAL_READ_PER_SECOND)

where the values of CPU_PER_SESSION and LOGICAL_READS_PER_SECOND are either the
values in the DEFAULT profile or in the profile of the user of this session.
This ALTER RESOURCE COST statement changes the formula in these ways:
■

■

The statement omits a weight for the CPU_PER_SESSION resource. That resource
was already assigned a weight, so the resource remains in the formula with its
original weight.
The statement assigns a weight to the LOGICAL_READS_PER_SESSION resource, so
this resource now appears in the formula.

11-36 Oracle Database SQL Language Reference

ALTER RESOURCE COST

■

■

The statement assigns a weight of 0 to the CONNECT_TIME resource, so this resource
no longer appears in the formula.
The statement omits a weight for the PRIVATE_SGA resource. That resource was not
already assigned a weight, so the resource still does not appear in the formula.

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-37

ALTER ROLE

ALTER ROLE
Purpose
11

Use the ALTER ROLE statement to change the authorization needed to enable a role.
See Also:
■
■

CREATE ROLE on page 15-59 for information on creating a role
SET ROLE on page 19-61 for information on enabling or disabling
a role for your session

Prerequisites
11

You must either have been granted the role with the ADMIN OPTION or have ALTER ANY
ROLE system privilege.
Before you alter a role to IDENTIFIED GLOBALLY, you must:
■

Revoke all grants of roles identified externally to the role and

■

Revoke the grant of the role from all users, roles, and PUBLIC.

The one exception to this rule is that you should not revoke the role from the user who
is currently altering the role.

Syntax
11

alter_role::=
NOT

IDENTIFIED
BY

ALTER

ROLE

password

role

;
schema
IDENTIFIED

USING

.
package

EXTERNALLY
GLOBALLY

Semantics
11

The keywords, parameters, and clauses in the ALTER ROLE statement all have the same
meaning as in the CREATE ROLE statement.
Restriction on Altering a Role You cannot alter a NOT IDENTIFIED role to any of the

IDENTIFIED types if it is granted to another role.
Notes on Altering a Role The following notes apply when altering a role:
■
■

User sessions in which the role is already enabled are not affected.
If you change a role identified by password to an application role (with the USING
package clause), then password information associated with the role is lost. Oracle
Database will use the new authentication mechanism the next time the role is to be
enabled.

11-38 Oracle Database SQL Language Reference

ALTER ROLE

■

If you have the ALTER ANY ROLE system privilege and you change a role that is
IDENTIFIED GLOBALLY to IDENTIFIED BY password, IDENTIFIED EXTERNALLY, or NOT
IDENTIFIED, then Oracle Database grants you the altered role with the ADMIN
OPTION, as it would have if you had created the role identified nonglobally.

For more information, refer to CREATE ROLE on page 15-59 and to the examples that
follow.

Examples
11

The following statement changes the role
warehouse_user (created in "Creating a Role: Example" on page 15-60) to NOT
IDENTIFIED:
Changing Role Identification: Example

ALTER ROLE warehouse_user NOT IDENTIFIED;

This statement changes the password on the
dw_manager role (created in "Creating a Role: Example" on page 15-60) to data:
Changing a Role Password: Example
ALTER ROLE dw_manager
IDENTIFIED BY data;

Users granted the dw_manager role must subsequently use the new password data to
enable the role.
The following example changes the dw_manager role to
an application role using the hr.admin package:

Application Roles: Example

ALTER ROLE dw_manager IDENTIFIED USING hr.admin;

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-39

ALTER ROLLBACK SEGMENT

ALTER ROLLBACK SEGMENT
Oracle strongly recommends that you run your database in
automatic undo management mode instead of using rollback
segments. Do not use rollback segments unless you must do so for
compatibility with earlier versions of Oracle Database. Refer to Oracle
Database Administrator's Guide for information on automatic undo
management.

Note:

Purpose
11

Use the ALTER ROLLBACK SEGMENT statement to bring a rollback segment online or
offline, change its storage characteristics, or shrink it to an optimal or specified size.
This section assumes that your database is running in rollback undo mode (the UNDO_
MANAGEMENT initialization parameter is set to MANUAL or not set at all). If your database
is running in automatic undo mode (the UNDO_MANAGEMENT initialization parameter is
set to AUTO, which is the default), then user-created rollback segments are irrelevant.
See Also:
■

■

CREATE ROLLBACK SEGMENT on page 15-62 for information
on creating a rollback segment
Oracle Database Reference for information on the UNDO_MANAGEMENT
parameter

Prerequisites
11

You must have the ALTER ROLLBACK SEGMENT system privilege.

Syntax
11

alter_rollback_segment::=
ONLINE
OFFLINE
ALTER

ROLLBACK

SEGMENT

rollback_segment

;

storage_clause
TO

size_clause

SHRINK

(storage_clause on page 8-48, size_clause::= on page 8-47)

Semantics
11

rollback_segment
Specify the name of an existing rollback segment.

ONLINE
Specify ONLINE to bring the rollback segment online. When you create a rollback
segment, it is initially offline and not available for transactions. This clause brings the

11-40 Oracle Database SQL Language Reference

ALTER ROLLBACK SEGMENT

rollback segment online, making it available for transactions by your instance. You can
also bring a rollback segment online when you start your instance with the
initialization parameter ROLLBACK_SEGMENTS.
See Also: "Bringing a Rollback Segment Online: Example" on
page 11-42

OFFLINE
Specify OFFLINE to take the rollback segment offline.
■

■

If the rollback segment does not contain any information needed to roll back an
active transaction, then Oracle Database takes it offline immediately.
If the rollback segment does contain information for active transactions, then the
database makes the rollback segment unavailable for future transactions and takes
it offline after all the active transactions are committed or rolled back.

When the rollback segment is offline, it can be brought online by any instance.
To see whether a rollback segment is online or offline, query STATUS column of the data
dictionary view DBA_ROLLBACK_SEGS. Online rollback segments have a value of IN_USE.
Offline rollback segments have a value of AVAILABLE.
Restriction on Taking Rollback Segments Offline

You cannot take the SYSTEM

rollback segment offline.

storage_clause
Use the storage_clause to change the storage characteristics of the rollback segment.
You cannot change the value of INITIAL
parameter. If the rollback segment is in a locally managed tablespace, then the only
storage parameter you can change is OPTIMAL. If the rollback segment is in a
dictionary-managed tablespace, then the only storage parameters you can change are
NEXT, MINEXTENTS, MAXEXTENTS and OPTIMAL.
Restrictions on Rollback Segment Storage

See Also: storage_clause on page 8-48 for syntax and additional
information

SHRINK Clause
Specify SHRINK if you want Oracle Database to attempt to shrink the rollback segment
to an optimal or specified size. The success and amount of shrinkage depend on the
available free space in the rollback segment and how active transactions are holding
space in the rollback segment.
If you do not specify TO size_clause, then the size defaults to the OPTIMAL value of the
storage_clause of the CREATE ROLLBACK SEGMENT statement that created the rollback
segment. If OPTIMAL was not specified, then the size defaults to the MINEXTENTS value
of the storage_clause of the CREATE ROLLBACK SEGMENT statement.
Regardless of whether you specify TO size_clause:
■

■

The value to which Oracle Database shrinks the rollback segment is valid for the
execution of the statement. Thereafter, the size reverts to the OPTIMAL value of the
CREATE ROLLBACK SEGMENT statement.
The rollback segment cannot shrink to less than two extents.

To determine the actual size of a rollback segment after attempting to shrink it, query
the BYTES, BLOCKS, and EXTENTS columns of the DBA_SEGMENTS view.
SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-41

ALTER ROLLBACK SEGMENT

Restriction on Shrinking Rollback Segments In an Oracle Real Application Clusters

environment, you can shrink only rollback segments that are online to your instance.
See Also: size_clause on page 8-47 for information on that clause, and
"Resizing a Rollback Segment: Example" on page 11-42

Examples
11

The following examples use the rbs_one rollback segment, which was created in
"Creating a Rollback Segment: Example" on page 15-64.
Bringing a Rollback Segment Online: Example

This statement brings the rollback

segment rbs_one online:
ALTER ROLLBACK SEGMENT rbs_one ONLINE;

Resizing a Rollback Segment: Example

rbs_one:
ALTER ROLLBACK SEGMENT rbs_one
SHRINK TO 100M;

11-42 Oracle Database SQL Language Reference

This statement shrinks the rollback segment

ALTER SEQUENCE

ALTER SEQUENCE
Purpose
11

Use the ALTER SEQUENCE statement to change the increment, minimum and maximum
values, cached numbers, and behavior of an existing sequence. This statement affects
only future sequence numbers.
CREATE SEQUENCE on page 15-67 for additional
information on sequences

See Also:

Prerequisites
11

The sequence must be in your own schema, or you must have the ALTER object
privilege on the sequence, or you must have the ALTER ANY SEQUENCE system privilege.

Syntax
11

alter_sequence::=
INCREMENT

BY

MAXVALUE

integer

integer

NOMAXVALUE
MINVALUE
schema
ALTER

NOMINVALUE

.

SEQUENCE

integer

sequence

;

CYCLE
NOCYCLE
CACHE

integer

NOCACHE
ORDER
NOORDER

Semantics
11

The keywords and parameters in this statement serve the same purposes they serve
when you create a sequence.
■
■

■

To restart the sequence at a different number, you must drop and re-create it.
If you change the INCREMENT BY value before the first invocation of NEXTVAL, then
some sequence numbers will be skipped. Therefore, if you want to retain the
original START WITH value, you must drop the sequence and re-create it with the
original START WITH value and the new INCREMENT BY value.
Oracle Database performs some validations. For example, a new MAXVALUE cannot
be imposed that is less than the current sequence number.
CREATE SEQUENCE on page 15-67 for information on
creating a sequence and DROP SEQUENCE on page 18-2 for
information on dropping and re-creating a sequence

See Also:

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-43

ALTER SEQUENCE

Examples
11

Modifying a Sequence: Examples This statement sets a new maximum value for the

customers_seq sequence, which was created in "Creating a Sequence: Example" on
page 15-70:
ALTER SEQUENCE customers_seq
MAXVALUE 1500;

This statement turns on CYCLE and CACHE for the customers_seq sequence:
ALTER SEQUENCE customers_seq
CYCLE
CACHE 5;

11-44 Oracle Database SQL Language Reference

ALTER SESSION

ALTER SESSION
Purpose
11

Use the ALTER SESSION statement to set or modify any of the conditions or parameters
that affect your connection to the database. The statement stays in effect until you
disconnect from the database.

Prerequisites
11

To enable and disable the SQL trace facility, you must have ALTER SESSION system
privilege.
To enable or disable resumable space allocation, you must have the RESUMABLE system
privilege.
You do not need any privileges to perform the other operations of this statement
unless otherwise indicated.

Syntax
11

alter_session::=
COMMIT
ADVISE

ROLLBACK
NOTHING

CLOSE

DATABASE

LINK

dblink

ENABLE
COMMIT

IN

PROCEDURE

DISABLE
ENABLE
GUARD
DISABLE
ALTER

SESSION

ENABLE

DML

DISABLE

PARALLEL

FORCE

integer

DDL
QUERY
TIMEOUT

ENABLE

RESUMABLE

DISABLE

RESUMABLE

SYNC

;
PARALLEL

WITH

integer

NAME

string

PRIMARY

alter_session_set_clause

alter_session_set_clause::=
parameter_name

=

parameter_value

SET
EDITION

=

edition_name

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-45

ALTER SESSION

Semantics
11

ADVISE Clause
The ADVISE clause sends advice to a remote database to force a distributed transaction.
The advice appears in the ADVICE column of the DBA_2PC_PENDING view on the remote
database (the values are 'C' for COMMIT, 'R' for ROLLBACK, and ' ' for NOTHING). If the
transaction becomes in doubt, then the administrator of that database can use this
advice to decide whether to commit or roll back the transaction.
You can send different advice to different remote databases by issuing multiple ALTER
SESSION statements with the ADVISE clause in a single transaction. Each such statement
sends advice to the databases referenced in the following statements in the transaction
until another such statement is issued.
See Also: "Forcing a Distributed Transaction: Example" on
page 11-55

CLOSE DATABASE LINK Clause
Specify CLOSE DATABASE LINK to close the database link dblink. When you issue a
statement that uses a database link, Oracle Database creates a session for you on the
remote database using that link. The connection remains open until you end your local
session or until the number of database links for your session exceeds the value of the
initialization parameter OPEN_LINKS. If you want to reduce the network overhead
associated with keeping the link open, then use this clause to close the link explicitly if
you do not plan to use it again in your session.
See Also:

Closing a Database Link: Example on page 11-55

ENABLE | DISABLE COMMIT IN PROCEDURE
Procedures and stored functions written in PL/SQL can issue COMMIT and ROLLBACK
statements. If your application would be disrupted by a COMMIT or ROLLBACK statement
not issued directly by the application itself, then specify DISABLE COMMIT IN PROCEDURE
clause to prevent procedures and stored functions called during your session from
issuing these statements.
You can subsequently allow procedures and stored functions to issue COMMIT and
ROLLBACK statements in your session by issuing the ENABLE COMMIT IN PROCEDURE.
Some applications automatically prohibit COMMIT and ROLLBACK statements in
procedures and stored functions. Refer to your application documentation for more
information.

ENABLE | DISABLE GUARD
The security_clause of ALTER DATABASE lets you prevent anyone other than the SYS
user from making any changes to data or database objects on the primary or standby
database. This clause lets you override that setting for the current session.
security_clause on page 10-42 for more information on the
GUARD setting
See Also:

PARALLEL DML | DDL | QUERY
The PARALLEL parameter determines whether all subsequent DML, DDL, or query
statements in the session will be considered for parallel execution. This clause enables
you to override the degree of parallelism of tables during the current session without

11-46 Oracle Database SQL Language Reference

ALTER SESSION

changing the tables themselves. Uncommitted transactions must either be committed
or rolled back prior to executing this clause for DML.
See Also:

"Enabling Parallel DML: Example" on page 11-55

ENABLE Clause
Specify ENABLE to execute subsequent statements in the session in parallel. This is the
default for DDL and query statements.
■

■
■

DML: DML statements are executed in parallel mode if a parallel hint or a parallel
clause is specified.
DDL: DDL statements are executed in parallel mode if a parallel clause is specified.
QUERY: Queries are executed in parallel mode if a parallel hint or a parallel clause is
specified.

Restriction on the ENABLE clause You cannot specify the optional PARALLEL

integer with ENABLE.
DISABLE Clause
Specify DISABLE to execute subsequent statements in the session serially. This is the
default for DML statements.
■

DML: DML statements are executed serially.

■

DDL: DDL statements are executed serially.

■

QUERY: Queries are executed serially.

Restriction on the DISABLE clause You cannot specify the optional PARALLEL

integer with DISABLE.
FORCE Clause
FORCE forces parallel execution of subsequent statements in the session. If no parallel
clause or hint is specified, then a default degree of parallelism is used. This clause
overrides any parallel_clause specified in subsequent statements in the session but
is overridden by a parallel hint.
■

■

DML: Provided no parallel DML restrictions are violated, subsequent DML
statements in the session are executed with the default degree of parallelism,
unless a degree is specified in this clause.
DDL: Subsequent DDL statements in the session are executed with the default
degree of parallelism, unless a degree is specified in this clause. Resulting database
objects will have associated with them the prevailing degree of parallelism.
Specifying FORCE DDL automatically causes all tables created in this session to be
created with a default level of parallelism. The effect is the same as if you had
specified the parallel_clause (with the default degree) in the CREATE TABLE
statement.

■

QUERY: Subsequent queries are executed with the default degree of parallelism,
unless a degree is specified in this clause.

PARALLEL integer
■

Specify an integer to explicitly specify a degree of parallelism:

For FORCE DDL, the degree overrides any parallel clause in subsequent DDL
statements.

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-47

ALTER SESSION

■

■

For FORCE DML and QUERY, the degree overrides the degree currently stored for the
table in the data dictionary.
A degree specified in a statement through a hint will override the degree being
forced.

The following types of DML operations are not parallelized regardless of this clause:
■
■

■
■

Operations on cluster tables
Operations with embedded functions that either write or read database or package
states
Operations on tables with triggers that could fire
Operations on tables or schema objects containing object types, or LONG or LOB
data types

RESUMABLE Clauses
These clauses let you enable and disable resumable space allocation. This feature
allows an operation to be suspended in the event of an out-of-space error condition
and to resume automatically from the point of interruption when the error condition is
fixed.
Resumable space allocation is fully supported for operations
on locally managed tablespaces. Some restrictions apply if you are
using dictionary-managed tablespaces. For information on these
restrictions, refer to Oracle Database Administrator's Guide.

Note:

ENABLE RESUMABLE
This clause enables resumable space allocation for the session.
TIMEOUT lets you specify (in seconds) the time during which an operation
can remain suspended while waiting for the error condition to be fixed. If the error
condition is not fixed within the TIMEOUT period, then Oracle Database aborts the
suspended operation.

TIMEOUT

NAME lets you specify a user-defined text string to help users identify the
statements issued during the session while the session is in resumable mode. Oracle
Database inserts the text string into the USER_RESUMABLE and DBA_RESUMABLE data
dictionary views. If you do not specify NAME, then Oracle Database inserts the default
string 'User username(userid), Session sessionid, Instance instanceid'.
NAME

Oracle Database Reference for information on the data
dictionary views

See Also:

DISABLE RESUMABLE
This clause disables resumable space allocation for the session.
SYNC WITH PRIMARY
Use this clause to synchronize redo apply on a physical standby database with the
primary database. An ALTER SESSION statement with this clause blocks until redo
apply has applied all redo data received by the standby at the time the statement is
issued. This clause returns an error, and synchronization does not occur, if the redo
transport state for the standby database is not SYNCHRONIZED or if redo apply is not
active.
11-48 Oracle Database SQL Language Reference

ALTER SESSION

See Also: Oracle Data Guard Concepts and Administration for more
information on this session parameter

alter_session_set_clause
Use the alter_session_set_clause to set initialization parameter values or to set an
edition for the current session.
Initialization Parameters You can set two types of parameters using this clause:
■

■

Initialization parameters that are dynamic in the scope of the ALTER SESSION
statement (listed in "Initialization Parameters and ALTER SESSION" on
page 11-50)
Session parameters (listed in "Session Parameters and ALTER SESSION" on
page 11-51)

You can set values for multiple parameters in the same alter_session_set_clause.
Edition Specify EDITION = edition to set the specified edition as the edition in the
database session. You must have the USE object privilege on edition, edition must
already have been created, and it must be USABLE.

When this statement is successful, the database discards PL/SQL package state
corresponding to editionable packages but retains package state corresponding to
packages that are not editionable.
You can also set the edition for the current session at startup with the EDITION
parameter of the SQL*Plus CONNECT command. However, you cannot specify an ALTER
SESSION SET EDITION statement in a recursive SQL or PL/SQL block.
You can determine the edition in use by the current session with the following query:
SELECT SYS_CONTEXT('USERENV', 'CURRENT_EDITION_NAME') FROM DUAL;

See Also: CREATE EDITION on page 14-51 for more information on
editions and Oracle Database PL/SQL Language Reference for
information on how editions are designated as USABLE

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-49

Initialization Parameters and ALTER SESSION

Initialization Parameters and ALTER SESSION
Some initialization parameter are dynamic in the scope of ALTER SESSION. When you
set these parameters using ALTER SESSION, the value you set persists only for the
duration of the current session.To determine whether a parameter can be altered using
an ALTER SESSION statement, query the ISSES_MODIFIABLE column of the V$PARAMETER
dynamic performance view.
Caution: Before changing the values of initialization parameters,
refer to their full description in Oracle Database Reference.

A number of parameters that can be set using ALTER SESSION are not initialization
parameters. You can set them only with ALTER SESSION, not in an initialization
parameter file. Those session parameters are described in "Session Parameters and
ALTER SESSION" on page 11-51.

11-50 Oracle Database SQL Language Reference

ALTER SESSION

Session Parameters and ALTER SESSION
The following parameters are session parameters only, not initialization parameters:

CONSTRAINT[S]
Syntax:
CONSTRAINT[S] = { IMMEDIATE | DEFERRED | DEFAULT }

The CONSTRAINT[S] parameter determines when conditions specified by a deferrable
constraint are enforced.
■

■

■

IMMEDIATE indicates that the conditions specified by the deferrable constraint are
checked immediately after each DML statement. This setting is equivalent to
issuing the SET CONSTRAINTS ALL IMMEDIATE statement at the beginning of each
transaction in your session.
DEFERRED indicates that the conditions specified by the deferrable constraint are
checked when the transaction is committed. This setting is equivalent to issuing
the SET CONSTRAINTS ALL DEFERRED statement at the beginning of each transaction
in your session.
DEFAULT restores all constraints at the beginning of each transaction to their initial
state of DEFERRED or IMMEDIATE.

CURRENT_SCHEMA
Syntax:
CURRENT_SCHEMA = schema

The CURRENT_SCHEMA parameter changes the current schema of the session to the
specified schema. Subsequent unqualified references to schema objects during the
session will resolve to objects in the specified schema. The setting persists for the
duration of the session or until you issue another ALTER SESSION SET CURRENT_SCHEMA
statement.
This setting offers a convenient way to perform operations on objects in a schema
other than that of the current user without having to qualify the objects with the
schema name. This setting changes the current schema, but it does not change the
session user or the current user, nor does it give the session user any additional system
or object privileges for the session.

ERROR_ON_OVERLAP_TIME
Syntax:
ERROR_ON_OVERLAP_TIME = {TRUE | FALSE}

The ERROR_ON_OVERLAP_TIME parameter determines how Oracle Database should
handle an ambiguous boundary datetime value—a case in which it is not clear
whether the datetime is in standard or daylight saving time.
■
■

Specify TRUE to return an error for the ambiguous overlap timestamp.
Specify FALSE to default the ambiguous overlap timestamp to the standard time.
This is the default.

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-51

Session Parameters and ALTER SESSION

Refer to "Support for Daylight Saving Times" on page 3-22 for more information on
boundary datetime values.

FLAGGER
Syntax:
FLAGGER = { ENTRY | OFF }

The FLAGGER parameter specifies FIPS flagging (as specified in Federal Information
Processing Standard 127-2), which causes an error message to be generated when a
SQL statement issued is an extension of the Entry Level of SQL-92 (officially, ANSI
X3.135-1992, a standard that is now superseded by SQL:2008). FLAGGER is a session
parameter only, not an initialization parameter.
After flagging is set in a session, a subsequent ALTER SESSION SET FLAGGER statement
will work, but generates the message, ORA-00097. This allows FIPS flagging to be
altered without disconnecting the session. OFF turns off flagging.
See Also: Appendix C, "Oracle and Standard SQL", for more
information about Oracle compliance with current ANSI SQL
standards

INSTANCE
Syntax:
INSTANCE = integer

Setting the INSTANCE parameter lets you access another instance as if you were
connected to your own instance. INSTANCE is a session parameter only, not an
initialization parameter. In an Oracle Real Application Clusters (Oracle RAC)
environment, each Oracle RAC instance retains static or dynamic ownership of disk
space for optimal DML performance based on the setting of this parameter.

ISOLATION_LEVEL
Syntax:
ISOLATION_LEVEL = {SERIALIZABLE | READ COMMITTED}

The ISOLATION_LEVEL parameter specifies how transactions containing database
modifications are handled. ISOLATION_LEVEL is a session parameter only, not an
initialization parameter.
■

■

SERIALIZABLE indicates that transactions in the session use the serializable
transaction isolation mode as specified in the SQL standard. If a serializable
transaction attempts to execute a DML statement that updates rows currently
being updated by another uncommitted transaction at the start of the serializable
transaction, then the DML statement fails. A serializable transaction can see its
own updates.
READ COMMITTED indicates that transactions in the session will use the default
Oracle Database transaction behavior. If the transaction contains DML that
requires row locks held by another transaction, then the DML statement will wait
until the row locks are released.

11-52 Oracle Database SQL Language Reference

ALTER SESSION

Serializable transactions do not work with deferred segment
creation or interval partitioning. Trying to insert data into an empty
table with no segment created, or into a partition of an interval
partitioned table that does not yet have a segment, causes an error.

Note:

STANDBY_MAX_DATA_DELAY
Syntax:
STANDBY_MAX_DATA_DELAY =

{ integer | NONE }

In an Active Data Guard environment, this session parameter can be used to specify a
session-specific apply lag tolerance, measured in seconds, for queries issued by
non-administrative users to a physical standby database that is in real-time query
mode. This capability allows queries to be safely offloaded from the primary database
to a physical standby database, because it is possible to detect if the standby database
has become unacceptably stale.
If STANDBY_MAX_DATA_DELAY is set to the default value of NONE, queries issued to a
physical standby database will be executed regardless of the apply lag on that
database.
If STANDBY_MAX_DATA_DELAY is set to a non-zero value, a query issued to a physical
standby database will be executed only if the apply lag is less than or equal to
STANDBY_MAX_DATA_DELAY. Otherwise, an ORA-3172 error is returned to alert the client
that the apply lag is too large.
If STANDBY_MAX_DATA_DELAY is set to 0, a query issued to a physical standby database is
guaranteed to return the exact same result as if the query were issued on the primary
database, unless the standby database is lagging behind the primary database, in
which case an ORA-3172 error is returned.
See Also: Oracle Data Guard Concepts and Administration for more
information on Active Data Guard and using this session parameter

TIME_ZONE
Syntax:
TIME_ZONE =

'[+ | -] hh:mi'
| LOCAL
| DBTIMEZONE
| 'time_zone_region'

The TIME_ZONE parameter specifies the default local time zone offset or region name
for the current SQL session. TIME_ZONE is a session parameter only, not an initialization
parameter. To determine the time zone of the current session, query the built-in
function SESSIONTIMEZONE (see SESSIONTIMEZONE on page 5-244).
■

■

Specify a format mask ('[+|-]hh:mi') indicating the hours and minutes before or
after UTC (Coordinated Universal Time—formerly Greenwich Mean Time). The
valid range for hh:mm is -12:00 to +14:00.
Specify LOCAL to set the default local time zone offset of the current SQL session to
the original default local time zone offset that was established when the current
SQL session was started.

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-53

Session Parameters and ALTER SESSION

■

■

Specify DBTIMEZONE to set the current session time zone to match the value set for
the database time zone. If you specify this setting, then the DBTIMEZONE function
will return the database time zone as a UTC offset or a time zone region,
depending on how the database time zone has been set.
Specify a valid time_zone_region. To see a listing of valid time zone region
names, query the TZNAME column of the V$TIMEZONE_NAMES dynamic performance
view. If you specify this setting, then the SESSIONTIMEZONE function will return the
region name.

Time zone region names are needed by the daylight saving
feature. These names are stored in two types of time zone files: one
large and one small. One of these files is the default file, depending
on your environment and the release of Oracle Database you are
using. For more information regarding time zone files and names,
see Oracle Database Globalization Support Guide.

Note:

See Also: Oracle Database Globalization Support Guide for a complete
listing of the time zone region names in both files

You can also set the default client session time zone using the
ORA_SDTZ environment variable. Refer to Oracle Database Globalization
Support Guide for more information on this variable.
Note:

USE_PRIVATE_OUTLINES
Syntax:
USE_PRIVATE_OUTLINES = { TRUE | FALSE | category_name }

The USE_PRIVATE_OUTLINES parameter lets you control the use of private outlines.
When this parameter is enabled and an outlined SQL statement is issued, the
optimizer retrieves the outline from the session private area rather than the public area
used when USE_STORED_OUTLINES is enabled. If no outline exists in the session private
area, then the optimizer will not use an outline to compile the statement. USE_
PRIVATE_OUTLINES is not an initialization parameter.
■

■

■

TRUE causes the optimizer to use private outlines stored in the DEFAULT category
when compiling requests.
FALSE specifies that the optimizer should not use stored private outlines. This is
the default. If USE_STORED_OUTLINES is enabled, then the optimizer will use stored
public outlines.
category_name causes the optimizer to use outlines stored in the category_name
category when compiling requests.

Restriction on USE_PRIVATE_OUTLINES

You cannot enable this parameter if USE_

STORED_OUTLINES is enabled.

USE_STORED_OUTLINES
Syntax:
USE_STORED_OUTLINES = { TRUE | FALSE | category_name }

11-54 Oracle Database SQL Language Reference

ALTER SESSION

The USE_STORED_OUTLINES parameter determines whether the optimizer will use
stored public outlines to generate execution plans. USE_STORED_OUTLINES is not an
initialization parameter.
■

■

■

TRUE causes the optimizer to use outlines stored in the DEFAULT category when
compiling requests.
FALSE specifies that the optimizer should not use stored outlines. This is the
default.
category_name causes the optimizer to use outlines stored in the category_name
category when compiling requests.

Restriction on USED_STORED_OUTLINES You cannot enable this parameter if USE_

PRIVATE_OUTLINES is enabled.

Examples
11

Issue the following statement to enable parallel
DML mode for the current session:

Enabling Parallel DML: Example

ALTER SESSION ENABLE PARALLEL DML;

Forcing a Distributed Transaction: Example The following transaction inserts an
employee record into the employees table on the database identified by the database
link remote and deletes an employee record from the employees table on the database
identified by local:
ALTER SESSION
ADVISE COMMIT;
INSERT INTO employees@remote
VALUES (8002, 'Juan', 'Fernandez', 'juanf@example.com', NULL,
TO_DATE('04-OCT-1992', 'DD-MON-YYYY'), 'SA_CLERK', 3000,
NULL, 121, 20);
ALTER SESSION
ADVISE ROLLBACK;
DELETE FROM employees@local
WHERE employee_id = 8002;
COMMIT;

This transaction has two ALTER SESSION statements with the ADVISE clause. If the
transaction becomes in doubt, then remote is sent the advice 'COMMIT' by virtue of the
first ALTER SESSION statement and local is sent the advice 'ROLLBACK' by virtue of the
second statement.
Closing a Database Link: Example This statement updates the jobs table on the

local database using a database link, commits the transaction, and explicitly closes the
database link:
UPDATE jobs@local SET min_salary = 3000
WHERE job_id = 'SH_CLERK';
COMMIT;
ALTER SESSION

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-55

Session Parameters and ALTER SESSION

CLOSE DATABASE LINK local;

The following statement
dynamically changes the default date format for your session to 'YYYY MM
DD-HH24:MI:SS':

Changing the Date Format Dynamically: Example

ALTER SESSION
SET NLS_DATE_FORMAT = 'YYYY MM DD HH24:MI:SS';

Oracle Database uses the new default date format:
SELECT TO_CHAR(SYSDATE) Today
FROM DUAL;
TODAY
------------------2001 04 12 12:30:38

The following statement
changes the language for date format elements to French:

Changing the Date Language Dynamically: Example
ALTER SESSION
SET NLS_DATE_LANGUAGE = French;
SELECT TO_CHAR(SYSDATE, 'Day DD Month YYYY') Today
FROM DUAL;
TODAY
--------------------------Jeudi
12 Avril
2001

Changing the ISO Currency: Example The following statement dynamically changes

the ISO currency symbol to the ISO currency symbol for the territory America:
ALTER SESSION
SET NLS_ISO_CURRENCY = America;
SELECT TO_CHAR( SUM(salary), 'C999G999D99') Total
FROM employees;
TOTAL
-----------------USD694,900.00

The following
statement dynamically changes the decimal character to comma (,) and the group
separator to period (.):

Changing the Decimal Character and Group Separator: Example

ALTER SESSION SET NLS_NUMERIC_CHARACTERS = ',.' ;

Oracle Database returns these new characters when you use their number format
elements:
ALTER SESSION SET NLS_CURRENCY = 'FF';
SELECT TO_CHAR( SUM(salary), 'L999G999D99') Total FROM employees;
TOTAL
--------------------FF694.900,00

11-56 Oracle Database SQL Language Reference

ALTER SESSION

The following statement dynamically
changes the local currency symbol to 'DM':

Changing the NLS Currency: Example
ALTER SESSION
SET NLS_CURRENCY = 'DM';

SELECT TO_CHAR( SUM(salary), 'L999G999D99') Total
FROM employees;
TOTAL
--------------------DM694.900,00

The following statement dynamically
changes to French the language in which error messages are displayed:

Changing the NLS Language: Example
ALTER SESSION
SET NLS_LANGUAGE = FRENCH;
Session modifiee.
SELECT * FROM DMP;
ORA-00942: Table ou vue inexistante

Changing the Linguistic Sort Sequence: Example The following statement
dynamically changes the linguistic sort sequence to Spanish:
ALTER SESSION
SET NLS_SORT = XSpanish;

Oracle Database sorts character values based on their position in the Spanish linguistic
sort sequence.
Enabling SQL Trace: Example To enable the SQL trace facility for your session, issue
the following statement:
ALTER SESSION
SET SQL_TRACE = TRUE;

Enabling Query Rewrite: Example This statement enables query rewrite in the
current session for all materialized views that have not been explicitly disabled:
ALTER SESSION SET QUERY_REWRITE_ENABLED = TRUE;

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-57

ALTER SYSTEM

ALTER SYSTEM
Purpose
11

Use the ALTER SYSTEM statement to dynamically alter your Oracle Database instance.
The settings stay in effect as long as the database is mounted.

Prerequisites
11

You must have ALTER SYSTEM system privilege.

Syntax
11

alter_system::=
archive_log_clause
checkpoint_clause
check_datafiles_clause
distributed_recov_clauses
SHARED_POOL
GLOBAL

CONTEXT

BUFFER_CACHE

FLUSH

NO
CONFIRM
REDO

TO

APPLY

target_db_name

end_session_clauses
ALTER

SYSTEM

SWITCH

LOGFILE

;

SUSPEND
RESUME
quiesce_clauses
rolling_migration_clauses
security_clauses
shutdown_dispatcher_clause
REGISTER
SET
RESET

alter_system_set_clause
alter_system_reset_clause

(archive_log_clause::= on page 11-59, checkpoint_clause::= on page 11-59, check_datafiles_
clause::= on page 11-59, distributed_recov_clauses::= on page 11-59, end_session_clauses::=
on page 11-59, quiesce_clauses::= on page 11-59, rolling_migration_clauses::= on
page 11-60, security_clauses::= on page 11-60, shutdown_dispatcher_clause::= on
page 11-60, alter_system_set_clause::= on page 11-60, alter_system_reset_clause::= on
page 11-61)

11-58 Oracle Database SQL Language Reference

ALTER SYSTEM

archive_log_clause::=
INSTANCE
ARCHIVE

’

instance_name

’

LOG

SEQUENCE

integer

CHANGE

integer
NOSWITCH

CURRENT
GROUP

TO

’

location

’

integer
USING

LOGFILE

’

filename

BACKUP

CONTROLFILE

’

NEXT
ALL

checkpoint_clause::=
GLOBAL
LOCAL
CHECKPOINT

check_datafiles_clause::=
GLOBAL
LOCAL
CHECK

DATAFILES

distributed_recov_clauses::=
ENABLE
DISTRIBUTED

RECOVERY

DISABLE

end_session_clauses::=
POST_TRANSACTION
DISCONNECT

SESSION

’

integer1

,

’

integer1

,

integer2

integer2
,

KILL

SESSION

’
@

IMMEDIATE
integer3
’

quiesce_clauses::=
QUIESCE

RESTRICTED

UNQUIESCE

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-59

ALTER SYSTEM

rolling_migration_clauses::=
START

ROLLING

STOP

ROLLING

MIGRATION

TO

’

ASM_version

’

MIGRATION

security_clauses::=
ENABLE
RESTRICTED

SESSION

DISABLE
wallet_password
SET

ENCRYPTION

WALLET

OPEN

IDENTIFIED

BY

’’

’’
HSM_auth_string
wallet_password

IDENTIFIED

BY

’’

’’
HSM_auth_string

SET

ENCRYPTION

WALLET

CLOSE

set_encryption_key

set_encryption_key::=
SET

ENCRYPTION

’’

KEY

certificate_id

’’
IDENTIFIED

BY

’’

wallet_password

MIGRATE
IDENTIFIED

BY

’’

HSM_auth_string

’’

shutdown_dispatcher_clause::=
IMMEDIATE
SHUTDOWN

dispatcher_name

alter_system_set_clause::=
set_parameter_clause
TRUE
USE_STORED_OUTLINES

=

FALSE
category_name
TRUE

GLOBAL_TOPIC_ENABLED

=
FALSE

11-60 Oracle Database SQL Language Reference

USING

’’
’’

wallet_password

’’

ALTER SYSTEM

set_parameter_clause::=
,
parameter_name

=

COMMENT

=

string

DEFERRED

parameter_value

MEMORY
SCOPE

=

SPFILE
BOTH

SID

’

sid

’

*

’

=
’

alter_system_reset_clause::=
SCOPE

SID

=

SPFILE
’

sid

’

*

’

=
’

parameter_name

Semantics
11

archive_log_clause
The archive_log_clause manually archives redo log files or enables or disables
automatic archiving. To use this clause, your instance must have the database
mounted. The database can be either open or closed unless otherwise noted.
INSTANCE Clause
This clause is relevant only if you are using Oracle Real Application Clusters (Oracle
RAC). Specify the name of the instance for which you want the redo log file group to
be archived. The instance name is a string of up to 80 characters. Oracle Database
automatically determines the thread that is mapped to the specified instance and
archives the corresponding redo log file group. If no thread is mapped to the specified
instance, then Oracle Database returns an error.
SEQUENCE Clause
Specify SEQUENCE to manually archive the online redo log file group identified by the
log sequence number integer in the specified thread. If you omit the THREAD
parameter, then Oracle Database archives the specified group from the thread assigned
to your instance.
CHANGE Clause
Specify CHANGE to manually archive the online redo log file group containing the redo
log entry with the system change number (SCN) specified by integer in the specified
thread. If the SCN is in the current redo log file group, then Oracle Database performs
a log switch. If you omit the THREAD parameter, then Oracle Database archives the
groups containing this SCN from all enabled threads.
You can use this clause only when your instance has the database open.

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-61

ALTER SYSTEM

CURRENT Clause
Specify CURRENT to manually archive the current redo log file group of the specified
thread, forcing a log switch. If you omit the THREAD parameter, then Oracle Database
archives all redo log file groups from all enabled threads, including logs previous to
current logs. You can specify CURRENT only when the database is open.
NOSWITCH Specify NOSWITCH if you want to manually archive the current redo log
file group without forcing a log switch. This setting is used primarily with standby
databases to prevent data divergence when the primary database shuts down.
Divergence implies the possibility of data loss in case of primary database failure.

You can use the NOSWITCH clause only when your instance has the database mounted
but not open. If the database is open, then this operation closes the database
automatically. You must then manually shut down the database before you can reopen
it.
GROUP Clause
Specify GROUP to manually archive the online redo log file group with the GROUP value
specified by integer. You can determine the GROUP value for a redo log file group by
querying the dynamic performance view V$LOG. If you specify both the THREAD and
GROUP parameters, then the specified redo log file group must be in the specified
thread.
LOGFILE Clause
Specify LOGFILE to manually archive the online redo log file group containing the redo
log file member identified by 'filename'. If you specify both the THREAD and LOGFILE
parameters, then the specified redo log file group must be in the specified thread.
If the database was mounted with a backup control file, then specify USING BACKUP
CONTROLFILE to permit archiving of all online logfiles, including the current logfile.
You must archive redo log file groups in the
order in which they are filled. If you specify a redo log file group for archiving with
the LOGFILE parameter, and earlier redo log file groups are not yet archived, then
Oracle Database returns an error.
Restriction on the LOGFILE clause

NEXT Clause
Specify NEXT to manually archive the next online redo log file group from the specified
thread that is full but has not yet been archived. If you omit the THREAD parameter,
then Oracle Database archives the earliest unarchived redo log file group from any
enabled thread.
ALL Clause
Specify ALL to manually archive all online redo log file groups from the specified
thread that are full but have not been archived. If you omit the THREAD parameter, then
Oracle Database archives all full unarchived redo log file groups from all enabled
threads.
TO location Clause
Specify TO 'location' to indicate the primary location to which the redo log file groups
are archived. The value of this parameter must be a fully specified file location
following the conventions of your operating system. If you omit this parameter, then
Oracle Database archives the redo log file group to the location specified by the
initialization parameters LOG_ARCHIVE_DEST or LOG_ARCHIVE_DEST_n.

11-62 Oracle Database SQL Language Reference

ALTER SYSTEM

checkpoint_clause
Specify CHECKPOINT to explicitly force Oracle Database to perform a checkpoint,
ensuring that all changes made by committed transactions are written to data files on
disk. You can specify this clause only when your instance has the database open.
Oracle Database does not return control to you until the checkpoint is complete.
GLOBAL In an Oracle Real Application Clusters (Oracle RAC) environment, this
setting causes Oracle Database to perform a checkpoint for all instances that have
opened the database. This is the default.

In an Oracle RAC environment, this setting causes Oracle Database to
perform a checkpoint only for the thread of redo log file groups for the instance from
which you issue the statement.

LOCAL

See Also:

"Forcing a Checkpoint: Example" on page 11-75

check_datafiles_clause
In a distributed database system, such as an Oracle RAC environment, this clause
updates an instance's SGA from the database control file to reflect information on all
online data files.
■

■

Specify GLOBAL to perform this synchronization for all instances that have opened
the database. This is the default.
Specify LOCAL to perform this synchronization only for the local instance.

Your instance should have the database open.

end_session_clauses
The end_session_clauses give you several ways to end the current session.
DISCONNECT SESSION Clause
Use the DISCONNECT SESSION clause to disconnect the current session by destroying the
dedicated server process (or virtual circuit if the connection was made by way of a
Shared Sever). To use this clause, your instance must have the database open. You
must identify the session with both of the following values from the V$SESSION view:
■

For integer1, specify the value of the SID column.

■

For integer2, specify the value of the SERIAL# column.

If system parameters are appropriately configured, then application failover will take
effect.
■

■

The POST_TRANSACTION setting allows ongoing transactions to complete before the
session is disconnected. If the session has no ongoing transactions, then this clause
has the same effect described for as KILL SESSION.
The IMMEDIATE setting disconnects the session and recovers the entire session state
immediately, without waiting for ongoing transactions to complete.
–

If you also specify POST_TRANSACTION and the session has ongoing
transactions, then the IMMEDIATE keyword is ignored.

–

If you do not specify POST_TRANSACTION, or you specify POST_TRANSACTION but
the session has no ongoing transactions, then this clause has the same effect as
described for KILL SESSION IMMEDIATE.
See Also:

"Disconnecting a Session: Example" on page 11-77

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-63

ALTER SYSTEM

KILL SESSION Clause
The KILL SESSION clause lets you mark a session as terminated, roll back ongoing
transactions, release all session locks, and partially recover session resources. To use
this clause, your instance must have the database open. Your session and the session to
be terminated must be on the same instance unless you specify integer3.You must
identify the session with the following values from the V$SESSION view:
■

For integer1, specify the value of the SID column.

■

For integer2, specify the value of the SERIAL# column.

■

For the optional integer3, specify the ID of the instance where the target session
to be killed exists. You can find the instance ID by querying the GV$ tables.

If the session is performing some activity that must be completed, such as waiting for
a reply from a remote database or rolling back a transaction, then Oracle Database
waits for this activity to complete, marks the session as terminated, and then returns
control to you. If the waiting lasts a minute, then Oracle Database marks the session to
be terminated and returns control to you with a message that the session is marked to
be terminated. The PMON background process then marks the session as terminated
when the activity is complete.
Whether or not the session has an ongoing transaction, Oracle Database does not
recover the entire session state until the session user issues a request to the session and
receives a message that the session has been terminated.
See Also:

"Terminating a Session: Example" on page 11-76

IMMEDIATE Specify IMMEDIATE to instruct Oracle Database to roll back ongoing

transactions, release all session locks, recover the entire session state, and return
control to you immediately.

distributed_recov_clauses
The DISTRIBUTED RECOVERY clause lets you enable or disable distributed recovery. To
use this clause, your instance must have the database open.
ENABLE Specify ENABLE to enable distributed recovery. In a single-process
environment, you must use this clause to initiate distributed recovery.

You may need to issue the ENABLE DISTRIBUTED RECOVERY statement more than once to
recover an in-doubt transaction if the remote node involved in the transaction is not
accessible. In-doubt transactions appear in the data dictionary view DBA_2PC_PENDING.
See Also:

"Enabling Distributed Recovery: Example" on page 11-76

DISABLE Specify DISABLE to disable distributed recovery.

FLUSH SHARED_POOL Clause
The FLUSH SHARED_POOL clause lets you clear data from the shared pool in the system
global area (SGA). The shared pool stores:
■
■

Cached data dictionary information and
Shared SQL and PL/SQL areas for SQL statements, stored procedures, function,
packages, and triggers.

This statement does not clear global application context information, nor does it clear
shared SQL and PL/SQL areas for items that are currently being executed. You can use

11-64 Oracle Database SQL Language Reference

ALTER SYSTEM

this clause regardless of whether your instance has the database dismounted or
mounted, open or closed.
See Also:

"Clearing the Shared Pool: Example" on page 11-75

FLUSH GLOBAL CONTEXT Clause
The FLUSH GLOBAL CONTEXT clause lets you flush all global application context
information from the shared pool in the system global area (SGA). You can use this
clause regardless of whether your instance has the database dismounted or mounted,
open or closed.

FLUSH BUFFER_CACHE Clause
The FLUSH BUFFER_CACHE clause lets you clear all data from the buffer cache in the
system global area (SGA), including the KEEP, RECYCLE, and DEFAULT buffer pools.
Caution: This clause is intended for use only on a test database. Do
not use this clause on a production database, because as a result of this
statement, subsequent queries will have no hits, only misses.

This clause is useful if you need to measure the performance of rewritten queries or a
suite of queries from identical starting points.

FLUSH REDO Clause
Use the FLUSH REDO clause to flush redo data from a primary database to a standby
database and to optionally wait for the flushed redo data to be applied to a physical or
logical standby database.
This clause can allow a failover to be performed on the target standby database
without data loss, even if the primary database is not in a zero data loss data
protection mode, provided that all redo data that has been generated by the primary
database can be flushed to the standby database.
The FLUSH REDO clause must be issued on a mounted, but not open, primary database.
target_db_name
For target_db_name, specify the DB_UNIQUE_NAME of the standby database that is to
receive the redo data flushed from the primary database.
The value of the LOG_ARCHIVE_DEST_n database initialization parameter that
corresponds to the target standby database must contain the DB_UNIQUE_NAME attribute,
and the value of that attribute must match the DB_UNIQUE_NAME of the target standby
database.
NO CONFIRM APPLY
If you specify this clause, then the ALTER SYSTEM statement will not complete until the
standby database has received all of the flushed redo data. You must specify this
clause if the target standby database is a snapshot standby database.
CONFIRM APPLY
If you specify this clause, then the ALTER SYSTEM statement will not complete until the
target standby database has received and applied all flushed redo data. This is the
default behavior unless you specify NO CONFIRM APPLY. You cannot specify this clause if
the target standby database is a snapshot standby database.

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-65

ALTER SYSTEM

See Also: Oracle Data Guard Concepts and Administration for more
information about the FLUSH REDO clause and failovers

SWITCH LOGFILE Clause
The SWITCH LOGFILE clause lets you explicitly force Oracle Database to begin writing to
a new redo log file group, regardless of whether the files in the current redo log file
group are full. When you force a log switch, Oracle Database begins to perform a
checkpoint but returns control to you immediately rather than when the checkpoint is
complete. To use this clause, your instance must have the database open.
See Also:

"Forcing a Log Switch: Example" on page 11-76

SUSPEND | RESUME
The SUSPEND clause lets you suspend all I/O (data file, control file, and file header) as
well as queries, in all instances, enabling you to make copies of the database without
having to handle ongoing transactions.
Restrictions on SUSPEND and RESUME SUSPEND and RESUME are subject to the
following restrictions:
■

■

■

Do not use this clause unless you have put the database tablespaces in hot backup
mode.
Do not terminate the session that issued the ALTER SYSTEM SUSPEND statement. An
attempt to reconnect while the system is suspended may fail because of recursive
SQL that is running during the SYS login.
If you start a new instance while the system is suspended, then that new instance
will not be suspended.

The RESUME clause lets you make the database available once again for queries and
I/O.

rolling_migration_clauses
Use these clauses in a clustered Oracle Automatic Storage Management (Oracle ASM)
environment to migrate one node at a time to a different Oracle ASM version without
affecting the overall availability of the Oracle ASM cluster or the database clusters
using Oracle ASM for storage.
When starting rolling upgrade, for ASM_version, you
must specify the following string:

START ROLLING MIGRATION

', , ,,'

ASM_version must be equal to or greater than 11.1.0.0.0. The surrounding single
quotation marks are required. Oracle ASM first verifies that the current release is
compatible for migration to the specified release, and then goes into limited
functionality mode. Oracle ASM then determines whether any rebalance operations
are under way anywhere in the cluster. If there are any such operations, then the
statement fails and must be reissued after the rebalance operations are complete.
Rolling upgrade mode is a cluster-wide in-memory persistent state. The cluster
continues to be in this state until there is at least one Oracle ASM instance running in
the cluster. Any new instance joining the cluster switches to migration mode
immediately upon startup. If all the instances in the cluster terminate, then subsequent
startup of any Oracle ASM instance will not be in rolling upgrade mode until you
reissue this statement to restart rolling upgrade of the Oracle ASM instances.

11-66 Oracle Database SQL Language Reference

ALTER SYSTEM

STOP ROLLING MIGRATION Use this clause to stop rolling upgrade and bring the
cluster back into normal operation. Specify this clause only after all instances in the
cluster have migrated to the same software version. The statement will fail if the
cluster is not in rolling upgrade mode.

When you specify this clause, the Oracle ASM instance validates that all the members
of the cluster are at the same software version, takes the instance out of rolling
upgrade mode, and returns to full functionality of the Oracle ASM cluster. If any
rebalance operations are pending because disks have gone offline, then those
operations are restarted if the ASM_POWER_LIMIT parameter would not be violated by
such a restart.
See Also: Oracle Automatic Storage Management Administrator's Guide
for more information about rolling upgrade

quiesce_clauses
Use the QUIESCE RESTRICTED and UNQUIESCE clauses to put the database in and take it
out of the quiesced state. This state enables database administrators to perform
administrative operations that cannot be safely performed in the presence of
concurrent transactions, queries, or PL/SQL operations.
Note: The QUIESCE RESTRICTED clause is valid only if the Database
Resource Manager is installed and only if the Resource Manager has
been on continuously since database startup in any instances that
have opened the database.

If multiple QUIESCE RESTRICTED or UNQUIESCE statements issue at the same time from
different sessions or instances, then all but one will receive an error.
QUIESCE RESTRICTED
Specify QUIESCE RESTRICTED to put the database in the quiesced state. For all instances
with the database open, this clause has the following effect:
■

■

■

Oracle Database instructs the Database Resource Manager in all instances to
prevent all inactive sessions (other than SYS and SYSTEM) from becoming active. No
user other than SYS and SYSTEM can start a new transaction, a new query, a new
fetch, or a new PL/SQL operation.
Oracle Database waits for all existing transactions in all instances that were
initiated by a user other than SYS or SYSTEM to finish (either commit or abort).
Oracle Database also waits for all running queries, fetches, and PL/SQL
procedures in all instances that were initiated by users other than SYS or SYSTEM
and that are not inside transactions to finish. If a query is carried out by multiple
successive OCI fetches, then Oracle Database does not wait for all fetches to finish.
It waits for the current fetch to finish and then blocks the next fetch. Oracle
Database also waits for all sessions (other than those of SYS or SYSTEM) that hold
any shared resources (such as enqueues) to release those resources. After all these
operations finish, Oracle Database places the database into quiesced state and
finishes executing the QUIESCE RESTRICTED statement.
If an instance is running in shared server mode, then Oracle Database instructs the
Database Resource Manager to block logins (other than SYS or SYSTEM) on that
instance. If an instance is running in non-shared-server mode, then Oracle
Database does not impose any restrictions on user logins in that instance.

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-67

ALTER SYSTEM

During the quiesced state, you cannot change the Resource Manager plan in any
instance.
UNQUIESCE
Specify UNQUIESCE to take the database out of quiesced state. Doing so permits
transactions, queries, fetches, and PL/SQL procedures that were initiated by users
other than SYS or SYSTEM to be undertaken once again. The UNQUIESCE statement does
not have to originate in the same session that issued the QUIESCE RESTRICTED
statement.

security_clauses
The security_clauses let you control access to the instance. They also allow you to
enable or disable access to the encrypted data in the instance.
RESTRICTED SESSION
The RESTRICTED SESSION clause lets you restrict logon to Oracle Database. You can use
this clause regardless of whether your instance has the database dismounted or
mounted, open or closed.
■

Specify ENABLE to allow only users with RESTRICTED SESSION system privilege to
log on to Oracle Database. Existing sessions are not terminated.
This clause applies only to the current instance. Therefore, in an Oracle RAC
environment, authorized users without the RESTRICTED SESSION system privilege
can still access the database by way of other instances.

■

Specify DISABLE to reverse the effect of the ENABLE RESTRICTED SESSION clause,
allowing all users with CREATE SESSION system privilege to log on to Oracle
Database. This is the default.
See Also:

"Restricting Sessions: Example" on page 11-74

SET ENCRYPTION WALLET Clause
Use this clause to manage database access to the Transparent Data Encryption (TDE)
master encryption key. The TDE master encryption key is stored in an external security
module, which can be an encryption wallet or Hardware Security Module (HSM).
Although this statement begins with the keyword ALTER, an ALTER SYSTEM SET
ENCRYPTION WALLET statement is not a DDL clause. However, you cannot roll back such
a statement.
Although this clause begins with the SET keyword, do not confuse it with the alter_
system_set_clause on page 11-71, which allows you to use the SET keyword to set the
value of initialization parameters. ENCRYPTION WALLET is not an initialization
parameter.
OPEN When you specify this clause, the database uses the specified password to

open the encryption wallet and load the TDE master key into database memory for the
duration of the instance, or establish a connection to the HSM in order to send the
encrypted table and tablespace keys to the HSM and receive then back decrypted.
■

Specify wallet_password to retrieve the master encryption key from the
encryption wallet. If the encryption wallet is not available or is already open, then
the database returns an error. The double quotation marks around wallet_
password are required.

11-68 Oracle Database SQL Language Reference

ALTER SYSTEM

■

Specify HSM_auth_string to make the HSM accessible. HSM_auth_string is of the
form "user_id:password" where:
–

user_id is the user ID created for the database using the HSM management
interface

–

password is the password created for the user ID using the HSM management
interface

The double quotation marks around HSM_auth_string are required
Use this clause to disable encryption and decryption in your database. The
wallet_password is required to close an encryption wallet. HSM_auth_string is
required to disable access to the HSM. Refer to OPEN on page 11-68 for details on
specifying HSM_auth_string.
CLOSE

A password is not required to close an auto-open wallet when only an auto-open
wallet is present. The password is required to close an auto-open wallet when both an
auto-open wallet and an encryption wallet are open. In this case, using CLOSE with a
password will close the auto-open wallet and the encryption wallet.
See Also: Oracle Real Application Clusters Administration and
Deployment Guide for information on setting encryption wallets in an
Oracle Real Application Clusters (Oracle RAC) environment

set_encryption_key
Use this clause to generate a new TDE master encryption key, if none exists. If there
are existing master keys in the HSM or wallet, then this clause rekeys the existing table
and tablespace keys, that is, it decrypts all table and tablespace keys with the old
master key and reencrypts them with the new master key.
An ALTER SYSTEM SET ENCRYPTION KEY statement is a DDL statement and will
automatically commit any pending transactions in the schema.
Although this clause begins with the SET keyword, do not confuse it with the alter_
system_set_clause on page 11-71, which allows you to use the SET keyword to set the
value of initialization parameters. ENCRYPTION KEY is not an initialization parameter.
IDENTIFIED BY wallet_password This clause loads the TDE master encryption key
from the encryption wallet into memory for access to encrypted data.
■

■

The certificate_id is required if you are using PKI asymmetric key pairs as
master encryption keys. Specify the integer that identifies the certificate. You can
find this value by querying the CERT_ID column of the V$WALLET dynamic
performance view. Do not specify certificate_id if you are using symmetric
keys, which are the default.
For wallet_password, specify the password used to connect to the security
module.

If you specify an invalid certificate_id or wallet_password, then the database
returns an error. The double quotation marks around certificate_id and wallet_
password are required.
PKI-based master keys, including
unified master encryption keys, can only be used with TDE column encryption and an
Oracle Wallet, not with HSM.

Restriction on IDENTIFIED BY wallet_password

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-69

ALTER SYSTEM

IDENTIFIED BY HSM_auth_string This clause creates a master encryption key that

will be stored inside the HSM. The master encryption key is used to encrypt or decrypt
table keys inside the HSM.
HSM_auth_string is of the form "user_id:password" where:
■

■

user_id is the user ID created for the database using the HSM management
interface
password is the password created for the user ID using the HSM management
interface

The double quotation marks around HSM_auth_string are required.
If you are already using Transparent Data Encryption with an Oracle Wallet and you
would like to migrate to an HSM, then specify the MIGRATE USING wallet_password
clause. This decrypts the existing table and tablespace keys, and then reencrypts them
with the newly created, HSM-based, master encryption key. Note that the encryption
wallet is still in use after you migrate to an HSM, because it may contain master
encryption keys that were used for export files, RMAN backups, or encrypted data in
temporary or undo tablespaces or redo log files. After migrating, perform one of the
following steps:
■

■

Change the wallet password to the HSM_auth_string using Oracle Wallet Manager
or the orapki command-line tool.
Create a local auto-open wallet from the encryption wallet and either rename the
encryption wallet, or move it out of the directory specified in ENCRYPTION_WALLET_
LOCATION in sqlnet.ora. Do not delete the encryption wallet and do not forget the
wallet password.
See Also:
■

■

■

Oracle Database Advanced Security Administrator's Guide for more
information on using the server wallet and encryption keys and
on Transparent Data Encryption
The description of the CREATE TABLE "encryption_spec" on
page 16-27 for information on using that feature to encrypt table
columns
"Establishing a Wallet and Encryption Key: Examples" on
page 11-74

shutdown_dispatcher_clause
The SHUTDOWN clause is relevant only if your system is using the shared server
architecture of Oracle Database. It shuts down a dispatcher identified by dispatcher_
name.
Do not confuse this clause with the SQL*Plus command
SHUTDOWN, which is used to shut down the entire database.
Note:

The dispatcher_name must be a string of the form 'Dxxx', where xxx indicates the
number of the dispatcher. For a listing of dispatcher names, query the NAME column of
the V$DISPATCHER dynamic performance view.
■

If you specify IMMEDIATE, then the dispatcher stops accepting new connections
immediately and Oracle Database terminates all existing connections through that
dispatcher. After all sessions are cleaned up, the dispatcher process shuts down.

11-70 Oracle Database SQL Language Reference

ALTER SYSTEM

■

If you do not specify IMMEDIATE, then the dispatcher stops accepting new
connections immediately but waits for all its users to disconnect and for all its
database links to terminate. Then it shuts down.

REGISTER Clause
Specify REGISTER to instruct the PMON background process to register the instance with
the listeners immediately. If you do not specify this clause, then registration of the
instance does not occur until the next time PMON executes the discovery routine. As a
result, clients may not be able to access the services for as long as 60 seconds after the
listener is started.
Oracle Database Concepts and Oracle Database Net Services
Administrator's Guide for information on the PMON background process
and listeners

See Also:

alter_system_set_clause
You can change the value of many initialization parameters for the current instance,
whether you have started the database with a traditional plain-text parameter file
(pfile) or with a server parameter file (spfile). Oracle Database Reference indicates these
parameters in the "Modifiable" category of each parameter description. If you are
using a pfile, then the change will persist only for the duration of the instance.
However, if you have started the database with an spfile, then you can change the
value of the parameter in the spfile itself, so that the new value will occur in
subsequent instances.
Oracle Database Reference documents all initialization parameters in full. The
parameters fall into three categories:
■

■

■

Basic parameters: Database administrators should be familiar with and consider
the setting for all of the basic parameters.
Functional categories: Oracle Database Reference also lists the initialization
parameters by their functional category.
Alphabetical listing: The Table of Contents of Oracle Database Reference contains all
initialization parameters in alphabetical order.

The ability to change initialization parameter values depends on whether you have
started up the database with a traditional plain-text initialization parameter file (pfile)
or with a server parameter file (spfile). To determine whether you can change the
value of a particular parameter, query the ISSYS_MODIFIABLE column of the
V$PARAMETER dynamic performance view.
set_parameter_clause
When setting a parameter value, you can specify additional settings as follows:
The COMMENT clause lets you associate a comment string with this change
in the value of the parameter. The comment string cannot contain control characters or
a line break. If you also specify SPFILE, then this comment will appear in the
parameter file to indicate the most recent change made to this parameter.
COMMENT

DEFERRED The DEFERRED keyword sets or modifies the value of the parameter for

future sessions that connect to the database. Current sessions retain the old value.
You must specify DEFERRED if the value of the ISSYS_MODIFIABLE column of
V$PARAMETER for this parameter is DEFERRED. If the value of that column is IMMEDIATE,

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-71

ALTER SYSTEM

then the DEFERRED keyword in this clause is optional. If the value of that column is
FALSE, then you cannot specify DEFERRED in this ALTER SYSTEM statement.
Oracle Database Reference for information on the
V$PARAMETER dynamic performance view
See Also:

The SCOPE clause lets you specify when the change takes effect. Scope
depends on whether you started up the database using a traditional plain-text
parameter file (pfile) or server parameter file (spfile).

SCOPE

■

■

■

MEMORY indicates that the change is made in memory, takes effect immediately, and
persists until the database is shut down. If you started up the database using a
parameter file (pfile), then this is the only scope you can specify.
SPFILE indicates that the change is made in the server parameter file. The new
setting takes effect when the database is next shut down and started up again. You
must specify SPFILE when changing the value of a static parameter that is
described as not modifiable in Oracle Database Reference.
BOTH indicates that the change is made in memory and in the server parameter file.
The new setting takes effect immediately and persists after the database is shut
down and started up again.

If a server parameter file was used to start up the database, then BOTH is the default. If
a parameter file was used to start up the database, then MEMORY is the default, as well
as the only scope you can specify.
SID The SID clause lets you specify the SID of the instance where the value will take
effect.
■

■

Specify SID = '*' if you want Oracle Database to change the value of the
parameter for all instances that do not already have an explicit setting for this
parameter.
Specify SID = 'sid' if you want Oracle Database to change the value of the
parameter only for the instance sid. This setting takes precedence over previous
and subsequent ALTER SYSTEM SET statements that specify SID = '*'.

If you do not specify this clause, then:
■

■

If the instance was started up with a pfile (traditional plain-text initialization
parameter file), then Oracle Database assumes the SID of the current instance.
If the instance was started up with an spfile (server parameter file), then Oracle
Database assumes SID = '*'.

If you specify an instance other than the current instance, then Oracle Database sends
a message to that instance to change the parameter value in the memory of that
instance.
Oracle Database Reference for information about the
V$PARAMETER view
See Also:

USE_STORED_OUTLINES Clause
USE_STORED_OUTLINES is a system parameter, not an initialization parameter. You
cannot set it in a pfile or spfile, but you can set it with an ALTER SYSTEM statement. This
parameter determines whether the optimizer will use stored public outlines to
generate execution plans.
■

TRUE causes the optimizer to use outlines stored in the DEFAULT category when
compiling requests.

11-72 Oracle Database SQL Language Reference

ALTER SYSTEM

■

■

FALSE specifies that the optimizer should not use stored outlines. This is the
default.
category_name causes the optimizer to use outlines stored in the category_name
category when compiling requests.

GLOBAL_TOPIC_ENABLED
GLOBAL_TOPIC_ENABLED is a system parameter, not an initialization parameter. You
cannot set it in a pfile or spfile, but you can set it with an ALTER SYSTEM statement. This
parameter determines whether all queues and topics created in Oracle Streams AQ are
automatically registered with the LDAP server. If GLOBAL_TOPIC_ENABLED = TRUE when
a queue table is created, altered, or dropped, then the corresponding Lightweight
Directory Access Protocol (LDAP) entry is also created, altered or dropped.
The parameter works the same way for the Java Message Service (JMS). If a database
has been configured to use LDAP and the GLOBAL_TOPIC_ENABLED parameter has been
set to TRUE, then all JMS queues and topics are automatically registered with the LDAP
server when they are created. The administrator can also create aliases to the queues
and topics registered in LDAP. Queues and topics that are registered in LDAP can be
looked up through JNDI using the name or alias of the queue or topic.
Shared Server Parameters
When you start your instance, Oracle Database creates shared server processes and
dispatcher processes for the shared server architecture based on the values of the
SHARED_SERVERS and DISPATCHERS initialization parameters. You can also set the
SHARED_SERVERS and DISPATCHERS parameters with ALTER SYSTEM to perform one of the
following operations while the instance is running:
■

■

■

■

Create additional shared server processes by increasing the minimum number of
shared server processes.
Terminate existing shared server processes after their current calls finish
processing.
Create more dispatcher processes for a specific protocol, up to a maximum across
all protocols specified by the initialization parameter MAX_DISPATCHERS.
Terminate existing dispatcher processes for a specific protocol after their current
user processes disconnect from the instance.

alter_system_reset_clause
This clause lets you remove the setting, for any instance, of any initialization
parameter in the spfile that was used to start the instance. Neither SCOPE=MEMORY nor
SCOPE=BOTH are allowed. The SCOPE = SPFILE clause is not required, but is included for
syntactic clarity. You can use this clause in a single-instance environment, but only if
the instance was started using an spfile rather than a pfile.
Use the SID clause to remove the spfile parameter setting for a specified instance. In a
non-Oracle RAC environment, you can omit this clause, because there is only one
instance. In an Oracle RAC environment, if you omit this clause, then the default of
SID = '*' is used, which means that the all settings of the parameter of the form
*.parameter = value are removed.

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-73

ALTER SYSTEM

See Also:
■

■

Oracle Real Application Clusters Administration and Deployment
Guide for information on setting parameter values for an
individual instance in an Oracle Real Application Clusters
environment
The following examples of using the ALTER SYSTEM statement:
"Changing Licensing Parameters: Examples" on page 11-75,
"Enabling Query Rewrite: Example" on page 11-74, "Enabling
Resource Limits: Example" on page 11-75, "Shared Server
Parameters" on page 11-73, and "Changing Shared Server Settings:
Examples" on page 11-75

Examples
11

The following statement manually
archives the redo log file group containing the redo log entry with the SCN 9356083:

Archiving Redo Logs Manually: Examples
ALTER SYSTEM ARCHIVE LOG CHANGE 9356083;

The following statement manually archives the redo log file group containing a
member named 'diskl:log6.log' to an archived redo log file in the location
'diska:[arch$]':
ALTER SYSTEM ARCHIVE LOG
LOGFILE 'diskl:log6.log'
TO 'diska:[arch$]';

Enabling Query Rewrite: Example This statement enables query rewrite in all
sessions for all materialized views for which query rewrite has not been explicitly
disabled:
ALTER SYSTEM SET QUERY_REWRITE_ENABLED = TRUE;

You might want to restrict sessions if you are
performing application maintenance and you want only application developers with
RESTRICTED SESSION system privilege to log on. To restrict sessions, issue the following
statement:
Restricting Sessions: Example

ALTER SYSTEM
ENABLE RESTRICTED SESSION;

You can then terminate any existing sessions using the KILL SESSION clause of the
ALTER SYSTEM statement.
After performing maintenance on your application, issue the following statement to
allow any user with CREATE SESSION system privilege to log on:
ALTER SYSTEM
DISABLE RESTRICTED SESSION;

The following statements load
information from the server wallet into memory and set the Transparent Data
Encryption master key:

Establishing a Wallet and Encryption Key: Examples

ALTER SYSTEM SET ENCRYPTION WALLET OPEN IDENTIFIED BY "password";
ALTER SYSTEM SET ENCRYPTION KEY IDENTIFIED BY "password";

11-74 Oracle Database SQL Language Reference

ALTER SYSTEM

These statements assume that you have initialized the security module and created a
wallet with password.
Closing a Wallet: Examples The following statement removes password-based

wallet information from memory:
ALTER SYSTEM SET ENCRYPTION WALLET CLOSE IDENTIFIED BY "password";

The following statement removes password-based wallet information and auto-login
information, if present, from memory:
ALTER SYSTEM SET ENCRYPTION WALLET CLOSE;

Clearing the Shared Pool: Example You might want to clear the shared pool before
beginning performance analysis. To clear the shared pool, issue the following
statement:
ALTER SYSTEM FLUSH SHARED_POOL;

Forcing a Checkpoint: Example The following statement forces a checkpoint:
ALTER SYSTEM CHECKPOINT;

Enabling Resource Limits: Example

This ALTER SYSTEM statement dynamically

enables resource limits:
ALTER SYSTEM SET RESOURCE_LIMIT = TRUE;

Changing Shared Server Settings: Examples The following statement changes the
minimum number of shared server processes to 25:
ALTER SYSTEM SET SHARED_SERVERS = 25;

If there are currently fewer than 25 shared server processes, then Oracle Database
creates more. If there are currently more than 25, then Oracle Database terminates
some of them when they are finished processing their current calls if the load could be
managed by the remaining 25.
The following statement dynamically changes the number of dispatcher processes for
the TCP/IP protocol to 5 and the number of dispatcher processes for the ipc protocol
to 10:
ALTER SYSTEM
SET DISPATCHERS =
'(INDEX=0)(PROTOCOL=TCP)(DISPATCHERS=5)',
'(INDEX=1)(PROTOCOL=ipc)(DISPATCHERS=10)';

If there are currently fewer than 5 dispatcher processes for TCP, then Oracle Database
creates new ones. If there are currently more than 5, then Oracle Database terminates
some of them after the connected users disconnect.
If there are currently fewer than 10 dispatcher processes for ipc, then Oracle Database
creates new ones. If there are currently more than 10, then Oracle Database terminates
some of them after the connected users disconnect.
If there are currently existing dispatchers for another protocol, then the preceding
statement does not affect the number of dispatchers for that protocol.
The following statement dynamically
changes the limit on sessions for your instance to 64 and the warning threshold for
sessions on your instance to 54:

Changing Licensing Parameters: Examples

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-75

ALTER SYSTEM

ALTER SYSTEM
SET LICENSE_MAX_SESSIONS = 64
LICENSE_SESSIONS_WARNING = 54;

If the number of sessions reaches 54, then Oracle Database writes a warning message
to the ALERT file for each subsequent session. Also, users with RESTRICTED SESSION
system privilege receive warning messages when they begin subsequent sessions.
If the number of sessions reaches 64, then only users with RESTRICTED SESSION system
privilege can begin new sessions until the number of sessions falls below 64 again.
The following statement dynamically disables the limit for sessions on your instance.
After you issue this statement, Oracle Database no longer limits the number of
sessions on your instance.
ALTER SYSTEM SET LICENSE_MAX_SESSIONS = 0;

The following statement dynamically changes the limit on the number of users in the
database to 200. After you issue the preceding statement, Oracle Database prevents the
number of users in the database from exceeding 200.
ALTER SYSTEM SET LICENSE_MAX_USERS = 200;

Forcing a Log Switch: Example You might want to force a log switch to drop or

rename the current redo log file group or one of its members, because you cannot drop
or rename a file while Oracle Database is writing to it. The forced log switch affects
only the redo log thread of your instance. The following statement forces a log switch:
ALTER SYSTEM SWITCH LOGFILE;

Enabling Distributed Recovery: Example The following statement enables
distributed recovery:
ALTER SYSTEM ENABLE DISTRIBUTED RECOVERY;

You might want to disable distributed recovery for demonstration or testing purposes.
You can disable distributed recovery in both single-process and multiprocess mode
with the following statement:
ALTER SYSTEM DISABLE DISTRIBUTED RECOVERY;

When your demonstration or testing is complete, you can then enable distributed
recovery again by issuing an ALTER SYSTEM statement with the ENABLE DISTRIBUTED
RECOVERY clause.
Terminating a Session: Example You might want to terminate the session of a user
that is holding resources needed by other users. The user receives an error message
indicating that the session has been terminated. That user can no longer make calls to
the database without beginning a new session. Consider this data from the V$SESSION
dynamic performance table, when the users SYS and oe both have open sessions:
SELECT sid, serial#, username
FROM V$SESSION;
SID
SERIAL# USERNAME
---------- ---------- -----------------------------29
85 SYS
33
1
35
8
39
23 OE
40
1

11-76 Oracle Database SQL Language Reference

ALTER SYSTEM

. . .

The following statement terminates the session of the user scott using the SID and
SERIAL# values from V$SESSION:
ALTER SYSTEM KILL SESSION '39, 23';

Disconnecting a Session: Example The following statement disconnects user

scott's session, using the SID and SERIAL# values from V$SESSION:
ALTER SYSTEM DISCONNECT SESSION '13, 8' POST_TRANSACTION;

SQL Statements: ALTER LIBRARY to ALTER SYSTEM 11-77

ALTER SYSTEM

11-78 Oracle Database SQL Language Reference

12
SQL Statements: ALTER TABLE to
ALTER TABLESPACE
12

This chapter contains the following SQL statements:
■

ALTER TABLE

■

ALTER TABLESPACE

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-1

ALTER TABLE

ALTER TABLE
Purpose
12

Use the ALTER TABLE statement to alter the definition of a nonpartitioned table, a
partitioned table, a table partition, or a table subpartition. For object tables or relational
tables with object columns, use ALTER TABLE to convert the table to the latest definition
of its referenced type after the type has been altered.

Oracle recommends that you use the ALTER MATERIALIZED VIEW
LOG statement, rather than ALTER TABLE, whenever possible for
operations on materialized view log tables.
Note:

See Also:
■
■

CREATE TABLE on page 16-6 for information on creating tables
Oracle Text Reference for information on ALTER TABLE statements in
conjunction with Oracle Text

Prerequisites
12

The table must be in your own schema, or you must have ALTER object privilege on the
table, or you must have ALTER ANY TABLE system privilege.
Additional Prerequisites for Partitioning Operations If you are not the owner of the
table, then you need the DROP ANY TABLE privilege in order to use the drop_table_
partition or truncate_table_partition clause.

You must also have space quota in the tablespace in which space is to be acquired in
order to use the add_table_partition, modify_table_partition, move_table_
partition, and split_table_partition clauses.
When a partitioning operation cascades to reference-partitioned child tables, privileges
are not required on the reference-partitioned child tables.
Additional Prerequisites for Constraints and Triggers To enable a unique or primary

key constraint, you must have the privileges necessary to create an index on the table.
You need these privileges because Oracle Database creates an index on the columns of
the unique or primary key in the schema containing the table.
To enable or disable triggers, the triggers must be in your schema or you must have
the ALTER ANY TRIGGER system privilege.
See Also: CREATE INDEX on page 14-60 for information on the
privileges needed to create indexes
Additional Prerequisites When Using Object Types To use an object type in a
column definition when modifying a table, either that object must belong to the same
schema as the table being altered, or you must have either the EXECUTE ANY TYPE
system privilege or the EXECUTE object privilege for the object type.

To use the
flashback_archive_clause to enable historical tracking for the table, you must have
the FLASHBACK ARCHIVE object privilege on the flashback data archive that will contain
Additional Prerequisites for Flashback Data Archive Operations

12-2 Oracle Database SQL Language Reference

ALTER TABLE

the historical data. To use the flashback_archive_clause to disable historical tracking
for the table, you must have the FLASHBACK ARCHIVE ADMINSTER system privilege or you
must be logged in as SYSDBA.

Syntax
12

alter_table::=
schema
ALTER

TABLE

.
table

alter_table_properties
column_clauses
constraint_clauses
alter_table_partitioning

enable_disable_clause

alter_external_table

ENABLE

TABLE

move_table_clause

DISABLE

ALL

LOCK
TRIGGERS
;

Note: You must specify some clause after table. None of the clauses
after table are required, but you must specify at least one of them.

Groups of ALTER TABLE syntax:
■

alter_table_properties::= on page 12-4

■

column_clauses::= on page 12-8

■

constraint_clauses::= on page 12-11

■

alter_table_partitioning::= on page 12-19

■

alter_external_table::= on page 12-18

■

move_table_clause::= on page 12-31

■

enable_disable_clause::= on page 12-32

After each clause you will find links to its component subclauses.

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-3

ALTER TABLE

alter_table_properties::=
physical_attributes_clause
logging_clause
table_compression
supplemental_table_logging
allocate_extent_clause
deallocate_unused_clause
CACHE
NOCACHE
DEFAULT
RESULT_CACHE

(

MODE

)
FORCE

upgrade_table_clause

alter_iot_clauses

alter_XMLSchema_clause

records_per_block_clause
parallel_clause
row_movement_clause
flashback_archive_clause
RENAME

TO

new_table_name

shrink_clause
READ

ONLY

READ

WRITE

REKEY

encryption_spec

(physical_attributes_clause::= on page 12-4, logging_clause::= on page 8-38, table_
compression::= on page 12-5, supplemental_table_logging::= on page 12-5, allocate_extent_
clause::= on page 12-6, deallocate_unused_clause::= on page 12-6, upgrade_table_clause::=
on page 12-6, records_per_block_clause::= on page 12-6, parallel_clause::= on page 12-6,
row_movement_clause::= on page 12-6, flashback_archive_clause::= on page 12-32, shrink_
clause::= on page 12-6, alter_iot_clauses::= on page 12-7, alter_XMLSchema_clause::= on
page 12-18)
physical_attributes_clause::=
PCTFREE

integer

PCTUSED

integer

INITRANS

integer

storage_clause

(storage_clause::= on page 8-50)

12-4 Oracle Database SQL Language Reference

ALTER TABLE

logging_clause::=
LOGGING
NOLOGGING
FILESYSTEM_LIKE_LOGGING

table_compression::=
BASIC
OLTP
LOW

FOR

HIGH

QUERY
ARCHIVE
COMPRESS
NOCOMPRESS

supplemental_table_logging::=
,
supplemental_log_grp_clause
ADD

SUPPLEMENTAL

LOG
supplemental_id_key_clause
,
supplemental_id_key_clause

DROP

SUPPLEMENTAL

LOG
GROUP

log_group

supplemental_log_grp_clause::=
,
NO
GROUP

log_group

(

LOG

column

ALWAYS
)

supplemental_id_key_clause::=
,
ALL
PRIMARY
DATA

KEY

(

)

COLUMNS

UNIQUE
FOREIGN

KEY

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-5

ALTER TABLE

allocate_extent_clause::=
SIZE
(

size_clause

DATAFILE

’

INSTANCE
ALLOCATE

filename

’

)

integer

EXTENT

(size_clause::= on page 8-47)
deallocate_unused_clause::=
KEEP
DEALLOCATE

size_clause

UNUSED

(size_clause::= on page 8-47)
shrink_clause::=
COMPACT
SHRINK

CASCADE

SPACE

upgrade_table_clause::=
NOT
INCLUDING

DATA

UPGRADE

(column_properties::= on page 12-11)
records_per_block_clause::=
MINIMIZE
RECORDS_PER_BLOCK
NOMINIMIZE

parallel_clause::=
NOPARALLEL
integer
PARALLEL

row_movement_clause::=
ENABLE
ROW

MOVEMENT

DISABLE

12-6 Oracle Database SQL Language Reference

column_properties

ALTER TABLE

alter_iot_clauses::=
index_org_table_clause
alter_overflow_clause
alter_mapping_table_clauses
COALESCE

(alter_overflow_clause::= on page 12-8, alter_mapping_table_clauses::= on page 12-8)
index_org_table_clause::=
mapping_table_clause
PCTTHRESHOLD

integer

key_compression

index_org_overflow_clause

mapping_table_clauses::=
MAPPING

TABLE

NOMAPPING

key_compression::=
integer
COMPRESS
NOCOMPRESS

index_org_overflow_clause::=
INCLUDING

column_name

segment_attributes_clause
OVERFLOW

(segment_attributes_clause::= on page 12-8)
partition_extended_name::=
PARTITION

partition
,

PARTITION

FOR

(

partition_key_value

)

subpartition_extended_name::=
SUBPARTITION

subpartition
,

SUBPARTITION

FOR

(

subpartition_key_value

)

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-7

ALTER TABLE

segment_attributes_clause::=
physical_attributes_clause
TABLESPACE

tablespace

logging_clause

(physical_attributes_clause::= on page 12-4, logging_clause::= on page 8-38)
alter_overflow_clause::=
add_overflow_clause
segment_attributes_clause
allocate_extent_clause
OVERFLOW
shrink_clause
deallocate_unused_clause

(segment_attributes_clause::= on page 12-8, allocate_extent_clause::= on page 12-6, shrink_
clause::= on page 12-6, deallocate_unused_clause::= on page 12-6)
add_overflow_clause::=
,
segment_attributes_clause
segment_attributes_clause
ADD

(

PARTITION

OVERFLOW

(segment_attributes_clause::= on page 12-8)
alter_mapping_table_clauses::=
allocate_extent_clause
MAPPING

TABLE
deallocate_unused_clause

(allocate_extent_clause::= on page 12-6, deallocate_unused_clause::= on page 12-6)
column_clauses::=
add_column_clause
modify_column_clauses
drop_column_clause
rename_column_clause
modify_collection_retrieval
modify_LOB_storage_clause
alter_varray_col_properties

12-8 Oracle Database SQL Language Reference

)

ALTER TABLE

(add_column_clause::= on page 12-9, modify_column_clauses::= on page 12-9, drop_
column_clause::= on page 12-10, rename_column_clause::= on page 12-10, modify_
collection_retrieval::= on page 12-10, modify_LOB_storage_clause::= on page 12-15, alter_
varray_col_properties::= on page 12-17, encryption_spec::= on page 12-10)
add_column_clause::=
,

,
column_properties

column_definition
ADD

(

out_of_line_part_storage

)
virtual_column_definition

(column_definition::= on page 12-9, column_properties::= on page 12-11)
column_definition::=
SORT
column

DEFAULT

expr

ENCRYPT

encryption_spec

datatype

inline_constraint
inline_ref_constraint

(encryption_spec::= on page 12-10, inline_constraint and inline_ref_constraint:
constraint::= on page 8-5)
virtual_column_definition::=
datatype

GENERATED

ALWAYS

column

AS

VIRTUAL

(

column_expression

)

inline_constraint

modify_column_clauses::=
,
(

modify_col_properties

)

MODIFY
modify_col_substitutable

modify_col_properties::=
ENCRYPT
datatype

DEFAULT

expr

encryption_spec

DECRYPT

column

inline_constraint

LOB_storage_clause

alter_XMLSchema_clause

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-9

ALTER TABLE

(encryption_spec::= on page 12-10, inline_constraint: constraint::= on page 8-5, LOB_
storage_clause::= on page 12-14)
encryption_spec::=
USING

’

encrypt_algorithm

’

IDENTIFIED

BY

password

ALL

LEVELS

NO
’

integrity_algorithm

’

SALT

modify_col_substitutable::=
NOT
COLUMN

FORCE

column

SUBSTITUTABLE

AT

drop_column_clause::=
CASCADE
COLUMN
SET

UNUSED

column

CONSTRAINTS

INVALIDATE

,
(

column

)
CASCADE

COLUMN

column

DROP

CONSTRAINTS

INVALIDATE

CHECKPOINT

,
(

column

UNUSED

)
CHECKPOINT

COLUMNS

integer

DROP
COLUMNS

CONTINUE

rename_column_clause::=
RENAME

COLUMN

old_name

TO

new_name

modify_collection_retrieval::=
LOCATOR
MODIFY

NESTED

TABLE

collection_item

RETURN

AS
VALUE

12-10 Oracle Database SQL Language Reference

integer

ALTER TABLE

constraint_clauses::=
out_of_line_constraint
ADD
out_of_line_ref_constraint
CONSTRAINT
PRIMARY

constraint_name
CASCADE

KEY

MODIFY

constraint_state
,
UNIQUE

RENAME

(

CONSTRAINT

column
old_name

)
TO

new_name

drop_constraint_clause

(out_of_line_constraint::= on page 8-5, out_of_line_ref_constraint::= on page 8-6,
constraint_state::= on page 8-6, drop_constraint_clause::= on page 12-11)
drop_constraint_clause::=
KEEP
PRIMARY

INDEX

KEY

CASCADE

DROP

,
DROP

UNIQUE

(

column

)
CASCADE

CONSTRAINT

constraint_name

column_properties::=
object_type_col_properties
nested_table_col_properties
,
varray_col_properties

(

LOB_partition_storage

)

LOB_storage_clause
XMLType_column_properties

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-11

ALTER TABLE

out_of_line_part_storage::=
nested_table_col_properties
PARTITION

partition

LOB_storage_clause
varray_col_properties

nested_table_col_properties
SUBPARTITION

subpartition

LOB_storage_clause
varray_col_properties

object_type_col_properties::=
COLUMN

column

substitutable_column_clause

substitutable_column_clause::=
ELEMENT

TYPE
IS

OF

(

ONLY

type

)

NOT
SUBSTITUTABLE

AT

ALL

LEVELS

nested_table_col_properties::=
LOCAL
substitutable_column_clause

nested_item
NESTED

GLOBAL

TABLE
COLUMN_VALUE

(
(

object_properties

physical_properties
column_properties

STORE

AS

storage_table

AS

LOCATOR

RETURN
VALUE

12-12 Oracle Database SQL Language Reference

)
)

ALTER TABLE

object_properties::=
inline_constraint
DEFAULT

column

expr

inline_ref_constraint

attribute
out_of_line_constraint
out_of_line_ref_constraint
supplemental_logging_props

(inline_constraint, inline_ref_constraint, out_of_line_constraint, out_of_
line_ref_constraint: constraint::= on page 8-5)
supplemental_logging_props::=
supplemental_log_grp_clause
SUPPLEMENTAL

LOG
supplemental_id_key_clause

(supplemental_log_grp_clause::= on page 12-5, supplemental_id_key_clause::= on
page 12-5)
physical_properties::=
deferred_segment_creation

table_compression
segment_attributes_clause
segment_attributes_clause

table_compression

HEAP
deferred_segment_creation

segment_attributes_clause

ORGANIZATION

INDEX

index_org_table_clause

EXTERNAL

external_table_clause

,
CLUSTER

cluster

(

column

)

(deferred_segment_creation::= on page 12-13, segment_attributes_clause::= on page 12-8,
index_org_table_clause::= on page 12-7, external_data_properties::= on page 12-18)
deferred_segment_creation::=
IMMEDIATE
SEGMENT

CREATION
DEFERRED

varray_col_properties::=
substitutable_column_clause
varray_storage_clause
VARRAY

varray_item
substitutable_column_clause

(substitutable_column_clause::= on page 12-12, varray_storage_clause::= on page 12-14)
SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-13

ALTER TABLE

varray_storage_clause::=
SECUREFILE
LOB_segname
BASICFILE
STORE

(

AS

LOB_storage_parameters

LOB
LOB_segname

(LOB_parameters::= on page 12-15)
LOB_storage_clause::=
SECUREFILE

,
(

LOB_item

)

STORE

BASICFILE

AS
(

LOB

LOB_storage_parameters

)

SECUREFILE
BASICFILE
(

LOB_item

)

STORE

AS

LOB_segname
(

LOB_storage_parameters

)

(LOB_storage_parameters::= on page 12-14)
LOB_storage_parameters::=
TABLESPACE

tablespace
storage_clause

LOB_parameters
storage_clause

(LOB_parameters::= on page 12-15, storage_clause::= on page 8-50)

12-14 Oracle Database SQL Language Reference

)

ALTER TABLE

LOB_parameters::=
ENABLE
STORAGE

IN

ROW

DISABLE
CHUNK

integer

PCTVERSION

integer

FREEPOOLS

integer

LOB_retention_clause
LOB_deduplicate_clause
LOB_compression_clause
ENCRYPT

encryption_spec

DECRYPT
CACHE

logging_clause

NOCACHE
CACHE

READS

(LOB_retention_clause::= on page 12-16, LOB_deduplicate_clause::= on page 12-16, LOB_
compression_clause::= on page 12-16, encryption_spec::= on page 12-10, logging_clause::=
on page 8-38)
modify_LOB_storage_clause::=
MODIFY

LOB

(

LOB_item

)

(

modify_LOB_parameters

)

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-15

ALTER TABLE

modify_LOB_parameters::=
storage_clause
PCTVERSION
integer
FREEPOOLS
REBUILD

FREEPOOLS

LOB_retention_clause
LOB_deduplicate_clause
LOB_compression_clause
ENCRYPT

encryption_spec

DECRYPT
CACHE
logging_clause

NOCACHE
CACHE

READS

allocate_extent_clause
shrink_clause
deallocate_unused_clause

(storage_clause::= on page 8-50, LOB_retention_clause::= on page 12-16, LOB_
compression_clause::= on page 12-16, encryption_spec::= on page 12-10, logging_clause::=
on page 8-38, allocate_extent_clause::= on page 12-6, shrink_clause::= on page 12-6,
deallocate_unused_clause::= on page 12-6)
LOB_retention_clause::=
MAX
MIN

integer

AUTO
NONE
RETENTION

LOB_deduplicate_clause::=
DEDUPLICATE
KEEP_DUPLICATES

LOB_compression_clause::=
MEDIUM
LOW
COMPRESS
NOCOMPRESS

12-16 Oracle Database SQL Language Reference

ALTER TABLE

alter_varray_col_properties::=
MODIFY

VARRAY

varray_item

(

modify_LOB_parameters

)

(modify_LOB_parameters::= on page 12-16)
LOB_partition_storage::=
LOB_storage_clause
PARTITION

partition
varray_col_properties

LOB_partitioning_storage
(

SUBPARTITION

subpartition

)
varray_col_properties

(LOB_storage_clause::= on page 12-14, varray_col_properties::= on page 12-13, LOB_
partitioning_storage::= on page 12-17)
LOB_partitioning_storage::=
LOB

(

LOB_item

)

(

STORE

BASICFILE

LOB_segname

SECUREFILE

(

TABLESPACE

TABLESPACE

tablespace

tablespace

)

)

AS

XMLType_column_properties::=
COLUMN
XMLTYPE

XMLType_storage

XMLSchema_spec

column

XMLType_storage::=
STORE

OBJECT

RELATIONAL
(

AS

SECUREFILE

LOB_parameters

)

LOB_segname

BASICFILE

(

CLOB
BINARY

LOB_parameters

)

XML

LOBS
ALL

VARRAYS

AS
TABLES

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-17

ALTER TABLE

XMLSchema_spec::=
XMLSCHEMA

XMLSchema_URL

element
ELEMENT
XMLSchema_URL

ALLOW

#

element

ALLOW
NONSCHEMA

ANYSCHEMA

DISALLOW

DISALLOW

alter_XMLSchema_clause::=
ANYSCHEMA
ALLOW
NONSCHEMA
DISALLOW

NONSCHEMA

alter_external_table::=
add_column_clause
modify_column_clauses
drop_column_clause
parallel_clause
external_data_properties
integer
REJECT

LIMIT
UNLIMITED
ALL

PROJECT

COLUMN
REFERENCED

(add_column_clause::= on page 12-9, modify_column_clauses::= on page 12-9, drop_
column_clause::= on page 12-10, drop_constraint_clause::= on page 12-11, parallel_clause::=
on page 12-6)
external_data_properties::=
(
ACCESS

opaque_format_spec

USING
DEFAULT

DIRECTORY

directory

,
directory
LOCATION

(

:
’

location_specifier

12-18 Oracle Database SQL Language Reference

)

PARAMETERS

’

)

CLOB

subquery

ALTER TABLE

alter_table_partitioning::=
modify_table_default_attrs
alter_interval_partitioning
set_subpartition_template
modify_table_partition
modify_table_subpartition
move_table_partition
move_table_subpartition
add_table_partition
coalesce_table_partition
coalesce_table_subpartition
drop_table_partition
drop_table_subpartition
rename_partition_subpart
truncate_partition_subpart
split_table_partition
split_table_subpartition
merge_table_partitions
merge_table_subpartitions
exchange_partition_subpart

(modify_table_default_attrs::= on page 12-20, alter_interval_partitioning::= on page 12-20,
set_subpartition_template::= on page 12-20, modify_table_partition::= on page 12-20,
modify_table_subpartition::= on page 12-22, move_table_partition::= on page 12-22, move_
table_subpartition::= on page 12-22, add_table_partition::= on page 12-22, coalesce_table_
partition::= on page 12-24, coalesce_table_subpartition::= on page 12-24, drop_table_
partition::= on page 12-24, drop_table_subpartition::= on page 12-25, rename_partition_
subpart::= on page 12-25, truncate_partition_subpart::= on page 12-25, split_table_
partition::= on page 12-25, split_table_subpartition::= on page 12-26, merge_table_
partitions::= on page 12-26, merge_table_subpartitions::= on page 12-27, exchange_
partition_subpart::= on page 12-27

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-19

ALTER TABLE

modify_table_default_attrs::=
FOR
MODIFY

DEFAULT

deferred_segment_creation

segment_attributes_clause

PCTTHRESHOLD

integer

LOB

LOB_item

(

table_compression

key_compression

alter_overflow_clause

)
(

VARRAY

partition_extended_name

ATTRIBUTES

LOB_parameters

)

varray

(partition_extended_name::= on page 12-7, deferred_segment_creation::= on page 12-13,
segment_attributes_clause::= on page 12-8, table_compression::= on page 12-5, key_
compression::= on page 12-7, alter_overflow_clause::= on page 12-8, LOB_parameters::= on
page 12-15)
alter_interval_partitioning::=
expr
SET

INTERVAL

(

)
,

SET

STORE

IN

(

tablespace

)

set_subpartition_template::=
,
range_subpartition_desc
,
list_subpartition_desc
,
individual_hash_subparts
(
SET

SUBPARTITION

)

TEMPLATE
hash_subpartition_quantity

(range_subpartition_desc::= on page 12-29, list_subpartition_desc::= on page 12-29,
individual_hash_subparts::= on page 12-29)
modify_table_partition::=
modify_range_partition
modify_hash_partition
modify_list_partition

12-20 Oracle Database SQL Language Reference

ALTER TABLE

(modify_range_partition::= on page 12-21, modify_hash_partition::= on page 12-21, modify_
list_partition::= on page 12-21)
modify_range_partition::=
partition_attributes
add_range_subpartition
add_hash_subpartition
add_list_subpartition
MODIFY

partition_extended_name

update_index_clauses
COALESCE

parallel_clause

SUBPARTITION

alter_mapping_table_clause
REBUILD
UNUSABLE

LOCAL

INDEXES

(partition_extended_name::= on page 12-7, partition_attributes::= on page 12-30, add_
range_subpartition::= on page 12-24, add_hash_subpartition::= on page 12-24, add_list_
subpartition::= on page 12-24, update_index_clauses::= on page 12-30, parallel_clause::= on
page 12-6, alter_mapping_table_clauses::= on page 12-8)
modify_hash_partition::=
partition_attributes
MODIFY

partition_extended_name

alter_mapping_table_clause
REBUILD
UNUSABLE

LOCAL

INDEXES

(partition_extended_name::= on page 12-7, partition_attributes::= on page 12-30, alter_
mapping_table_clauses::= on page 12-8)
modify_list_partition::=
partition_attributes
,
ADD
VALUES

(

literal

)

DROP
add_range_subpartition
MODIFY

partition_extended_name

add_list_subpartition
add_hash_subpartition
update_index_clauses
COALESCE

parallel_clause

SUBPARTITION

REBUILD
UNUSABLE

LOCAL

INDEXES

(partition_extended_name::= on page 12-7, partition_attributes::= on page 12-30, add_
range_subpartition::= on page 12-24, add_list_subpartition::= on page 12-24, add_hash_
subpartition::= on page 12-24)

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-21

ALTER TABLE

modify_table_subpartition::=
allocate_extent_clause
deallocate_unused_clause
shrink_clause
LOB
MODIFY

LOB_item
(

subpartition_extended_name

VARRAY

modify_LOB_parameters

)

varray

REBUILD
UNUSABLE

LOCAL

INDEXES

,
ADD
VALUES

(

literal

)

DROP

(subpartition_extended_name::= on page 12-7, allocate_extent_clause::= on page 12-6,
deallocate_unused_clause::= on page 12-6, shrink_clause::= on page 12-6, modify_LOB_
parameters::= on page 12-16)
move_table_partition::=
MAPPING
MOVE

TABLE

table_partition_description

partition_extended_name

update_index_clauses

parallel_clause

(partition_extended_name::= on page 12-7, table_partition_description::= on page 12-28,
update_index_clauses::= on page 12-30, parallel_clause::= on page 12-6)
move_table_subpartition::=
partitioning_storage_clause
MOVE

update_index_clauses

subpartition_extended_name

parallel_clause

(subpartition_extended_name::= on page 12-7, partitioning_storage_clause::= on page 12-29,
update_index_clauses::= on page 12-30, parallel_clause::= on page 12-6)
add_table_partition::=
add_range_partition_clause
partition
ADD

add_hash_partition_clause

dependent_tables_clause

PARTITION
add_list_partition_clause
add_system_partition_clause

(add_range_partition_clause::= on page 12-23, add_hash_partition_clause::= on page 12-23,
add_list_partition_clause::= on page 12-23, add_system_partition_clause::= on page 12-23,
dependent_tables_clause:= on page 12-24
12-22 Oracle Database SQL Language Reference

ALTER TABLE

add_range_partition_clause::=
,
range_subpartition_desc
,
(

list_subpartition_desc

)

,
individual_hash_subparts
table_partition_description

hash_subparts_by_quantity

update_index_clauses

range_values_clause

(range_values_clause::= on page 12-27, table_partition_description::= on page 12-28, range_
subpartition_desc::= on page 12-29, list_subpartition_desc::= on page 12-29, individual_
hash_subparts::= on page 12-29, hash_subparts_by_quantity::= on page 12-29, update_
index_clauses::= on page 12-30)
add_hash_partition_clause::=
update_index_clauses

parallel_clause

partitioning_storage_clause

(partitioning_storage_clause::= on page 12-29, update_index_clauses::= on page 12-30,
parallel_clause::= on page 12-6)
add_list_partition_clause::=
,
range_subpartition_desc
,
(

list_subpartition_desc

)

,
individual_hash_subparts
table_partition_description

hash_subparts_by_quantity

update_index_clauses

list_values_clause

(list_values_clause::= on page 12-27, table_partition_description::= on page 12-28, range_
subpartition_desc::= on page 12-29, list_subpartition_desc::= on page 12-29, individual_
hash_subparts::= on page 12-29, hash_subparts_by_quantity::= on page 12-29, update_
index_clauses::= on page 12-30)
add_system_partition_clause::=
partition_name
BEFORE
partition_number

table_partition_description

update_index_clauses

(table_partition_description::= on page 12-28, update_index_clauses::= on page 12-30)

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-23

ALTER TABLE

add_range_subpartition::=
dependent_tables_clause
ADD

update_index_clauses

range_subpartition_desc

(range_subpartition_desc::= on page 12-29, update_index_clauses::= on page 12-30,
parallel_clause::= on page 12-6)
add_hash_subpartition::=
dependent_tables_clause
ADD

update_index_clauses

parallel_clause

individual_hash_subparts

(individual_hash_subparts::= on page 12-29, dependent_tables_clause:= on page 12-24,
update_index_clauses::= on page 12-30, parallel_clause::= on page 12-6)
add_list_subpartition::=
dependent_tables_clause
ADD

update_index_clauses

list_subpartition_desc

(list_subpartition_desc::= on page 12-29, update_index_clauses::= on page 12-30)
dependent_tables_clause:=
,
,
DEPENDENT

TABLES

(

table

(

partition_spec

)

)

(partition_spec::= on page 12-30)
coalesce_table_partition::=
update_index_clauses
COALESCE

parallel_clause

PARTITION

(update_index_clauses::= on page 12-30, parallel_clause::= on page 12-6)
coalesce_table_subpartition::=
update_index_clauses
COALESCE

SUBPARTITION

parallel_clause

subpartition

(update_index_clauses::= on page 12-30, parallel_clause::= on page 12-6)
drop_table_partition::=
parallel_clause
update_index_clauses
DROP

partition_extended_name

(partition_extended_name::= on page 12-7, update_index_clauses::= on page 12-30, parallel_
clause::= on page 12-6)

12-24 Oracle Database SQL Language Reference

ALTER TABLE

drop_table_subpartition::=
parallel_clause
update_index_clauses
DROP

subpartition_extended_name

(subpartition_extended_name::= on page 12-7, update_index_clauses::= on page 12-30,
parallel_clause::= on page 12-6)
rename_partition_subpart::=
partition_extended_name
RENAME

TO

new_name

subpartition_extended_name

(partition_extended_name::= on page 12-7, subpartition_extended_name::= on page 12-7)
truncate_partition_subpart::=
partition_extended_name
TRUNCATE
subpartition_extended_name

ALL
DROP

parallel_clause

STORAGE
REUSE

update_index_clauses

(partition_extended_name::= on page 12-7, subpartition_extended_name::= on page 12-7,
update_index_clauses::= on page 12-30, parallel_clause::= on page 12-6)
split_table_partition::=
SPLIT

partition_extended_name

,
AT

(

INTO

literal
,

VALUES

(

(

range_partition_desc

,

range_partition_desc

)

list_partition_desc

)

)

literal

split_nested_table_part

INTO

(

list_partition_desc

,

)

dependent_tables_clause

update_index_clauses

parallel_clause

(partition_extended_name::= on page 12-7, range_partition_desc::= on page 12-28, list_
partition_desc::= on page 12-28, dependent_tables_clause:= on page 12-24, update_index_
clauses::= on page 12-30, parallel_clause::= on page 12-6)

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-25

ALTER TABLE

split_nested_table_part::=
NESTED

TABLE

column

INTO

split_nested_table_part
(

nested_table_partition_spec

,

split_nested_table_part

nested_table_partition_spec

)

nested_table_partition_spec::=
segment_attributes_clause
PARTITION

partition

split_table_subpartition::=
SPLIT

subpartition_extended_name

,
AT

(

INTO

literal

(

range_subpartition_desc

,

range_subpartition_desc

)

)
,
INTO

literal
VALUES

(

(

list_subpartition_desc

,

list_subpartition_desc

)

)
NULL

dependent_tables_clause

update_index_clauses

parallel_clause

(subpartition_extended_name::= on page 12-7, range_subpartition_desc::= on page 12-29,
list_subpartition_desc::= on page 12-29, dependent_tables_clause:= on page 12-24, update_
index_clauses::= on page 12-30, parallel_clause::= on page 12-6)
merge_table_partitions::=
partition
MERGE

PARTITIONS

,
FOR

INTO

partition

partition_spec

(

partition_key_value

dependent_tables_clause

,
)

,
FOR

update_index_clauses

(

partition_key_value

)

parallel_clause

(partition_spec::= on page 12-30, dependent_tables_clause:= on page 12-24, update_index_
clauses::= on page 12-30, parallel_clause::= on page 12-6)

12-26 Oracle Database SQL Language Reference

ALTER TABLE

merge_table_subpartitions::=
subpartition
MERGE

subpartition

SUBPARTITIONS

,
FOR

(

,

subpartition_key_value

)

,
FOR

(

subpartition_key_value

range_subpartition_desc
INTO
list_subpartition_desc

dependent_tables_clause

update_index_clauses

parallel_clause

(range_subpartition_desc::= on page 12-29, list_subpartition_desc::= on page 12-29, update_
index_clauses::= on page 12-30, parallel_clause::= on page 12-6)
exchange_partition_subpart::=
schema

partition_extended_name
EXCHANGE

WITH

TABLE

.
table

subpartition_extended_name

INCLUDING

WITH
INDEXES

VALIDATION

EXCLUDING

WITHOUT

parallel_clause
exceptions_clause

update_index_clauses

(partition_extended_name::= on page 12-7, subpartition_extended_name::= on page 12-7,
exceptions_clause::= on page 12-27, update_index_clauses::= on page 12-30, parallel_
clause::= on page 12-6)
exceptions_clause::=
schema
EXCEPTIONS

.

INTO

table

range_values_clause::=
,
literal
VALUES

LESS

THAN

(

)
MAXVALUE

list_values_clause::=
,
literal
VALUES

(

NULL

)

DEFAULT

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-27

)

ALTER TABLE

table_partition_description::=
table_compression
deferred_segment_creation

segment_attributes_clause

key_compression

LOB_storage_clause
varray_col_properties

segment_attributes_clause
OVERFLOW

nested_table_col_properties

(deferred_segment_creation::= on page 12-13, segment_attributes_clause::= on page 12-8,
table_compression::= on page 12-5, key_compression::= on page 12-7, LOB_storage_
clause::= on page 12-14, varray_col_properties::= on page 12-13)
range_partition_desc::=
,
range_subpartition_desc
,
(

list_subpartition_desc

)

,
individual_hash_subparts
partition
PARTITION

hash_subparts_by_quantity
range_values_clause

table_partition_description

(range_values_clause::= on page 12-27, table_partition_description::= on page 12-28, range_
subpartition_desc::= on page 12-29, list_subpartition_desc::= on page 12-29)
list_partition_desc::=
,
range_subpartition_desc
,
(

list_subpartition_desc
,
individual_hash_subparts

partition
PARTITION

hash_subparts_by_quantity
list_values_clause

table_partition_description

(list_values_clause::= on page 12-27, table_partition_description::= on page 12-28, range_
subpartition_desc::= on page 12-29, list_subpartition_desc::= on page 12-29)

12-28 Oracle Database SQL Language Reference

)

ALTER TABLE

range_subpartition_desc::=
subpartition

partitioning_storage_clause

SUBPARTITION

range_values_clause

(range_values_clause::= on page 12-27, partitioning_storage_clause::= on page 12-29)
list_subpartition_desc::=
subpartition

partitioning_storage_clause

SUBPARTITION

list_values_clause

(list_values_clause::= on page 12-27, partitioning_storage_clause::= on page 12-29)
individual_hash_subparts::=
subpartition

partitioning_storage_clause

SUBPARTITION

hash_subparts_by_quantity::=
,
STORE
SUBPARTITIONS

IN

(

tablespace

)

integer

partitioning_storage_clause::=
TABLESPACE

tablespace
TABLESPACE

tablespace

OVERFLOW
table_compression
key_compression
LOB_partitioning_storage
SECUREFILE
BASICFILE
VARRAY

varray_item

STORE

AS

LOB

LOB_segname

table_compression::= on page 12-5, LOB_partitioning_storage::= on page 12-29
LOB_partitioning_storage::=
LOB

(

LOB_item

)

(

STORE

BASICFILE

LOB_segname

SECUREFILE

(

TABLESPACE

TABLESPACE

tablespace

tablespace

)

)

AS

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-29

ALTER TABLE

partition_attributes::=
physical_attributes_clause
logging_clause

physical_attributes_clause

allocate_extent_clause

logging_clause
OVERFLOW

deallocate_unused_clause

allocate_extent_clause

shrink_clause

deallocate_unused_clause

LOB

LOB_item
(

table_compression

VARRAY

modify_LOB_parameters

)

varray

(physical_attributes_clause::= on page 12-4, logging_clause::= on page 8-38, allocate_
extent_clause::= on page 12-6, deallocate_unused_clause::= on page 12-6, shrink_clause::=
on page 12-6, table_compression::= on page 12-5, modify_LOB_parameters::= on
page 12-16)
partition_spec::=
partition

table_partition_description

PARTITION

(table_partition_description::= on page 12-28)
update_index_clauses::=
update_global_index_clause
update_all_indexes_clause

(update_global_index_clause::= on page 12-30, update_all_indexes_clause::= on page 12-30)
update_global_index_clause::=
UPDATE
GLOBAL

INDEXES

INVALIDATE

update_all_indexes_clause::=
,
update_index_partition
(

index

(

)

)

update_index_subpartition
UPDATE

INDEXES

(update_index_partition::= on page 12-31, update_index_subpartition::= on page 12-31)

12-30 Oracle Database SQL Language Reference

ALTER TABLE

update_index_partition::=
,
index_subpartition_clause
index_partition_description

(index_partition_description::= on page 12-31, index_subpartition_clause::= on page 12-31)
update_index_subpartition::=
,
subpartition

TABLESPACE

tablespace

SUBPARTITION

index_partition_description::=
segment_attributes_clause
key_compression
PARAMETERS

(

’

ODCI_parameters

’

)

UNUSABLE

partition
PARTITION

(segment_attributes_clause::= on page 12-8, key_compression::= on page 12-7)
index_subpartition_clause::=
,
STORE

IN

(

tablespace

)
,

subpartition
(

TABLESPACE

tablespace

key_compression

UNUSABLE

SUBPARTITION

)

parallel_clause::=
NOPARALLEL
integer
PARALLEL

move_table_clause::=
ONLINE

segment_attributes_clause

table_compression

MOVE

LOB_storage_clause
index_org_table_clause

varray_col_properties

parallel_clause

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-31

ALTER TABLE

(segment_attributes_clause::= on page 12-8, table_compression::= on page 12-5, index_org_
table_clause::= on page 12-7, LOB_storage_clause::= on page 12-14, varray_col_
properties::= on page 12-13)
flashback_archive_clause::=
flashback_archive
FLASHBACK
NO

ARCHIVE

FLASHBACK

ARCHIVE

enable_disable_clause::=
,

VALIDATE
UNIQUE

NOVALIDATE

ENABLE

(

PRIMARY

column

)

KEY

DISABLE
CONSTRAINT

constraint_name

KEEP
INDEX
using_index_clause

exceptions_clause

CASCADE

DROP

(using_index_clause::= on page 12-32, exceptions_clause::= on page 12-27,)
using_index_clause::=
schema

.
index

USING

INDEX

(

create_index_statement

)

index_properties

(create_index::= on page 14-61, index_properties::= on page 12-32)
index_properties::=
global_partitioned_index
local_partitioned_index
index_attributes
domain_index_clause
INDEXTYPE

IS
XMLIndex_clause

(global_partitioned_index::= on page 14-64, local_partitioned_index::= on page 14-65—part
of CREATE INDEX, index_attributes::= on page 12-33, domain_index_clause: not
supported in using_index_clause)

12-32 Oracle Database SQL Language Reference

ALTER TABLE

index_attributes::=
physical_attributes_clause
logging_clause
ONLINE
tablespace
TABLESPACE
DEFAULT
key_compression
SORT
NOSORT
REVERSE
VISIBLE
INVISIBLE
parallel_clause

(physical_attributes_clause::= on page 12-4, logging_clause::= on page 8-38, key_
compression::= on page 12-7, parallel_clause: not supported in using_index_clause)

Semantics
12

Many clauses of the ALTER TABLE statement have the same functionality they have in a
CREATE TABLE statement. For more information on such clauses, see CREATE TABLE
on page 16-6.
Operations performed by the ALTER TABLE statement can cause
Oracle Database to invalidate procedures and stored functions that
access the table. For information on how and when the database
invalidates such objects, see Oracle Database Advanced Application
Developer's Guide.
Note:

schema
Specify the schema containing the table. If you omit schema, then Oracle Database
assumes the table is in your own schema.

table
Specify the name of the table to be altered.
If you alter a table that is a master table for one or more
materialized views, then Oracle Database marks the materialized
views INVALID. Invalid materialized views cannot be used by query
rewrite and cannot be refreshed. For information on revalidating a
materialized view, see ALTER MATERIALIZED VIEW on page 11-3.

Note:

See Also: Oracle Database Data Warehousing Guide for more
information on materialized views in general

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-33

ALTER TABLE

Restrictions on Altering Temporary Tables You can modify, drop columns from, or
rename a temporary table. However, for a temporary table you cannot:
■
■

■

■

Add columns of nested table type. You can add columns of other types.
Specify referential integrity (foreign key) constraints for an added or modified
column.
Specify the following clauses of the LOB_storage_clause for an added or modified
LOB column: TABLESPACE, storage_clause, logging_clause, allocate_extent_
clause, or deallocate_unused_clause.
Specify the physical_attributes_clause, nested_table_col_properties,
parallel_clause, allocate_extent_clause, deallocate_unused_clause, or any
of the index-organized table clauses.

■

Exchange partitions between a partition and a temporary table.

■

Specify the logging_clause.

■

Specify MOVE.

You can add, drop, or modify the columns
of an external table. However, for an external table you cannot:

Restrictions on Altering External Tables
■

Add a LONG, LOB, or object type column or change the data type of an external
table column to any of these data types.

■

Add a constraint to an external table.

■

Modify the storage parameters of an external table.

■

Specify the logging_clause.

■

Specify MOVE.

alter_table_properties
Use the alter_table_clauses to modify a database table.

physical_attributes_clause
The physical_attributes_clause lets you change the value of the PCTFREE, PCTUSED,
and INITRANS parameters and storage characteristics. Refer to physical_attributes_clause
on page 8-44 and storage_clause on page 8-46 for a full description of these parameters
and characteristics.
Restrictions on Altering Table Physical Attributes

Altering physical attributes is

subject to the following restrictions:
■

■

■

You cannot specify the PCTUSED parameter for the index segment of an
index-organized table.
If you attempt to alter the storage attributes of tables in locally managed
tablespaces, then Oracle Database raises an error. However, if some segments of a
partitioned table reside in a locally managed tablespace and other segments reside
in a dictionary-managed tablespace, then the database alters the storage attributes
of the segments in the dictionary-managed tablespace but does not alter the
attributes of the segments in the locally managed tablespace, and does not raise an
error.
For segments with automatic segment-space management, the database ignores
attempts to change the PCTUSED setting. If you alter the PCTFREE setting, then you

12-34 Oracle Database SQL Language Reference

ALTER TABLE

must subsequently run the DBMS_REPAIR.SEGMENT_FIX_STATUS procedure to
implement the new setting on blocks already allocated to the segment.
Cautions on Altering Tables Physical Attributes

The values you specify in this clause

affect the table as follows:
■

■

■

For a nonpartitioned table, the values you specify override any values specified
for the table at create time.
For a range-, list-, or hash-partitioned table, the values you specify are the default
values for the table and the actual values for every existing partition, overriding
any values already set for the partitions. To change default table attributes without
overriding existing partition values, use the modify_table_default_attrs clause.
For a composite-partitioned table, the values you specify are the default values for
the table and all partitions of the table and the actual values for all subpartitions of
the table, overriding any values already set for the subpartitions. To change
default partition attributes without overriding existing subpartition values, use the
modify_table_default_attrs clause with the FOR PARTITION clause.

logging_clause
Use the logging_clause to change the logging attribute of the table. The logging_
clause specifies whether subsequent ALTER TABLE ... MOVE and ALTER TABLE ... SPLIT
operations will be logged or not logged.
When used with the modify_table_default_attrs clause, this clause affects the
logging attribute of a partitioned table.
See Also:
■
■

logging_clause on page 8-38 for a full description of this clause
Oracle Database VLDB and Partitioning Guide for more information
about the logging_clause and parallel DML

table_compression
The table_compression clause is valid only for heap-organized tables. Use this clause
to instruct Oracle Database whether to compress data segments to reduce disk and
memory use. Refer to the CREATE TABLE table_compression on page 16-34 for the full
semantics of this clause and for information on creating objects with table
compression.
The first time a table is altered in such a way that compressed
data will be added, all bitmap indexes and bitmap index partitions on
that table must be marked UNUSABLE.

Note:

See Also: Oracle Database Data Warehousing Guide for information on
table compression usage scenarios

supplemental_table_logging
Use the supplemental_table_logging clause to add or drop a redo log group or one
or more supplementally logged columns in a redo log group.
■

In the ADD clause, use supplemental_log_grp_clause to create named
supplemental log group. Use the supplemental_id_key_clause to create a
system-generated log group.

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-35

ALTER TABLE

■

On the DROP clause, use GROUP log_group syntax to drop a named supplemental
log group and use the supplemental_id_key_clause to drop a system-generated
log group.

The supplemental_log_grp_clause and the supplemental_id_key_clause have the
same semantics in CREATE TABLE and ALTER TABLE statements. For full information on
these clauses, refer to supplemental_log_grp_clause on page 16-31 and supplemental_id_
key_clause on page 16-31 in the documentation on CREATE TABLE.
See Also: Oracle Data Guard Concepts and Administration for
information on supplemental redo log groups

allocate_extent_clause
Use the allocate_extent_clause to explicitly allocate a new extent for the table, the
partition or subpartition, the overflow data segment, the LOB data segment, or the
LOB index.
You cannot allocate an extent for a
temporary table or for a range- or composite-partitioned table.

Restriction on Allocating Table Extents

See Also: allocate_extent_clause on page 8-2 for a full description of
this clause and "Allocating Extents: Example" on page 12-86

deallocate_unused_clause
Use the deallocate_unused_clause to explicitly deallocate unused space at the end of
the table, partition or subpartition, overflow data segment, LOB data segment, or LOB
index and make the space available for other segments in the tablespace.
See Also: deallocate_unused_clause on page 8-27 for a full description
of this clause and "Deallocating Unused Space: Example" on
page 12-81

shrink_clause
The shrink clause lets you manually shrink space in a table, index-organized table or
its overflow segment, index, partition, subpartition, LOB segment, materialized view,
or materialized view log. This clause is valid only for segments in tablespaces with
automatic segment management. By default, Oracle Database compacts the segment,
adjusts the high water mark, and releases the recuperated space immediately.
Compacting the segment requires row movement. Therefore, you must enable row
movement for the object you want to shrink before specifying this clause. Further, if
your application has any rowid-based triggers, you should disable them before issuing
this clause.
Do not attempt to enable row movement for an
index-organized table before specifying the shrink_clause. The ROWID
of an index-organized table is its primary key, which never changes.
Therefore, row movement is neither relevant nor valid for such tables.
Note:

If you specify COMPACT, then Oracle Database only defragments the
segment space and compacts the table rows for subsequent release. The database does
not readjust the high water mark and does not release the space immediately. You
must issue another ALTER TABLE ... SHRINK SPACE statement later to complete the

COMPACT

12-36 Oracle Database SQL Language Reference

ALTER TABLE

operation. This clause is useful if you want to accomplish the shrink operation in two
shorter steps rather than one longer step.
For an index or index-organized table, specifying ALTER [INDEX | TABLE] ... SHRINK
SPACE COMPACT is equivalent to specifying ALTER [INDEX | TABLE ... COALESCE. The
shrink_clause can be cascaded (refer to the CASCADE clause, which follows) and
compacts the segment more densely than does a coalesce operation, which can
improve performance. However, if you do not want to release the unused space, then
you can use the appropriate COALESCE clause.
If you specify CASCADE, then Oracle Database performs the same
operations on all dependent objects of table, including secondary indexes on
index-organized tables.

CASCADE

Restrictions on the shrink_clause The shrink_clause is subject to the following
restrictions:
■

You cannot combine this clause with any other clauses in the same ALTER TABLE
statement.
You cannot specify this clause for a cluster, a clustered table, or any object with a
LONG column.

■

■

■
■

Segment shrink is not supported for tables with function-based indexes, domain
indexes, or bitmap join indexes.
This clause does not shrink mapping tables of index-organized tables, even if you
specify CASCADE.
You cannot specify this clause for a compressed table.
You cannot shrink a table that is the master table of an ON COMMIT materialized
view. Rowid materialized views must be rebuilt after the shrink operation.

CACHE | NOCACHE
The CACHE and NOCACHE clauses have the same semantics in CREATE TABLE and ALTER
TABLE statements. For complete information on these clauses, refer to "CACHE |
NOCACHE | CACHE READS" on page 16-62 in the documentation on CREATE TABLE.
If you omit both of these clauses in an ALTER TABLE statement, then the existing value
is unchanged.

RESULT_CACHE
The RESULT_CACHE clause has the same semantics in CREATE TABLE and ALTER TABLE
statements. For complete information on this clause, refer to "RESULT_CACHE
Clause" on page 16-63 in the documentation on CREATE TABLE. If you omit this clause in
an ALTER TABLE statement, then the existing setting is unchanged.
upgrade_table_clause
The upgrade_table_clause is relevant for object tables and for relational tables with
object columns. It lets you instruct Oracle Database to convert the metadata of the
target table to conform with the latest version of each referenced type. If table is
already valid, then the table metadata remains unchanged.
Restriction on Upgrading Object Tables and Columns Within this clause, you cannot
specify object_type_col_properties as a clause of column_properties.
INCLUDING DATA Specify INCLUDING DATA if you want Oracle Database to convert
the data in the table to the latest type version format. You can define the storage for
SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-37

ALTER TABLE

any new column while upgrading the table by using the column_properties and the
LOB_partition_storage. This is the default.
You can convert data in the table at the time you upgrade the type by specifying
CASCADE INCLUDING TABLE DATA in the dependent_handling_clause of the ALTER TYPE
statement. See Oracle Database PL/SQL Language Reference for information on this
clause. For information on whether a table contains data based on an older type
version, refer to the DATA_UPGRADED column of the USER_TAB_COLUMNS data dictionary
view.
Specify NOT INCLUDING DATA if you want Oracle Database to
leave column data unchanged.

NOT INCLUDING DATA

Restriction on NOT INCLUDING DATA You cannot specify NOT INCLUDING DATA if the

table contains columns in Oracle8 release 8.0.x image format. To determine whether
the table contains such columns, refer to the V80_FMT_IMAGE column of the USER_TAB_
COLUMNS data dictionary view.
See Also:
■

■

Oracle Database Reference for information on the data dictionary
views
ALTER TYPE on page 13-4 for information on converting
dependent table data when modifying a type upon which the
table depends

records_per_block_clause
The records_per_block_clause lets you specify whether Oracle Database restricts the
number of records that can be stored in a block. This clause ensures that any bitmap
indexes subsequently created on the table will be as compressed as possible.
Restrictions on Records in a Block

The record_per_block_clause is subject to the

following restrictions:
■

■

You cannot specify either MINIMIZE or NOMINIMIZE if a bitmap index has already
been defined on table. You must first drop the bitmap index.
You cannot specify this clause for an index-organized table or a nested table.

MINIMIZE Specify MINIMIZE to instruct Oracle Database to calculate the largest
number of records in any block in the table and to limit future inserts so that no block
can contain more than that number of records.

Oracle recommends that a representative set of data already exist in the table before
you specify MINIMIZE. If you are using table compression (see table_compression on
page 12-35), then a representative set of compressed data should already exist in the
table.
Restriction on MINIMIZE You cannot specify MINIMIZE for an empty table.
NOMINIMIZE

Specify NOMINIMIZE to disable the MINIMIZE feature. This is the default.

row_movement_clause
You cannot disable row movement in a reference-partitioned table unless row
movement is also disabled in the parent table. Otherwise, this clause has the same
semantics in CREATE TABLE and ALTER TABLE statements. For complete information on
these clauses, refer to row_movement_clause on page 16-65 in the documentation on
12-38 Oracle Database SQL Language Reference

ALTER TABLE

CREATE TABLE.
flashback_archive_clause
You must have the FLASHBACK ARCHIVE object privilege on the specified flashback data
archive to specify this clause. Use this clause to enable or disable historical tracking for
the table.
■

Specify FLASHBACK ARCHIVE to enable tracking for the table. You can specify
flashback_archive to designate a particular flashback data archive for this table.
The flashback data archive you specify much already exist.
If you omit the archive name, then the database uses the default flashback data
archive designated for the system. If no default flashback data archive has been
designated for the system, then you must specify flashback_archive.
You cannot specify FLASHBACK ARCHIVE to change the flashback data archive for this
table. Instead you must first issue an ALTER TABLE statement with the NO FLASHBACK
ARCHIVE clause and then issue an ALTER TABLE statement with the FLASHBACK
ARCHIVE clause.

■

Specify NO FLASHBACK ARCHIVE to disable tracking for the table.
See Also: The CREATE TABLE flashback_archive_clause on page 16-66
for information on creating a table with tracking enabled and CREATE
FLASHBACK ARCHIVE on page 14-55 for information on creating
default flashback data archives

RENAME TO
Use the RENAME clause to rename table to new_table_name.
Using this clause invalidates any dependent materialized views. For more information
on materialized views, see CREATE MATERIALIZED VIEW on page 15-4 and Oracle
Database Data Warehousing Guide.
If a domain index is defined on the table, then the database invokes the
ODCIIndexAlter() method with the RENAME option. This operation establishes
correspondence between the indextype metadata and the base table.

READ ONLY | READ WRITE
Specify READ ONLY to put the table in read-only mode. When the table is in READ ONLY
mode, you cannot issue any DML statements that affect the table or any SELECT ... FOR
UPDATE statements. You can issue DDL statements as long as they do not modify any
table data. Operations on indexes associated with the table are allowed when the table
is in READ ONLY mode.
Specify READ WRITE to return a read-only table to read/write mode.

REKEY encryption_spec
Use the REKEY clause to generate a new encryption key or to switch between different
algorithms. This operation returns only after all encrypted columns in the table,
including LOB columns, have been reencrypted.
alter_iot_clauses
index_org_table_clause
This clause lets you alter some of the characteristics of an existing index-organized
table. Index-organized tables keep data sorted on the primary key and are therefore
SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-39

ALTER TABLE

best suited for primary-key-based access and manipulation. See index_org_table_clause
on page 16-37 in the context of CREATE TABLE.
See Also:

"Modifying Index-Organized Tables: Examples" on

page 12-82
key_compression
This clause is relevant only if table is index organized. Specify COMPRESS to instruct
Oracle Database to combine the primary key index blocks of the index-organized table
where possible to free blocks for reuse. You can specify this clause with the parallel_
clause.
PCTTHRESHOLD integer Refer to "PCTTHRESHOLD integer" on page 16-38 in the
documentation of CREATE TABLE.

Refer to "INCLUDING column_name" on page 16-39 in
the documentation of CREATE TABLE.

INCLUDING column_name

overflow_attributes
The overflow_attributes let you specify the overflow data segment physical storage
and logging attributes to be modified for the index-organized table. Parameter values
specified in this clause apply only to the overflow data segment.
See Also:

CREATE TABLE on page 16-6

add_overflow_clause
The add_overflow_clause lets you add an overflow data segment to the specified
index-organized table. You can also use this clause to explicitly allocate an extent to or
deallocate unused space from an existing overflow segment.
Use the STORE IN tablespace clause to specify tablespace storage for the entire
overflow segment. Use the PARTITION clause to specify tablespace storage for the
segment by partition.
For a partitioned index-organized table:
■

■

If you do not specify PARTITION, then Oracle Database automatically allocates an
overflow segment for each partition. The physical attributes of these segments are
inherited from the table level.
If you want to specify separate physical attributes for one or more partitions, then
you must specify such attributes for every partition in the table. You need not
specify the name of the partitions, but you must specify their attributes in the
order in which they were created.

You can find the order of the partitions by querying the PARTITION_NAME and
PARTITION_POSITION columns of the USER_IND_PARTITIONS view.
If you do not specify TABLESPACE for a particular partition, then the database uses the
tablespace specified for the table. If you do not specify TABLESPACE at the table level,
then the database uses the tablespace of the partition primary key index segment.
Restrictions on Overflow Attributes
■
■

Within the segment_attributes_clause:

You cannot specify the OPTIMAL parameter of the physical_attributes_clause.
You cannot specify tablespace storage for the overflow segment using this clause.
For a nonpartitioned table, you can use ALTER TABLE ... MOVE ... OVERFLOW to move
the segment to a different tablespace. For a partitioned table, use ALTER TABLE ...

12-40 Oracle Database SQL Language Reference

ALTER TABLE

MODIFY DEFAULT ATTRIBUTES ... OVERFLOW to change the default tablespace of the
overflow segment.
Additional restrictions apply if table is in a locally managed tablespace, because in
such tablespaces several segment attributes are managed automatically by the
database.
See Also: allocate_extent_clause on page 8-2 and deallocate_unused_
clause on page 8-27 for full descriptions of these clauses of the add_
overflow_clause

alter_overflow_clause
The alter_overflow_clause lets you change the definition of the overflow segment of
an existing index-organized table.
The restrictions that apply to the add_overflow_clause also apply to the alter_
overflow_clause.
When you add a column to an index-organized table, Oracle
Database evaluates the maximum size of each column to estimate the
largest possible row. If an overflow segment is needed but you have
not specified OVERFLOW, then the database raises an error and does not
execute the ALTER TABLE statement. This checking function guarantees
that subsequent DML operations on the index-organized table will not
fail because an overflow segment is lacking.
Note:

alter_mapping_table_clauses
The alter_mapping_table_clauses is valid only if table is index organized and has a
mapping table.
allocate_extent_clause Use the allocate_extent_clause to allocate a new extent at
the end of the mapping table for the index-organized table. Refer to allocate_extent_
clause on page 8-2 for a full description of this clause.
deallocate_unused_clause Specify the deallocate_unused_clause to deallocate

unused space at the end of the mapping table of the index-organized table. Refer to
deallocate_unused_clause on page 8-27 for a full description of this clause.
Oracle Database automatically maintains all other attributes of the mapping table or
its partitions.

COALESCE Clause
Specify COALESCE to instruct Oracle Database to merge the contents of index blocks of
the index the database uses to maintain the index-organized table where possible to
free blocks for reuse. Refer to the shrink_clause on page 12-36 for information on the
relationship between these two clauses.
alter_XMLSchema_clause
This clause is valid as part of alter_table_properties only if you are modifying an
XMLType table with BINARY XML storage. Refer to XMLSchema_spec on page 16-70 in the
documentation on CREATE TABLE for more information on the ALLOW and DISALLOW
clauses.

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-41

ALTER TABLE

column_clauses
Use these clauses to add, drop, or otherwise modify a column.

add_column_clause
The add_column_clause lets you add a column to a table.
See Also: CREATE TABLE on page 16-6 for a description of the
keywords and parameters of this clause and "Adding a Table Column:
Example" on page 12-85

column_definition
Unless otherwise noted in this section, the elements of column_definition have the
same behavior when adding a column to an existing table as they do when creating a
new table. Refer to column_definition on page 16-26 for information.
The SORT parameter is valid only when creating a
new table. You cannot specify SORT in the column_definition of an ALTER TABLE ... ADD
statement.
Restriction on column_definition

When you add a column, the initial value of each row for the new column is null.
■

If you specify the DEFAULT clause for a NOT NULL column, then the default value is
stored as metadata but the column itself is not populated with data. However,
subsequent queries that specify the new column are rewritten so that the default
value is returned in the result set.
This optimized behavior differs from earlier releases, when as part of the ALTER
TABLE operation Oracle Database updated each row in the newly created column
with the default value, and then fired any update triggers defined on the table. In
this release, no triggers are fired because the default is stored only as metadata.
The optimized behavior is subject to the following restrictions:
■

■

■

■

The table cannot have any LOB columns. It cannot be index-organized,
temporary, or part of a cluster. It also cannot be a queue table, an object table,
or the container table of a materialized view.
If the table has a Virtual Private Database (VPD) policy on it, then the
optimized behavior will not take place unless the user who issues the ALTER
TABLE ... ADD statement has the EXEMPT ACCESS POLICY system privilege.
The column being added cannot be encrypted, and cannot be an object
column, nested table column, or a LOB column.

If you specify the DEFAULT clause for a nullable column, then the default value is
added to existing rows as part of this ALTER TABLE statement, and any update
triggers defined on the table are fired. This behavior also results if you change a
NOT NULL column with a default value to be nullable.

You can add an overflow data segment to each partition of a partitioned
index-organized table.
You can add LOB columns to nonpartitioned and partitioned tables. You can specify
LOB storage at the table and at the partition or subpartition level.
If you previously created a view with a query that used the SELECT * syntax to select
all columns from table, and you now add a column to table, then the database does
not automatically add the new column to the view. To add the new column to the
view, re-create the view using the CREATE VIEW statement with the OR REPLACE clause.
Refer to CREATE VIEW on page 17-14 for more information.

12-42 Oracle Database SQL Language Reference

ALTER TABLE

Restrictions on Adding Columns The addition of columns is subject to the following
restrictions:
■
■

■

■

You cannot add a LOB column to a cluster table.
If you add a LOB column to a hash-partitioned table, then the only attribute you
can specify for the new partition is TABLESPACE.
You cannot add a column with a NOT NULL constraint if table has any rows unless
you also specify the DEFAULT clause.
If you specify this clause for an index-organized table, then you cannot specify any
other clauses in the same statement.

DEFAULT
Use the DEFAULT clause to specify a default for a new column or a new default for an
existing column. Oracle Database assigns this value to the column if a subsequent
INSERT statement omits a value for the column. If you are adding a new column to the
table and specify the default value, then the database inserts the default column value
into all rows of the table.
The data type of the default value must match the data type specified for the column.
The column must also be large enough to hold the default value.
Restrictions on Default Column Values

Default column values are subject to the

following restrictions:
■

■

A DEFAULT expression cannot contain references to other columns, the
pseudocolumns CURRVAL, NEXTVAL, LEVEL, and ROWNUM, or date constants that are
not fully specified.
The expression can be of any form except a scalar subquery expression.
See Also: "Specifying Default Column Value: Examples" on
page 12-86

inline_constraint
Use inline_constraint to add a constraint to the new column.
inline_ref_constraint
This clause lets you describe a new column of type REF. Refer to constraint on page 8-4
for syntax and description of this type of constraint, including restrictions.
virtual_column_definition
The virtual_column_definition has the same semantics when you add a column that
it has when you create a column.
See Also: The CREATE TABLE virtual_column_definition on page 16-29
and "Adding a Virtual Table Column: Example" on page 12-85 for
more information

You cannot add a virtual column when the
SQL expression for the virtual column involves a column on which an Oracle Data
Redaction policy is defined.

Restriction on Adding a Virtual Column

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-43

ALTER TABLE

column_properties
The clauses of column_properties determine the storage characteristics of an object
type, nested table, varray, or LOB column.
object_type_col_properties This clause is valid only when you are adding a new
object type column or attribute. To modify the properties of an existing object type
column, use the modify_column_clauses. The semantics of this clause are the same as
for CREATE TABLE unless otherwise noted.

Use the object_type_col_properties clause to specify storage characteristics for a
new object column or attribute or an element of a collection column or attribute.
For complete information on this clause, refer to object_type_col_properties on
page 16-42 in the documentation on CREATE TABLE.
nested_table_col_properties The nested_table_col_properties clause lets you
specify separate storage characteristics for a nested table, which in turn lets you to
define the nested table as an index-organized table. You must include this clause when
creating a table with columns or column attributes whose type is a nested table.
(Clauses within this clause that function the same way they function for parent object
tables are not repeated here. See the CREATE TABLE clause nested_table_col_properties on
page 16-48 for more information about these clauses.)
■

For nested_item, specify the name of a column (or a top-level attribute of the
nested table object type) whose type is a nested table.
If the nested table is a multilevel collection, and the inner nested table does not
have a name, then specify COLUMN_VALUE in place of the nested_item name.

■

For storage_table, specify the name of the table where the rows of nested_item
reside. The storage table is created in the same schema and the same tablespace as
the parent table.

Restrictions on Nested Table Column Properties Nested table column properties are
subject to the following restrictions:
■

You cannot specify the parallel_clause.

■

You cannot specify CLUSTER as part of the physical_properties clause.
See Also:

"Nested Tables: Examples" on page 12-88

varray_col_properties The varray_col_properties clause lets you specify separate
storage characteristics for the LOB in which a varray will be stored. If you specify this
clause, then Oracle Database will always store the varray in a LOB, even if it is small
enough to be stored inline. If varray_item is a multilevel collection, then the database
stores all collection items nested within varray_item in the same LOB in which
varray_item is stored.
Restriction on Varray Column Properties You cannot specify TABLESPACE as part of

LOB_parameters for a varray column. The LOB tablespace for a varray defaults to the
tablespace of the containing table.
out_of_line_part_storage
This clause lets you specify storage attributes the newly added column for each
partition or subpartition in a partitioned table. For any partition or subpartition you
do not name in this clause, the storage attributes for the new column are the same as
those specified in the nested_table_col_properties at the table level.

12-44 Oracle Database SQL Language Reference

ALTER TABLE

LOB_storage_clause
Use the LOB_storage_clause to specify the LOB storage characteristics for a newly
added LOB column, LOB partition, or LOB subpartition, or when you are converting a
LONG column into a LOB column. You cannot use this clause to modify an existing
LOB. Instead, you must use the modify_LOB_storage_clause on page 12-53.
Unless otherwise noted in this section, all LOB parameters, in both the LOB_storage_
clause and the modify_LOB_storage_clause, have the same semantics in an ALTER
TABLE statement that they have in a CREATE TABLE statement. Refer to the CREATE TABLE
LOB_storage_clause on page 16-42 for complete information on this clause.
Restriction on LOB Parameters The only parameter of LOB_parameters you can
specify for a hash partition or hash subpartition is TABLESPACE.
CACHE READS Clause When you add a new LOB column, you can specify the
logging attribute with CACHE READS, as you can when defining a LOB column at create
time. Refer to the CREATE TABLE clause CACHE READS on page 16-62 for full
information on this clause.
ENABLE | DISABLE STORAGE IN ROW You cannot change STORAGE IN ROW once it is
set. Therefore, you cannot specify this clause as part of the modify_col_properties
clause. However, you can change this setting when adding a new column (add_
column_clause) or when moving the table (move_table_clause). Refer to the CREATE TABLE
clause ENABLE STORAGE IN ROW on page 16-44 for complete information on this
clause.
CHUNK integer You use cannot use the modify_col_properties clause to change the
value of CHUNK after it has been set. If you require a different CHUNK value for a column
after it has been created, use ALTER TABLE … MOVE. Refer to the CREATE TABLE clause
CHUNK integer on page 16-44 for more information.
RETENTION For BasicFiles LOBs, if the database is in automatic undo mode, then

you can specify RETENTION instead of PCTVERSION to instruct Oracle Database to retain
old versions of this LOB. This clause overrides any prior setting of PCTVERSION. Refer
to the CREATE TABLE clause LOB_retention_clause on page 16-44 for a full description of
this parameter.
For BasicFiles LOBs, if the database is in automatic undo mode,
then you can use this clause to specify the number of freelist groups for this LOB. This
clause overrides any prior setting of FREELIST GROUPS. Refer to the CREATE TABLE clause
FREEPOOLS integer on page 16-45 for a full description of this parameter. The
database ignores this parameter for SecureFiles LOBs.

FREEPOOLS integer

LOB_partition_storage
You can specify only one list of LOB_partition_storage clauses in a single ALTER
TABLE statement, and all LOB_storage_clauses and varray_col_properties clause
must precede the list of LOB_partition_storage clauses. Refer to the CREATE TABLE
clause LOB_partition_storage on page 16-46 for full information on this clause,
including restrictions.
XMLType_column_properties Refer to the CREATE TABLE clause XMLType_column_

properties on page 16-49 for a full description of this clause.

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-45

ALTER TABLE

See Also:
■

■

■

LOB_storage_clause on page 12-45 for information on the LOB_
segname and LOB_parameters clauses
"XMLType Column Examples" on page 16-76 for an example of
XMLType columns in object-relational tables and "Using XML in
SQL Statements" on page F-8 for an example of creating an
XMLSchema
Oracle XML DB Developer's Guide for more information on XMLType
columns and tables and on creating an XMLSchema

modify_column_clauses
Use the modify_column_clauses to modify the properties of an existing column or the
substitutability of an existing object type column.
See Also:

"Modifying Table Columns: Examples" on page 12-86

modify_col_properties
Use this clause to modify the properties of the column. Any of the optional parts of the
column definition (data type, default value, or constraint) that you omit from this
clause remain unchanged.
datatype You can change the data type of any column if all rows of the column
contain nulls. However, if you change the data type of a column in a materialized view
container table, then Oracle Database invalidates the corresponding materialized view.

You can omit the data type only if the statement also designates the column as part of
the foreign key of a referential integrity constraint. The database automatically assigns
the column the same data type as the corresponding column of the referenced key of
the referential integrity constraint.
You can always increase the size of a character or raw column or the precision of a
numeric column, whether or not all the rows contain nulls. You can reduce the size of a
data type of a column as long as the change does not require data to be modified. The
database scans existing data and returns an error if data exists that exceeds the new
length limit.
You can change a DATE column to a TIMESTAMP or TIMESTAMP WITH LOCAL TIME ZONE
column, and you can change a TIMESTAMP or TIMESTAMP WITH LOCAL TIME ZONE column
to a DATE column. The following rules apply:
■

■

When you change a TIMESTAMP or TIMESTAMP WITH LOCAL TIME ZONE column to a
DATE column, Oracle Database updates each column value that has non-zero
fractional seconds by rounding the value to the nearest second. If, while updating
such a value, Oracle Database encounters a minute field greater than or equal to 60
(which can occur in a boundary case when the daylight saving rule switches), then
it updates the minute field by subtracting 60 from it.
After you change a TIMESTAMP WITH LOCAL TIME ZONE column to a DATE column, the
values in the column still represent the local time that they represented in the
database time zone. However, the database time zone is no longer associated with
the values. When queried in SQL*Plus, the values are no longer automatically
adjusted to the session time zone. It is now the responsibility of applications
processing the column values to interpret them in a particular time zone.

If the table is empty, then you can increase or decrease the leading field or the
fractional second value of a datetime or interval column. If the table is not empty, then

12-46 Oracle Database SQL Language Reference

ALTER TABLE

you can only increase the leading field or fractional second of a datetime or interval
column.
You can use the TO_LOB function to change a LONG column to a CLOB or NCLOB column,
and a LONG RAW column to a BLOB column. However, you cannot use the TO_LOB
function from within a PL/SQL package. Instead use the TO_CLOB or TO_BLOB
functions.
■

■

■

The modified LOB column inherits all constraints and triggers that were defined
on the original LONG column. If you want to change any constraints, then you must
do so in a subsequent ALTER TABLE statement.
If any domain indexes are defined on the LONG column, then you must drop them
before modifying the column to a LOB.
After the modification, you will have to rebuild all other indexes on all columns of
the table.

You can use the TO_CLOB function to convert NCLOB columns CLOB columns.
See Also:
■

■

Oracle Database SecureFiles and Large Objects Developer's Guide for
information on LONG to LOB migration
ALTER INDEX on page 10-78 for information on dropping and
rebuilding indexes

For CHAR and VARCHAR2 columns, you can change the length semantics by specifying
CHAR (to indicate character semantics for a column that was originally specified in
bytes) or BYTE (to indicate byte semantics for a column that was originally specified in
characters). To learn the length semantics of existing columns, query the CHAR_USED
column of the ALL_, USER_, or DBA_TAB_COLUMNS data dictionary view.
See Also:
■

■

Oracle Database Globalization Support Guide for information on byte
and character semantics
Oracle Database Reference for information on the data dictionary
views

ENCRYPT encryption_spec | DECRYPT Use this clause to decrypt an encrypted

column, to encrypt an unencrypted column, or to change the integrity algorithm or the
SALT option of an encrypted column.
When encrypting an existing column, if you specify encryption_spec, it must match
the encryption specification of any other encrypted columns in the same table. Refer to
the CREATE TABLE clause encryption_spec on page 16-27 for additional information and
restrictions on the encryption_spec.
If a materialized view log is defined on the table, then Oracle Database encrypts or
decrypts in the materialized view log any columns you encrypt or decrypt in this
clause.
Restrictions on ENCRYPT encryption_spec | DECRYPT: This clause is subject to the
following restrictions:
■

If the new or existing column is a LOB column, then it must be stored as a
SecureFiles LOB, and you cannot specify the SALT option.

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-47

ALTER TABLE

■

You cannot encrypt or decrypt a column on which a fine-grained audit policy for
the UPDATE statement is enabled. However, you can disable the fine-grained audit
policy, encrypt or decrypt the column, and then enable the fine-grained audit
policy.
See Also:

"Data Encryption: Examples" on page 12-86

inline_constraint This clause lets you add a constraint to a column you are

modifying. To change the state of existing constraints on existing columns, use the
constraint_clauses.
The LOB_storage_clause is permitted within modify_col_
properties only if you are converting a LONG column to a LOB column. In this case
only, you can specify LOB storage for the column using the LOB_storage_clause.
However, you can specify only the single column as a LOB_item. Default LOB storage
attributes are used for any attributes you omit in the LOB_storage_clause.
LOB_storage_clause

alter_XMLSchema_clause This clause is valid within modify_col_properties only
for XMLType tables with BINARY XML storage. Refer to XMLSchema_spec on page 16-70 in
the documentation on CREATE TABLE for more information on the ALLOW and DISALLOW
clauses.
Restrictions on Modifying Column Properties The modification of column
properties is subject to the following restrictions:
■
■

■

■

■

■

■

■

You cannot change the data type of a LOB column.
You cannot modify a column of a table if a domain index is defined on the column.
You must first drop the domain index and then modify the column.
You cannot modify the data type or length of a column that is part of the
partitioning or subpartitioning key of a table or index.
You can change a CHAR column to VARCHAR2 (or VARCHAR) and a VARCHAR2 (or
VARCHAR) column to CHAR only if the BLANK_TRIMMING initialization parameter is set
to TRUE and the column size stays the same or increases. If the BLANK_TRIMMING
initialization parameter is set to TRUE, then you can also reduce the column size to
any size greater than or equal to the maximum trimmed data value.
You cannot change a LONG or LONG RAW column to a LOB if the table is part of a
cluster. If you do change a LONG or LONG RAW column to a LOB, then the only other
clauses you can specify in this ALTER TABLE statement are the DEFAULT clause and
the LOB_storage_clause.
You can specify the LOB_storage_clause as part of modify_col_properties only
when you are changing a LONG or LONG RAW column to a LOB.
You cannot specify a column of data type ROWID for an index-organized table, but
you can specify a column of type UROWID.
You cannot change the data type of a column to REF.
See Also: ALTER MATERIALIZED VIEW on page 11-3 for
information on revalidating a materialized view

modify_col_substitutable
Use this clause to set or change the substitutability of an existing object type column.

12-48 Oracle Database SQL Language Reference

ALTER TABLE

The FORCE keyword drops any hidden columns containing typeid information or data
for subtype attributes. You must specify FORCE if the column or any attributes of its
type are not FINAL.
Restrictions on Modifying Column Substitutability The modification of column
substitutability is subject to the following restrictions:
■
■

■

■

You can specify this clause only once in any ALTER TABLE statement.
You cannot modify the substitutability of a column in an object table if the
substitutability of the table itself has been set.
You cannot specify this clause if the column was created or added using the IS OF
TYPE syntax, which limits the range of subtypes permitted in an object column or
attribute to a particular subtype. Refer to substitutable_column_clause on page 16-42
in the documentation on CREATE TABLE for information on the IS OF TYPE syntax.
You cannot change a varray column to NOT SUBSTITUTABLE, even by specifying
FORCE, if any of its attributes are nested object types that are not FINAL.

drop_column_clause
The drop_column_clause lets you free space in the database by dropping columns you
no longer need or by marking them to be dropped at a future time when the demand
on system resources is less.
■
■

■

■

If you drop a nested table column, then its storage table is removed.
If you drop a LOB column, then the LOB data and its corresponding LOB index
segment are removed.
If you drop a BFILE column, then only the locators stored in that column are
removed, not the files referenced by the locators.
If you drop or mark unused a column defined as an INCLUDING column, then the
column stored immediately before this column will become the new INCLUDING
column.

SET UNUSED Clause
Specify SET UNUSED to mark one or more columns as unused. For an internal
heap-organized table, specifying this clause does not actually remove the target
columns from each row in the table. It does not restore the disk space used by these
columns. Therefore, the response time is faster than when you execute the DROP clause.
When you specify this clause for a column in an external table, the clause is
transparently converted to an ALTER TABLE ... DROP COLUMN statement. The reason for
this is that any operation on an external table is a metadata-only operation, so there is
no difference in the performance of the two commands.
You can view all tables with columns marked UNUSED in the data dictionary views
USER_UNUSED_COL_TABS, DBA_UNUSED_COL_TABS, and ALL_UNUSED_COL_TABS.
Oracle Database Reference for information on the data
dictionary views

See Also:

Unused columns are treated as if they were dropped, even though their column data
remains in the table rows. After a column has been marked UNUSED, you have no access
to that column. A SELECT * query will not retrieve data from unused columns. In
addition, the names and types of columns marked UNUSED will not be displayed during

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-49

ALTER TABLE

a DESCRIBE, and you can add to the table a new column with the same name as an
unused column.
Until you actually drop these columns, they continue to count
toward the absolute limit of 1000 columns in a single table. However,
as with all DDL statements, you cannot roll back the results of this
clause. You cannot issue SET USED counterpart to retrieve a column
that you have SET UNUSED. Refer to CREATE TABLE on page 16-6 for
more information on the 1000-column limit.
Note:

Also, if you mark a LONG column as UNUSED, then you cannot add
another LONG column to the table until you actually drop the unused
LONG column.
Restriction on Marking Columns Unused You cannot mark as unused a column in a
table owned by SYS.

DROP Clause
Specify DROP to remove the column descriptor and the data associated with the target
column from each row in the table. If you explicitly drop a particular column, then all
columns currently marked UNUSED in the target table are dropped at the same time.
When the column data is dropped:
■

All indexes defined on any of the target columns are also dropped.

■

All constraints that reference a target column are removed.

■

If any statistics types are associated with the target columns, then Oracle Database
disassociates the statistics from the column with the FORCE option and drops any
statistics collected using the statistics type.
If the target column is a parent key of a nontarget column, or if
a check constraint references both the target and nontarget columns,
then Oracle Database returns an error and does not drop the column
unless you have specified the CASCADE CONSTRAINTS clause. If you
have specified that clause, then the database removes all constraints
that reference any of the target columns.

Note:

See Also: DISASSOCIATE STATISTICS on page 17-34 for more
information on disassociating statistics types

DROP UNUSED COLUMNS Clause
Specify DROP UNUSED COLUMNS to remove from the table all columns currently marked as
unused. Use this statement when you want to reclaim the extra disk space from
unused columns in the table. If the table contains no unused columns, then the
statement returns with no errors.
Specify one or more columns to be set as unused or dropped. Use the COLUMN
keyword only if you are specifying only one column. If you specify a column list, then
it cannot contain duplicates.

column

CASCADE CONSTRAINTS Specify CASCADE CONSTRAINTS if you want to drop all
foreign key constraints that refer to the primary and unique keys defined on the

12-50 Oracle Database SQL Language Reference

ALTER TABLE

dropped columns as well as all multicolumn constraints defined on the dropped
columns. If any constraint is referenced by columns from other tables or remaining
columns in the target table, then you must specify CASCADE CONSTRAINTS. Otherwise,
the statement aborts and an error is returned.
The INVALIDATE keyword is optional. Oracle Database automatically
invalidates all dependent objects, such as views, triggers, and stored program units.
Object invalidation is a recursive process. Therefore, all directly dependent and
indirectly dependent objects are invalidated. However, only local dependencies are
invalidated, because the database manages remote dependencies differently from local
dependencies.

INVALIDATE

An object invalidated by this statement is automatically revalidated when next
referenced. You must then correct any errors that exist in that object before referencing
it.
Oracle Database Concepts for more information on
dependencies

See Also:

CHECKPOINT Specify CHECKPOINT if you want Oracle Database to apply a checkpoint
for the DROP COLUMN operation after processing integer rows; integer is optional and
must be greater than zero. If integer is greater than the number of rows in the table,
then the database applies a checkpoint after all the rows have been processed. If you
do not specify integer, then the database sets the default of 512. Checkpointing cuts
down the amount of undo logs accumulated during the DROP COLUMN operation to
avoid running out of undo space. However, if this statement is interrupted after a
checkpoint has been applied, then the table remains in an unusable state. While the
table is unusable, the only operations allowed on it are DROP TABLE, TRUNCATE TABLE,
and ALTER TABLE DROP ... COLUMNS CONTINUE (described in sections that follow).

You cannot use this clause with SET UNUSED, because that clause does not remove
column data.
DROP COLUMNS CONTINUE Clause
Specify DROP COLUMNS CONTINUE to continue the drop column operation from the point
at which it was interrupted. Submitting this statement while the table is in an invalid
state results in an error.
Restrictions on Dropping Columns

Dropping columns is subject to the following

restrictions:
■

Each of the parts of this clause can be specified only once in the statement and
cannot be mixed with any other ALTER TABLE clauses. For example, the following
statements are not allowed:
ALTER TABLE t1 DROP COLUMN f1 DROP (f2);
ALTER TABLE t1 DROP COLUMN f1 SET UNUSED (f2);
ALTER TABLE t1 DROP (f1) ADD (f2 NUMBER);
ALTER TABLE t1 SET UNUSED (f3)
ADD (CONSTRAINT ck1 CHECK (f2 > 0));

■

You can drop an object type column only as an entity. To drop an attribute from an
object type column, use the ALTER TYPE ... DROP ATTRIBUTE statement with the
CASCADE INCLUDING TABLE DATA clause. Be aware that dropping an attribute affects
all dependent objects. See Oracle Database PL/SQL Language Reference for more
information.

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-51

ALTER TABLE

■

■

■

■
■

■

You can drop a column from an index-organized table only if it is not a primary
key column. The primary key constraint of an index-organized table can never be
dropped, so you cannot drop a primary key column even if you have specified
CASCADE CONSTRAINTS.
You can export tables with dropped or unused columns. However, you can import
a table only if all the columns specified in the export files are present in the table
(none of those columns has been dropped or marked unused). Otherwise, Oracle
Database returns an error.
You can set unused a column from a table that uses COMPRESS BASIC, but you
cannot drop the column. However, all clauses of the drop_column_clause are valid
for tables that use COMPRESS FOR OLTP. See the semantics for table_compression on
page 16-34 for more information.
You cannot drop a column on which a domain index has been built.
You cannot drop a SCOPE table constraint or a WITH ROWID constraint on a REF
column.
You cannot use this clause to drop:
–

A pseudocolumn, cluster column, or partitioning column. You can drop
nonpartitioning columns from a partitioned table if all the tablespaces where
the partitions were created are online and in read/write mode.

–

A column from a nested table, an object table, or a table owned by SYS.
See Also:

"Dropping a Column: Example" on page 12-81

rename_column_clause
Use the rename_column_clause to rename a column of table. The new column name
must not be the same as any other column name in table.
When you rename a column, Oracle Database handles dependent objects as follows:
■

■

■

Function-based indexes and check constraints that depend on the renamed
column remain valid.
Dependent views, triggers, functions, procedures, and packages are invalidated.
Oracle Database attempts to revalidate them when they are next accessed, but you
may need to alter these objects with the new column name if revalidation fails.
If a domain index is defined on the column being renamed, then the database
invokes the ODCIIndexAlter method with the RENAME option. This operation
establishes correspondence between the indextype metadata and the base table

Restrictions on Renaming Columns

Renaming columns is subject to the following

restrictions:
■

■

You cannot combine this clause with any of the other column_clauses in the same
statement.
You cannot rename a column that is used to define a join index. Instead you must
drop the index, rename the column, and re-create the index.
See Also:

"Renaming a Column: Example" on page 12-81

modify_collection_retrieval
Use the modify_collection_retrieval clause to change what Oracle Database returns
when a collection item is retrieved from the database.

12-52 Oracle Database SQL Language Reference

ALTER TABLE

collection_item Specify the name of a column-qualified attribute whose type is
nested table or varray.
RETURN AS

Specify what Oracle Database should return as the result of a query:

■

LOCATOR specifies that a unique locator for the nested table is returned.

■

VALUE specifies that a copy of the nested table itself is returned.
See Also:

"Collection Retrieval: Example" on page 12-79

modify_LOB_storage_clause
The modify_LOB_storage_clause lets you change the physical attributes of LOB_item.
You can specify only one LOB_item for each modify_LOB_storage_clause.
The sections that follow describe the semantics of parameters specific to modify_LOB_
parameters. Unless otherwise documented in this section, the remaining LOB
parameters have the same semantics when altering a table that they have when you
are creating a table. Refer to the restrictions at the end of this section and to the CREATE
TABLE clause LOB_storage_parameters on page 16-43 for more information.

Notes:
■

■

You can modify LOB storage with an ALTER TABLE statement or
with online redefinition by using the DBMS_REDEFINITION package.
If you have not enabled LOB encryption, compression, or
deduplication at create time, Oracle recommends that you use
online redefinition to enable them after creation, as this process is
more disk space efficient for changes to these three parameters.
See Oracle Database PL/SQL Packages and Types Reference for more
information on DBMS_REDEFINITION.
You cannot convert a LOB from one type of storage to the other.
Instead you must migrate to SecureFiles or BasicFiles by using
online redefinition or partition exchange.

PCTVERSION integer Refer to the CREATE TABLE clause PCTVERSION integer on
page 16-44 for information on this clause.
LOB_retention_clause If the database is in automatic undo mode, then you can
specify RETENTION instead of PCTVERSION to instruct Oracle Database to retain old
versions of this LOB. This clause overrides any prior setting of PCTVERSION.

For BasicFiles LOBs, if the database is in automatic undo mode,
then you can use this clause to specify the number of freelist groups for this LOB. This
clause overrides any prior setting of FREELIST GROUPS. Refer to the CREATE TABLE clause
FREEPOOLS integer on page 16-45 for a full description of this parameter. The
database ignores this parameter for SecureFiles LOBs.

FREEPOOLS integer

REBUILD FREEPOOLS This clause applies only to BasicFiles LOBs, not to
SecureFiles LOBs. The REBUILD FREEPOOLS clause removes all the old versions of data
from the LOB column. This clause is useful for removing all retained old version space
in a LOB segment, freeing that space to be used immediately by new LOB data.
LOB_deduplicate_clause This clause is valid only for SecureFiles LOBs. KEEP_

DUPLICATES disables LOB deduplication. DEDUPLICATE enables LOB deduplication. All
SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-53

ALTER TABLE

lobs in the segment are read, and any matching LOBs are deduplicated before
returning.
LOB_compression_clause This clause is valid only for SecureFiles LOBs. COMPRESS
compresses all LOBs in the segment and then returns. NOCOMPRESS uncompresses all
LOBs in the segment and then returns.

LOB encryption has the same semantics as column encryption
in general. See "ENCRYPT encryption_spec | DECRYPT" on page 12-47 for more
information.

ENCRYPT | DECRYPT

When you modify a LOB column from CACHE or
NOCACHE to CACHE READS, or from CACHE READS to CACHE or NOCACHE, you can change the
logging attribute. If you do not specify LOGGING or NOLOGGING, then this attribute
defaults to the current logging attribute of the LOB column. If you do not specify
CACHE, NOCACHE, or CACHE READS, then Oracle Database retains the existing values of the
LOB attributes.
CACHE, NOCACHE, CACHE READS

Restrictions on Modifying LOB Storage Modifying LOB storage is subject to the
following restrictions:
■

■

You cannot modify the value of the INITIAL parameter in the storage_clause
when modifying the LOB storage attributes.
You cannot specify both the allocate_extent_clause and the deallocate_
unused_clause in the same statement.

■

You cannot specify both the PCTVERSION and RETENTION parameters.

■

You cannot specify the shrink_clause for SecureFiles LOBs.
See Also: LOB_storage_clause on page 16-42 (in CREATE TABLE) for
information on setting LOB parameters and "LOB Columns:
Examples" on page 12-87

alter_varray_col_properties
The alter_varray_col_properties clause lets you change the storage characteristics
of an existing LOB in which a varray is stored.
Restriction on Altering Varray Column Properties You cannot specify the

TABLESPACE clause of LOB_parameters as part of this clause. The LOB tablespace for a
varray defaults to the tablespace of the containing table.
REKEY encryption_spec
The REKEY clause causes the database to generate a new encryption key. All encrypted
columns in the table are reencrypted using the new key and, if you specify the USING
clause of the encryption_spec, a new encryption algorithm. You cannot combine this
clause with any other clauses in this ALTER TABLE statement.
See Also: Oracle Database Advanced Security Administrator's Guide for
more information on transparent column encryption

constraint_clauses
Use the constraint_clauses to add a new constraint using out-of-line declaration,
modify the state of an existing constraint, or drop a constraint. Refer to constraint on
page 8-4 for a description of all the keywords and parameters of out-of-line constraints
and constraint_state.
12-54 Oracle Database SQL Language Reference

ALTER TABLE

Adding a Constraint
The ADD clause lets you add a new out-of-line constraint or out-of-line REF constraint to
the table.
See Also: "Disabling a CHECK Constraint: Example" on page 12-81,
"Specifying Object Identifiers: Example" on page 12-85, and "REF
Columns: Examples" on page 12-89

Modifying a Constraint
The MODIFY CONSTRAINT clause lets you change the state of an existing constraint.
The CASCADE keyword is valid only when you are disabling a unique or primary key
constraint on which a foreign key constraint is defined. In this case, you must specify
CASCADE so that the unique or primary key constraint and all of its dependent foreign
key constraints are disabled.
Restrictions on Modifying Constraints

Modifying constraints is subject to the

following restrictions:
■

■

■

You cannot change the state of a NOT DEFERRABLE constraint to INITIALLY
DEFERRED.
If you specify this clause for an index-organized table, then you cannot specify any
other clauses in the same statement.
You cannot change the NOT NULL constraint on a foreign key column of a
reference-partitioned table, and you cannot change the state of a partitioning
referential constraint of a reference-partitioned table.
See Also: "Changing the State of a Constraint: Examples" on
page 12-80

Renaming a Constraint
The RENAME CONSTRAINT clause lets you rename any existing constraint on table. The
new constraint name cannot be the same as any existing constraint on any object in the
same schema. All objects that are dependent on the constraint remain valid.
See Also:

"Renaming Constraints: Example" on page 12-87

drop_constraint_clause
The drop_constraint_clause lets you drop an integrity constraint from the database.
Oracle Database stops enforcing the constraint and removes it from the data
dictionary. You can specify only one constraint for each drop_constraint_clause, but
you can specify multiple drop_constraint_clause in one statement.
Restrictions on Dropping Constraints You cannot drop the NOT NULL constraint on a

foreign key column of a reference-partitioned table, and you cannot drop a
partitioning referential constraint of a reference-partitioned table.
PRIMARY KEY
UNIQUE

Specify PRIMARY KEY to drop the primary key constraint of table.

Specify UNIQUE to drop the unique constraint on the specified columns.

If you drop the primary key or unique constraint from a column on which a bitmap
join index is defined, then Oracle Database invalidates the index. See CREATE INDEX
on page 14-60 for information on bitmap join indexes.

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-55

ALTER TABLE

Specify CONSTRAINT constraint_name to drop an integrity constraint
other than a primary key or unique constraint.

CONSTRAINT

Specify CASCADE if you want all other integrity constraints that depend on
the dropped integrity constraint to be dropped as well.

CASCADE

Specify KEEP INDEX or DROP INDEX to indicate whether
Oracle Database should preserve or drop the index it has been using to enforce the
PRIMARY KEY or UNIQUE constraint.
KEEP INDEX | DROP INDEX

Restrictions on Dropping Constraints

Dropping constraints is subject to the

following restrictions:
■

■

■

■

You cannot drop a primary key or unique key constraint that is part of a referential
integrity constraint without also dropping the foreign key. To drop the referenced
key and the foreign key together, use the CASCADE clause. If you omit CASCADE, then
Oracle Database does not drop the primary key or unique constraint if any foreign
key references it.
You cannot drop a primary key constraint (even with the CASCADE clause) on a
table that uses the primary key as its object identifier (OID).
If you drop a referential integrity constraint on a REF column, then the REF column
remains scoped to the referenced table.
You cannot drop the scope of a REF column.
See Also:

"Dropping Constraints: Examples" on page 12-87

alter_external_table
Use the alter_external_table clauses to change the characteristics of an external
table. This clause has no affect on the external data itself. The syntax and semantics of
the parallel_clause, enable_disable_clause, external_data_properties, and
REJECT LIMIT clause are the same as described for CREATE TABLE. See the external_table_
clause on page 16-40 (in CREATE TABLE).
PROJECT COLUMN Clause This clause lets you determine how the access driver
validates the rows of an external table in subsequent queries. The default is PROJECT
COLUMN ALL, which means that the access driver processes all column values, regardless
of which columns are selected, and validates only those rows with fully valid column
entries. If any column value would raise an error, such as a data type conversion error,
then the row is rejected even if that column was not referenced in the select list. If you
specify PROJECT COLUMN REFERENCED, then the access driver processes only those
columns in the select list.

The ALL setting guarantees consistent result sets. The REFERENCED setting can result in
different numbers of rows returned, depending on the columns referenced in
subsequent queries, but is faster than the ALL setting. If a subsequent query selects all
columns of the external table, then the settings behave identically.
Restrictions on Altering External Tables

Altering external tables is subject to the

following restrictions:
■
■

■

You cannot modify an external table using any clause outside of this clause.
You cannot add a LONG, varray, or object type column to an external table, nor can
you change the data type of an external table column to any of these data types.
You cannot add a constraint to an external table.

12-56 Oracle Database SQL Language Reference

ALTER TABLE

■

You cannot modify the storage parameters of an external table.

alter_table_partitioning
The clauses in this section apply only to partitioned tables. You cannot combine
partition operations with other partition operations or with operations on the base
table in the same ALTER TABLE statement.
Notes on Changing Table Partitioning
table partitioning:
■

■

■

■

■

■

■

The following notes apply when changing

If you drop, exchange, truncate, move, modify, or split a partition on a table that is
a master table for one or more materialized views, then existing bulk load
information about the table will be deleted. Therefore, be sure to refresh all
dependent materialized views before performing any of these operations.
If a bitmap join index is defined on table, then any operation that alters a
partition of table causes Oracle Database to mark the index UNUSABLE.
The only alter_table_partitioning clauses you can specify for a
reference-partitioned table are modify_table_default_attrs, move_table_
[sub]partition, truncate_partition_subpart, and exchange_partition_
subpart. None of these operations cascade to any child table of the
reference-partitioned table. No other partition maintenance operations are valid on
a reference-partitioned table, but you can specify the other partition maintenance
operations on the parent table of a reference-partitioned table, and the operation
will cascade to the child reference-partitioned table.
When adding partitions and subpartitions, bear in mind that you can specify up to
a total of 1024K-1 partitions and subpartitions for each table.
When you add a table partition or subpartition and you omit the partition name,
the database generates a name using the rules described in "Notes on Partitioning
in General" on page 16-51.
When you move, add (hash only), coalesce, drop, split, merge, rename, or truncate
a table partition or subpartition, the procedures, functions, packages, package
bodies, views, type bodies, and triggers that reference the table remain valid. All
other dependent objects are invalidated.
Deferred segment creation is not supported for partition maintenance operations
that create new segments on tables with LOB columns; segments will always be
created for the involved (sub)partitions.

For additional information on partition operations on tables with an associated
CONTEXT domain index, refer to Oracle Text Reference.
The storage of partitioned database entities in tablespaces of different block sizes is
subject to several restrictions. Refer to Oracle Database VLDB and Partitioning Guide for
a discussion of these restrictions.

modify_table_default_attrs
The modify_table_default_attrs clause lets you specify new default values for the
attributes of table. Only attributes named in the statement are affected. Partitions and
LOB partitions you create subsequently will inherit these values unless you override
them explicitly when creating the partition or LOB partition. Existing partitions and
LOB partitions are not affected by this clause.

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-57

ALTER TABLE

Only attributes named in the statement are affected, and the default values specified
are overridden by any attributes specified at the individual partition or LOB partition
level.
■

■

■

■

FOR partition_extended_name applies only to composite-partitioned tables. This
clause specifies new default values for the attributes of the partition identified in
partition_extended_name. Subpartitions and LOB subpartitions of that partition
that you create subsequently will inherit these values unless you override them
explicitly when creating the subpartition or LOB subpartition. Existing
subpartitions are not affected by this clause.
PCTTHRESHOLD, key_compression, and the alter_overflow_clause are valid only
for partitioned index-organized tables.
You can specify the key compression only if key compression is already specified
at the table level. Further, you cannot specify an integer after the COMPRESS
keyword. Key compression length can be specified only when you create the table.
You cannot specify the PCTUSED parameter in segment_attributes for the index
segment of an index-organized table.

alter_interval_partitioning
Use this clause:
■

■

■
■

To convert an existing range-partitioned table to interval partitioning. The
database automatically creates partitions of the specified numeric range or
datetime interval as needed for data beyond the highest value allowed for the last
range partition.
To change the interval of an existing interval-partitioned table. The database first
converts existing interval partitions to range partitions and determines the high
value of the defined range partitions. The database then automatically creates
partitions of the specified numeric range or datetime interval as needed for data
that is beyond that high value.
To change the tablespace storage for an existing interval-partitioned table.
To change an interval-partitioned table back to a range-partitioned table. Use SET
INTERVAL () to disable interval partitioning. The database converts existing
interval partitions to range partitions, using the higher boundaries of created
interval partitions as upper boundaries for the range partitions to be created.

For expr, specify a valid number or interval expression.
See Also: The CREATE TABLE "INTERVAL Clause" on page 16-52 and
Oracle Database VLDB and Partitioning Guide for more information on
interval partitioning

set_subpartition_template
Use the set_subpartition_template clause to create or replace existing default range,
list, or hash subpartition definitions for each table partition. This clause is valid only
for composite-partitioned tables. It replaces the existing subpartition template or
creates a new template if you have not previously created one. Existing subpartitions
are not affected, nor are existing local and global indexes. However, subsequent
partitioning operations (such as add and merge operations) will use the new template.
You can drop an existing subpartition template by specifying ALTER TABLE table SET
SUBPARTITION TEMPLATE ().

12-58 Oracle Database SQL Language Reference

ALTER TABLE

The set_subpartition_template clause has the same semantics as the subpartition_
template clause of CREATE TABLE. Refer to the subpartition_template clause of CREATE
TABLE on page 16-58 for more information.

modify_table_partition
The modify_table_partition clause lets you change the real physical attributes of a
range, hash, list partition, or system partition. This clause optionally modifies the
storage attributes of one or more LOB items for the partition. You can specify new
values for physical attributes (with some restrictions, as noted in the sections that
follow), logging, and storage parameters.
For all types of partitions, you can also specify how Oracle Database should handle
local indexes that become unusable as a result of the modification to the partition. See
"UNUSABLE LOCAL INDEXES Clauses" on page 12-74.
For partitioned index-organized tables, you can also update the mapping table in
conjunction with partition changes. See the alter_mapping_table_clauses on page 12-41.
Notes on Modifying Table Partitions The following notes apply to operations on
range, list, and hash table partitions:
■

■

■

■

■

For all types of table partition, in the partition_attributes clause, the shrink_
clause lets you compact an individual partition segment. Refer to shrink_clause on
page 12-36 for additional information on this clause.
The syntax and semantics for modifying a system partition are the same as those
for modifying a hash partition. Refer to modify_hash_partition on page 12-60.
If table is composite partitioned, then:
–

If you specify the allocate_extent_clause, then Oracle Database allocates an
extent for each subpartition of partition.

–

If you specify the deallocate_unused_clause, then Oracle Database
deallocates unused storage from each subpartition of partition.

–

Any other attributes changed in this clause will be changed in subpartitions of
partition as well, overriding existing values. To avoid changing the attributes
of existing subpartitions, use the FOR PARTITION clause of modify_table_
default_attrs.

When you modify the partition_attributes of a table partition with
equipartitioned nested tables, the changes do not apply to the nested table
partitions corresponding to the table partition being modified. However, you can
modify the storage table of the nested table partition directly with an ALTER TABLE
statement.
Unless otherwise documented, the remaining clauses of partition_attributes
have the same behavior they have when you are creating a partitioned table. Refer
to the CREATE TABLE table_partitioning_clauses on page 16-51 for more information.
See Also:

"Modifying Table Partitions: Examples" on page 12-84

modify_range_partition
Use this clause to modify the characteristics of a range partition.
add_range_subpartition This clause is valid only for range-range composite
partitions. It lets you add a range subpartition to partition.

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-59

ALTER TABLE

This clause is valid only for range-hash composite
partitions. The add_hash_subpartition clause lets you add a hash subpartition to
partition. Oracle Database populates the new subpartition with rows rehashed from
the other subpartition(s) of partition as determined by the hash function. For optimal
load balancing, the total number of subpartitions should be a power of 2.
add_hash_subpartition

In the partitioning_storage_clause, the only clause you can specify for
subpartitions is the TABLESPACE clause. If you do not specify TABLESPACE, then the new
subpartition will reside in the default tablespace of partition.
Oracle Database adds local index partitions corresponding to the selected partition.
Oracle Database marks UNUSABLE the local index partitions corresponding to the added
partitions. The database invalidates any indexes on heap-organized tables. You can
update these indexes during this operation using the update_index_clauses.
add_list_subpartition This clause is valid only for range-list and list-list composite
partitions. It lets you add a list subpartition to partition, and only if you have not
already created a DEFAULT subpartition.
■

■

The list_values_clause is required in this operation, and the values you specify
in the list_values_clause cannot exist in any other subpartition of partition.
However, these values can duplicate values found in subpartitions of other
partitions.
In the partitioning_storage_clause, the only clauses you can specify for
subpartitions are the TABLESPACE clause and table compression.

Oracle Database also adds a subpartition with the same value list to all local index
partitions of the table. The status of existing local and global index partitions of table
are not affected.
Restriction on Adding List Subpartitions You cannot specify this clause if you have

already created a DEFAULT subpartition for this partition. Instead you must split the
DEFAULT partition using the split_list_subpartition clause.
COALESCE SUBPARTITION applies only to hash
subpartitions. Use the COALESCE SUBPARTITION clause if you want Oracle Database to
select the last hash subpartition, distribute its contents into one or more remaining
subpartitions (determined by the hash function), and then drop the last subpartition.
COALESCE SUBPARTITION

■

■

Oracle Database drops local index partitions corresponding to the selected
partition.
Oracle Database marks UNUSABLE the local index partitions corresponding to one
or more absorbing partitions. The database invalidates any global indexes on
heap-organized tables. You can update these indexes during this operation using
the update_index_clauses.

modify_hash_partition
When modifying a hash partition, in the partition_attributes clause, you can
specify only the allocate_extent_clause and deallocate_unused_clause. All other
attributes of the partition are inherited from the table-level defaults except
TABLESPACE, which stays the same as it was at create time.
modify_list_partition
Clauses available to you when modifying a list partition have the same semantics as
when you are modifying a range partition. When modifying a list partition, the
following additional clauses are available:
12-60 Oracle Database SQL Language Reference

ALTER TABLE

ADD | DROP VALUES Clauses These clauses are valid only when you are modifying
composite partitions. Local and global indexes on the table are not affected by either of
these clauses.
■

■

Use the ADD VALUES clause to extend the partition_key_value list of partition to
include additional values. The added partition values must comply with all rules
and restrictions listed in the CREATE TABLE clause list_partitions on page 16-54.
Use the DROP VALUES clause to reduce the partition_key_value list of partition
by eliminating one or more partition_key_value. When you specify this clause,
Oracle Database checks to ensure that no rows with this value exist. If such rows
do exist, then Oracle Database returns an error.
ADD VALUES and DROP VALUES operations on a table with a
DEFAULT list partition are enhanced if you have defined a local
prefixed index on the table.
Note:

Restrictions on Adding and Dropping List Values

Adding and dropping list values

are subject to the following restrictions:
■
■

You cannot add values to or drop values from a DEFAULT list partition.
If table contains a DEFAULT partition and you attempt to add values to a
nondefault partition, then Oracle Database will check that the values being added
do not already exist in the DEFAULT partition. If the values do exist in the DEFAULT
partition, then Oracle Database returns an error.

modify_table_subpartition
This clause applies only to composite-partitioned tables. Its subclauses let you modify
the characteristics of an individual range, list, or hash subpartition.
The shrink_clause lets you compact an individual subpartition segment. Refer to
shrink_clause on page 12-36 for additional information on this clause.
You can also specify how Oracle Database should handle local indexes that become
unusable as a result of the modification to the partition. See "UNUSABLE LOCAL
INDEXES Clauses" on page 12-74.
Restriction on Modifying Hash Subpartitions The only modify_LOB_parameters you
can specify for subpartition are the allocate_extent_clause and deallocate_
unused_clause.
ADD | DROP VALUES Clauses These clauses are valid only when you are modifying
list subpartitions. Local and global indexes on the table are not affected by either of
these clauses.
■

■

Use the ADD VALUES clause to extend the subpartition_key_value list of
subpartition to include additional values. The added partition values must
comply with all rules and restrictions listed in the CREATE TABLE clause list_
partitions on page 16-54.
Use the DROP VALUES clause to reduce the subpartition_key_value list of
subpartition by eliminating one or more subpartition_key_value. When you
specify this clause, Oracle Database checks to ensure that no rows with this value
exist. If such rows do exist, then Oracle Database returns an error.

You can also specify how Oracle Database should handle local indexes that become
unusable as a result of the modification to the partition. See "UNUSABLE LOCAL

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-61

ALTER TABLE

INDEXES Clauses" on page 12-74.
Restriction on Modifying List Subpartitions The only modify_LOB_parameters you
can specify for subpartition are the allocate_extent_clause and deallocate_
unused_clause.

move_table_partition
Use the move_table_partition clause to move partition to another segment. You can
move partition data to another tablespace, recluster data to reduce fragmentation, or
change create-time physical attributes.
If the table contains LOB columns, then you can use the LOB_storage_clause to move
the LOB data and LOB index segments associated with this partition. Only the LOBs
named are affected. If you do not specify the LOB_storage_clause for a particular LOB
column, then its LOB data and LOB index segments are not moved.
If the table contains nested table columns, then you can use the nested_table_col_
properties clause of the table_partition_description to move the nested table
segments associated with this partition. Only the nested table items named are
affected. If you do not specify the nested_table_col_properties clause of the
table_partition_description for a particular nested table column, then its segments
are not moved.
Oracle Database moves local index partitions corresponding to the specified partition.
If the moved partitions are not empty, then the database marks them UNUSABLE. The
database invalidates global indexes on heap-organized tables. You can update these
indexes during this operation using the update_index_clauses.
When you move a LOB data segment, Oracle Database drops the old data segment
and corresponding index segment and creates new segments even if you do not
specify a new tablespace.
The move operation obtains its parallel attribute from the parallel_clause, if
specified. When it is not specified, the default parallel attributes of the table, if any, are
used. If neither is specified, then Oracle Database performs the move serially.
Specifying the parallel_clause in MOVE PARTITION does not change the default
parallel attributes of table.
For index-organized tables, Oracle Database uses the address
of the primary key, as well as its value, to construct logical rowids.
The logical rowids are stored in the secondary index of the table. If
you move a partition of an index-organized table, then the address
portion of the rowids will change, which can hamper performance. To
ensure optimal performance, rebuild the secondary index(es) on the
moved partition to update the rowids.

Note:

See Also:

"Moving Table Partitions: Example" on page 12-84

MAPPING TABLE The MAPPING TABLE clause is relevant only for an index-organized
table that already has a mapping table defined for it. Oracle Database moves the
mapping table along with the moved index-organized table partition. The mapping
table partition inherits the physical attributes of the moved index-organized table
partition. This is the only way you can change the attributes of the mapping table
partition. If you omit this clause, then the mapping table partition retains its original
attributes.

12-62 Oracle Database SQL Language Reference

ALTER TABLE

Oracle Database marks UNUSABLE all corresponding bitmap index partitions.
Refer to the mapping_table_clauses on page 16-38 (in CREATE TABLE) for more
information on this clause.
Restrictions on Moving Table Partitions Moving table partitions is subject to the
following restrictions:
■

■

If partition is a hash partition, then the only attribute you can specify in this
clause is TABLESPACE.
You cannot specify this clause for a partition containing subpartitions. However,
you can move subpartitions using the move_table_subpartition clause.

move_table_subpartition
Use the move_table_subpartition clause to move the subpartition identified by
subpartition_extended_name to another segment. If you do not specify TABLESPACE,
then the subpartition remains in the same tablespace.
If the subpartition is not empty, then Oracle Database marks UNUSABLE all local index
subpartitions corresponding to the subpartition being moved. You can update all
indexes on heap-organized tables during this operation using the update_index_clauses.
If the table contains LOB columns, then you can use the LOB_storage_clause to move
the LOB data and LOB index segments associated with this subpartition. Only the
LOBs specified are affected. If you do not specify the LOB_storage_clause for a
particular LOB column, then its LOB data and LOB index segments are not moved.
When you move a LOB data segment, Oracle Database drops the old data segment
and corresponding index segment and creates new segments even if you do not
specify a new tablespace.
The only clauses of the partitioning_
storage_clause you can specify are the TABLESPACE clause and table_compression.

Restriction on Moving Table Subpartitions

add_table_partition
Use the add_table_partition clause to add a hash, range, list, or system partition to
table.
Oracle Database adds to any local index defined on table a new partition with the
same name as that of the base table partition. If the index already has a partition with
such a name, then Oracle Database generates a partition name of the form SYS_Pn.
If table is index organized, then Oracle Database adds a partition to any mapping
table and overflow area defined on the table as well.
If table is the parent table of a reference-partitioned table, then you can use the
dependent_tables_clause to propagate the partition maintenance operation you are
specifying in this statement to all the reference-partitioned child tables.
For composite-partitioned tables, Oracle Database adds a new index partition with the
same subpartition descriptions to all local indexes defined on table. Global indexes
defined on table are not affected.
See Also: "Adding a Table Partition with a LOB and Nested Table
Storage: Examples" on page 12-83

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-63

ALTER TABLE

add_range_partition_clause
The add_range_partition_clause lets you add a new range partition to the high end
of a range-partitioned or composite range-partitioned table (after the last existing
partition).
If a domain index is defined on table, then the index must not be marked IN_
PROGRESS or FAILED.
Restrictions on Adding Range Partitions Adding range partitions is subject to the
following restrictions:
■

■

■

If the upper partition bound of each partitioning key in the existing high partition
is MAXVALUE, then you cannot add a partition to the table. Instead, use the split_
table_partition clause to add a partition at the beginning or the middle of the
table.
The key_compression and OVERFLOW clauses are valid only for a partitioned
index-organized table. You can specify OVERFLOW only if the partitioned table
already has an overflow segment. You can specify key compression only if key
compression is enabled at the table level.
You cannot specify the PCTUSED parameter for the index segment of an
index-organized table.

range_values_clause Specify the upper bound for the new partition. The value_list

is a comma-delimited, ordered list of literal values corresponding to the partitioning
key columns. The value_list must collate greater than the partition bound for the
highest existing partition in the table.
table_partition_description Use this clause specify any create-time physical
attributes for the new partition. If the table contains LOB columns, then you can also
specify partition-level attributes for one or more LOB items.
Subpartition Descriptions These clauses are valid only for composite-partitioned
tables. Use the range_subpartition_desc, list_subpartition_desc, individual_
hash_subparts, or hash_subparts_by_quantity clause, as appropriate, if you want to
specify subpartitions for the new partition. This clause overrides any subpartition
descriptions defined in subpartition_template at the table level.

add_hash_partition_clause
The add_hash_partition_clause lets you add a new hash partition to the high end of
a hash-partitioned table. Oracle Database populates the new partition with rows
rehashed from other partitions of table as determined by the hash function. For
optimal load balancing, the total number of partitions should be a power of 2.
You can specify a name for the partition, and optionally a tablespace where it should
be stored. If you do not specify a name, then the database assigns a partition name of
the form SYS_Pn. If you do not specify TABLESPACE, then the new partition is stored in
the default tablespace of the table. Other attributes are always inherited from
table-level defaults.
If this operation causes data to be rehashed among partitions, then the database marks
UNUSABLE any corresponding local index partitions. You can update all indexes on
heap-organized tables during this operation using the update_index_clauses.
Use the parallel_clause to specify whether to parallelize the creation of the new
partition.

12-64 Oracle Database SQL Language Reference

ALTER TABLE

CREATE TABLE on page 16-6 and Oracle Database VLDB
and Partitioning Guide for more information on hash partitioning

See Also:

add_list_partition_clause
The add_list_partition_clause lets you add a new partition to table using a new
set of partition values. You can specify any create-time physical attributes for the new
partition. If the table contains LOB columns, then you can also specify partition-level
attributes for one or more LOB items.
You cannot add a list partition if you have
already defined a DEFAULT partition for the table. Instead, you must use the split_
table_partition clause to split the DEFAULT partition.

Restrictions on Adding List Partitions

See Also:
■

■

list_partitions of CREATE TABLE on page 16-54 for more information
and restrictions on list partitions
"Working with Default List Partitions: Example" on page 12-83

add_system_partition_clause
Use this clause to add a partition to a system-partitioned table. Oracle Database adds a
corresponding index partition to all local indexes defined on the table.
The BEFORE clause lets you specify where the new partition should be added in relation
to existing partitions. You cannot split a system partition. Therefore, this clause is
useful if you want to divide the contents of one existing partition among multiple new
partitions. If you omit this clause, then the database adds the new partition after the
existing partitions.
The table_partition_description lets you specify partition-level attributes of the
new partition. The values of any unspecified attributes are inherited from the
table-level values.
Restriction on Adding System Partitions You cannot specify the OVERFLOW clause
when adding a system partition.
See Also: The CREATE TABLE clause system_partitioning on page 16-61
for more information on system partitions

coalesce_table_partition
COALESCE applies only to hash partitions. Use the coalesce_table_partition clause to
indicate that Oracle Database should select the last hash partition, distribute its
contents into one or more remaining partitions as determined by the hash function,
and then drop the last partition.
Oracle Database drops local index partitions corresponding to the selected partition.
The database marks UNUSABLE the local index partitions corresponding to one or more
absorbing partitions. The database invalidates any indexes on heap-organized tables.
You can update all indexes during this operation using the update_index_clauses.
If you update global indexes using the
update_all_indexes_clause, then you can specify only the keywords UPDATE INDEXES,
not the subclause.
Restriction on Coalescing Table Partitions

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-65

ALTER TABLE

drop_table_partition
The drop_table_partition clause removes the partition identified by partition_
extended_name, and the data in that partition, from a partitioned table. If you want to
drop a partition but keep its data in the table, then you must merge the partition into
one of the adjacent partitions.
See Also:
■

■

■

■

merge_table_partitions on page 12-70

If table has LOB columns, then Oracle Database also drops the LOB data and LOB
index partitions and any subpartitions corresponding to partition.
If table has equipartitioned nested table columns, then Oracle Database also
drops the nested table partitions corresponding to the table partition being
dropped.
If table is index organized and has a mapping table defined on it, then the
database drops the corresponding mapping table partition as well.
Oracle Database drops local index partitions and subpartitions corresponding to
the dropped partition, even if they are marked UNUSABLE.

You can update indexes on table during this operation using the update_index_clauses.
If you specify the parallel_clause with the update_index_clauses, then the database
parallelizes the index update, not the drop operation.
If you drop a range partition and later insert a row that would have belonged to the
dropped partition, then the database stores the row in the next higher partition.
However, if that partition is the highest partition, then the insert will fail, because the
range of values represented by the dropped partition is no longer valid for the table.
Restrictions on Dropping Table Partitions

Dropping table partitions is subject to the

following restrictions:
■

■

■

You cannot drop a partition of a hash-partitioned table. Instead, use the coalesce_
table_partition clause.
If table contains only one partition, then you cannot drop that partition. Instead,
drop the table.
If you update global indexes using the update_index_clauses, then you can specify
only the UPDATE INDEXES keywords but not the subclause.
See Also:

"Dropping a Table Partition: Example" on page 12-84

drop_table_subpartition
Use this clause to drop a range or list subpartition from a range, list, or hash
composite-partitioned table. Oracle Database deletes any rows in the dropped
subpartition.
Oracle Database drops the corresponding subpartition of any local index. Other index
subpartitions are not affected. Any global indexes are marked UNUSABLE unless you
specify the update_global_index_clause or update_all_indexes_clause.
Restrictions on Dropping Table Subpartitions Dropping table subpartitions is
subject to the following restrictions:
■

You cannot drop a hash subpartition. Instead use the MODIFY PARTITION ...
COALESCE SUBPARTITION syntax.

12-66 Oracle Database SQL Language Reference

ALTER TABLE

■

■

If a partition contains only one subpartition, then you cannot drop that
subpartition. Instead, use the drop_table_partition clause.
If you update the global indexes, then you cannot specify the optional subclause of
the update_all_indexes_clause.

rename_partition_subpart
Use the rename_partition_subpart clause to rename a table partition or subpartition
to new_name. For both partitions and subpartitions, new_name must be different from all
existing partitions and subpartitions of the same table.
If table is index organized, then Oracle Database assigns the same name to the
corresponding primary key index partition as well as to any existing overflow
partitions and mapping table partitions.
See Also:

"Renaming Table Partitions: Examples" on page 12-84

truncate_partition_subpart
Specify TRUNCATE PARTITION to remove all rows from the partition identified by
partition_extended_name or, if the table is composite partitioned, all rows from the
subpartitions of that partition. Specify TRUNCATE SUBPARTITION to remove all rows
from an individual subpartition. If table is index organized, then Oracle Database also
truncates any corresponding mapping table partitions and overflow area partitions.
■

■

■

■

If the partition or subpartition to be truncated contains data, then you must first
disable any referential integrity constraints on the table. Alternatively, you can
delete the rows and then truncate the partition.
If table contains any LOB columns, then the LOB data and LOB index segments
for this partition are also truncated. If table is composite partitioned, then the
LOB data and LOB index segments for the subpartitions of the partition are
truncated.
If table contains any equipartitioned nested tables, then you cannot truncate the
parent partition unless its corresponding nested table partition is empty.
If a domain index is defined on table, then the index must not be marked IN_
PROGRESS or FAILED, and the index partition corresponding to the table partition
being truncated must not be marked IN_PROGRESS.

For each partition or subpartition truncated, Oracle Database also truncates
corresponding local index partitions and subpartitions. If those index partitions or
subpartitions are marked UNUSABLE, then the database truncates them and resets the
UNUSABLE marker to VALID.
You can update global indexes on table during this operation using the update_global_
index_clause or the update_all_indexes_clause. If you specify the parallel_clause with
one of these clauses, then the database parallelizes the index update, not the truncate
operation.
Specify DROP STORAGE to deallocate all space from the deleted rows,
except the space allocated by the MINEXTENTS parameter. This space can subsequently
be used by other objects in the tablespace.
DROP STORAGE

DROP ALL STORAGE Specify DROP ALL STORAGE to deallocate all space from the

deleted rows, including the space allocated by the MINEXTENTS parameter. All
segments for the partition or subpartition, as well as all segments for its dependent
objects, will be deallocated.

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-67

ALTER TABLE

Restrictions on DROP ALL STORAGE This clause is subject to the same restrictions
as described in "Restrictions on Deferred Segment Creation" on page 16-33.
Note: The DROP ALL STORAGE clause is available starting with Oracle
Database 11g Release 2 (11.2.0.2).

Specify REUSE STORAGE to keep space from the deleted rows
allocated to the partition or subpartition. The space is subsequently available only for
inserts and updates to the same partition or subpartition.

REUSE STORAGE

See Also:

"Truncating Table Partitions: Example" on page 12-85

Restriction on Truncating Table Partitions and Subpartitions If you update global
indexes using the update_all_indexes_clause, then you can specify only the UPDATE
INDEXES keywords, not the subclause.

split_table_partition
The split_table_partition clause lets you create, from the partition identified by
partition_extended_name, two new partitions, each with a new segment, new
physical attributes, and new initial extents. The segment associated with the current
partition is discarded.
The new partitions inherit all unspecified physical attributes from the current
partition.
Oracle Database can optimize and speed up SPLIT PARTITION
and SPLIT SUBPARTITION operations if specific conditions are met.
Refer to Oracle Database VLDB and Partitioning Guide for information
on optimizing these operations.
Note:

■

■

■

■

If you split a DEFAULT list partition, then the first of the resulting partitions will
have the split values, and the second resulting partition will have the DEFAULT
value.
If table is index organized, then Oracle Database splits any corresponding
mapping table partition and places it in the same tablespace as the parent
index-organized table partition. The database also splits any corresponding
overflow area, and you can use the OVERFLOW clause to specify segment attributes
for the new overflow areas.
If table contains LOB columns, then you can use the LOB_storage_clause to
specify separate LOB storage attributes for the LOB data segments resulting from
the split. The database drops the LOB data and LOB index segments of the current
partition and creates new segments for each LOB column, for each partition, even
if you do not specify a new tablespace.
If table contains nested table columns, then you can use the split_nested_
table_part clause to specify the storage table names and segment attributes of the
nested table segments resulting from the split. The database drops the nested table
segments of the current partition and creates new segments for each nested table
column, for each partition. This clause allows for multiple nested table columns in
the parent table as well as multilevel nested table columns.

Oracle Database splits the corresponding local index partition, even if it is marked
UNUSABLE. The database marks UNUSABLE, and you must rebuild the local index
12-68 Oracle Database SQL Language Reference

ALTER TABLE

partitions corresponding to the split partitions. The new index partitions inherit their
attributes from the partition being split. The database stores the new index partitions
in the default tablespace of the index partition being split. If that index partition has no
default tablespace, then the database uses the tablespace of the new underlying table
partitions.
AT Clause The AT clause applies only to range partitions. Specify the new
noninclusive upper bound for the first of the two new partitions. The value list must
compare less than the original partition bound for the current partition and greater
than the partition bound for the next lowest partition (if there is one).
VALUES Clause The VALUES clause applies only to list partitions. Specify the
partition values you want to include in the first of the two new partitions. Oracle
Database creates the first new partition using the partition value list you specify and
creates the second new partition using the remaining partition values from the current
partition. Therefore, the value list cannot contain all of the partition values of the
current partition, nor can it contain any partition values that do not already exist for
the current partition.

The INTO clause lets you describe the two partitions resulting from the
split. In range_partition_desc or list_partition_desc, as appropriate, the keyword
PARTITION is required even if you do not specify the optional names and physical
attributes of the two partitions resulting from the split. If you do not specify new
partition names, then Oracle Database assigns names of the form SYS_Pn. Any
attributes you do not specify are inherited from the current partition.
INTO Clause

For range-hash composite-partitioned tables, if you specify subpartitioning for the
new partitions, then you can specify only TABLESPACE and table compression for the
subpartitions. All other attributes are inherited from the current partition. If you do
not specify subpartitioning for the new partitions, then their tablespace is also
inherited from the current partition.
For range-list and list-list composite-partitioned tables, you cannot specify
subpartitions for the new partitions at all. The list subpartitions of the split partition
inherit the number of subpartitions and value lists from the current partition.
For all composite-partitioned tables for which you do not specify subpartition names
for the newly created subpartitions, the newly created subpartitions inherit their
names from the parent partition as follows:
■

■

For those subpartitions in the parent partition with names of the form partition_
name underscore (_) subpartition_name (for example, P1_SUBP1), Oracle Database
generates corresponding names in the newly created subpartitions using the new
partition names (for example P1A_SUB1 and P1B_SUB1).
For those subpartitions in the parent partition with names of any other form,
Oracle Database generates subpartition names of the form SYS_SUBPn.

Oracle Database splits the corresponding partition in each local index defined on
table, even if the index is marked UNUSABLE.
Oracle Database invalidates any indexes on heap-organized tables. You can update
these indexes during this operation using the update_index_clauses.
If table is the parent table of a reference-partitioned table, then you can use the
dependent_tables_clause to propagate the partition maintenance operation you are
specifying in this statement to all the reference-partitioned child tables.
The parallel_clause lets you parallelize the split operation but does not change the
default parallel attributes of the table.
SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-69

ALTER TABLE

Restrictions on Splitting Table Partitions You cannot specify this clause for a hash

partition.

split_table_subpartition
Use this clause to split a list subpartition into two separate subpartitions with
nonoverlapping value lists.
Oracle Database can optimize and speed up SPLIT PARTITION
and SPLIT SUBPARTITION operations if specific conditions are met.
Refer to Oracle Database VLDB and Partitioning Guide for information
on optimizing these operations.
Note:

AT Clause The AT clause is valid only for range subpartitions. Specify the new
noninclusive upper bound for the first of the two new subpartitions. The value list
must compare less than the original subpartition bound for the subpartition identified
by subpartition_extended_name and greater than the partition bound for the next
lowest subpartition (if there is one).
VALUES Clause The VALUES clause is valid only for list subpartitions. Specify the
subpartition values you want to include in the first of the two new subpartitions. You
can specify NULL if you have not already specified NULL for another subpartition in the
same partition. Oracle Database creates the first new subpartition using the
subpartition value list you specify and creates the second new partition using the
remaining partition values from the current subpartition. Therefore, the value list
cannot contain all of the partition values of the current subpartition, nor can it contain
any partition values that do not already exist for the current subpartition.

For both range and list subpartitions, the INTO clause lets you describe
the two subpartitions resulting from the split. In range_subpartition_desc or list_
subpartition_desc, as appropriate, the keyword SUBPARTITION is required even if you
do not specify the optional names and attributes of the two new subpartitions. Any
attributes you do not specify are inherited from the current subpartition.
INTO Clause

Oracle Database splits any corresponding local subpartition index, even if it is marked
UNUSABLE. The new index subpartitions inherit the names of the new table
subpartitions unless those names are already held by index subpartitions. In that case,
the database assigns new index subpartition names of the form SYS_SUBPn. The new
index subpartitions inherit physical attributes from the parent subpartition. However,
if the parent subpartition does not have a default TABLESPACE attribute, then the new
subpartitions inherit the tablespace of the corresponding new table subpartitions.
Oracle Database invalidates indexes on heap-organized tables. You can update these
indexes by using the update_index_clauses.
Restrictions on Splitting Table Subpartitions

Splitting table subpartitions is subject

to the following restrictions:
■
■

You cannot specify this clause for a hash subpartition.
In subpartition descriptions, the only clauses of partitioning_storage_clause
you can specify are TABLESPACE and table compression.

merge_table_partitions
The merge_table_partitions clause lets you merge the contents of two range
partitions, two list partitions, or two system partitions of table into one new partition

12-70 Oracle Database SQL Language Reference

ALTER TABLE

and then drop the original two partitions. This clause is not valid for hash partitions.
Use the coalesce_table_partition clause instead.
For each partition, use partition to specify a partition name or the FOR clause to
specify a partition without using its name. See "References to Partitioned Tables and
Indexes" on page 3-119 for more information on the FOR clause.
■

■

■

■

The two partitions to be merged must be adjacent if they are range partitions. List
partitions and system partitions need not be adjacent in order to be merged.
When you merge two range partitions, the new partition inherits the partition
bound of the higher of the two original partitions.
When you merge two list partitions, the resulting partition value list is the union
of the set of the two partition values lists of the partitions being merged. If you
merge a DEFAULT list partition with another list partition, then the resulting
partition will be the DEFAULT partition and will have the DEFAULT value.
When you merge two composite range partitions or two composite list partitions,
range-list or list-list composite partitions, you cannot specify subpartition
descriptions. Oracle Database obtains the subpartitioning information from the
subpartition template. If you have not specified a subpartition template, then the
database creates one MAXVALUE subpartition from range subpartitions or one
DEFAULT subpartition from list subpartitions.

Any attributes you do not specify explicitly for the new partition are inherited from
table-level defaults. However, if you reuse one of the partition names for the new
partition, then the new partition inherits values from the partition whose name is
being reused rather than from table-level default values.
Oracle Database drops local index partitions corresponding to the selected partitions
and marks UNUSABLE the local index partition corresponding to merged partition. The
database also marks UNUSABLE any global indexes on heap-organized tables. You can
update all these indexes during this operation using the update_index_clauses.
If table is the parent table of a reference-partitioned table, then you can use the
dependent_tables_clause to propagate the partition maintenance operation you are
specifying in this statement to all the reference-partitioned child tables.
See Also: "Merging Two Table Partitions: Example" on page 12-83
and "Working with Default List Partitions: Example" on page 12-83

merge_table_subpartitions
The merge_table_subpartitions clause lets you merge the contents of two range or
list subpartitions of table into one new subpartition and then drop the original two
subpartitions. This clause is not valid for hash subpartitions. Use the coalesce_hash_
subpartition clause instead.
For each subpartition, use subpartition to specify a subpartition name or the FOR
clause to specify a subpartition without using its name. See "References to Partitioned
Tables and Indexes" on page 3-119 for more information on the FOR clause.
The two subpartitions to be merged must belong to the same partition. If they are
range subpartitions, then they must be adjacent. If they are list subpartitions, then they
need not be adjacent. The data in the resulting subpartition consists of the combined
data from the merged subpartitions.
If you specify the INTO clause, then in the range_subpartition_desc or list_
subpartition_desc you cannot specify the range_values_clause or list_values_

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-71

ALTER TABLE

clause, respectively. Further, the only clauses you can specify in the partitioning_
storage_clause are the TABLESPACE clause and table_compression.
Any attributes you do not specify explicitly for the new subpartition are inherited
from partition-level values. However, if you reuse one of the subpartition names for
the new subpartition, then the new subpartition inherits values from the subpartition
whose name is being reused rather than from partition-level default values.
Oracle Database merges corresponding local index subpartitions and marks the
resulting index subpartition UNUSABLE. The database also marks UNUSABLE both
partitioned and nonpartitioned global indexes on heap-organized tables. You can
update all indexes during this operation using the update_index_clauses.

exchange_partition_subpart
Use the EXCHANGE PARTITION or EXCHANGE SUBPARTITION clause to exchange the data
and index segments of:
■

■

■

■

One nonpartitioned table with:
–

one range, list, or hash partition

–

one range, list, or hash subpartition

One range-partitioned table with the range subpartitions of a range-range or
list-range composite-partitioned table partition
One hash-partitioned table with the hash subpartitions of a range-hash or list-hash
composite-partitioned table partition
One list-partitioned table with the list subpartitions of a range-list or hash-list
composite-partitioned table partition

In all cases, the structure of the table and the partition or subpartition being
exchanged, including their partitioning keys, must be identical. In the case of list
partitions and subpartitions, the corresponding value lists must also match.
This clause facilitates high-speed data loading when used with transportable
tablespaces.
See Also: Oracle Database Administrator's Guide for information on
transportable tablespaces

If table contains LOB columns, then for each LOB column Oracle Database exchanges
LOB data and LOB index partition or subpartition segments with corresponding LOB
data and LOB index segments of table.
If table has nested table columns, then for each such column Oracle Database
exchanges nested table partition segments with corresponding nested table segments
of the nonpartitioned table.
All of the segment attributes of the two objects (including tablespace and logging) are
also exchanged.
Existing statistics for the table being exchanged into the partitioned table will be
exchanged. However, the global statistics for the partitioned table will not be altered.
Use the DBMS_STATS.GATHER_TABLE_STATS procedure to re-create global statistics. You
can set the GRANULARITY attribute equal to 'APPROX_GLOBAL AND PARTITION' to speed
up the process and aggregate new global statistics based on the existing partition
statistics. See Oracle Database PL/SQL Packages and Types Reference for more information
on this packaged procedure.

12-72 Oracle Database SQL Language Reference

ALTER TABLE

Oracle Database invalidates any global indexes on the objects being exchanged. You
can update the global indexes on the table whose partition is being exchanged by
using either the update_global_index_clause or the update_all_indexes_clause clause. For
the update_all_indexes_clause, you can specify only the keywords UPDATE INDEXES,
not the subclause. Global indexes on the table being exchanged remain invalidated.
The update_global_index_clause and update_all_indexes_clause do not update
local indexes during an exchange operation. You can specify local index maintenance
by using the INCLUDING | EXCLUDING INDEXES clause. If you specify the
parallel_clause with either of these clauses, then the database parallelizes the index
update, not the exchange operation.
See Also: "Notes on Exchanging Partitions and Subpartitions" on
page 12-73
WITH TABLE Specify the table with which the partition or subpartition will be
exchanged. If you omit schema, then Oracle Database assumes that table is in your
own schema.

Specify INCLUDING INDEXES if you want local
index partitions or subpartitions to be exchanged with the corresponding table index
(for a nonpartitioned table) or local indexes (for a hash-partitioned table). Specify
EXCLUDING INDEXES if you want all index partitions or subpartitions corresponding to
the partition and all the regular indexes and index partitions on the exchanged table to
be marked UNUSABLE. If you omit this clause, then the default is EXCLUDING INDEXES.
INCLUDING | EXCLUDING INDEXES

WITH | WITHOUT VALIDATION Specify WITH VALIDATION if you want Oracle Database
to return an error if any rows in the exchanged table do not map into partitions or
subpartitions being exchanged. Specify WITHOUT VALIDATION if you do not want Oracle
Database to check the proper mapping of rows in the exchanged table. If you omit this
clause, then the default is WITH VALIDATION.

See "Handling Constraint Exceptions" on page 8-18 for
information on this clause. In the context of exchanging partitions, this clause is valid
only if the partitioned table has been defined with a UNIQUE constraint, and that
constraint must be in DISABLE VALIDATE state. This clause is valid only for exchanging
partition, not subpartitions.
exceptions_clause

See Also:
■

■

■

The DBMS_IOT package in Oracle Database PL/SQL Packages and
Types Reference for information on the SQL scripts
Oracle Database Administrator's Guide for information on
eliminating migrated and chained rows
constraint on page 8-4 for more information on constraint checking
and "Creating an Exceptions Table for Index-Organized Tables:
Example" on page 12-80

Notes on Exchanging Partitions and Subpartitions The following notes apply when
exchanging partitions and subpartitions:
■

■

Both tables involved in the exchange must have the same primary key, and no
validated foreign keys can be referencing either of the tables unless the referenced
table is empty.
When exchanging partitioned index-organized tables:

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-73

ALTER TABLE

–

The source and target table or partition must have their primary key set on the
same columns, in the same order.

–

If key compression is enabled, then it must be enabled for both the source and
the target, and with the same prefix length.

–

Both the source and target must be index organized.

–

Both the source and target must have overflow segments, or neither can have
overflow segments. Also, both the source and target must have mapping
tables, or neither can have a mapping table.

–

Both the source and target must have identical storage attributes for any LOB
columns.
See Also:

"Exchanging Table Partitions: Example" on page 12-84

dependent_tables_clause
This clause is valid only when you are altering the parent table of a
reference-partitioned table. The clause lets you specify attributes of partitions that are
created by the operation for reference-partitioned child tables of the parent table.
■

■

If the parent table is not composite partitioned, then specify one or more child
tables, and for each child table specify one partition_spec for each partition
created in the parent table.
If the parent table is composite, then specify one or more child tables, and for each
child table specify one partition_spec for each subpartition created in the parent
table.
See Also: The CREATE TABLE clause reference_partitioning on
page 16-60 for information on creating reference-partitioned tables
and Oracle Database VLDB and Partitioning Guide for information on
partitioning by reference in general

UNUSABLE LOCAL INDEXES Clauses
These two clauses modify the attributes of local index partitions and index
subpartitions corresponding to partition, depending on whether you are modifying a
partition or subpartition.
■

■

UNUSABLE LOCAL INDEXES marks UNUSABLE the local index partition or index
subpartition associated with partition.
REBUILD UNUSABLE LOCAL INDEXES rebuilds the unusable local index partition or
index subpartition associated with partition.

Restrictions on UNUSABLE LOCAL INDEXES This clause is subject to the following
restrictions:
■

■

You cannot specify this clause with any other clauses of the modify_table_
partition clause.
You cannot specify this clause in the modify_table_partition clause for a
partition that has subpartitions. However, you can specify this clause in the
modify_table_subpartition clause.

update_index_clauses
Use the update_index_clauses to update the indexes on table as part of the table
partitioning operation. When you perform DDL on a table partition, if an index is

12-74 Oracle Database SQL Language Reference

ALTER TABLE

defined on table, then Oracle Database invalidates the entire index, not just the
partitions undergoing DDL. This clause lets you update the index partition you are
changing during the DDL operation, eliminating the need to rebuild the index after the
DDL.
The update_index_clauses are not needed, and are not valid, for partitioned
index-organized tables. Index-organized tables are primary key based, so Oracle can
keep global indexes USABLE during operations that move data but do not change its
value.
update_global_index_clause
Use this clause to update global indexes on table.
update_all_indexes_clause
Use this clause to update all indexes on table.
update_index_partition This clause is valid only for operations on table partitions
and affects only local indexes.
■

The index_partition_description lets you specify physical attributes, tablespace
storage, and logging for each partition of each local index. If you specify only the
PARTITION keyword, then Oracle Database updates the index partition as follows:
–

For operations on a single table partition (such as MOVE PARTITION and SPLIT
PARTITION), the corresponding index partition inherits the attributes of the
affected index table partition, Oracle Database does not generate names for
new index partitions, so any new index partitions resulting from this
operation inherit their names from the corresponding new table partition.

–

For MERGE PARTITION operations, the resulting local index partition inherits its
name from the resulting table partition and inherits its attributes from the
local index.

For a domain index, you can use the PARAMETERS clause to specify the parameter
string that is passed uninterpreted to the appropriate ODCI indextype routine.
The PARAMETERS clause is valid only for domain indexes, and is the only part of the
index_partition_description you can specify for a domain index.
For more information on the UNUSABLE clause, refer to ALTER INDEX ... UNUSABLE
on page 10-90.
See Also: Oracle Database Data Cartridge Developer's Guide for more
information on domain indexes
■

For a composite-partitioned index, the index_subpartition_clause lets you
specify tablespace storage for each subpartition. Refer to the index_subpartition_
clause on page 14-77 (in CREATE INDEX) for more information on this component of
the update_index_partition clause.

update_index_subpartition This clause is valid only for operations on subpartitions
of composite-partitioned tables and affects only local indexes on composite-partitioned
tables. It lets you specify tablespace storage for one or more subpartitions.
Restrictions on Updating All Indexes

The following restrictions apply to the update_

all_indexes_clause:
■

You cannot specify this clause for index-organized tables.

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-75

ALTER TABLE

■

When you exchange a partition or subpartition with the exchange_partition_
subpart clause, the update_all_indexes_clause is applicable only to global
indexes. Therefore, you cannot specify the update_index_partition or update_
index_subpartition clauses. You can, however, specify local index maintenance
during an exchange operation by using the INCLUDING | EXCLUDING
INDEXES clause.

update_global_index_clause
Use this clause to update only global indexes on table. Oracle Database marks
UNUSABLE all local indexes on table.
UPDATE GLOBAL INDEXES Specify UPDATE GLOBAL INDEXES to update the global

indexes defined on table.
Restriction on Updating Global Indexes If the global index is a global domain index
defined on a LOB column, then Oracle Database marks the domain index UNUSABLE
instead of updating it.
INVALIDATE GLOBAL INDEXES Specify INVALIDATE GLOBAL INDEXES to invalidate the
global indexes defined on table.

If you specify neither, then Oracle Database invalidates the global indexes.
Restrictions on Invalidating Global Indexes This clause is supported only for global

indexes. It is not supported for index-organized tables. In addition, this clause updates
only indexes that are USABLE and VALID. UNUSABLE indexes are left unusable, and
INVALID global indexes are ignored.
See Also: "Updating Global Indexes: Example" on page 12-85 and
"Updating Partitioned Indexes: Example" on page 12-85

parallel_clause
The parallel_clause lets you change the default degree of parallelism for queries and
DML on the table.
For complete information on this clause, refer to parallel_clause on page 16-63 in the
documentation on CREATE TABLE.
Restrictions on Changing Table Parallelization

Changing parallelization is subject to

the following restrictions:
■

■

If table contains any columns of LOB or user-defined object type, then subsequent
INSERT, UPDATE, and DELETE operations on table are executed serially without
notification. Subsequent queries, however, are executed in parallel.
If you specify the parallel_clause in conjunction with the move_table_clause,
then the parallelism applies only to the move, not to subsequent DML and query
operations on the table.
See Also:

"Specifying Parallel Processing: Example" on page 12-80

move_table_clause
The move_table_clause lets you relocate data of a nonpartitioned table or of a
partition of a partitioned table into a new segment, optionally in a different tablespace,
and optionally modify any of its storage attributes.

12-76 Oracle Database SQL Language Reference

ALTER TABLE

You can also move any LOB data segments associated with the table or partition using
the LOB_storage_clause and varray_col_properties clause. LOB items not specified
in this clause are not moved.
If you move the table to a different tablespace and the COMPATIBLE parameter is set to
10.0 or higher, then Oracle Database leaves the storage table of any nested table
columns in the tablespace in which it was created. If COMPATIBLE is set to any value less
than 10.0, then the database silently moves the storage table to the new tablespace
along with the table.
ONLINE Clause This clause is valid only for top-level index-organized tables and for
nested table storage tables that are index organized. Specify ONLINE if you want DML
operations on the index-organized table to be allowed during rebuilding of the
primary key index of the table.
Restrictions on Moving Tables Online Moving tables online is subject to the
following restrictions:
■

You cannot combine this clause with any other clause in the same statement.

■

You cannot specify this clause for a partitioned index-organized table.

■

■

Parallel DML and direct path INSERT operations require an exclusive lock on the
table. Therefore, these operations are not supported concurrently with an ongoing
online table MOVE, due to conflicting locks.
You cannot specify this clause if the index-organized table contains any LOB,
VARRAY, Oracle-supplied type, or user-defined object type columns.

index_org_table_clause
For an index-organized table, the index_org_table_clause of the move_table_clause
lets you additionally specify overflow segment attributes. The move_table_clause
rebuilds the primary key index of the index-organized table. The overflow data
segment is not rebuilt unless the OVERFLOW keyword is explicitly stated, with two
exceptions:
■

■

If you alter the values of PCTTHRESHOLD or the INCLUDING column as part of this
ALTER TABLE statement, then the overflow data segment is rebuilt.
If you explicitly move any of out-of-line columns (LOBs, varrays, nested table
columns) in the index-organized table, then the overflow data segment is also
rebuilt.

The index and data segments of LOB columns are not rebuilt unless you specify the
LOB columns explicitly as part of this ALTER TABLE statement.
mapping_table_clause Specify MAPPING TABLE if you want Oracle Database to create
a mapping table if one does not already exist. If it does exist, then the database moves
the mapping table along with the index-organized table, and marks any bitmapped
indexes UNUSABLE. The new mapping table is created in the same tablespace as the
parent table.

Specify NOMAPPING to instruct the database to drop an existing mapping table.
Refer to mapping_table_clauses on page 16-38 (in CREATE TABLE) for more information on
this clause.
Restriction on Mapping Tables You cannot specify NOMAPPING if any bitmapped

indexes have been defined on table.

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-77

ALTER TABLE

Use the key_compression clause to enable or disable key
compression in an index-organized table.

key_compression
■

COMPRESS enables key compression, which eliminates repeated occurrence 1of
primary key column values in index-organized tables. Use integer to specify the
prefix length (number of prefix columns to compress).
The valid range of prefix length values is from 1 to the number of primary key
columns minus 1. The default prefix length is the number of primary key columns
minus 1.

■

NOCOMPRESS disables key compression in index-organized tables. This is the
default.

TABLESPACE tablespace Specify the tablespace into which the rebuilt
index-organized table is to be stored.
LOB_storage_clause Use this clause to move a LOB segment to a different
tablespace. You cannot use this clause to move a LOB segment if the table contains a
LONG column. Instead, you must either convert the LONG column to a LOB, or you must
export the table, re-create the table specifying the desired tablespace storage for the
LOB column, and re-import the table data.
Restrictions on Moving Tables
■

■
■

Moving tables is subject to the following restrictions:

If you specify MOVE, then it must be the first clause in the ALTER TABLE statement,
and the only clauses outside this clause that are allowed are the physical_
attributes_clause, the parallel_clause, and the LOB_storage_clause.
You cannot move a table containing a LONG or LONG RAW column.
You cannot MOVE an entire partitioned table (either heap or index organized). You
must move individual partitions or subpartitions.
Notes Regarding LOBs: For any LOB columns you specify in a

move_table_clause:
■

■

Oracle Database drops the old LOB data segment and
corresponding index segment and creates new segments, even if
you do not specify a new tablespace.
If the LOB index in table resided in a different tablespace from
the LOB data, then Oracle Database collocates the LOB index in
the same tablespace with the LOB data after the move.

See Also: move_table_partition on page 12-62 and move_table_
subpartition on page 12-63

enable_disable_clause
The enable_disable_clause lets you specify whether and how Oracle Database
should apply an integrity constraint. The DROP and KEEP clauses are valid only when
you are disabling a unique or primary key constraint.

12-78 Oracle Database SQL Language Reference

ALTER TABLE

See Also:
■

■

The enable_disable_clause on page 16-64 (in CREATE TABLE) for a
complete description of this clause, including notes and
restrictions that relate to this statement
"Using Indexes to Enforce Constraints" on page 8-17 for
information on using indexes to enforce constraints

TABLE LOCK Oracle Database permits DDL operations on a table only if the table

can be locked during the operation. Such table locks are not required during DML
operations.
Note:

■

Table locks are not acquired on temporary tables.

Specify ENABLE TABLE LOCK to enable table locks, thereby allowing DDL operations
on the table. All currently executing transactions must commit or roll back before
Oracle Database enables the table lock.
Caution: Oracle Database waits until active DML transactions in the
database have completed before locking the table. Sometimes the
resulting delay is considerable.

■

Specify DISABLE TABLE LOCK to disable table locks, thereby preventing DDL
operations on the table.

Parallel DML operations are not performed when the table
lock of the target table is disabled.

Note:

ALL TRIGGERS Use the ALL TRIGGERS clause to enable or disable all triggers
associated with the table.
■

Specify ENABLE ALL TRIGGERS to enable all triggers associated with the table. Oracle
Database fires the triggers whenever their triggering condition is satisfied.
To enable a single trigger, use the enable_clause of ALTER TRIGGER.
See Also: CREATE TRIGGER on page 16-98, ALTER TRIGGER on
page 13-2, and "Enabling Triggers: Example" on page 12-81

■

Specify DISABLE ALL TRIGGERS to disable all triggers associated with the table.
Oracle Database does not fire a disabled trigger even if the triggering condition is
satisfied.

Examples
12

Collection Retrieval: Example The following statement modifies nested table column

ad_textdocs_ntab in the sample table sh.print_media so that when queried it returns
actual values instead of locators:
ALTER TABLE print_media MODIFY NESTED TABLE ad_textdocs_ntab
RETURN AS VALUE;

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-79

ALTER TABLE

Specifying Parallel Processing: Example The following statement specifies parallel

processing for queries to the sample table oe.customers:
ALTER TABLE customers
PARALLEL;

Changing the State of a Constraint: Examples The following statement places in

ENABLE VALIDATE state an integrity constraint named emp_manager_fk in the employees
table:
ALTER TABLE employees
ENABLE VALIDATE CONSTRAINT emp_manager_fk
EXCEPTIONS INTO exceptions;

Each row of the employees table must satisfy the constraint for Oracle Database to
enable the constraint. If any row violates the constraint, then the constraint remains
disabled. The database lists any exceptions in the table exceptions. You can also
identify the exceptions in the employees table with the following statement:
SELECT e.*
FROM employees e, exceptions ex
WHERE e.rowid = ex.row_id
AND ex.table_name = 'EMPLOYEES'
AND ex.constraint = 'EMP_MANAGER_FK';

The following statement tries to place in ENABLE NOVALIDATE state two constraints on
the employees table:
ALTER TABLE employees
ENABLE NOVALIDATE PRIMARY KEY
ENABLE NOVALIDATE CONSTRAINT emp_last_name_nn;

This statement has two ENABLE clauses:
■
■

The first places a primary key constraint on the table in ENABLE NOVALIDATE state.
The second places the constraint named emp_last_name_nn in ENABLE NOVALIDATE
state.

In this case, Oracle Database enables the constraints only if both are satisfied by each
row in the table. If any row violates either constraint, then the database returns an
error and both constraints remain disabled.
Consider the foreign key constraint on the location_id column of the departments
table, which references the primary key of the locations table. The following
statement disables the primary key of the locations table:
ALTER TABLE locations
MODIFY PRIMARY KEY DISABLE CASCADE;

The unique key in the locations table is referenced by the foreign key in the
departments table, so you must specify CASCADE to disable the primary key. This clause
disables the foreign key as well.
Creating an Exceptions Table for Index-Organized Tables: Example The following
example creates the except_table table to hold rows from the index-organized table
hr.countries that violate the primary key constraint:
EXECUTE DBMS_IOT.BUILD_EXCEPTIONS_TABLE ('hr', 'countries', 'except_table');
ALTER TABLE countries
ENABLE PRIMARY KEY
EXCEPTIONS INTO except_table;

12-80 Oracle Database SQL Language Reference

ALTER TABLE

To specify an exception table, you must have the privileges necessary to insert rows
into the table. To examine the identified exceptions, you must have the privileges
necessary to query the exceptions table.
See Also: INSERT on page 18-54 and SELECT on page 19-4 for
information on the privileges necessary to insert rows into tables
Disabling a CHECK Constraint: Example The following statement defines and

disables a CHECK constraint on the employees table:
ALTER TABLE employees ADD CONSTRAINT check_comp
CHECK (salary + (commission_pct*salary) <= 5000)
DISABLE;

The constraint check_comp ensures that no employee's total compensation exceeds
$5000. The constraint is disabled, so you can increase an employee's compensation
above this limit.
Enabling Triggers: Example

The following statement enables all triggers associated

with the employees table:
ALTER TABLE employees
ENABLE ALL TRIGGERS;

Deallocating Unused Space: Example The following statement frees all unused
space for reuse in table employees, where the high water mark is above MINEXTENTS:
ALTER TABLE employees
DEALLOCATE UNUSED;

Renaming a Column: Example The following example renames the credit_limit

column of the sample table oe.customers to credit_amount:
ALTER TABLE customers
RENAME COLUMN credit_limit TO credit_amount;

This statement illustrates the drop_column_clause
with CASCADE CONSTRAINTS. Assume table t1 is created as follows:

Dropping a Column: Example

CREATE TABLE t1 (
pk NUMBER PRIMARY KEY,
fk NUMBER,
c1 NUMBER,
c2 NUMBER,
CONSTRAINT ri FOREIGN KEY (fk) REFERENCES t1,
CONSTRAINT ck1 CHECK (pk > 0 and c1 > 0),
CONSTRAINT ck2 CHECK (c2 > 0)
);

An error will be returned for the following statements:
/* The next two statements return errors:
ALTER TABLE t1 DROP (pk); -- pk is a parent key
ALTER TABLE t1 DROP (c1); -- c1 is referenced by multicolumn
-- constraint ck1

Submitting the following statement drops column pk, the primary key constraint, the
foreign key constraint, ri, and the check constraint, ck1:
ALTER TABLE t1 DROP (pk) CASCADE CONSTRAINTS;

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-81

ALTER TABLE

If all columns referenced by the constraints defined on the dropped columns are also
dropped, then CASCADE CONSTRAINTS is not required. For example, assuming that no
other referential constraints from other tables refer to column pk, then it is valid to
submit the following statement without the CASCADE CONSTRAINTS clause:
ALTER TABLE t1 DROP (pk, fk, c1);

This statement modifies the INITRANS
parameter for the index segment of index-organized table countries_demo, which is
based on hr.countries:
Modifying Index-Organized Tables: Examples

ALTER TABLE countries_demo INITRANS 4;

The following statement adds an overflow data segment to index-organized table
countries:
ALTER TABLE countries_demo ADD OVERFLOW;

This statement modifies the INITRANS parameter for the overflow data segment of
index-organized table countries:
ALTER TABLE countries_demo OVERFLOW INITRANS 4;

The following statement splits the old partition
sales_q4_2000 in the sample table sh.sales, creating two new partitions, naming one
sales_q4_2000b and reusing the name of the old partition for the other:
Splitting Table Partitions: Examples

ALTER TABLE sales SPLIT PARTITION SALES_Q4_2000
AT (TO_DATE('15-NOV-2000','DD-MON-YYYY'))
INTO (PARTITION SALES_Q4_2000, PARTITION SALES_Q4_2000b);

The following statements create a partitioned version of the pm.print_media table. The
LONG column in the print_media table has been converted to LOB. The table is stored in
tablespaces created in "Creating Oracle Managed Files: Examples" on page 16-97. The
object types underlying the ad_textdocs_ntab and ad_header columns are created in
the script that creates the pm sample schema:
CREATE TABLE print_media_part (
product_id NUMBER(6),
ad_id
NUMBER(6),
ad_composite
BLOB,
ad_sourcetext
CLOB,
ad_finaltext
CLOB,
ad_fltextn
NCLOB,
ad_textdocs_ntab
TEXTDOC_TAB,
ad_photo
BLOB,
ad_graphic
BFILE,
ad_header
ADHEADER_TYP)
NESTED TABLE ad_textdocs_ntab STORE AS textdoc_nt
PARTITION BY RANGE (product_id)
(PARTITION p1 VALUES LESS THAN (100),
PARTITION p2 VALUES LESS THAN (200));

The following statement splits partition p2 of that table into partitions p2a and p2b:
ALTER TABLE print_media_part
SPLIT PARTITION p2 AT (150) INTO
(PARTITION p2a TABLESPACE omf_ts1
LOB (ad_photo, ad_composite) STORE AS (TABLESPACE omf_ts2),
PARTITION p2b

12-82 Oracle Database SQL Language Reference

ALTER TABLE

LOB (ad_photo, ad_composite) STORE AS (TABLESPACE omf_ts2))
NESTED TABLE ad_textdocs_ntab INTO (PARTITION nt_p2a, PARTITION nt_p2b);

In both partitions p2a and p2b, Oracle Database creates the LOB segments for columns
ad_photo and ad_composite in tablespace omf_ts2. The LOB segments for the
remaining columns in partition p2a are stored in tablespace omf_ts1. The LOB
segments for the remaining columns in partition p2b remain in the tablespaces in
which they resided prior to this ALTER statement. However, the database creates new
segments for all the LOB data and LOB index segments, even if they are not moved to
a new tablespace.
The database also creates new segments for nested table column ad_textdocs_ntab.
The storage tables is those new segments are nt_p2a and nt_p2b.
The following statement merges back into
one partition the partitions created in "Splitting Table Partitions: Examples" on
page 12-82:

Merging Two Table Partitions: Example

ALTER TABLE sales
MERGE PARTITIONS sales_q4_2000, sales_q4_2000b
INTO PARTITION sales_q4_2000;

The next statement reverses the example in "Splitting Table Partitions: Examples" on
page 12-82:
ALTER TABLE print_media_part
MERGE PARTITIONS p2a, p2b INTO PARTITION p2ab TABLESPACE example
NESTED TABLE ad_textdocs_ntab STORE AS nt_p2ab;

Adding a Table Partition with a LOB and Nested Table Storage: Examples The
following statement adds a partition p3 to the print_media_part table (see preceding
example) and specifies storage characteristics for the BLOB, CLOB, and nested table
columns of that table:
ALTER TABLE print_media_part ADD PARTITION p3 VALUES LESS THAN (400)
LOB(ad_photo, ad_composite) STORE AS (TABLESPACE omf_ts1)
LOB(ad_sourcetext, ad_finaltext) STORE AS (TABLESPACE omf_ts2)
NESTED TABLE ad_textdocs_ntab STORE AS nt_p3;

The LOB data and LOB index segments for columns ad_photo and ad_composite in
partition p3 will reside in tablespace omf_ts1. The remaining attributes for these LOB
columns will be inherited first from the table-level defaults, and then from the
tablespace defaults.
The LOB data segments for columns ad_source_text and ad_finaltext will reside in
the omf_ts2 tablespace, and will inherit all other attributes first from the table-level
defaults, and then from the tablespace defaults.
The partition for the storage table for nested table storage column ad_textdocs_ntab
corresponding to partition p3 of the base table is named nt_p3 and inherits all other
attributes first from the table-level defaults, and then from the tablespace defaults.
The following statements use the list
partitioned table created in "List Partitioning Example" on page 16-77. The first
statement splits the existing default partition into a new south partition and a default
partition:

Working with Default List Partitions: Example

ALTER TABLE list_customers SPLIT PARTITION rest
VALUES ('MEXICO', 'COLOMBIA')
INTO (PARTITION south, PARTITION rest);

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-83

ALTER TABLE

The next statement merges the resulting default partition with the asia partition:
ALTER TABLE list_customers
MERGE PARTITIONS asia, rest INTO PARTITION rest;

The next statement re-creates the asia partition by splitting the default partition:
ALTER TABLE list_customers SPLIT PARTITION rest
VALUES ('CHINA', 'THAILAND')
INTO (PARTITION asia, PARTITION rest);

Dropping a Table Partition: Example The following statement drops partition p3
created in "Adding a Table Partition with a LOB and Nested Table Storage: Examples"
on page 12-83:
ALTER TABLE print_media_part DROP PARTITION p3;

This example creates the table exchange_
table with the same structure as the partitions of the list_customers table created in
"List Partitioning Example" on page 16-77. It then replaces partition rest of table list_
customers with table exchange_table without exchanging local index partitions with
corresponding indexes on exchange_table and without verifying that data in
exchange_table falls within the bounds of partition rest:
Exchanging Table Partitions: Example

CREATE TABLE exchange_table (
customer_id
NUMBER(6),
cust_first_name VARCHAR2(20),
cust_last_name VARCHAR2(20),
cust_address
CUST_ADDRESS_TYP,
nls_territory
VARCHAR2(30),
cust_email
VARCHAR2(30));
ALTER TABLE list_customers
EXCHANGE PARTITION rest WITH TABLE exchange_table
WITHOUT VALIDATION;

The following statement marks all the local
index partitions corresponding to the asia partition of the list_customers table
UNUSABLE:
Modifying Table Partitions: Examples

ALTER TABLE list_customers MODIFY PARTITION asia
UNUSABLE LOCAL INDEXES;

The following statement rebuilds all the local index partitions that were marked
UNUSABLE:
ALTER TABLE list_customers MODIFY PARTITION asia
REBUILD UNUSABLE LOCAL INDEXES;

Moving Table Partitions: Example The following statement moves partition p2b

(from "Splitting Table Partitions: Examples" on page 12-82) to tablespace omf_ts1:
ALTER TABLE print_media_part
MOVE PARTITION p2b TABLESPACE omf_ts1;

Renaming Table Partitions: Examples

The following statement renames a partition

of the sh.sales table:
ALTER TABLE sales RENAME PARTITION sales_q4_2003 TO sales_currentq;

12-84 Oracle Database SQL Language Reference

ALTER TABLE

The following statement uses the print_
media_demo table created in "Partitioned Table with LOB Columns Example" on
page 16-78. It deletes all the data in the p1 partition and deallocates the freed space:

Truncating Table Partitions: Example

ALTER TABLE print_media_demo
TRUNCATE PARTITION p1 DROP STORAGE;

The following statement splits partition sales_
q1_2000 of the sample table sh.sales and updates any global indexes defined on it:

Updating Global Indexes: Example

ALTER TABLE sales SPLIT PARTITION sales_q1_2000
AT (TO_DATE('16-FEB-2000','DD-MON-YYYY'))
INTO (PARTITION q1a_2000, PARTITION q1b_2000)
UPDATE GLOBAL INDEXES;

The following statement splits partition
costs_Q4_2003 of the sample table sh.costs and updates the local index defined on it.
It uses the tablespaces created in "Creating Basic Tablespaces: Examples" on
page 16-96.
Updating Partitioned Indexes: Example

CREATE INDEX cost_ix ON costs(channel_id) LOCAL;
ALTER TABLE costs
SPLIT PARTITION costs_q4_2003 at
(TO_DATE('01-Nov-2003','dd-mon-yyyy'))
INTO (PARTITION c_p1, PARTITION c_p2)
UPDATE INDEXES (cost_ix (PARTITION c_p1 tablespace tbs_02,
PARTITION c_p2 tablespace tbs_03));

The following statements create an object
type, a corresponding object table with a primary-key-based object identifier, and a
table having a user-defined REF column:

Specifying Object Identifiers: Example

CREATE TYPE emp_t AS OBJECT (empno NUMBER, address CHAR(30));
CREATE TABLE emp OF emp_t (
empno PRIMARY KEY)
OBJECT IDENTIFIER IS PRIMARY KEY;
CREATE TABLE dept (dno NUMBER, mgr_ref REF emp_t SCOPE is emp);

The next statements add a constraint and a user-defined REF column, both of which
reference table emp
ALTER TABLE dept ADD CONSTRAINT mgr_cons FOREIGN KEY (mgr_ref)
REFERENCES emp;
ALTER TABLE dept ADD sr_mgr REF emp_t REFERENCES emp;

Adding a Table Column: Example The following statement adds to the countries
table a column named duty_pct of data type NUMBER and a column named visa_
needed of data type VARCHAR2 with a size of 3 and a CHECK integrity constraint:
ALTER TABLE countries
ADD (duty_pct
NUMBER(2,2) CHECK (duty_pct < 10.5),
visa_needed VARCHAR2(3));

The following statement adds to a copy of
the hr.employees table a column named income, which is a combination of salary plus
commission. Both salary and commission are NUMBER columns, so the database creates
the virtual column as a NUMBER column even though the data type is not specified in
the statement:
Adding a Virtual Table Column: Example

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-85

ALTER TABLE

CREATE TABLE emp2 AS SELECT * FROM employees;
ALTER TABLE emp2 ADD (income AS (salary + (salary*commission_pct)));

Modifying Table Columns: Examples The following statement increases the size of
the duty_pct column:
ALTER TABLE countries
MODIFY (duty_pct NUMBER(3,2));

Because the MODIFY clause contains only one column definition, the parentheses
around the definition are optional.
The following statement changes the values of the PCTFREE and PCTUSED parameters
for the employees table to 30 and 60, respectively:
ALTER TABLE employees
PCTFREE 30
PCTUSED 60;

The following statement encrypts the salary column of
the hr.employees table using the encryption algorithm AES256. As described in
"Semantics" above, you must first enable Transparent Data Encryption:

Data Encryption: Examples

ALTER TABLE employees
MODIFY (salary ENCRYPT USING 'AES256' 'NOMAC');

The following statement adds a new encrypted column online_acct_pw to the
oe.customers table, using the default encryption algorithm AES192. Specifying NO SALT
will allow a B-tree index to be created on the column, if desired.
ALTER TABLE customers
ADD (online_acct_pw VARCHAR2(8) ENCRYPT NO SALT 'NOMAC');

The following example decrypts the customer.online_acct_pw column:
ALTER TABLE customers
MODIFY (online_acct_pw DECRYPT);

Allocating Extents: Example The following statement allocates an extent of 5
kilobytes for the employees table and makes it available to instance 4:
ALTER TABLE employees
ALLOCATE EXTENT (SIZE 5K INSTANCE 4);

Because this statement omits the DATAFILE parameter, Oracle Database allocates the
extent in one of the data files belonging to the tablespace containing the table.
Specifying Default Column Value: Examples This statement modifies the min_price
column of the product_information table so that it has a default value of 10:
ALTER TABLE product_information
MODIFY (min_price DEFAULT 10);

If you subsequently add a new row to the product_information table and do not
specify a value for the min_price column, then the value of the min_price column is
automatically 10:
INSERT INTO product_information (product_id, product_name,
list_price)
VALUES (300, 'left-handed mouse', 40.50);

12-86 Oracle Database SQL Language Reference

ALTER TABLE

SELECT product_id, product_name, list_price, min_price
FROM product_information
WHERE product_id = 300;
PRODUCT_ID PRODUCT_NAME
LIST_PRICE MIN_PRICE
---------- -------------------- ---------- ---------300 left-handed mouse
40.5
10

To discontinue previously specified default values, so that they are no longer
automatically inserted into newly added rows, replace the values with NULL, as shown
in this statement:
ALTER TABLE product_information
MODIFY (min_price DEFAULT NULL);

The MODIFY clause need only specify the column name and the modified part of the
definition, rather than the entire column definition. This statement has no effect on any
existing values in existing rows.
Adding a Constraint to an XMLType Table: Example The following example adds a
primary key constraint to the xwarehouses table, created in "XMLType Examples" on
page 16-75:
ALTER TABLE xwarehouses
ADD (PRIMARY KEY(XMLDATA."WarehouseID"));

Refer to XMLDATA Pseudocolumn on page 2-11 for information about this
pseudocolumn.
The following statement renames the cust_fname_
nn constraint on the sample table oe.customers to cust_firstname_nn:

Renaming Constraints: Example

ALTER TABLE customers RENAME CONSTRAINT cust_fname_nn
TO cust_firstname_nn;

Dropping Constraints: Examples

The following statement drops the primary key of

the departments table:
ALTER TABLE departments
DROP PRIMARY KEY CASCADE;

If you know that the name of the PRIMARY KEY constraint is pk_dept, then you could
also drop it with the following statement:
ALTER TABLE departments
DROP CONSTRAINT pk_dept CASCADE;

The CASCADE clause causes Oracle Database to drop any foreign keys that reference the
primary key.
The following statement drops the unique key on the email column of the employees
table:
ALTER TABLE employees
DROP UNIQUE (email);

The DROP clause in this statement omits the CASCADE clause. Because of this omission,
Oracle Database does not drop the unique key if any foreign key references it.
The following statement adds CLOB column resume to the
employee table and specifies LOB storage characteristics for the new column:
LOB Columns: Examples

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-87

ALTER TABLE

ALTER TABLE employees ADD (resume CLOB)
LOB (resume) STORE AS resume_seg (TABLESPACE example);

To modify the LOB column resume to use caching, enter the following statement:
ALTER TABLE employees MODIFY LOB (resume) (CACHE);

The following statement adds a SecureFiles CLOB column resume to the employee table
and specifies LOB storage characteristics for the new column. SecureFiles LOBs must
be stored in tablespaces with automatic segment-space management. Therefore, the
LOB data in this example is stored in the auto_seg_ts tablespace, which was created
in "Specifying Segment Space Management for a Tablespace: Example" on page 16-97:
ALTER TABLE employees ADD (resume CLOB)
LOB (resume) STORE AS SECUREFILE resume_seg (TABLESPACE auto_seg_ts);

To modify the LOB column resume so that it does not use caching, enter the following
statement:
ALTER TABLE employees MODIFY LOB (resume) (NOCACHE);

The following statement adds the nested table column
skills to the employee table:
Nested Tables: Examples

ALTER TABLE employees ADD (skills skill_table_type)
NESTED TABLE skills STORE AS nested_skill_table;

You can also modify nested table storage characteristics. Use the name of the storage
table specified in the nested_table_col_properties to make the modification. You
cannot query or perform DML statements on the storage table. Use the storage table
only to modify the nested table column storage characteristics.
The following statement creates table vet_service with nested table column client
and storage table client_tab. Nested table client_tab is modified to specify
constraints:
CREATE TYPE pet_t AS OBJECT
(pet_id NUMBER, pet_name VARCHAR2(10), pet_dob DATE);
/
CREATE TYPE pet AS TABLE OF pet_t;
/
CREATE TABLE vet_service (vet_name VARCHAR2(30),
client
pet)
NESTED TABLE client STORE AS client_tab;
ALTER TABLE client_tab ADD UNIQUE (pet_id);

The following statement alters the storage table for a nested table of REF values to
specify that the REF is scoped:
CREATE TYPE emp_t AS OBJECT (eno number, ename char(31));
CREATE TYPE emps_t AS TABLE OF REF emp_t;
CREATE TABLE emptab OF emp_t;
CREATE TABLE dept (dno NUMBER, employees emps_t)
NESTED TABLE employees STORE AS deptemps;
ALTER TABLE deptemps ADD (SCOPE FOR (COLUMN_VALUE) IS emptab);

Similarly, to specify storing the REF with rowid:
ALTER TABLE deptemps ADD (REF(column_value) WITH ROWID);

12-88 Oracle Database SQL Language Reference

ALTER TABLE

In order to execute these ALTER TABLE statements successfully, the storage table
deptemps must be empty. Also, because the nested table is defined as a table of scalar
values (REF values), Oracle Database implicitly provides the column name COLUMN_
VALUE for the storage table.
See Also:
■

■

CREATE TABLE on page 16-6 for more information about nested
table storage
Oracle Database Object-Relational Developer's Guide for more
information about nested tables

REF Columns: Examples

The following statement creates an object type dept_t and

then creates table staff:
CREATE TYPE dept_t AS OBJECT
(deptno NUMBER, dname VARCHAR2(20));
/
CREATE TABLE staff
(name
VARCHAR(100),
salary NUMBER,
dept
REF dept_t);

An object table offices is created as:
CREATE TABLE offices OF dept_t;

The dept column can store references to objects of dept_t stored in any table. If you
would like to restrict the references to point only to objects stored in the departments
table, then you could do so by adding a scope constraint on the dept column as
follows:
ALTER TABLE staff
ADD (SCOPE FOR (dept) IS offices);

The preceding ALTER TABLE statement will succeed only if the staff table is empty.
If you want the REF values in the dept column of staff to also store the rowids, then
issue the following statement:
ALTER TABLE staff
ADD (REF(dept) WITH ROWID);

For examples of defining integrity constraints with the ALTER
TABLE statement, see the constraint on page 8-4.
Additional Examples

For examples of changing the storage parameters of a table, see the storage_clause on
page 8-46.

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-89

ALTER TABLESPACE

ALTER TABLESPACE
Purpose
12

Use the ALTER TABLESPACE statement to alter an existing tablespace or one or more of
its data files or temp files.
You cannot use this statement to convert a dictionary-managed tablespace to a locally
managed tablespace. For that purpose, use the DBMS_SPACE_ADMIN package, which is
documented in Oracle Database PL/SQL Packages and Types Reference.
See Also: Oracle Database Administrator's Guide and CREATE
TABLESPACE on page 16-83 for information on creating a tablespace

Prerequisites
12

To alter the SYSAUX tablespace, you must have the SYSDBA system privilege.
If you have the ALTER TABLESPACE system privilege, then you can perform any ALTER
TABLESPACE operation. If you have the MANAGE TABLESPACE system privilege, then you
can only perform the following operations:
■

Take a tablespace online or offline

■

Begin or end a backup

■

Make a tablespace read only or read write

■

Change the state of a tablespace to PERMANENT or TEMPORARY

■

Set the default logging mode of a tablespace to LOGGING or NOLOGGING

■

Put a tablespace in force logging mode or take it out of force logging mode

■

Rename a tablespace or a tablespace data file

■

Specify RETENTION GUARANTEE or RETENTION NOGUARANTEE for an undo tablespace

■

Resize a data file for a tablespace

■

Enable or disable autoextension of a data file for a tablespace

■

Shrink the amount of space a temporary tablespace or a temp file is taking

Before you can make a tablespace read only, the following conditions must be met:
■
■

■

The tablespace must be online.
The tablespace must not contain any active rollback segments. For this reason, the
SYSTEM tablespace can never be made read only, because it contains the SYSTEM
rollback segment. Additionally, because the rollback segments of a read-only
tablespace are not accessible, Oracle recommends that you drop the rollback
segments before you make a tablespace read only.
The tablespace must not be involved in an open backup, because the end of a
backup updates the header file of all data files in the tablespace.

Performing this function in restricted mode may help you meet these restrictions,
because only users with RESTRICTED SESSION system privilege can be logged on.

12-90 Oracle Database SQL Language Reference

ALTER TABLESPACE

Syntax
12

alter_tablespace::=
table_compression

storage_clause

DEFAULT
MINIMUM
RESIZE

EXTENT

size_clause

size_clause

COALESCE
KEEP
SHRINK
RENAME

size_clause

SPACE
TO

new_tablespace_name

BEGIN
ALTER

TABLESPACE

tablespace

BACKUP
END

;

datafile_tempfile_clauses
tablespace_logging_clauses
tablespace_group_clause
tablespace_state_clauses
autoextend_clause
flashback_mode_clause
tablespace_retention_clause

If you specify the DEFAULT clause, then you must specify at
least one of the clauses table_compression or storage_clause.

Note:

(table_compression::= on page 12-5—part of ALTER TABLE, storage_clause::= on page 8-50,
size_clause::= on page 8-47, datafile_tempfile_clauses::= on page 12-92, tablespace_logging_
clauses::= on page 12-92, tablespace_group_clause::= on page 12-92, tablespace_state_
clauses::= on page 12-92, autoextend_clause::= on page 12-93, flashback_mode_clause::= on
page 12-93, tablespace_retention_clause::= on page 12-93)

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-91

ALTER TABLESPACE

datafile_tempfile_clauses::=
,
file_specification

DATAFILE
ADD
TEMPFILE
DATAFILE

’

filename

TEMPFILE

file_number

’

DROP

’
SHRINK

filename

KEEP

’

size_clause

TEMPFILE
file_number
,

RENAME

DATAFILE

’

DATAFILE

ONLINE

TEMPFILE

OFFLINE

filename

,
’

TO

(file_specification::= on page 8-29).
tablespace_logging_clauses::=
logging_clause
NO
FORCE

LOGGING

(logging_clause::= on page 8-38)
tablespace_group_clause::=
tablespace_group_name
TABLESPACE

GROUP
’

’

tablespace_state_clauses::=
ONLINE
NORMAL
TEMPORARY
IMMEDIATE
OFFLINE
ONLY
READ
WRITE
PERMANENT
TEMPORARY

12-92 Oracle Database SQL Language Reference

’

filename

’

ALTER TABLESPACE

autoextend_clause::=
OFF
AUTOEXTEND

NEXT

size_clause

maxsize_clause

ON

(size_clause::= on page 8-47)
maxsize_clause::=
UNLIMITED
MAXSIZE
size_clause

(size_clause::= on page 8-47)
flashback_mode_clause::=
ON
FLASHBACK
OFF

tablespace_retention_clause::=
GUARANTEE
RETENTION
NOGUARANTEE

Semantics
12

tablespace
Specify the name of the tablespace to be altered.
Restrictions on Altering Tablespaces

Altering tablespaces is subject to the following

restrictions:
■

■

■

If tablespace is an undo tablespace, then the only other clauses you can specify in
this statement are ADD DATAFILE, RENAME DATAFILE, RENAME TO (renaming the
tablespace), DATAFILE ... ONLINE, DATAFILE ... OFFLINE, BEGIN BACKUP, and END
BACKUP.
You cannot make the SYSTEM tablespace read only or temporary and you cannot
take it offline.
For locally managed temporary tablespaces, the only clause you can specify in this
statement is the ADD clause.
See Also: Oracle Database Administrator's Guide for information on
automatic undo management and undo tablespaces

DEFAULT Clause
This clause lets you specify new default parameters for the tablespace. The new
default parameters apply to objects subsequently created in the tablespace.
The clauses table_compression and storage_clause have the same semantics in
CREATE TABLESPACE and ALTER TABLESPACE. For complete information on these clauses,
SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-93

ALTER TABLESPACE

refer to DEFAULT Clause on page 16-90 in the documentation on CREATE TABLESPACE.

MINIMUM EXTENT
This clause is valid only for permanent dictionary-managed tablespaces. The MINIMUM
EXTENT clause lets you control free space fragmentation in the tablespace by ensuring
that every used or free extent in a tablespace is at least as large as, and is a multiple of,
the value specified in the size_clause.
You cannot specify this clause for a locally
managed tablespace or for a dictionary-managed temporary tablespace.

Restriction on MINIMUM EXTENT

See Also: size_clause on page 8-47 for information about that clause,
Oracle Database Administrator's Guide for more information about using
MINIMUM EXTENT to control space fragmentation

RESIZE Clause
This clause is valid only for bigfile tablespaces. It lets you increase or decrease the size
of the single data file to an absolute size. Use K, M, G, or T to specify the size in
kilobytes, megabytes, gigabytes, or terabytes, respectively.
To change the size of a newly added data file or temp file in smallfile tablespaces, use
the ALTER DATABASE ... autoextend_clause (see database_file_clauses on page 10-25).
See Also: BIGFILE | SMALLFILE on page 16-86 for information on
bigfile tablespaces

COALESCE
For each data file in the tablespace, this clause combines all contiguous free extents
into larger contiguous extents.

SHRINK SPACE Clause
This clause is valid only for temporary tablespaces. It lets you reduce the amount of
space the tablespace is taking. In the optional KEEP clause, the size_clause defines the
lower bound that a tablespace can be shrunk to. It is the opposite of MAXSIZE for an
autoextensible tablespace. If you omit the KEEP clause, then the database will attempt
to shrink the tablespace as much as possible as long as other tablespace storage
attributes are satisfied.

RENAME Clause
Use this clause to rename tablespace. This clause is valid only if tablespace and all
its data files are online and the COMPATIBLE parameter is set to 10.0.0 or greater. You
can rename both permanent and temporary tablespaces.
If tablespace is read only, then Oracle Database does not update the data file headers
to reflect the new name. The alert log will indicate that the data file headers have not
been updated.
If you re-create the control file, and if the data files that Oracle
Database uses for this purpose are restored backups whose headers
reflect the old tablespace name, then the re-created control file will
also reflect the old tablespace name. However, after the database is
fully recovered, the control file will reflect the new name.

Note:

12-94 Oracle Database SQL Language Reference

ALTER TABLESPACE

If tablespace has been designated as the undo tablespace for a single-instance
database, or for any instance in an Oracle Real Application Clusters (Oracle RAC)
environment, and if a server parameter file was used to start up the database, then
Oracle Database changes the value of the UNDO_TABLESPACE parameter for that instance
in the server parameter file (spfile) to reflect the new tablespace name. If a
single-instance database is using a parameter file (pfile) instead of an spfile, then the
database puts a message in the alert log advising the database administrator to change
the value manually in the pfile.

The RENAME clause does not change the value of the UNDO_
TABLESPACE parameter in the running instance. Although this does not
affect the functioning of the undo tablespace, Oracle recommends that
you issue the following statement to manually change the value of
UNDO_TABLESPACE to the new tablespace name for the duration of the
instance:
Note:

ALTER SYSTEM SET UNDO_TABLESPACE = new_tablespace_name SCOPE =
MEMORY;

You only need to issue this statement once. If the UNDO_TABLESPACE
parameter is set to the new tablespace name in the pfile or spfile, then
the parameter will be set correctly when the instance is next restarted.
Restriction on Renaming Tablespaces You cannot rename the SYSTEM or SYSAUX
tablespaces.

BACKUP Clauses
Use these clauses to move all data files in a tablespace into or out of online (sometimes
called hot) backup mode.
See Also:
■

■

■

Oracle Database Administrator's Guide for information on restarting
the database without media recovery
ALTER DATABASE "BACKUP Clauses" on page 10-25 for information
on moving all data files in the database into and out of online
backup mode
ALTER DATABASE alter_datafile_clause on page 10-27 for information
on taking individual data files out of online backup mode

BEGIN BACKUP
Specify BEGIN BACKUP to indicate that an open backup is to be performed on the data
files that make up this tablespace. This clause does not prevent users from accessing
the tablespace. You must use this clause before beginning an open backup.
Restrictions on Beginning Tablespace Backup Beginning tablespace backup is
subject to the following restrictions:
■

■

You cannot specify this clause for a read-only tablespace or for a temporary locally
managed tablespace.
While the backup is in progress, you cannot take the tablespace offline normally,
shut down the instance, or begin another backup of the tablespace.

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-95

ALTER TABLESPACE

See Also:

"Backing Up Tablespaces: Examples" on page 12-101

END BACKUP
Specify END BACKUP to indicate that an online backup of the tablespace is complete. Use
this clause as soon as possible after completing an online backup. Otherwise, if an
instance failure or SHUTDOWN ABORT occurs, then Oracle Database assumes that media
recovery (possibly requiring archived redo log) is necessary at the next instance
startup.
Restriction on Ending Tablespace Backup

You cannot use this clause on a read-only

tablespace.

datafile_tempfile_clauses
The tablespace file clauses let you add or modify a data file or temp file.
ADD Clause
Specify ADD to add to the tablespace a data file or temp file specified by file_
specification. Use the datafile_tempfile_spec form of file_specification (see
file_specification on page 8-29) to list regular data files and temp files in an operating
system file system or to list Oracle Automatic Storage Management disk group files.
For locally managed temporary tablespaces, this is the only clause you can specify at
any time.
If you omit file_specification, then Oracle Database creates an Oracle Managed File
of 100M with AUTOEXTEND enabled.
You can add a data file or temp file to a locally managed tablespace that is online or to
a dictionary managed tablespace that is online or offline. Ensure the file is not in use
by another database.
Restriction on Adding Data Files and Temp Files You cannot specify this clause for a

bigfile (single-file) tablespace, as such a tablespace has only one data file or temp file.

On some operating systems, Oracle does not allocate space
for a temp file until the temp file blocks are actually accessed. This
delay in space allocation results in faster creation and resizing of
temp files, but it requires that sufficient disk space is available
when the temp files are later used. To avoid potential problems,
before you create or resize a temp file, ensure that the available disk
space exceeds the size of the new temp file or the increased size of a
resized temp file. The excess space should allow for anticipated
increases in disk space use by unrelated operations as well. Then
proceed with the creation or resizing operation.

Note:

See Also: file_specification on page 8-29, "Adding and Dropping Data
Files and Temp Files: Examples" on page 12-101, and "Adding an
Oracle-managed Data File: Example" on page 12-102

DROP Clause
Specify DROP to drop from the tablespace an empty data file or temp file specified by
filename or file_number. This clause causes the data file or temp file to be removed

12-96 Oracle Database SQL Language Reference

ALTER TABLESPACE

from the data dictionary and deleted from the operating system. The database must be
open at the time this clause is specified.
The ALTER TABLESPACE ... DROP TEMPFILE statement is equivalent to specifying the
ALTER DATABASE TEMPFILE ... DROP INCLUDING DATAFILES.
Restrictions on Dropping Files

To drop a data file or temp file, the data file or temp

file:
■
■

■

■

Must be empty.
Cannot be the first file that was created in the tablespace. In such cases, drop the
tablespace instead.
Cannot be in a read-only tablespace that was migrated from dictionary managed
to locally managed. Dropping a data file from all other read-only tablespaces is
supported.
Cannot be offline.
See Also:
■

■

■

ALTER DATABASE alter_tempfile_clause on page 10-27 for additional
information on dropping temp files
Oracle Database Administrator's Guide for information on data file
numbers and for guidelines on managing data files
"Adding and Dropping Data Files and Temp Files: Examples" on
page 12-101

SHRINK TEMPFILE Clause
This clause is valid only when altering a temporary tablespace. It lets you reduce the
amount of space the specified temp file is taking. In the optional KEEP clause, the size_
clause defines the lower bound that the temp file can be shrunk to. It is the opposite
of MAXSIZE for an autoextensible tablespace. If you omit the KEEP clause, then the
database will attempt to shrink the temp file as much as possible as long as other
storage attributes are satisfied.
RENAME DATAFILE Clause
Specify RENAME DATAFILE to rename one or more of the tablespace data files. The
database must be open, and you must take the tablespace offline before renaming it.
Each filename must fully specify a data file using the conventions for filenames on
your operating system.
This clause merely associates the tablespace with the new file rather than the old one.
This clause does not actually change the name of the operating system file. You must
change the name of the file through your operating system.
See Also: "Moving and Renaming Tablespaces: Example" on
page 12-101

ONLINE | OFFLINE Clauses
Use these clauses to take all data files or temp files in the tablespace offline or put
them online. These clauses have no effect on the ONLINE or OFFLINE status of the
tablespace itself.
The database must be mounted. If tablespace is SYSTEM, or an undo tablespace, or the
default temporary tablespace, then the database must not be open.

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-97

ALTER TABLESPACE

tablespace_logging_clauses
Use these clauses to set or change the logging characteristics of the tablespace.
logging_clause
Specify LOGGING if you want logging of all tables, indexes, and partitions within the
tablespace. The tablespace-level logging attribute can be overridden by logging
specifications at the table, index, and partition levels.
When an existing tablespace logging attribute is changed by an ALTER TABLESPACE
statement, all tables, indexes, and partitions created after the statement will have the
new default logging attribute (which you can still subsequently override). The logging
attribute of existing objects is not changed.
If the tablespace is in FORCE LOGGING mode, then you can specify NOLOGGING in this
statement to set the default logging mode of the tablespace to NOLOGGING, but this will
not take the tablespace out of FORCE LOGGING mode.

[NO] FORCE LOGGING
Use this clause to put the tablespace in force logging mode or take it out of force
logging mode. The database must be open and in READ WRITE mode. Neither of these
settings changes the default LOGGING or NOLOGGING mode of the tablespace.
Restriction on Force Logging Mode You cannot specify FORCE LOGGING for an undo
or a temporary tablespace.
See Also: Oracle Database Administrator's Guide for information on
when to use FORCE LOGGING mode and "Changing Tablespace Logging
Attributes: Example" on page 12-102

tablespace_group_clause
This clause is valid only for locally managed temporary tablespaces. Use this clause to
add tablespace to or remove it from the tablespace_group_name tablespace group.
■

■

Specify a group name to indicate that tablespace is a member of this tablespace
group. If tablespace_group_name does not already exist, then Oracle Database
implicitly creates it when you alter tablespace to be a member of it.
Specify an empty string (' ') to remove tablespace from the tablespace_group_
name tablespace group.

Restriction on Tablespace Groups You cannot specify a tablespace group for a

permanent tablespace or for a dictionary-managed temporary tablespace.
See Also: Oracle Database Administrator's Guide for more information
on tablespace groups and "Assigning a Tablespace Group: Example"
on page 13-12

tablespace_state_clauses
Use these clauses to set or change the state of the tablespace.
ONLINE | OFFLINE
Specify ONLINE to bring the tablespace online. Specify OFFLINE to take the tablespace
offline and prevent further access to its segments. When you take a tablespace offline,
all of its data files are also offline.

12-98 Oracle Database SQL Language Reference

ALTER TABLESPACE

Before taking a tablespace offline for a long time,
consider changing the tablespace allocation of any users who have
been assigned the tablespace as either a default or temporary
tablespace. While the tablespace is offline, such users cannot allocate
space for objects or sort areas in the tablespace. See ALTER USER on
page 13-6 for more information on allocating tablespace quota to
users.
Suggestion:

Restriction on Taking Tablespaces Offline You cannot take a temporary tablespace

offline.
Specify NORMAL to flush all blocks in all data files in the tablespace
out of the system global area (SGA). You need not perform media recovery on this
tablespace before bringing it back online. This is the default.

OFFLINE NORMAL

OFFLINE TEMPORARY If you specify TEMPORARY, then Oracle Database performs a
checkpoint for all online data files in the tablespace but does not ensure that all files
can be written. Files that are offline when you issue this statement may require media
recovery before you bring the tablespace back online.

If you specify IMMEDIATE, then Oracle Database does not
ensure that tablespace files are available and does not perform a checkpoint. You must
perform media recovery on the tablespace before bringing it back online.

OFFLINE IMMEDIATE

The FOR RECOVER setting for ALTER TABLESPACE ... OFFLINE has
been deprecated. The syntax is supported for backward compatibility.
However, Oracle recommends that you use the transportable
tablespaces feature for tablespace recovery.

Note:

See Also: Oracle Database Backup and Recovery User's Guide for
information on using transportable tablespaces to perform media
recovery

READ ONLY | READ WRITE
Specify READ ONLY to place the tablespace in transition read-only mode. In this state,
existing transactions can complete (commit or roll back), but no further DML
operations are allowed to the tablespace except for rollback of existing transactions
that previously modified blocks in the tablespace. You cannot make the SYSAUX,
SYSTEM, or temporary tablespaces READ ONLY.
When a tablespace is read only, you can copy its files to read-only media. You must
then rename the data files in the control file to point to the new location by using the
SQL statement ALTER DATABASE ... RENAME.
See Also:
■

■

Oracle Database Concepts for more information on read-only
tablespaces
ALTER DATABASE on page 10-8

Specify READ WRITE to indicate that write operations are allowed on a previously
read-only tablespace.

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-99

ALTER TABLESPACE

PERMANENT | TEMPORARY
Specify PERMANENT to indicate that the tablespace is to be converted from a temporary
to a permanent tablespace. A permanent tablespace is one in which permanent
database objects can be stored. This is the default when a tablespace is created.
Specify TEMPORARY to indicate that the tablespace is to be converted from a permanent
to a temporary tablespace. A temporary tablespace is one in which no permanent
database objects can be stored. Objects in a temporary tablespace persist only for the
duration of the session.
Restrictions on Temporary Tablespaces

Temporary tablespaces are subject to the

following restrictions:
■
■

■

You cannot specify TEMPORARY for the SYSAUX tablespace.
If tablespace was not created with a standard block size, then you cannot change
it from permanent to temporary.
You cannot specify TEMPORARY for a tablespace in FORCE LOGGING mode.

autoextend_clause
This clause is valid only for bigfile (single-file) tablespaces. Use this clause to enable or
disable autoextension of the single data file in the tablespace. To enable or disable
autoextension of a newly added data file or temp file in smallfile tablespaces, use the
autoextend_clause of the database_file_clauses on page 10-25 in the ALTER DATABASE
statement.
See Also:
■

■

Oracle Database Administrator's Guide for information about bigfile
(single-file) tablespaces
file_specification on page 8-29 for more information about the
autoextend_clause

flashback_mode_clause
Use this clause to specify whether this tablespace should participate in any subsequent
FLASHBACK DATABASE operation.
■
■

For you to turn FLASHBACK mode on, the database must be mounted and closed.
For you to turn FLASHBACK mode off, the database must be mounted, either open
READ WRITE or closed.

This clause is not valid for temporary tablespaces.
Refer to CREATE TABLESPACE on page 16-83 for more complete information on this
clause.
See Also: Oracle Database Backup and Recovery User's Guide for more
information about Flashback Database

tablespace_retention_clause
This clause has the same semantics in CREATE TABLESPACE and ALTER TABLESPACE
statements. Refer to tablespace_retention_clause on page 16-94 in the documentation on
CREATE TABLESPACE.

12-100 Oracle Database SQL Language Reference

ALTER TABLESPACE

Examples
12

Backing Up Tablespaces: Examples The following statement signals to the database

that a backup is about to begin:
ALTER TABLESPACE tbs_01
BEGIN BACKUP;

The following statement signals to the database that the backup is finished:
ALTER TABLESPACE tbs_01
END BACKUP;

Moving and Renaming Tablespaces: Example This example moves and renames a

data file associated with the tbs_02 tablespace, created in "Enabling Autoextend for a
Tablespace: Example" on page 16-96, from diskb:tbs_f5.dbf to diska:tbs_f5.dbf:
1.

Take the tablespace offline using an ALTER TABLESPACE statement with the OFFLINE
clause:
ALTER TABLESPACE tbs_02 OFFLINE NORMAL;

2.

Copy the file from diskb:tbs_f5.dbf to diska:tbs_f5.dbf using your operating
system commands.

3.

Rename the data file using an ALTER TABLESPACE statement with the RENAME
DATAFILE clause:
ALTER TABLESPACE tbs_02
RENAME DATAFILE 'diskb:tbs_f5.dbf'
TO
'diska:tbs_f5.dbf';

4.

Bring the tablespace back online using an ALTER TABLESPACE statement with the
ONLINE clause:
ALTER TABLESPACE tbs_02 ONLINE;

The following statement
adds a data file to the tablespace. When more space is needed, new 10-kilobytes
extents will be added up to a maximum of 100 kilobytes:

Adding and Dropping Data Files and Temp Files: Examples

ALTER TABLESPACE tbs_03
ADD DATAFILE 'tbs_f04.dbf'
SIZE 100K
AUTOEXTEND ON
NEXT 10K
MAXSIZE 100K;

The following statement drops the empty data file:
ALTER TABLESPACE tbs_03
DROP DATAFILE 'tbs_f04.dbf';

The following statements add a temp file to the temporary tablespace created in
"Creating a Temporary Tablespace: Example" on page 16-95 and then drops the temp
file:
ALTER TABLESPACE temp_demo ADD TEMPFILE 'temp05.dbf' SIZE 5 AUTOEXTEND ON;
ALTER TABLESPACE temp_demo DROP TEMPFILE 'temp05.dbf';

SQL Statements: ALTER TABLE to ALTER TABLESPACE 12-101

ALTER TABLESPACE

The following statement
manages the space in the temporary tablespace created in "Creating a Temporary
Tablespace: Example" on page 16-95 using the SHRINK SPACE clause. The KEEP clause is
omitted, so the database will attempt to shrink the tablespace as much as possible as
long as other tablespace storage attributes are satisfied.
Managing Space in a Temporary Tablespace: Example

ALTER TABLESPACE temp_demo SHRINK SPACE;

Adding an Oracle-managed Data File: Example The following example adds an

Oracle-managed data file to the omf_ts1 tablespace (see "Creating Oracle Managed
Files: Examples" on page 16-97 for the creation of this tablespace). The new data file is
100M and is autoextensible with unlimited maximum size:
ALTER TABLESPACE omf_ts1 ADD DATAFILE;

The following example changes
the default logging attribute of a tablespace to NOLOGGING:

Changing Tablespace Logging Attributes: Example
ALTER TABLESPACE tbs_03 NOLOGGING;

Altering a tablespace logging attribute has no affect on the logging attributes of the
existing schema objects within the tablespace. The tablespace-level logging attribute
can be overridden by logging specifications at the table, index, and partition levels.
Changing Undo Data Retention: Examples The following statement changes the
undo data retention for tablespace undots1 to normal undo data behavior:
ALTER TABLESPACE undots1
RETENTION NOGUARANTEE;

The following statement changes the undo data retention for tablespace undots1 to
behavior that preserves unexpired undo data:
ALTER TABLESPACE undots1
RETENTION GUARANTEE;

12-102 Oracle Database SQL Language Reference

13

13
SQL Statements: ALTER TRIGGER to COMMIT
This chapter contains the following SQL statements:
■

ALTER TRIGGER

■

ALTER TYPE

■

ALTER USER

■

ALTER VIEW

■

ANALYZE

■

ASSOCIATE STATISTICS

■

AUDIT

■

CALL

■

COMMENT

■

COMMIT

SQL Statements: ALTER TRIGGER to COMMIT

13-1

ALTER TRIGGER

ALTER TRIGGER
Purpose
13

Triggers are defined using PL/SQL. Therefore, this section provides some general
information but refers to Oracle Database PL/SQL Language Reference for details of
syntax and semantics.
Use the ALTER TRIGGER statement to enable, disable, or compile a database trigger.
This statement does not change the declaration or definition of
an existing trigger. To redeclare or redefine a trigger, use the CREATE
TRIGGER statement with the OR REPLACE keywords.
Note:

See Also:
■

■

■

CREATE TRIGGER on page 16-98 for information on creating a
trigger
DROP TRIGGER on page 18-12 for information on dropping a
trigger
Oracle Database Concepts for general information on triggers

Prerequisites
13

The trigger must be in your own schema or you must have ALTER ANY TRIGGER system
privilege.
In addition, to alter a trigger on DATABASE, you must have the ADMINISTER DATABASE
TRIGGER privilege.
CREATE TRIGGER on page 16-98 for more information on
triggers based on DATABASE triggers

See Also:

Syntax
13

alter_trigger::=
ENABLE
schema
ALTER

TRIGGER

DISABLE

.
trigger

RENAME

TO

new_name

;

trigger_compile_clause

(trigger_compile_clause: See Oracle Database PL/SQL Language Reference for the
syntax of this clause.)

Semantics
13

schema
Specify the schema containing the trigger. If you omit schema, then Oracle Database
assumes the trigger is in your own schema.

13-2 Oracle Database SQL Language Reference

ALTER TRIGGER

trigger
Specify the name of the trigger to be altered.

ENABLE | DISABLE
Specify ENABLE to enable the trigger. You can also use the ENABLE ALL TRIGGERS clause
of ALTER TABLE to enable all triggers associated with a table. See ALTER TABLE on
page 12-2.
Specify DISABLE to disable the trigger. You can also use the DISABLE ALL TRIGGERS
clause of ALTER TABLE to disable all triggers associated with a table.

RENAME Clause
Specify RENAME TO new_name to rename the trigger. Oracle Database renames the trigger
and leaves it in the same state it was in before being renamed.
When you rename a trigger, the database rebuilds the remembered source of the
trigger in the USER_SOURCE, ALL_SOURCE, and DBA_SOURCE data dictionary views. As a
result, comments and formatting may change in the TEXT column of those views even
though the trigger source did not change.

trigger_compile_clause
See Oracle Database PL/SQL Language Reference for the syntax and semantics of this
clause and for complete information on creating and compiling triggers.

SQL Statements: ALTER TRIGGER to COMMIT

13-3

ALTER TYPE

ALTER TYPE
Purpose
13

Object types are defined using PL/SQL. Therefore, this section provides some general
information but refers to Oracle Database PL/SQL Language Reference for details of
syntax and semantics.
Use the ALTER TYPE statement to add or drop member attributes or methods. You can
change the existing properties (FINAL or INSTANTIABLE) of an object type, and you can
modify the scalar attributes of the type.
You can also use this statement to recompile the specification or body of the type or to
change the specification of an object type by adding new object member subprogram
specifications.

Prerequisites
13

The object type must be in your own schema and you must have CREATE TYPE or
CREATE ANY TYPE system privilege, or you must have ALTER ANY TYPE system privileges.

Syntax
13

alter_type::=
schema
ALTER

TYPE

.
type

alter_type_clauses

(alter_type_clauses: See Oracle Database PL/SQL Language Reference for the syntax of
this clause.)

Semantics
13

schema
Specify the schema that contains the type. If you omit schema, then Oracle Database
assumes the type is in your current schema.

type
Specify the name of an object type, a nested table type, or a varray type.
You cannot evolve an editioned object type. The ALTER
TYPE statement fails with ORA-22348 if either of the following is true:
Restriction on type_name
■

■

The type is an editioned object type and the ALTER TYPE statement has no compile_
type_clause. You can use the ALTER TYPE statement to recompile an editioned
object type, but not for any other purpose.
The type has a dependent that is an editioned object type and the ALTER TYPE
statement has a CASCADE clause.

Refer to Oracle Database PL/SQL Language Reference for more information on the
compile_type_clause and the CASCADE clause.

13-4 Oracle Database SQL Language Reference

ALTER TYPE

alter_type_clauses
See Oracle Database PL/SQL Language Reference for the syntax and semantics of this
clause and for complete information on creating and compiling object types.

SQL Statements: ALTER TRIGGER to COMMIT

13-5

ALTER USER

ALTER USER
Purpose
13

Use the ALTER USER statement:
■

■
■

To change the authentication or database resource characteristics of a database
user
To permit a proxy server to connect as a client without authentication
In an Oracle Automatic Storage Management (Oracle ASM) cluster, to change the
password of a user in the password file that is local to the Oracle ASM instance of
the current node
See Also: Oracle Database Security Guide for detailed information
about user authentication methods

Prerequisites
13

In general, you must have the ALTER USER system privilege. However, the current user
can change his or her own password without this privilege.
You must be authenticated AS SYSASM to change the password of a user other than
yourself in an Oracle ASM instance password file.

13-6 Oracle Database SQL Language Reference

ALTER USER

Syntax
13

alter_user::=
REPLACE
BY

old_password

password
certificate_DN
AS

’

’
kerberos_principal_name

IDENTIFIED

EXTERNALLY
directory_DN
AS

’

’

GLOBALLY
DEFAULT

TABLESPACE

tablespace
tablespace

TEMPORARY

TABLESPACE
tablespace_group_name
size_clause

QUOTA

ON

tablespace

UNLIMITED
PROFILE

profile

user

,
role
,
DEFAULT

ROLE

EXCEPT

role

ALL
NONE
ALTER

USER

PASSWORD

;

EXPIRE
LOCK

ACCOUNT
UNLOCK
FORCE
ENABLE

EDITIONS

,
user

proxy_clause

(size_clause::= on page 8-47)
proxy_clause::=
GRANT

CONNECT

ENTERPRISE

USERS

db_user_proxy

db_user_proxy_clauses

THROUGH

ENTERPRISE
REVOKE

CONNECT

USERS

THROUGH
db_user_proxy

SQL Statements: ALTER TRIGGER to COMMIT

13-7

ALTER USER

db_user_proxy_clauses::=
,
role_name
ROLE
WITH

,
ALL

NO

EXCEPT

role_name

ROLES

AUTHENTICATION

REQUIRED

Semantics
13

The keywords, parameters, and clauses described in this section are unique to ALTER
USER or have different semantics than they have in CREATE USER. Keywords,
parameters, and clauses that do not appear here have the same meaning as in the
CREATE USER statement.

Oracle recommends that user names and passwords be
encoded in ASCII or EBCDIC characters only, depending on your
platform.

Note:

See Also: CREATE USER on page 17-7 for information on the
keywords and parameters and CREATE PROFILE on page 15-50 for
information on assigning limits on database resources to a user

IDENTIFIED Clause
BY password Specify BY password to specify a new password for the user.
Passwords are case sensitive. Any subsequent CONNECT string used to connect this user
to the database must specify the password using the same case (upper, lower, or
mixed) that is used in this ALTER USER statement. Passwords can contain single-byte, or
multibyte characters, or both from your database character set.

Oracle Database expects a different timestamp for each
resetting of a particular password. If you reset one password multiple
times within one second (for example, by cycling through a set of
passwords using a script), then the database may return an error
message that the password cannot be reused. For this reason, Oracle
recommends that you avoid using scripts to reset passwords.

Note:

You can omit the REPLACE clause if you are setting your own password or you have the
ALTER USER system privilege and you are changing another user's password. However,
unless you have the ALTER USER system privilege, you must always specify the REPLACE
clause if a password complexity verification function has been enabled, either by
running the UTLPWDMG.SQL script or by specifying such a function in the PASSWORD_
VERIFY_FUNCTION parameter of a profile that has been assigned to the user.
In an Oracle ASM cluster, you can use this clause to change the password of a user in
the password file that is local to an Oracle ASM instance of the current node. You must
be authenticated AS SYSASM to specify IDENTIFIED BY password without the REPLACE

13-8 Oracle Database SQL Language Reference

ALTER USER

old_password clause. If you are not authenticated AS SYSASM, then you can only change
your own password by specifying REPLACE old_password.
Oracle Database does not check the old password, even if you provide it in the
REPLACE clause, unless you are changing your own existing password.
See Also:

Oracle Database Security Guide for guidelines on creating

passwords
GLOBALLY Refer to CREATE USER on page 17-7 for more information on this

clause.
You can change a user's access verification method from IDENTIFIED GLOBALLY to either
IDENTIFIED BY password or IDENTIFIED EXTERNALLY. You can change a user's access
verification method to IDENTIFIED GLOBALLY from one of the other methods only if all
external roles granted explicitly to the user are revoked.
EXTERNALLY Refer to CREATE USER on page 17-7 for more information on this
clause.
See Also: Oracle Database Enterprise User Security Administrator's
Guide for more information on globally and externally identified
users, "Changing User Identification: Example" on page 13-11, and
"Changing User Authentication: Examples" on page 13-12

DEFAULT TABLESPACE Clause
Use this clause to assign or reassign a tablespace for the user's permanent segments.
This clause overrides any default tablespace that has been specified for the database.
You cannot specify a locally managed temporary
tablespace, including an undo tablespace, or a dictionary-managed temporary
tablespace, as a user's default tablespace.

Restriction on Default Tablespaces

TEMPORARY TABLESPACE Clause
Use this clause to assign or reassign a tablespace or tablespace group for the user's
temporary segments.
■
■

Specify tablespace to indicate the user's temporary tablespace.
Specify tablespace_group_name to indicate that the user can save temporary
segments in any tablespace in the tablespace group specified by tablespace_
group_name.

Any individual tablespace you assign or
reassign as the user's temporary tablespace must be a temporary tablespace and must
have a standard block size.

Restriction on User Temporary Tablespace

See Also:

"Assigning a Tablespace Group: Example" on page 13-12

DEFAULT ROLE Clause
Specify the roles granted by default to the user at logon. This clause can contain only
roles that have been granted directly to the user with a GRANT statement, or roles
created by the user with the CREATE ROLE privilege. You cannot use the DEFAULT ROLE
clause to specify:
■

Roles not granted to the user

SQL Statements: ALTER TRIGGER to COMMIT

13-9

ALTER USER

■
■

■

Roles granted through other roles
Roles managed by an external service (such as the operating system), or by the
Oracle Internet Directory
Roles that are enabled by the SET ROLE statement, such as password-authenticated
roles and secure application roles
See Also:

CREATE ROLE on page 15-59

ENABLE EDITIONS
This clause is not reversible. Specify ENABLE EDITIONS to allow the user to create
multiple versions of editionable objects in this schema using editions. Editionable
objects in non-editions-enabled schemas cannot be editioned.
If the schema to be editions-enabled contains any objects that are not editionable and
that depend on editionable type objects in the schema, then you must specify FORCE to
enable editions for this schema. In this case, all the objects that are not editionable and
that depend on the editionable type objects in the schema being editions-enabled
become invalid.

proxy_clause
The proxy_clause lets you control the ability of an enterprise user (a user outside the
database) or a database proxy (another database user) to connect as the database user
being altered.
GRANT CONNECT THROUGH
Specify GRANT CONNECT THROUGH to allow the connection.
REVOKE CONNECT THROUGH
Specify REVOKE CONNECT THROUGH to prohibit the connection.
ENTERPRISE USER
This clause lets you expose user to proxy use by enterprise users. The administrator
working in Oracle Internet Directory must then grant privileges for appropriate
enterprise users to act on behalf of user.
db_user_proxy
This clause lets you expose user to proxy use by database user db_user_proxy (the
proxy).
■
■

The proxy will have all privileges that were directly granted to user.
The proxy will have all roles associated with user, unless you specify the WITH
clauses of db_user_proxy_clauses to limit the proxy to some or none of the roles
of user. For each role associated with the proxy, if the role is enabled by default for
user at login, then that role will also be enabled by default for the proxy at login.

db_user_proxy_clauses
Specify the WITH clauses to limit the proxy to some or none of the roles associated with
user, and the AUTHENTICATION RQUIRED clause to specify whether authentication is
required.

13-10 Oracle Database SQL Language Reference

ALTER USER

WITH ROLE role_name permits the proxy to connect as the specified user
and to activate only the roles that are specified by role_name. This clause can contain
only roles that are associated with user.

WITH ROLE

WITH ROLE ALL EXCEPT role_name permits the proxy to
connect as the specified user and to activate all roles associated with that user except
those specified for role_name. This clause can contain only roles that are associated
with user.

WITH ROLE ALL EXCEPT

WITH NO ROLES WITH NO ROLES permits the proxy to connect as the specified user,

but prohibits the proxy from activating any of that user's roles after connecting.
Oracle Database does not expect the proxy to
authenticate the user unless you specify the AUTHENTICATION REQUIRED clause. This
clause ensures that authentication credentials for the user must be presented when the
user is authenticated through the specified proxy. The credential is a password.

AUTHENTICATION REQUIRED

AUTHENTICATED USING The AUTHENTICATED USING clauses, which appeared in the
syntax of earlier releases, have been deprecated and are no longer needed. If you
specify the AUTHENTICATED USING PASSWORD clause, then Oracle Database converts it to
the AUTHENTICATION REQUIRED clause. Specifying the AUTHENTICATED USING
CERTIFICATE clause or the AUTHENTICATED USING DISTINGUISHED NAME clause is
equivalent to omitting the AUTHENTICATION REQUIRED clause.
See Also:
■

■

■

Oracle Database Security Guide for more information on proxies and
their use of the database and "Proxy Users: Examples" on
page 13-12
Oracle Security Overview for an overview of database security and
for information on middle-tier systems and proxy authentication
Oracle Database Advanced Application Developer's Guide for
information on proxy authentication of application users

Examples
13

The following statement changes the
password of the user sidney (created in "Creating a Database User: Example" on
page 17-12) second_2nd_pwd and default tablespace to the tablespace example:
Changing User Identification: Example

ALTER USER sidney
IDENTIFIED BY second_2nd_pwd
DEFAULT TABLESPACE example;

The following statement assigns the new_profile profile (created in "Creating a
Profile: Example" on page 15-54) to the sample user sh:
ALTER USER sh
PROFILE new_profile;

In subsequent sessions, sh is restricted by limits in the new_profile profile.
The following statement makes all roles granted directly to sh default roles, except the
dw_manager role:
ALTER USER sh
DEFAULT ROLE ALL EXCEPT dw_manager;

SQL Statements: ALTER TRIGGER to COMMIT 13-11

ALTER USER

At the beginning of sh's next session, Oracle Database enables all roles granted directly
to sh except the dw_manager role.
Changing User Authentication: Examples The following statement changes the
authentication mechanism of user app_user1 (created in "Creating a Database User:
Example" on page 17-12):
ALTER USER app_user1 IDENTIFIED GLOBALLY AS 'CN=tom,O=oracle,C=US';

The following statement causes user sidney's password to expire:
ALTER USER sidney PASSWORD EXPIRE;

If you cause a database user's password to expire with PASSWORD EXPIRE, then the user
(or the DBA) must change the password before attempting to log in to the database
following the expiration. However, tools such as SQL*Plus allow the user to change
the password on the first attempted login following the expiration.
Assigning a Tablespace Group: Example The following statement assigns tbs_grp_

01 (created in "Adding a Temporary Tablespace to a Tablespace Group: Example" on
page 16-95) as the tablespace group for user sh:
ALTER USER sh
TEMPORARY TABLESPACE tbs_grp_01;

Proxy Users: Examples The following statement alters the user app_user1. The
example permits the app_user1 to connect through the proxy user sh. The example
also allows app_user1 to enable its warehouse_user role (created in "Creating a Role:
Example" on page 15-60) when connected through the proxy sh:
ALTER USER app_user1
GRANT CONNECT THROUGH sh
WITH ROLE warehouse_user;

To show basic syntax, this example uses the sample database Sales History user (sh) as
the proxy. Normally a proxy user would be an application server or middle-tier entity.
For information on creating the interface between an application user and a database
by way of an application server, refer to Oracle Call Interface Programmer's Guide.
See Also:
■

■

"Creating External Database Users: Examples" on page 17-13 to
see how to create the app_user user
"Creating a Role: Example" on page 15-60 to see how to create the
dw_user role

The following statement takes away the right of user app_user1 to connect through the
proxy user sh:
ALTER USER app_user1 REVOKE CONNECT THROUGH sh;

The following hypothetical examples shows another method of proxy authentication:
ALTER USER sully GRANT CONNECT THROUGH OAS1
AUTHENTICATED USING PASSWORD;

The following example exposes the user app_user1 to proxy use by enterprise users.
The enterprise users cannot act on behalf of app_user1 until the Oracle Internet
Directory administrator has granted them appropriate privileges:
13-12 Oracle Database SQL Language Reference

ALTER USER

ALTER USER app_user1
GRANT CONNECT THROUGH ENTERPRISE USERS;

SQL Statements: ALTER TRIGGER to COMMIT 13-13

ALTER VIEW

ALTER VIEW
Purpose
13

Use the ALTER VIEW statement to explicitly recompile a view that is invalid or to modify
view constraints. Explicit recompilation lets you locate recompilation errors before run
time. You may want to recompile a view explicitly after altering one of its base tables
to ensure that the alteration does not affect the view or other objects that depend on it.
You can also use ALTER VIEW to define, modify, or drop view constraints.
You cannot use this statement to change the definition of an existing view. Further, if
DDL changes to the view's base tables invalidate the view, then you cannot use this
statement to compile the invalid view. In these cases, you must redefine the view using
CREATE VIEW with the OR REPLACE keywords.
When you issue an ALTER VIEW statement, Oracle Database recompiles the view
regardless of whether it is valid or invalid. The database also invalidates any local
objects that depend on the view.
If you alter a view that is referenced by one or more materialized views, then those
materialized views are invalidated. Invalid materialized views cannot be used by
query rewrite and cannot be refreshed.
See Also:
■

■

■

CREATE VIEW on page 17-14 for information on redefining a
view and ALTER MATERIALIZED VIEW on page 11-3 for
information on revalidating an invalid materialized view
Oracle Database Data Warehousing Guide for general information on
data warehouses
Oracle Database Concepts for more about dependencies among
schema objects

Prerequisites
13

The view must be in your own schema or you must have ALTER ANY TABLE system
privilege.

13-14 Oracle Database SQL Language Reference

ALTER VIEW

Syntax
13

alter_view::=
ADD

out_of_line_constraint
RELY

MODIFY

CONSTRAINT

constraint
NORELY

CONSTRAINT
schema
ALTER

.

VIEW

view

PRIMARY

constraint
KEY

;

DROP
,
UNIQUE

(

column

)

COMPILE
ONLY
READ
WRITE

(out_of_line_constraint::= on page 8-5—part of constraint::= syntax)

Semantics
13

schema
Specify the schema containing the view. If you omit schema, then Oracle Database
assumes the view is in your own schema.

view
Specify the name of the view to be recompiled.

ADD Clause
Use the ADD clause to add a constraint to view. Refer to constraint on page 8-4 for
information on view constraints and their restrictions.

MODIFY CONSTRAINT Clause
Use the MODIFY CONSTRAINT clause to change the RELY or NORELY setting of an existing
view constraint. Refer to "RELY Clause" on page 8-17 for information on the uses of
these settings and to "Notes on View Constraints" on page 8-19 for general information
on view constraints.
Restriction on Modifying Constraints You cannot change the setting of a unique or

primary key constraint if it is part of a referential integrity constraint without
dropping the foreign key or changing its setting to match that of view.

DROP Clause
Use the DROP clause to drop an existing view constraint.
Restriction on Dropping Constraints You cannot drop a unique or primary key
constraint if it is part of a referential integrity constraint on a view.

SQL Statements: ALTER TRIGGER to COMMIT 13-15

ALTER VIEW

COMPILE
The COMPILE keyword directs Oracle Database to recompile the view.

{ READ ONLY | READ WRITE }
These clauses are valid only for editioning views.
■

Specify READ ONLY to indicate that the editioning view cannot be updated.

■

Specify READ WRITE to return a read-only editioning view to read/write status.

When you specify these clauses, the database does not invalidate dependant objects,
but it may invalidate cursors.
See Also: CREATE VIEW on page 17-14 for information about
editioning views

Examples
13

To recompile the view customer_ro (created in "Creating a
Read-Only View: Example" on page 17-24), issue the following statement:

Altering a View: Example
ALTER VIEW customer_ro
COMPILE;

If Oracle Database encounters no compilation errors while recompiling customer_ro,
then customer_ro becomes valid. If recompiling results in compilation errors, then the
database returns an error and customer_ro remains invalid.
Oracle Database also invalidates all dependent objects. These objects include any
procedures, functions, package bodies, and views that reference customer_ro. If you
subsequently reference one of these objects without first explicitly recompiling it, then
the database recompiles it implicitly at run time.

13-16 Oracle Database SQL Language Reference

ANALYZE

ANALYZE
Purpose
13

Use the ANALYZE statement to collect statistics, for example, to:
■

■

■

Collect or delete statistics about an index or index partition, table or table
partition, index-organized table, cluster, or scalar object attribute.
Validate the structure of an index or index partition, table or table partition,
index-organized table, cluster, or object reference (REF).
Identify migrated and chained rows of a table or cluster.
For the collection of most statistics, use the DBMS_STATS
package, which lets you collect statistics in parallel, collect global
statistics for partitioned objects, and fine tune your statistics
collection in other ways. See Oracle Database PL/SQL Packages and
Types Reference for more information on the DBMS_STATS package.

Note:

Use the ANALYZE statement (rather than DBMS_STATS) for statistics
collection not related to the cost-based optimizer:
■

To use the VALIDATE or LIST CHAINED ROWS clauses

■

To collect information on freelist blocks

Prerequisites
13

The schema object to be analyzed must be local, and it must be in your own schema or
you must have the ANALYZE ANY system privilege.
If you want to list chained rows of a table or cluster into a list table, then the list table
must be in your own schema, or you must have INSERT privilege on the list table, or
you must have INSERT ANY TABLE system privilege.
If you want to validate a partitioned table, then you must have the INSERT object
privilege on the table into which you list analyzed rowids, or you must have the
INSERT ANY TABLE system privilege.

SQL Statements: ALTER TRIGGER to COMMIT 13-17

ANALYZE

Syntax
13

analyze::=
schema

.

TABLE

table

ANALYZE

schema

partition_extension_clause

.

INDEX

index

schema

.

CLUSTER

cluster

validation_clauses
into_clause
LIST

CHAINED

ROWS

;

SYSTEM
DELETE

STATISTICS

partition_extension_clause::=
(

partition

)

PARTITION

,
FOR

(

(

partition_key_value

subpartition

)

)

SUBPARTITION

,
FOR

(

subpartition_key_value

)

validation_clauses::=
SET
VALIDATE

REF

DANGLING

TO

NULL

UPDATE
FAST
CASCADE

OFFLINE

into_clause

COMPLETE
ONLINE
VALIDATE

STRUCTURE

into_clause::=
schema

.

INTO

table

Semantics
13

schema
Specify the schema containing the table, index, or cluster. If you omit schema, then
Oracle Database assumes the table, index, or cluster is in your own schema.

13-18 Oracle Database SQL Language Reference

ANALYZE

TABLE table
Specify a table to be analyzed. When you analyze a table, the database collects
statistics about expressions occurring in any function-based indexes as well. Therefore,
be sure to create function-based indexes on the table before analyzing the table. Refer
to CREATE INDEX on page 14-60 for more information about function-based indexes.
When analyzing a table, the database skips all domain indexes marked LOADING or
FAILED.
For an index-organized table, the database also analyzes any mapping table and
calculates its PCT_ACCESSS_DIRECT statistics. These statistics estimate the accuracy of
guess data block addresses stored as part of the local rowids in the mapping table.
Oracle Database collects the following statistics for a table. Statistics marked with an
asterisk are always computed exactly. Table statistics, including the status of domain
indexes, appear in the data dictionary views USER_TABLES, ALL_TABLES, and DBA_
TABLES in the columns shown in parentheses.
■
■

■

Number of rows (NUM_ROWS)
* Number of data blocks below the high water mark—the number of data blocks
that have been formatted to receive data, regardless whether they currently
contain data or are empty (BLOCKS)
* Number of data blocks allocated to the table that have never been used (EMPTY_
BLOCKS)

■

Average available free space in each data block in bytes (AVG_SPACE)

■

Number of chained rows (CHAIN_COUNT)

■

Average row length, including the row overhead, in bytes (AVG_ROW_LEN)

Restrictions on Analyzing Tables Analyzing tables is subject to the following

restrictions:
■
■

■

■

You cannot use ANALYZE to collect statistics on data dictionary tables.
You cannot use ANALYZE to collect statistics on an external table. Instead, you must
use the DBMS_STATS package.
You cannot use ANALYZE to collect default statistics on a temporary table. However,
if you have already created an association between one or more columns of a
temporary table and a user-defined statistics type, then you can use ANALYZE to
collect the user-defined statistics on the temporary table.
You cannot compute or estimate statistics for the following column types: REF
column types, varrays, nested tables, LOB column types (LOB column types are
not analyzed, they are skipped), LONG column types, or object types. However, if a
statistics type is associated with such a column, then Oracle Database collects
user-defined statistics.
See Also:
■
■

ASSOCIATE STATISTICS on page 13-25
Oracle Database Reference for information on the data dictionary
views

partition_extension_clause
Specify the partition or subpartition, or the partition or subpartition value, on which
you want statistics to be gathered. You cannot use this clause when analyzing clusters.

SQL Statements: ALTER TRIGGER to COMMIT 13-19

ANALYZE

If you specify PARTITION and table is composite-partitioned, then Oracle Database
analyzes all the subpartitions within the specified partition.

INDEX index
Specify an index to be analyzed.
Oracle Database collects the following statistics for an index. Statistics marked with an
asterisk are always computed exactly. For conventional indexes, when you compute or
estimate statistics, the statistics appear in the data dictionary views USER_INDEXES,
ALL_INDEXES, and DBA_INDEXES in the columns shown in parentheses.
■

* Depth of the index from its root block to its leaf blocks (BLEVEL)

■

Number of leaf blocks (LEAF_BLOCKS)

■

Number of distinct index values (DISTINCT_KEYS)

■

Average number of leaf blocks for each index value (AVG_LEAF_BLOCKS_PER_KEY)

■

■

Average number of data blocks for each index value (for an index on a table) (AVG_
DATA_BLOCKS_PER_KEY)
Clustering factor (how well ordered the rows are about the indexed values)
(CLUSTERING_FACTOR)

For domain indexes, this statement invokes the user-defined statistics collection
function specified in the statistics type associated with the index (see ASSOCIATE
STATISTICS on page 13-25). If no statistics type is associated with the domain index,
then the statistics type associated with its indextype is used. If no statistics type exists
for either the index or its indextype, then no user-defined statistics are collected.
User-defined index statistics appear in the STATISTICS column of the data dictionary
views USER_USTATS, ALL_USTATS, and DBA_USTATS.
When you analyze an index from which a substantial number
of rows has been deleted, Oracle Database sometimes executes a
COMPUTE statistics operation (which can entail a full table scan) even if
you request an ESTIMATE statistics operation. Such an operation can be
quite time consuming.
Note:

Restriction on Analyzing Indexes You cannot analyze a domain index that is marked

IN_PROGRESS or FAILED.
See Also:
■

■

■

CREATE INDEX on page 14-60 for more information on domain
indexes
Oracle Database Reference for information on the data dictionary
views
"Analyzing an Index: Example" on page 13-23

CLUSTER cluster
Specify a cluster to be analyzed. When you collect statistics for a cluster, Oracle
Database also automatically collects the statistics for all the tables in the cluster and all
their indexes, including the cluster index.

13-20 Oracle Database SQL Language Reference

ANALYZE

For both indexed and hash clusters, the database collects the average number of data
blocks taken up by a single cluster key (AVG_BLOCKS_PER_KEY). These statistics appear
in the data dictionary views ALL_CLUSTERS, USER_CLUSTERS, and DBA_CLUSTERS.
Oracle Database Reference for information on the data
dictionary views and "Analyzing a Cluster: Example" on page 13-24

See Also:

validation_clauses
The validation clauses let you validate REF values and the structure of the analyzed
object.
See Also: Oracle Database Administrator's Guide for more information
about validating tables, indexes, clusters, and materialized views

VALIDATE REF UPDATE Clause
Specify VALIDATE REF UPDATE to validate the REF values in the specified table, check the
rowid portion in each REF, compare it with the true rowid, and correct it, if necessary.
You can use this clause only when analyzing a table.
If the owner of the table does not have SELECT object privilege on the referenced
objects, then Oracle Database will consider them invalid and set them to null.
Subsequently these REF values will not be available in a query, even if it is issued by a
user with appropriate privileges on the objects.
SET DANGLING TO NULL SET DANGLING TO NULL sets to null any REF values (whether
or not scoped) in the specified table that are found to point to an invalid or nonexistent
object.

VALIDATE STRUCTURE
Specify VALIDATE STRUCTURE to validate the structure of the analyzed object. The
statistics collected by this clause are not used by the Oracle Database optimizer.
See Also:
■

■

■

■

■

"Validating a Table: Example" on page 13-23

For a table, Oracle Database verifies the integrity of each of the data blocks and
rows. For an index-organized table, the database also generates compression
statistics (optimal prefix compression count) for the primary key index on the
table.
For a cluster, Oracle Database automatically validates the structure of the cluster
tables.
For a partitioned table, Oracle Database also verifies that each row belongs to the
correct partition. If a row does not collate correctly, then its rowid is inserted into
the INVALID_ROWS table.
For a temporary table, Oracle Database validates the structure of the table and its
indexes during the current session.
For an index, Oracle Database verifies the integrity of each data block in the index
and checks for block corruption. This clause does not confirm that each row in the
table has an index entry or that each index entry points to a row in the table. You
can perform these operations by validating the structure of the table with the
CASCADE clause.
Oracle Database also computes compression statistics (optimal prefix compression
count) for all normal indexes.

SQL Statements: ALTER TRIGGER to COMMIT 13-21

ANALYZE

Oracle Database stores statistics about the index in the data dictionary views
INDEX_STATS and INDEX_HISTOGRAM.
See Also:

Oracle Database Reference for information on these views

If Oracle Database encounters corruption in the structure of the object, then an error
message is returned. In this case, drop and re-create the object.
Specify CASCADE if you want Oracle Database to validate the structure of
the indexes associated with the table or cluster. If you use this clause when validating
a table, then the database also validates the indexes defined on the table. If you use
this clause when validating a cluster, then the database also validates all the cluster
tables indexes, including the cluster index.

CASCADE

By default, CASCADE performs a COMPLETE validation, which can be resource intensive.
Specify FAST if you want the database to check for the existence of corruptions without
reporting details about the corruption. If the FAST check finds a corruption, you can
then use the CASCADE option without the FAST clause to locate and learn details about
it.
If you use this clause to validate an enabled (but previously disabled) function-based
index, then validation errors may result. In this case, you must rebuild the index.
ONLINE | OFFLINE Specify ONLINE to enable Oracle Database to run the validation
while DML operations are ongoing within the object. The database reduces the
amount of validation performed to allow for concurrency.
Note: When you validate the structure of an object ONLINE, Oracle
Database does not collect any statistics, as it does when you validate
the structure of the object OFFLINE.

Specify OFFLINE, to maximize the amount of validation performed. This setting
prevents INSERT, UPDATE, and DELETE statements from concurrently accessing the
object during validation but allows queries. This is the default.
Restriction on ONLINE

You cannot specify ONLINE when analyzing a cluster.

The INTO clause of VALIDATE STRUCTURE is valid only for partitioned tables.
Specify a table into which Oracle Database lists the rowids of the partitions whose
rows do not collate correctly. If you omit schema, then the database assumes the list is
in your own schema. If you omit this clause altogether, then the database assumes that
the table is named INVALID_ROWS. The SQL script used to create this table is
UTLVALID.SQL.
INTO

LIST CHAINED ROWS
LIST CHAINED ROWS lets you identify migrated and chained rows of the analyzed table
or cluster. You cannot use this clause when analyzing an index.
In the INTO clause, specify a table into which Oracle Database lists the migrated and
chained rows. If you omit schema, then the database assumes the chained-rows table is
in your own schema. If you omit this clause altogether, then the database assumes that
the table is named CHAINED_ROWS. The chained-rows table must be on your local
database.
You can create the CHAINED_ROWS table using one of these scripts:

13-22 Oracle Database SQL Language Reference

ANALYZE

■

■

UTLCHAIN.SQL uses physical rowids. Therefore it can accommodate rows from
conventional tables but not from index-organized tables. (See the Note that
follows.)
UTLCHN1.SQL uses universal rowids, so it can accommodate rows from both
conventional and index-organized tables.

If you create your own chained-rows table, then it must follow the format prescribed
by one of these two scripts.
If you are analyzing index-organized tables based on primary keys (rather than
universal rowids), then you must create a separate chained-rows table for each
index-organized table to accommodate its primary-key storage. Use the SQL scripts
DBMSIOTC.SQL and PRVTIOTC.PLB to define the BUILD_CHAIN_ROWS_TABLE procedure,
and then execute this procedure to create an IOT_CHAINED_ROWS table for each such
index-organized table.
See Also:
■

■

■

Oracle Database Performance Tuning Guide information about these
scripts and about eliminating chained rows
The DBMS_IOT package in Oracle Database PL/SQL Packages and
Types Reference for information on the packaged SQL scripts
"Listing Chained Rows: Example" on page 13-24

DELETE STATISTICS
Specify DELETE STATISTICS to delete any statistics about the analyzed object that are
currently stored in the data dictionary. Use this statement when you no longer want
Oracle Database to use the statistics.
When you use this clause on a table, the database also automatically removes statistics
for all the indexes defined on the table. When you use this clause on a cluster, the
database also automatically removes statistics for all the cluster tables and all their
indexes, including the cluster index.
Specify SYSTEM if you want Oracle Database to delete only system (not user-defined)
statistics. If you omit SYSTEM, and if user-defined column or index statistics were
collected for an object, then the database also removes the user-defined statistics by
invoking the statistics deletion function specified in the statistics type that was used to
collect the statistics.
See Also:

"Deleting Statistics: Example" on page 13-23

Examples
13

Deleting Statistics: Example The following statement deletes statistics about the
sample table oe.orders and all its indexes from the data dictionary:
ANALYZE TABLE orders DELETE STATISTICS;

The following statement validates the structure of the
sample index oe.inv_product_ix:

Analyzing an Index: Example

ANALYZE INDEX inv_product_ix VALIDATE STRUCTURE;

The following statement analyzes the sample table
hr.employees and all of its indexes:
Validating a Table: Example

ANALYZE TABLE employees VALIDATE STRUCTURE CASCADE;
SQL Statements: ALTER TRIGGER to COMMIT 13-23

ANALYZE

For a table, the VALIDATE REF UPDATE clause verifies the REF values in the specified
table, checks the rowid portion of each REF, and then compares it with the true rowid.
If the result is an incorrect rowid, then the REF is updated so that the rowid portion is
correct.
The following statement validates the REF values in the sample table oe.customers:
ANALYZE TABLE customers VALIDATE REF UPDATE;

The following statement validates the structure of the sample table oe.customers
while allowing simultaneous DML:
ANALYZE TABLE customers VALIDATE STRUCTURE ONLINE;

Analyzing a Cluster: Example The following statement analyzes the personnel

cluster (created in "Creating a Cluster: Example" on page 14-7), all of its tables, and all
of their indexes, including the cluster index:
ANALYZE CLUSTER personnel
VALIDATE STRUCTURE CASCADE;

Listing Chained Rows: Example The following statement collects information about
all the chained rows in the table orders:
ANALYZE TABLE orders
LIST CHAINED ROWS INTO chained_rows;

The preceding statement places the information into the table chained_rows. You can
then examine the rows with this query (no rows will be returned if the table contains
no chained rows):
SELECT owner_name, table_name, head_rowid, analyze_timestamp
FROM chained_rows
ORDER BY owner_name, table_name, head_rowid, analyze_timestamp;
OWNER_NAME
---------OE

TABLE_NAME
---------ORDERS

HEAD_ROWID
ANALYZE_TIMESTAMP
------------------ ----------------AAAAZzAABAAABrXAAA 25-SEP-2000

13-24 Oracle Database SQL Language Reference

ASSOCIATE STATISTICS

ASSOCIATE STATISTICS
Purpose
13

Use the ASSOCIATE STATISTICS statement to associate a statistics type (or default
statistics) containing functions relevant to statistics collection, selectivity, or cost with
one or more columns, standalone functions, packages, types, domain indexes, or
indextypes.
For a listing of all current statistics type associations, query the USER_ASSOCIATIONS
data dictionary view. If you analyze the object with which you are associating
statistics, then you can also query the associations in the USER_USTATS view.
ANALYZE on page 13-17 for information on the order of
precedence with which ANALYZE uses associations

See Also:

Prerequisites
13

To issue this statement, you must have the appropriate privileges to alter the base
object (table, function, package, type, domain index, or indextype). In addition, unless
you are associating only default statistics, you must have execute privilege on the
statistics type. The statistics type must already have been defined.
See Also: CREATE TYPE on page 17-3 for information on defining
types

Syntax
13

associate_statistics::=
column_association
ASSOCIATE

STATISTICS

storage_table_clause

WITH

;
function_association

column_association::=
,
schema
COLUMNS

.
table

.

column

using_statistics_type

SQL Statements: ALTER TRIGGER to COMMIT 13-25

ASSOCIATE STATISTICS

function_association::=
,
schema

.

FUNCTIONS

function
,
schema

.

PACKAGES

package
,
using_statistics_type
schema

.

TYPES

,

type

default_selectivity_clause

default_cost_clause
,
,
schema

.

default_cost_clause

default_selectivity_clause

INDEXES

index
,
schema

.

INDEXTYPES

indextype

using_statistics_type::=
schema

.
statistics_type

USING
NULL

default_cost_clause::=
DEFAULT

COST

(

cpu_cost

,

io_cost

,

network_cost

)

default_selectivity_clause::=
DEFAULT

SELECTIVITY

default_selectivity

storage_table_clause::=
SYSTEM
WITH

MANAGED

STORAGE

TABLES

USER

Semantics
13

column_association
Specify one or more table columns. If you do not specify schema, then Oracle Database
assumes the table is in your own schema.

function_association
Specify one or more standalone functions, packages, user-defined data types, domain
indexes, or indextypes. If you do not specify schema, then Oracle Database assumes
the object is in your own schema.
13-26 Oracle Database SQL Language Reference

ASSOCIATE STATISTICS

■

■

FUNCTIONS refers only to standalone functions, not to method types or to built-in
functions.
TYPES refers only to user-defined types, not to built-in SQL data types.

You cannot specify an object for which you
have already defined an association. You must first disassociate the statistics from this
object.

Restriction on function_association

DISASSOCIATE STATISTICS on page 17-34 "Associating
Statistics: Example" on page 13-28

See Also:

using_statistics_type
Specify the statistics type (or a synonym for the type) being associated with column,
function, package, type, domain index, or indextype. The statistics_type must
already have been created.
The NULL keyword is valid only when you are associating statistics with a column or
an index. When you associate a statistics type with an object type, columns of that
object type inherit the statistics type. Likewise, when you associate a statistics type
with an indextype, index instances of the indextype inherit the statistics type.You can
override this inheritance by associating a different statistics type for the column or
index. Alternatively, if you do not want to associate any statistics type for the column
or index, then you can specify NULL in the using_statistics_type clause.
Restriction on Specifying Statistics Type

You cannot specify NULL for functions,

packages, types, or indextypes.
See Also: Oracle Database Data Cartridge Developer's Guide for
information on creating statistics collection functions

default_cost_clause
Specify default costs for standalone functions, packages, types, domain indexes, or
indextypes. If you specify this clause, then you must include one number each for CPU
cost, I/O cost, and network cost, in that order. Each cost is for a single execution of the
function or method or for a single domain index access. Accepted values are integers
of zero or greater.

default_selectivity_clause
Specify as a percent the default selectivity for predicates with standalone functions,
types, packages, or user-defined operators. The default_selectivity_clause must be
a number between 0 and 100. Values outside this range are ignored.
You cannot specify DEFAULT
SELECTIVITY for domain indexes or indextypes.
Restriction on the default_selectivity_clause

See Also:

"Specifying Default Cost: Example" on page 13-28

storage_table_clause
This clause is relevant only for statistics on INDEXTYPE.
■

Specify WITH SYSTEM MANAGED STORAGE TABLES to indicate that the storage of
statistics data is to be managed by the system. The type you specify in
statistics_type should be storing the statistics related information in tables that
are maintained by the system. Also, the indextype you specify must already have
been created or altered to support the WITH SYSTEM MANAGED STORAGE TABLES clause.
SQL Statements: ALTER TRIGGER to COMMIT 13-27

ASSOCIATE STATISTICS

■

Specify WITH USER MANAGED STORAGE TABLES to indicate that the tables that store the
user-defined statistics will be managed by the user. This is the default behavior.

Examples
13

This statement creates an association for the
standalone package emp_mgmt. See Oracle Database PL/SQL Language Reference for the
example that creates this package.

Associating Statistics: Example

ASSOCIATE STATISTICS WITH PACKAGES emp_mgmt DEFAULT SELECTIVITY 10;

Specifying Default Cost: Example This statement specifies that using the domain

index salary_index, created in "Using Extensible Indexing" on page F-1, to implement
a given predicate always has a CPU cost of 100, I/O cost of 5, and network cost of 0.
ASSOCIATE STATISTICS WITH INDEXES salary_index DEFAULT COST (100,5,0);

The optimizer will use these default costs instead of calling a cost function.

13-28 Oracle Database SQL Language Reference

AUDIT

AUDIT
Purpose
13

Use the AUDIT statement to:
■

■

Track the issuance of SQL statements in subsequent user sessions. You can track
the issuance of a specific SQL statement or of all SQL statements authorized by a
particular system privilege. Auditing operations on SQL statements apply only to
subsequent sessions, not to current sessions.
Track operations on a specific schema object. Auditing operations on schema
objects apply to current sessions as well as to subsequent sessions.
See Also:
■

■

Oracle Database PL/SQL Packages and Types Reference for
information on the DBMS_FGA package, which lets you create and
administer value-based auditing policies
NOAUDIT on page 18-79 for information on disabling auditing

Prerequisites
13

To audit issuances of a SQL statement, you must have the AUDIT SYSTEM system
privilege. However, the AUDIT SYSTEM system privilege is not required when you use
the IN SESSION CURRENT clause.
To collect auditing results, you must enable auditing by setting the initialization
parameter AUDIT_TRAIL to a value other than the default setting of NONE. You can
specify auditing options regardless of whether auditing is enabled. However, Oracle
Database does not generate audit records until you enable auditing.
To audit operations on a schema object, the object you choose for auditing must be in
your own schema or you must have AUDIT ANY system privilege. In addition, if the
object you choose for auditing is a directory object, even if you created it, then you
must have AUDIT ANY system privilege.
The AUDIT ANY system privileges allows the grantee to audit
any object in any schema except the SYS schema. You can allow such a
grantee to audit objects in the SYS schema by setting the O7_
DICTIONARY_ACCESSIBILITY initialization parameter to TRUE. For
security reasons, Oracle recommends that you use this setting only
with great caution.
Note:

Oracle Database Reference for information on the AUDIT_
TRAIL parameter

See Also:

SQL Statements: ALTER TRIGGER to COMMIT 13-29

AUDIT

Syntax
13

audit::=
auditing_by_clause
IN

SESSION

CURRENT

audit_operation_clause
AUDIT

audit_schema_object_clause
NETWORK

SESSION

NOT

BY
ACCESS

WHENEVER

SUCCESSFUL
;

audit_operation_clause::=
,
sql_statement_shortcut
ALL
ALL

STATEMENTS
,

system_privilege
ALL

PRIVILEGES

auditing_by_clause::=
,
BY

user

audit_schema_object_clause::=
,
sql_operation
auditing_on_clause
ALL

auditing_on_clause::=
schema

.
object

DIRECTORY

directory_name

ON

schema
MINING

.

MODEL

DEFAULT

13-30 Oracle Database SQL Language Reference

model

AUDIT

Semantics
13

audit_operation_clause
Use the audit_operation_clause to audit specified operations, regardless of the
schema objects affected by the operations.
sql_statement_shortcut
Specify a shortcut to audit the use of specific SQL statements. Table 13–1 on page 13-34
and Table 13–2 on page 13-37 list the shortcuts and the SQL statements they audit.
Do not confuse SQL statement shortcuts with system
privileges. For example:

Note:
■

■

An AUDIT USER statement specifies the USER shortcut for auditing of
all CREATE USER, ALTER USER, and DROP USER SQL statements.
Auditing in this case includes an operation in which a user
changes his or her own password with an ALTER USER statement.
An AUDIT ALTER USER statement specifies the ALTER USER system
privilege for auditing of all operations that make use of that system
privilege. Auditing in this case does not include an operation in
which a user changes his or her own password, because that
operation does not require the ALTER USER system privilege.

For each audited operation, Oracle Database produces an audit record containing this
information:
■

The user performing the operation

■

The type of operation

■

The object involved in the operation

■

The date and time of the operation

Oracle Database writes audit records to the audit trail, which is a database table
containing audit records. You can review database activity by examining the audit trail
through data dictionary views.
See Also:
■

■

■

Oracle Database Security Guide for a listing of the audit trail data
dictionary views
Oracle Database Reference for detailed descriptions of the data
dictionary views
"Auditing SQL Statements Relating to Roles: Example" on
page 13-39

system_privilege
Specify a system privilege to audit SQL statements and other operations that are
authorized by the specified system privilege.

SQL Statements: ALTER TRIGGER to COMMIT 13-31

AUDIT

Auditing the use of a system privilege containing the ANY
keyword is more restrictive than auditing the use of the same
privilege without the ANY keyword. For example:

Note:

■

■

AUDIT CREATE PROCEDURE audits the statements issued using either
the CREATE PROCEDURE or CREATE ANY PROCEDURE privilege.
AUDIT CREATE ANY PROCEDURE audits only those statements issued
using the CREATE ANY PROCEDURE privilege.

Rather than specifying many individual system privileges, you can specify the roles
CONNECT, RESOURCE, and DBA. Doing so is equivalent to auditing all of the system
privileges granted to those roles.
Oracle Database also provides three shortcuts for specifying groups of system
privileges and statement options at once:
ALL Specify ALL to audit all statements options shown in Table 13–1 but not the
additional statement options shown in Table 13–2.
ALL STATEMENTS Specify ALL STATEMENTS to audit all top-level SQL statements
executed. Top-level SQL statements are issued directly by a user. SQL statements run
from within a PL/SQL procedure or function are not considered top-level statements.
Therefore, this clause does not audit the statements executed within PL/SQL
procedures or functions. However, the execution of the PL/SQL procedure or function
itself is audited. This clause is useful if you want to audit all the statements in a
specific environment, regardless of other auditing configurations that are system wide
or user specific.
ALL PRIVILEGES

Specify ALL PRIVILEGES to audit system privileges.

Oracle recommends that you specify individual system
privileges and statement options for auditing rather than roles or
shortcuts. The specific system privileges and statement options
encompassed by roles and shortcuts change from one release to the
next and may not be supported in future versions of Oracle Database.

Note:

See Also:
■

■

■

Table 18–1, " System Privileges (Organized by the Database Object
Operated Upon)" on page 18-40 for a list of all system privileges
and the SQL statements that they authorize
Oracle Database Security Guide for more information on the
CONNECT, RESOURCE, and DBA roles
"Auditing Query and Update SQL Statements: Example" on
page 13-39, "Auditing Deletions: Example" on page 13-39, and
"Auditing Statements Relating to Directories: Examples" on
page 13-39

auditing_by_clause
Specify the auditing_by_clause to restrict auditing to only SQL statements issued by
the specified users. If you omit this clause, then Oracle Database audits all users'
statements.
13-32 Oracle Database SQL Language Reference

AUDIT

IN SESSION CURRENT
Use this clause to limit auditing to the current session.

audit_schema_object_clause
Use the audit_schema_object_clause to audit operations on specific schema objects.
sql_operation
Specify the SQL operation to be audited. Table 13–3 on page 13-38 shows the types of
objects that can be audited, and for each object the SQL statements that can be audited.
For example, if you choose to audit a table with the ALTER operation, then Oracle
Database audits all ALTER TABLE statements issued against the table. If you choose to
audit a sequence with the SELECT operation, then the database audits all statements
that use any values of the sequence.
ALL
Specify ALL as a shortcut equivalent to specifying all SQL operations applicable for the
type of object.
auditing_on_clause
The auditing_on_clause lets you specify the particular schema object to be audited.
See Also: "Auditing Queries on a Table: Example" on page 13-40,
"Auditing Inserts and Updates on a Table: Example" on page 13-40,
and "Auditing Operations on a Sequence: Example" on page 13-40

Specify the schema containing the object chosen for auditing. If you omit
schema, then Oracle Database assumes the object is in your own schema.
schema

object Specify the name of the object to be audited. The object must be a table, view,
sequence, stored procedure, function, package, materialized view, mining model, or
library.

You can also specify a synonym for a table, view, sequence, procedure, stored function,
package, materialized view, or user-defined type.
ON DEFAULT Specify ON DEFAULT to establish the specified object options as default
object options for subsequently created objects. After you have established these
default auditing options, any subsequently created object is automatically audited
with those options. The default auditing options for a view are always the union of the
auditing options for the base tables of the view. You can see the current default
auditing options by querying the ALL_DEF_AUDIT_OPTS data dictionary view.

When you change the default auditing options, the auditing options for previously
created objects remain the same. You can change the auditing options for an existing
object only by specifying the object in the ON clause of the AUDIT statement.
See Also: "Setting Default Auditing Options: Example" on
page 13-40
ON DIRECTORY The ON DIRECTORY clause lets you specify the name of a directory
chosen for auditing.
ON MINING MODEL The ON MINING MODEL clause lets you specify the name of a
mining model to be audited.

SQL Statements: ALTER TRIGGER to COMMIT 13-33

AUDIT

NETWORK
Use this clause to detect internal failures in the network layer.
See Also: Oracle Database Security Guide for information on network
auditing

BY SESSION
In earlier releases, BY SESSION caused the database to write a single record for all SQL
statements or operations of the same type executed on the same schema objects in the
same session. Beginning with this release of Oracle Database, both BY SESSION and BY
ACCESS cause Oracle Database to write one audit record for each audited statement
and operation. BY SESSION continues to populate different values to the audit trail
compared with BY ACCESS. Oracle recommends that you include the BY ACCESS clause
for all AUDIT statements, which results in a more detailed audit record. If you specify
neither clause, then BY SESSION is the default.
Note: This change applies only to schema object audit options,
statement options and system privileges that audit SQL statements
other than data definition language (DDL) statements. The database
has always audited BY ACCESS all SQL statements and system
privileges that audit a DDL statement.

BY ACCESS
Specify BY ACCESS if you want Oracle Database to write one record for each audited
statement and operation.
If you specify either a SQL statement shortcut or a system
privilege that audits a data definition language (DDL) statement, then
the database always audits by access. In all other cases, the database
honors the BY SESSION or BY ACCESS specification.

Note:

For statement options and system privileges that audit SQL statements other than
DDL, you can specify either BY SESSION or BY ACCESS. BY SESSION is the default.

WHENEVER [NOT] SUCCESSFUL
Specify WHENEVER SUCCESSFUL to audit only SQL statements and operations that
succeed.
Specify WHENEVER NOT SUCCESSFUL to audit only statements and operations that fail or
result in errors.
If you omit this clause, then Oracle Database performs the audit regardless of success
or failure.

Tables of Auditing Options
13

Table 13–1

SQL Statement Shortcuts for Auditing

SQL Statement Shortcut

SQL Statements and Operations Audited

ALTER SYSTEM

ALTER SYSTEM

13-34 Oracle Database SQL Language Reference

AUDIT

Table 13–1 (Cont.) SQL Statement Shortcuts for Auditing
SQL Statement Shortcut

SQL Statements and Operations Audited

CLUSTER

CREATE CLUSTER
ALTER CLUSTER
DROP CLUSTER
TRUNCATE CLUSTER

CONTEXT

CREATE CONTEXT
DROP CONTEXT

DATABASE LINK

CREATE DATABASE LINK
ALTER DATABASE LINK
DROP DATABASE LINK

DIMENSION

CREATE DIMENSION
ALTER DIMENSION
DROP DIMENSION

DIRECTORY

CREATE DIRECTORY
DROP DIRECTORY

INDEX

CREATE INDEX
ALTER INDEX
ANALYZE INDEX
DROP INDEX

MATERIALIZED VIEW

CREATE MATERIALIZED VIEW
ALTER MATERIALIZED VIEW
DROP MATERIALIZED VIEW

NOT EXISTS

All SQL statements that fail because a specified object does
not exist.

OUTLINE

CREATE OUTLINE
ALTER OUTLINE
DROP OUTLINE

PROCEDURE (See note at end
of table)

CREATE FUNCTION
CREATE LIBRARY
CREATE PACKAGE
CREATE PACKAGE BODY
CREATE PROCEDURE
DROP FUNCTION
DROP LIBRARY
DROP PACKAGE
DROP PROCEDURE

PROFILE

CREATE PROFILE
ALTER PROFILE
DROP PROFILE

PUBLIC DATABASE LINK

CREATE PUBLIC DATABASE LINK
ALTER PUBLIC DATABASE LINK
DROP PUBLIC DATABASE LINK

SQL Statements: ALTER TRIGGER to COMMIT 13-35

AUDIT

Table 13–1 (Cont.) SQL Statement Shortcuts for Auditing
SQL Statement Shortcut

SQL Statements and Operations Audited

PUBLIC SYNONYM

CREATE PUBLIC SYNONYM
DROP PUBLIC SYNONYM

ROLE

CREATE ROLE
ALTER ROLE
DROP ROLE
SET ROLE

ROLLBACK SEGMENT

CREATE ROLLBACK SEGMENT
ALTER ROLLBACK SEGMENT
DROP ROLLBACK SEGMENT

SEQUENCE

CREATE SEQUENCE
DROP SEQUENCE

SESSION

Logons

SYNONYM

CREATE SYNONYM
DROP SYNONYM

SYSTEM AUDIT

AUDIT sql_statements
NOAUDIT sql_statements

SYSTEM GRANT

GRANT system_privileges_and_roles
REVOKE system_privileges_and_roles

TABLE

CREATE TABLE
DROP TABLE
TRUNCATE TABLE

TABLESPACE

CREATE TABLESPACE
ALTER TABLESPACE
DROP TABLESPACE

TRIGGER

CREATE TRIGGER
ALTER TRIGGER
with ENABLE and DISABLE clauses
DROP TRIGGER
ALTER TABLE
with ENABLE ALL TRIGGERS clause
and DISABLE ALL TRIGGERS clause

TYPE

CREATE TYPE
CREATE TYPE BODY
ALTER TYPE
DROP TYPE
DROP TYPE BODY

13-36 Oracle Database SQL Language Reference

AUDIT

Table 13–1 (Cont.) SQL Statement Shortcuts for Auditing
SQL Statement Shortcut

SQL Statements and Operations Audited

USER

CREATE USER
ALTER USER
DROP USER
Notes:
■

■

VIEW

AUDIT USER audits these three SQL statements. Use
AUDIT ALTER USER to audit statements that require the
ALTER USER system privilege.
An AUDIT ALTER USER statement does not audit a user
changing his or her own password, as this activity
does not require the ALTER USER system privilege.

CREATE VIEW
DROP VIEW

Java schema objects (sources, classes, and resources) are
considered the same as procedures for purposes of auditing SQL
statements.

Note:

Table 13–2

Additional SQL Statement Shortcuts for Auditing

SQL Statement Shortcut

SQL Statements and Operations Audited

ALTER SEQUENCE

ALTER SEQUENCE

ALTER TABLE

ALTER TABLE

COMMENT TABLE

COMMENT ON TABLE table, view, materialized view
COMMENT ON COLUMN table.column, view.column, materialized
view.column

DELETE TABLE

DELETE FROM table, view

EXECUTE PROCEDURE

CALL
Execution of any procedure or function or access to any
variable, library, or cursor inside a package.

GRANT DIRECTORY

GRANT privilege ON directory
REVOKE privilege ON directory

GRANT PROCEDURE

GRANT privilege ON procedure, function, package
REVOKE privilege ON procedure, function, package

GRANT SEQUENCE

GRANT privilege ON sequence
REVOKE privilege ON sequence

GRANT TABLE

GRANT privilege ON table, view, materialized view
REVOKE privilege ON table, view, materialized view

GRANT TYPE

GRANT privilege ON TYPE
REVOKE privilege ON TYPE

INSERT TABLE

INSERT INTO table, view

LOCK TABLE

LOCK TABLE table, view

SQL Statements: ALTER TRIGGER to COMMIT 13-37

AUDIT

Table 13–2 (Cont.) Additional SQL Statement Shortcuts for Auditing
SQL Statement Shortcut

SQL Statements and Operations Audited

SELECT SEQUENCE

Any statement containing sequence.CURRVAL or
sequence.NEXTVAL

SELECT TABLE

SELECT FROM table, view, materialized view

UPDATE TABLE

UPDATE table, view

Table 13–3

Schema Object Auditing Options

Object

SQL Operations

Table

ALTER
AUDIT
COMMENT
DELETE
FLASHBACK (Note 3)
GRANT
INDEX
INSERT
LOCK
RENAME
SELECT
UPDATE

View

AUDIT
COMMENT
DELETE
FLASHBACK (Note 3)
GRANT
INSERT
LOCK
RENAME
SELECT
UPDATE

Sequence

ALTER
AUDIT
GRANT
SELECT

Procedure, Function,
Package (Note 1)

AUDIT
EXECUTE
GRANT

Materialized View (Note 2)

ALTER
AUDIT
COMMENT
DELETE
INDEX
INSERT
LOCK
SELECT
UPDATE

Mining Model

AUDIT
COMMENT
GRANT
RENAME
SELECT

Directory

AUDIT
GRANT
READ

13-38 Oracle Database SQL Language Reference

AUDIT

Table 13–3 (Cont.) Schema Object Auditing Options
Object

SQL Operations

Library

EXECUTE
GRANT

Object Type

ALTER
AUDIT
GRANT

Note 1: Java schema objects (sources, classes, and resources) are considered the same
as procedures, functions, and packages for purposes of auditing options.
Note 2: You can audit INSERT, UPDATE, and DELETE operations only on updatable
materialized views.
Note 3: The FLASHBACK audit object option applies only to flashback queries.

Examples
13

To choose auditing for every
SQL statement that creates, alters, drops, or sets a role, regardless of whether the
statement completes successfully, issue the following statement:

Auditing SQL Statements Relating to Roles: Example

AUDIT ROLE;

To choose auditing for every statement that successfully creates, alters, drops, or sets a
role, issue the following statement:
AUDIT ROLE
WHENEVER SUCCESSFUL;

To choose auditing for every CREATE ROLE, ALTER ROLE, DROP ROLE, or SET ROLE
statement that results in an Oracle Database error, issue the following statement:
AUDIT ROLE
WHENEVER NOT SUCCESSFUL;

Auditing Query and Update SQL Statements: Example To choose auditing for any
statement that queries or updates any table, issue the following statement:
AUDIT SELECT TABLE, UPDATE TABLE;

To choose auditing for statements issued by the users hr and oe that query or update a
table or view, issue the following statement
AUDIT SELECT TABLE, UPDATE TABLE
BY hr, oe;

To choose auditing for statements issued using the
DELETE ANY TABLE system privilege, issue the following statement:
Auditing Deletions: Example
AUDIT DELETE ANY TABLE;

Auditing Statements Relating to Directories: Examples To choose auditing for

statements issued using the CREATE ANY DIRECTORY system privilege, issue the
following statement:
AUDIT CREATE ANY DIRECTORY;

SQL Statements: ALTER TRIGGER to COMMIT 13-39

AUDIT

To choose auditing for CREATE DIRECTORY (and DROP DIRECTORY) statements that do not
use the CREATE ANY DIRECTORY system privilege, issue the following statement:
AUDIT DIRECTORY;

To choose auditing for every statement that reads files from the bfile_dir directory,
issue the following statement:
AUDIT READ ON DIRECTORY bfile_dir;

Auditing Queries on a Table: Example To choose auditing for every SQL statement
that queries the employees table in the schema hr, issue the following statement:
AUDIT SELECT
ON hr.employees;

To choose auditing for every statement that successfully queries the employees table in
the schema hr, issue the following statement:
AUDIT SELECT
ON hr.employees
WHENEVER SUCCESSFUL;

To choose auditing for every statement that queries the employees table in the schema
hr and results in an Oracle Database error, issue the following statement:
AUDIT SELECT
ON hr.employees
WHENEVER NOT SUCCESSFUL;

Auditing Inserts and Updates on a Table: Example To choose auditing for every
statement that inserts or updates a row in the customers table in the schema oe, issue
the following statement:
AUDIT INSERT, UPDATE
ON oe.customers;

To choose auditing for every
statement that performs any operation on the employees_seq sequence in the schema
hr, issue the following statement:
Auditing Operations on a Sequence: Example

AUDIT ALL
ON hr.employees_seq;

The preceding statement uses the ALL shortcut to choose auditing for the following
statements that operate on the sequence:
■

ALTER SEQUENCE

■

AUDIT

■

GRANT

■

any statement that accesses the values of the sequence using the pseudocolumns
CURRVAL or NEXTVAL

Setting Default Auditing Options: Example The following statement specifies default
auditing options for objects created in the future:
AUDIT ALTER, GRANT, INSERT, UPDATE, DELETE
ON DEFAULT;

13-40 Oracle Database SQL Language Reference

AUDIT

Any objects created later are automatically configured for audit with the specified
options that apply to them.
■

■

■

■

If you create a table, then Oracle Database automatically configures audit options
ALTER, GRANT, INSERT, UPDATE, or DELETE issued against the table.
If you create a view, then Oracle Database automatically configures audit options
GRANT, INSERT, UPDATE, or DELETE against the view.
If you create a sequence, then Oracle Database automatically configures audit
options ALTER or GRANT against the sequence.
If you create a procedure, package, or function, then Oracle Database
automatically configures audit options ALTER or GRANT against it.

SQL Statements: ALTER TRIGGER to COMMIT 13-41

CALL

CALL
Purpose
13

Use the CALL statement to execute a routine (a standalone procedure or function, or a
procedure or function defined within a type or package) from within SQL.
The restrictions on user-defined function expressions specified
in "Function Expressions" on page 6-10 apply to the CALL statement as
well.

Note:

Oracle Database PL/SQL Language Reference for information
on creating such routine

See Also:

Prerequisites
13

You must have EXECUTE privilege on the standalone routine or on the type or package
in which the routine is defined.

Syntax
13

call::=
routine_clause
CALL
object_access_expression

INDICATOR
:
INTO

:

indicator_variable

host_variable
;

routine_clause::=
type
schema

.

package

.
function

.

procedure
method

,
@

dblink_name

argument
(

13-42 Oracle Database SQL Language Reference

)

CALL

object_access_expression::=
,
argument

.
table_alias

.

column

object_table_alias
(

expr

)

.

.

(

)

attribute
,

.
.

method

argument
method

(

)

Semantics
13

You can execute a routine in two ways. You can issue a call to the routine itself by
name, by using the routine_clause, or you can invoke a routine inside the type of an
expression, by using an object_access_expression.
schema
Specify the schema in which the standalone routine, or the package or type containing
the routine, resides. If you do not specify schema, then Oracle Database assumes the
routine is in your own schema.
type or package
Specify the type or package in which the routine is defined.
routine_clause
Specify the name of the function or procedure being called, or a synonym that resolves
to a function or procedure.
When you call a member function or procedure of a type, if the first argument (SELF) is
a null IN OUT argument, then Oracle Database returns an error. If SELF is a null IN
argument, then the database returns null. In both cases, the function or procedure is
not invoked.
Restriction on Functions

If the routine is a function, then the INTO clause is required.

@dblink
In a distributed database system, specify the name of the database containing the
standalone routine, or the package or function containing the routine. If you omit
dblink, then Oracle Database looks in your local database.
See Also: "Calling a Procedure: Example" on page 13-44 for an
example of calling a routine directly

object_access_expression
If you have an expression of an object type, such as a type constructor or a bind
variable, then you can use this form of expression to call a routine defined within the
type. In this context, the object_access_expression is limited to method invocations.
See Also: "Object Access Expressions" on page 6-13 for syntax and
semantics of this form of expression, and "Calling a Procedure Using
an Expression of an Object Type: Example" on page 13-44 for an
example of calling a routine using an expression of an object type

SQL Statements: ALTER TRIGGER to COMMIT 13-43

CALL

argument
Specify one or more arguments to the routine, if the routine takes arguments. You can
use positional, named, or mixed notation for argument. For example, all of the
following notations are correct:
CALL my_procedure(arg1 => 3, arg2 => 4)
CALL my_procedure(3, 4)
CALL my_procedure(3, arg2 => 4)

Restrictions on Applying Arguments to Routines The argument is subject to the
following restrictions:
■

■

■

■
■

The data types of the parameters passed by the CALL statement must be SQL data
types. They cannot be PL/SQL-only data types such as BOOLEAN.
An argument cannot be a pseudocolumn or either of the object reference functions
VALUE or REF.
Any argument that is an IN OUT or OUT argument of the routine must correspond to
a host variable expression.
The number of arguments, including any return argument, is limited to 1000.
You cannot bind arguments of character and raw data types (CHAR, VARCHAR2,
NCHAR, NVARCHAR2, RAW, LONG RAW) that are larger than 4K.

INTO :host_variable
The INTO clause applies only to calls to functions. Specify which host variable will
store the return value of the function.

:indicator_variable
Specify the value or condition of the host variable.
Pro*C/C++ Programmer's Guide for more information on
host variables and indicator variables

See Also:

Examples
13

The following statement removes the Entertainment
department (created in "Inserting Sequence Values: Example" on page 18-67) using
uses the remove_dept procedure. See Oracle Database PL/SQL Language Reference for the
example that creates this procedure.
Calling a Procedure: Example

CALL emp_mgmt.remove_dept(162);

The following
examples show how call a procedure by using an expression of an object type in the
CALL statement. The example uses the warehouse_typ object type in the order entry
sample schema OE:
Calling a Procedure Using an Expression of an Object Type: Example

ALTER TYPE warehouse_typ
ADD MEMBER FUNCTION ret_name
RETURN VARCHAR2
CASCADE;
CREATE OR REPLACE TYPE BODY warehouse_typ
AS MEMBER FUNCTION ret_name

13-44 Oracle Database SQL Language Reference

CALL

RETURN VARCHAR2
IS
BEGIN
RETURN self.warehouse_name;
END;
END;
/
VARIABLE x VARCHAR2(25);
CALL warehouse_typ(456, 'Warehouse 456', 2236).ret_name()
INTO :x;
PRINT x;
X
-------------------------------Warehouse 456

The next example shows how to use an external function to achieve the same thing:
CREATE OR REPLACE FUNCTION ret_warehouse_typ(x warehouse_typ)
RETURN warehouse_typ
IS
BEGIN
RETURN x;
END;
/
CALL ret_warehouse_typ(warehouse_typ(234, 'Warehouse 234',
2235)).ret_name()
INTO :x;
PRINT x;
X
-------------------------------Warehouse 234

SQL Statements: ALTER TRIGGER to COMMIT 13-45

COMMENT

COMMENT
Purpose
13

Use the COMMENT statement to add to the data dictionary a comment about a table or
table column, view, materialized view, operator, indextype, mining model, or edition.
To drop a comment from the database, set it to the empty string ' '.
See Also:
■

■

"Comments" on page 3-72 for more information on associating
comments with SQL statements and schema objects
Oracle Database Reference for information on the data dictionary
views that display comments

Prerequisites
13

The object about which you are adding a comment must be in your own schema or:
■

■

■

■

To add a comment to a table, view, or materialized view, you must have COMMENT
ANY TABLE system privilege.
To add a comment to an indextype, you must have the CREATE ANY INDEXTYPE
system privilege.
To add a comment to an operator, you must have the CREATE ANY OPERATOR system
privilege.
To add a comment to an edition, you must have the CREATE ANY EDITION system
privilege, granted either directly or through a role.

Syntax
13

comment::=
schema

table

.

COLUMN

view

.

column

materialized_view
EDITION

edition_name
schema

.

INDEXTYPE
C0MMENT

ON

indextype

MATERIALIZED

VIEW

materialized_view
schema

MINING

IS

.

MODEL

model
schema

.

OPERATOR

operator
schema

.

table

TABLE
view

13-46 Oracle Database SQL Language Reference

string

;

COMMENT

Semantics
13

COLUMN Clause
Specify the name of the column of a table, view, or materialized view to be
commented. If you omit schema, then Oracle Database assumes the table, view, or
materialized view is in your own schema.
You can view the comments on a particular table or column by querying the data
dictionary views USER_TAB_COMMENTS, DBA_TAB_COMMENTS, or ALL_TAB_COMMENTS or
USER_COL_COMMENTS, DBA_COL_COMMENTS, or ALL_COL_COMMENTS.

EDITION Clause
Specify the name of an existing edition to be commented.
You can query the data dictionary view ALL_EDITION_COMMENTS to view comments
associated with editions that are accessible to the current user. You can query DBA_
EDITION_COMMENTS to view comments associated with all editions in the database.

TABLE Clause
Specify the schema and name of the table or materialized view to be commented. If
you omit schema, then Oracle Database assumes the table or materialized view is in
your own schema.
In earlier releases, you could use this clause to create a
comment on a materialized view. You should now use the COMMENT ON
MATERIALIZED VIEW clause for materialized views.
Note:

INDEXTYPE Clause
Specify the name of the indextype to be commented. If you omit schema, then Oracle
Database assumes the indextype is in your own schema.
You can view the comments on a particular indextype by querying the data dictionary
views USER_INDEXTYPE_COMMENTS, DBA_INDEXTYPE_COMMENTS, or ALL_INDEXTYPE_
COMMENTS.

MATERIALIZED VIEW Clause
Specify the name of the materialized view to be commented. If you omit schema, then
Oracle Database assumes the materialized view is in your own schema.
You can view the comments on a particular materialized view by querying the data
dictionary views USER_MVIEW_COMMENTS, DBA_MVIEW_COMMENTS, or ALL_MVIEW_
COMMENTS.

MINING MODEL
Specify the name of the mining model to be commented. You must have the COMMENT
ANY MINING MODEL system privilege to specify this clause.

OPERATOR Clause
Specify the name of the operator to be commented. If you omit schema, then Oracle
Database assumes the operator is in your own schema.
You can view the comments on a particular operator by querying the data dictionary
views USER_OPERATOR_COMMENTS, DBA_OPERATOR_COMMENTS, or ALL_OPERATOR_COMMENTS.

SQL Statements: ALTER TRIGGER to COMMIT 13-47

COMMENT

IS 'string'
Specify the text of the comment. Refer to "Text Literals" on page 3-45 for a syntax
description of 'string'.

Example
13

Creating Comments: Example To insert an explanatory remark on the job_id
column of the employees table, you might issue the following statement:
COMMENT ON COLUMN employees.job_id
IS 'abbreviated job title';

To drop this comment from the database, issue the following statement:
COMMENT ON COLUMN employees.job_id IS ' ';

13-48 Oracle Database SQL Language Reference

COMMIT

COMMIT
Purpose
13

Use the COMMIT statement to end your current transaction and make permanent all
changes performed in the transaction. A transaction is a sequence of SQL statements
that Oracle Database treats as a single unit. This statement also erases all savepoints in
the transaction and releases transaction locks.
Until you commit a transaction:
■

■

You can see any changes you have made during the transaction by querying the
modified tables, but other users cannot see the changes. After you commit the
transaction, the changes are visible to other users' statements that execute after the
commit.
You can roll back (undo) any changes made during the transaction with the
ROLLBACK statement (see ROLLBACK on page 18-96).

Oracle Database issues an implicit COMMIT under the following circumstances:
■

■

Before any syntactically valid data definition language (DDL) statement, even if
the statement results in an error
After any data definition language (DDL) statement that completes without an
error

You can also use this statement to:
■

Commit an in-doubt distributed transaction manually

■

Terminate a read-only transaction begun by a SET TRANSACTION statement

Oracle recommends that you explicitly end every transaction in your application
programs with a COMMIT or ROLLBACK statement, including the last transaction, before
disconnecting from Oracle Database. If you do not explicitly commit the transaction
and the program terminates abnormally, then the last uncommitted transaction is
automatically rolled back.
A normal exit from most Oracle utilities and tools causes the current transaction to be
committed. A normal exit from an Oracle precompiler program does not commit the
transaction and relies on Oracle Database to roll back the current transaction.
See Also:
■
■

Oracle Database Concepts for more information on transactions
SET TRANSACTION on page 19-64 for more information on
specifying characteristics of a transaction

Prerequisites
13

You need no privileges to commit your current transaction.
To manually commit a distributed in-doubt transaction that you originally committed,
you must have FORCE TRANSACTION system privilege. To manually commit a distributed
in-doubt transaction that was originally committed by another user, you must have
FORCE ANY TRANSACTION system privilege.

SQL Statements: ALTER TRIGGER to COMMIT 13-49

COMMIT

Syntax
13

commit::=
WORK
COMMIT

COMMENT

string

IMMEDIATE

NOWAIT

BATCH

WRITE
,

FORCE

WAIT

integer

string
;

Semantics
13

COMMIT
All clauses after the COMMIT keyword are optional. If you specify only COMMIT, then the
default is COMMIT WORK WRITE WAIT IMMEDIATE.

WORK
The WORK keyword is supported for compliance with standard SQL. The statements
COMMIT and COMMIT WORK are equivalent.

COMMENT Clause
This clause is supported for backward compatibility. Oracle recommends that you use
named transactions instead of commit comments.
See Also: SET TRANSACTION on page 19-64 and Oracle Database
Concepts for more information on named transactions

Specify a comment to be associated with the current transaction. The 'text' is a quoted
literal of up to 255 bytes that Oracle Database stores in the data dictionary view DBA_
2PC_PENDING along with the transaction ID if a distributed transaction becomes in
doubt. This comment can help you diagnose the failure of a distributed transaction.
COMMENT on page 13-46 for more information on
adding comments to SQL statements

See Also:

WRITE Clause
Use this clause to specify the priority with which the redo information generated by
the commit operation is written to the redo log. This clause can improve performance
by reducing latency, thus eliminating the wait for an I/O to the redo log. Use this
clause to improve response time in environments with stringent response time
requirements where the following conditions apply:
■

■

The volume of update transactions is large, requiring that the redo log be written
to disk frequently.
The application can tolerate the loss of an asynchronously committed transaction.

13-50 Oracle Database SQL Language Reference

COMMIT

■

The latency contributed by waiting for the redo log write to occur contributes
significantly to overall response time.

You can specify the WAIT | NOWAIT and IMMEDIATE | BATCH clauses in any order.

If you omit this clause, then the behavior of the commit
operation is controlled by the COMMIT_LOGGING and COMMIT_WAIT
initialization parameters, if they have been set.

Note:

WAIT | NOWAIT
■

■

Use these clauses to specify when control returns to the user.

The WAIT parameter ensures that the commit will return only after the
corresponding redo is persistent in the online redo log. Whether in BATCH or
IMMEDIATE mode, when the client receives a successful return from this COMMIT
statement, the transaction has been committed to durable media. A crash
occurring after a successful write to the log can prevent the success message from
returning to the client. In this case the client cannot tell whether or not the
transaction committed.
The NOWAIT parameter causes the commit to return to the client whether or not the
write to the redo log has completed. This behavior can increase transaction
throughput. With the WAIT parameter, if the commit message is received, then you
can be sure that no data has been lost.
With NOWAIT, a crash occurring after the commit message is
received, but before the redo log record(s) are written, can falsely
indicate to a transaction that its changes are persistent.

Caution:

If you omit this clause, then the transaction commits with the WAIT behavior.
IMMEDIATE | BATCH
■

■

Use these clauses to specify when the redo is written to the log.

The IMMEDIATE parameter causes the log writer process (LGWR) to write the
transaction's redo information to the log. This operation option forces a disk I/O,
so it can reduce transaction throughput.
The BATCH parameter causes the redo to be buffered to the redo log, along with
other concurrently executing transactions. When sufficient redo information is
collected, a disk write of the redo log is initiated. This behavior is called "group
commit", as redo for multiple transactions is written to the log in a single I/O
operation.

If you omit this clause, then the transaction commits with the IMMEDIATE behavior.
See Also: Oracle Database Advanced Application Developer's Guide for
more information on asynchronous commit

FORCE Clause
The FORCE string [, integer] clause lets you manually commit an in-doubt distributed
transaction. The transaction is identified by the 'string' containing its local or global
transaction ID. To find the IDs of such transactions, query the data dictionary view
DBA_2PC_PENDING. You can use integer to specifically assign the transaction a system
change number (SCN). If you omit integer, then the transaction is committed using
the current SCN.

SQL Statements: ALTER TRIGGER to COMMIT 13-51

COMMIT

Note: A COMMIT statement with a FORCE clause commits only the
specified transactions. Such a statement does not affect your current
transaction.

See Also: Oracle Database Administrator's Guide for more information
on these topics

Examples
13

Committing an Insert: Example

This statement inserts a row into the hr.regions

table and commits this change:
INSERT INTO regions VALUES (5, 'Antarctica');
COMMIT WORK;

To commit the same insert operation and instruct the database to buffer the change to
the redo log, without initiating disk I/O, use the following COMMIT statement:
COMMIT WRITE BATCH;

Commenting on COMMIT: Example The following statement commits the current
transaction and associates a comment with it:
COMMIT
COMMENT 'In-doubt transaction Code 36, Call (415) 555-2637';

If a network or machine failure prevents this distributed transaction from committing
properly, then Oracle Database stores the comment in the data dictionary along with
the transaction ID. The comment indicates the part of the application in which the
failure occurred and provides information for contacting the administrator of the
database where the transaction was committed.
The following statement manually
commits a hypothetical in-doubt distributed transaction. Query the V$CORRUPT_XID_
LIST data dictionary view to find the transaction IDs of corrupt transactions. You must
have DBA privileges to view the V$CORRUPT_XID_LIST and to issue this statement.
Forcing an In-Doubt Transaction: Example

COMMIT FORCE '22.57.53';

13-52 Oracle Database SQL Language Reference

14

14
SQL Statements: CREATE CLUSTER to
CREATE JAVA
This chapter contains the following SQL statements:
■

CREATE CLUSTER

■

CREATE CONTEXT

■

CREATE CONTROLFILE

■

CREATE DATABASE

■

CREATE DATABASE LINK

■

CREATE DIMENSION

■

CREATE DIRECTORY

■

CREATE DISKGROUP

■

CREATE EDITION

■

CREATE FLASHBACK ARCHIVE

■

CREATE FUNCTION

■

CREATE INDEX

■

CREATE INDEXTYPE

■

CREATE JAVA

SQL Statements: CREATE CLUSTER to CREATE JAVA 14-1

CREATE CLUSTER

CREATE CLUSTER
Purpose
14

Use the CREATE CLUSTER statement to create a cluster. A cluster is a schema object that
contains data from one or more tables.
■

■

An indexed cluster must contain more than one table, and all of the tables in the
cluster have one or more columns in common. Oracle Database stores together all
the rows from all the tables that share the same cluster key.
In a hash cluster, which can contain one or more tables, Oracle Database stores
together rows that have the same hash key value.

For information on existing clusters, query the USER_CLUSTERS, ALL_CLUSTERS, and
DBA_CLUSTERS data dictionary views.
See Also:
■
■

■

Oracle Database Concepts for general information on clusters
Oracle Database Performance Tuning Guide for suggestions on when
to use clusters
Oracle Database Reference for information on the data dictionary
views

Prerequisites
14

To create a cluster in your own schema, you must have CREATE CLUSTER system
privilege. To create a cluster in another user's schema, you must have CREATE ANY
CLUSTER system privilege. Also, the owner of the schema to contain the cluster must
have either space quota on the tablespace containing the cluster or the UNLIMITED
TABLESPACE system privilege.
Oracle Database does not automatically create an index for a cluster when the cluster
is initially created. Data manipulation language (DML) statements cannot be issued
against cluster tables in an indexed cluster until you create a cluster index with a
CREATE INDEX statement.

14-2 Oracle Database SQL Language Reference

CREATE CLUSTER

Syntax
14

create_cluster::=
,
schema
CREATE

CLUSTER

.

SORT
cluster

(

column

datatype

)

physical_attributes_clause
SIZE

size_clause

TABLESPACE

tablespace

INDEX
SINGLE

TABLE

HASH
HASHKEYS

parallel_clause

IS

expr

integer

N0ROWDEPENDENCIES

CACHE

ROWDEPENDENCIES

N0CACHE
;

(physical_attributes_clause::= on page 14-3, size_clause::= on page 8-47)
physical_attributes_clause::=
PCTFREE

integer

PCTUSED

integer

INITRANS

integer

storage_clause

(storage_clause::= on page 8-50)
parallel_clause::=
NOPARALLEL
integer
PARALLEL

Semantics
14

schema
Specify the schema to contain the cluster. If you omit schema, then Oracle Database
creates the cluster in your current schema.

cluster
Specify is the name of the cluster to be created.

SQL Statements: CREATE CLUSTER to CREATE JAVA 14-3

CREATE CLUSTER

After you create a cluster, you add tables to it. A cluster can contain a maximum of 32
tables. Object tables and tables containing LOB columns or columns of the Any*
Oracle-supplied types cannot be part of a cluster. After you create a cluster and add
tables to it, the cluster is transparent. You can access clustered tables with SQL
statements just as you can access nonclustered tables.
See Also: CREATE TABLE on page 16-6 for information on adding
tables to a cluster, "Creating a Cluster: Example" on page 14-7, and
"Adding Tables to a Cluster: Example" on page 14-7

column
Specify one or more names of columns in the cluster key. You can specify up to 16
cluster key columns. These columns must correspond in both data type and size to
columns in each of the clustered tables, although they need not correspond in name.
You cannot specify integrity constraints as part of the definition of a cluster key
column. Instead, you can associate integrity constraints with the tables that belong to
the cluster.
See Also:

"Cluster Keys: Example" on page 14-7

datatype
Specify the data type of each cluster key column.
Restrictions on Cluster Data Types

Cluster data types are subject to the following

restrictions:
■

■

You cannot specify a cluster key column of data type LONG, LONG RAW, REF, nested
table, varray, BLOB, CLOB, BFILE, the Any* Oracle-supplied types, or user-defined
object type.
You can specify a column of type ROWID, but Oracle Database does not guarantee
that the values in such columns are valid rowids.
See Also:

"Data Types" on page 3-1 for information on data types

SORT
The SORT keyword is valid only if you are creating a hash cluster. This clause instructs
Oracle Database to sort the rows of the cluster on this column after applying the hash
function when performing a DML operation. Doing so may improve response time
during subsequent queries on the clustered data. See "HASHKEYS Clause" on
page 14-5 for information on creating a hash cluster.
Restriction on Sorted Hash Clusters Row dependency is not supported for sorted
hash clusters.

physical_attributes_clause
The physical_attributes_clause lets you specify the storage characteristics of the
cluster. Each table in the cluster uses these storage characteristics as well. If you do not
specify values for these parameters, then Oracle Database uses the following defaults:
■

PCTFREE: 10

■

PCTUSED: 40

■

INITRANS: 2 or the default value of the tablespace to contain the cluster, whichever
is greater

14-4 Oracle Database SQL Language Reference

CREATE CLUSTER

See Also: physical_attributes_clause on page 8-44 and storage_clause on
page 8-46 for a complete description of these clauses

SIZE
Specify the amount of space in bytes reserved to store all rows with the same cluster
key value or the same hash value. This space determines the maximum number of
cluster or hash values stored in a data block. If SIZE is not a divisor of the data block
size, then Oracle Database uses the next largest divisor. If SIZE is larger than the data
block size, then the database uses the operating system block size, reserving at least
one data block for each cluster or hash value.
The database also considers the length of the cluster key when determining how much
space to reserve for the rows having a cluster key value. Larger cluster keys require
larger sizes. To see the actual size, query the KEY_SIZE column of the USER_CLUSTERS
data dictionary view. (This value does not apply to hash clusters, because hash values
are not actually stored in the cluster.)
If you omit this parameter, then the database reserves one data block for each cluster
key value or hash value.

TABLESPACE
Specify the tablespace in which the cluster is to be created.

INDEX Clause
Specify INDEX to create an indexed cluster. In an indexed cluster, Oracle Database
stores together rows having the same cluster key value. Each distinct cluster key value
is stored only once in each data block, regardless of the number of tables and rows in
which it occurs. If you specify neither INDEX nor HASHKEYS, then Oracle Database
creates an indexed cluster by default.
After you create an indexed cluster, you must create an index on the cluster key before
you can issue any data manipulation language (DML) statements against a table in the
cluster. This index is called the cluster index.
You cannot create a cluster index for a hash cluster, and you need not create an index
on a hash cluster key.
See Also: CREATE INDEX on page 14-60 for information on
creating a cluster index and Oracle Database Concepts for general
information in indexed clusters

HASHKEYS Clause
Specify the HASHKEYS clause to create a hash cluster and specify the number of hash
values for the hash cluster. In a hash cluster, Oracle Database stores together rows that
have the same hash key value. The hash value for a row is the value returned by the
hash function of the cluster.
Oracle Database rounds up the HASHKEYS value to the nearest prime number to obtain
the actual number of hash values. The minimum value for this parameter is 2. If you
omit both the INDEX clause and the HASHKEYS parameter, then the database creates an
indexed cluster by default.
When you create a hash cluster, the database immediately allocates space for the
cluster based on the values of the SIZE and HASHKEYS parameters.

SQL Statements: CREATE CLUSTER to CREATE JAVA 14-5

CREATE CLUSTER

Oracle Database Concepts for more information on how
Oracle Database allocates space for clusters and "Hash Clusters:
Examples" on page 14-7

See Also:

SINGLE TABLE SINGLE TABLE indicates that the cluster is a type of hash cluster
containing only one table. This clause can provide faster access to rows in the table.
Restriction on Single-table Clusters Only one table can be present in the cluster at a

time. However, you can drop the table and create a different table in the same cluster.
See Also:

"Single-Table Hash Clusters: Example" on page 14-8

HASH IS expr Specify an expression to be used as the hash function for the hash

cluster. The expression:
■
■

Must evaluate to a positive value
Must contain at least one column, with referenced columns of any data type as
long as the entire expression evaluates to a number of scale 0. For example:
number_column * LENGTH(varchar2_column)

■

Cannot reference user-defined PL/SQL functions

■

Cannot reference the pseudocolumns LEVEL or ROWNUM

■

Cannot reference the user-related functions USERENV, UID, or USER or the datetime
functions CURRENT_DATE, CURRENT_TIMESTAMP, DBTIMEZONE, EXTRACT (datetime),
FROM_TZ, LOCALTIMESTAMP, NUMTODSINTERVAL, NUMTOYMINTERVAL, SESSIONTIMEZONE,
SYSDATE, SYSTIMESTAMP, TO_DSINTERVAL, TO_TIMESTAMP, TO_DATE, TO_TIMESTAMP_
TZ, TO_YMINTERVAL, and TZ_OFFSET.

■

Cannot evaluate to a constant

■

Cannot be a scalar subquery expression

■

Cannot contain columns qualified with a schema or object name (other than the
cluster name)

If you omit the HASH IS clause, then Oracle Database uses an internal hash function for
the hash cluster.
For information on existing hash functions, query the USER_, ALL_, and DBA_CLUSTER_
HASH_EXPRESSIONS data dictionary tables.
The cluster key of a hash column can have one or more columns of any data type.
Hash clusters with composite cluster keys or cluster keys made up of noninteger
columns must use the internal hash function.
Oracle Database Reference for information on the data
dictionary views

See Also:

parallel_clause
The parallel_clause lets you parallelize the creation of the cluster.
For complete information on this clause, refer to parallel_clause on page 16-63 in the
documentation on CREATE TABLE.

14-6 Oracle Database SQL Language Reference

CREATE CLUSTER

NOROWDEPENDENCIES | ROWDEPENDENCIES
This clause has the same behavior for a cluster that it has for a table. Refer to
"NOROWDEPENDENCIES | ROWDEPENDENCIES" in CREATE TABLE on
page 16-64 for information.

CACHE | NOCACHE
Specify CACHE if you want the blocks retrieved for this cluster to be placed at
the most recently used end of the least recently used (LRU) list in the buffer cache
when a full table scan is performed. This clause is useful for small lookup tables.

CACHE

NOCACHE Specify NOCACHE if you want the blocks retrieved for this cluster to be
placed at the least recently used end of the LRU list in the buffer cache when a full
table scan is performed. This is the default behavior.

NOCACHE has no effect on clusters for which you specify KEEP in the storage_clause.

Examples
14

The following statement creates a cluster named
personnel with the cluster key column department, a cluster size of 512 bytes, and
storage parameter values:
Creating a Cluster: Example

CREATE CLUSTER personnel
(department NUMBER(4))
SIZE 512
STORAGE (initial 100K next 50K);

Cluster Keys: Example

The following statement creates the cluster index on the

cluster key of personnel:
CREATE INDEX idx_personnel ON CLUSTER personnel;

After creating the cluster index, you can add tables to the index and perform DML
operations on those tables.
The following statements create some
departmental tables from the sample hr.employees table and add them to the
personnel cluster created in the earlier example:
Adding Tables to a Cluster: Example

CREATE TABLE dept_10
CLUSTER personnel (department_id)
AS SELECT * FROM employees WHERE department_id = 10;
CREATE TABLE dept_20
CLUSTER personnel (department_id)
AS SELECT * FROM employees WHERE department_id = 20;

The following statement creates a hash cluster named
language with the cluster key column cust_language, a maximum of 10 hash key
values, each of which is allocated 512 bytes, and storage parameter values:
Hash Clusters: Examples

CREATE CLUSTER language (cust_language VARCHAR2(3))
SIZE 512 HASHKEYS 10
STORAGE (INITIAL 100k next 50k);

Because the preceding statement omits the HASH IS clause, Oracle Database uses the
internal hash function for the cluster.

SQL Statements: CREATE CLUSTER to CREATE JAVA 14-7

CREATE CLUSTER

The following statement creates a hash cluster named address with the cluster key
made up of the columns postal_code and country_id, and uses a SQL expression
containing these columns for the hash function:
CREATE CLUSTER address
(postal_code NUMBER, country_id CHAR(2))
HASHKEYS 20
HASH IS MOD(postal_code + country_id, 101);

Single-Table Hash Clusters: Example The following statement creates a single-table
hash cluster named cust_orders with the cluster key customer_id and a maximum of
100 hash key values, each of which is allocated 512 bytes:
CREATE CLUSTER cust_orders (customer_id NUMBER(6))
SIZE 512 SINGLE TABLE HASHKEYS 100;

14-8 Oracle Database SQL Language Reference

CREATE CONTEXT

CREATE CONTEXT
Purpose
14

Use the CREATE CONTEXT statement to:
■

■

Create a namespace for a context (a set of application-defined attributes that
validates and secures an application)
Associate the namespace with the externally created package that sets the context

You can use the DBMS_SESSION.SET_CONTEXT procedure in your designated package to
set or reset the attributes of the context.
See Also:
■
■

Oracle Database Security Guide for a discussion of contexts
Oracle Database PL/SQL Packages and Types Reference for
information on the DBMS_SESSION.SET_CONTEXT procedure

Prerequisites
14

To create a context namespace, you must have CREATE ANY CONTEXT system privilege.

Syntax
14

create_context::=
OR

REPLACE

CREATE

schema
CONTEXT

namespace

USING

.
package

EXTERNALLY
INITIALIZED
GLOBALLY
ACCESSED

GLOBALLY
;

Semantics
14

OR REPLACE
Specify OR REPLACE to redefine an existing context namespace using a different
package.

namespace
Specify the name of the context namespace to create or modify. Context namespaces
are always stored in the schema SYS.
See Also: "Database Object Naming Rules" on page 3-111 for
guidelines on naming a context namespace

schema
Specify the schema owning package. If you omit schema, then Oracle Database uses
the current schema.

SQL Statements: CREATE CLUSTER to CREATE JAVA 14-9

CREATE CONTEXT

package
Specify the PL/SQL package that sets or resets the context attributes under the
namespace for a user session.
To provide some design flexibility, Oracle Database does not verify the existence of the
schema or the validity of the package at the time you create the context.

INITIALIZED Clause
The INITIALIZED clause lets you specify an entity other than Oracle Database that can
initialize the context namespace.
EXTERNALLY EXTERNALLY indicates that the namespace can be initialized using an
OCI interface when establishing a session.
See Also: Oracle Call Interface Programmer's Guide for information on
using OCI to establish a session
GLOBALLY GLOBALLY indicates that the namespace can be initialized by the LDAP
directory when a global user connects to the database.

After the session is established, only the designated PL/SQL package can issue
commands to write to any attributes inside the namespace.
See Also:
■

■

Oracle Database Security Guide for information on establishing
globally initialized contexts
Oracle Internet Directory Administrator's Guide for information on
the connecting to the database through the LDAP directory

ACCESSED GLOBALLY
This clause indicates that any application context set in namespace is accessible
throughout the entire instance. This setting lets multiple sessions share application
attributes.

Examples
14

This example uses a PL/SQL package
emp_mgmt, which validates and secures a human resources application. See Oracle
Database PL/SQL Language Reference for the example that creates that package. The
following statement creates the context namespace hr_context and associates it with
the package emp_mgmt:
Creating an Application Context: Example

CREATE CONTEXT hr_context USING emp_mgmt;

You can control data access based on this context using the SYS_CONTEXT function. For
example, the emp_mgmt package has defined an attribute department_id as a particular
department identifier. You can secure the base table employees by creating a view that
restricts access based on the value of department_id, as follows:
CREATE VIEW hr_org_secure_view AS
SELECT * FROM employees
WHERE department_id = SYS_CONTEXT('hr_context', 'department_id');

14-10 Oracle Database SQL Language Reference

CREATE CONTEXT

SYS_CONTEXT on page 5-279 and Oracle Database Security
Guide for more information on using application contexts to retrieve
user information

See Also:

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-11

CREATE CONTROLFILE

CREATE CONTROLFILE
Caution: Oracle recommends that you perform a full backup of all
files in the database before using this statement. For more information,
see Oracle Database Backup and Recovery User's Guide.

Purpose
14

The CREATE CONTROLFILE statement should be used in only a few cases. Use this
statement to re-create a control file if all control files being used by the database are
lost and no backup control file exists. You can also use this statement to change the
maximum number of redo log file groups, redo log file members, archived redo log
files, data files, or instances that can concurrently have the database mounted and
open.
To change the name of the database, Oracle recommends that you use the DBNEWID
utility rather than the CREATE CONTROLFILE statement. DBNEWID is preferable because no
OPEN RESETLOGS operation is required after changing the database name.
See Also:
■

■

Oracle Database Utilities for more information about the DBNEWID
utility
ALTER DATABASE "BACKUP CONTROLFILE Clause" on page 10-33
for information creating a script based on an existing database
control file

Prerequisites
14

To create a control file, you must have the SYSDBA system privilege.
The database must not be mounted by any instance. After successfully creating the
control file, Oracle mounts the database in the mode specified by the CLUSTER_
DATABASE parameter. The DBA must then perform media recovery before opening the
database. If you are using the database with Oracle Real Application Clusters (Oracle
RAC), then you must then shut down and remount the database in SHARED mode (by
setting the value of the CLUSTER_DATABASE initialization parameter to TRUE) before
other instances can start up.

14-12 Oracle Database SQL Language Reference

CREATE CONTROLFILE

Syntax
14

create_controlfile::=
REUSE
CREATE

SET

CONTROLFILE

DATABASE

database

,
logfile_clause

RESETL0GS

DATAFILE

file_specification

NORESETL0GS

MAXLOGFILES

integer

MAXLOGMEMBERS

integer

MAXLOGHISTORY

integer

MAXDATAFILES

integer

MAXINSTANCES

integer

ARCHIVELOG
NOARCHIVELOG
FORCE

LOGGING

character_set_clause
;

(storage_clause::= on page 8-50)
logfile_clause::=
,
GROUP

integer

LOGFILE

file_specification

(file_specification::= on page 8-29)
character_set_clause::=
CHARACTER

SET

character_set

Semantics
14

When you issue a CREATE CONTROLFILE statement, Oracle Database creates a new
control file based on the information you specify in the statement. The control file
resides in the location specified in the CONTROL_FILES initialization parameter. If that
parameter does not have a value, then the database creates an Oracle-managed control
file in the default control file destination, which is one of the following (in order of
precedence):
1.

One or more control files as specified in the DB_CREATE_ONLINE_LOG_DEST_n
initialization parameter. The file in the first directory is the primary control file.
When DB_CREATE_ONLINE_LOG_DEST_n is specified, the database does not create a

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-13

CREATE CONTROLFILE

control file in DB_CREATE_FILE_DEST or in DB_RECOVERY_FILE_DEST (the fast
recovery area).
2.

If no value is specified for DB_CREATE_ONLINE_LOG_DEST_n, but values are set for
both the DB_CREATE_FILE_DEST and DB_RECOVERY_FILE_DEST, then the database
creates one control file in each location. The location specified in DB_CREATE_FILE_
DEST is the primary control file.

3.

If a value is specified only for DB_CREATE_FILE_DEST, then the database creates one
control file in that location.

4.

If a value is specified only for DB_RECOVERY_FILE_DEST, then the database creates
one control file in that location.

If no values are set for any of these parameters, then the database creates a control file
in the default location for the operating system on which the database is running. This
control file is not an Oracle Managed File.
If you omit any clauses, then Oracle Database uses the default values rather than the
values for the previous control file. After successfully creating the control file, Oracle
Database mounts the database in the mode specified by the initialization parameter
CLUSTER_DATABASE. If that parameter is not set, then the default value is FALSE, and the
database is mounted in EXCLUSIVE mode. Oracle recommends that you then shut
down the instance and take a full backup of all files in the database.
See Also:

Oracle Database Backup and Recovery User's Guide

REUSE
Specify REUSE to indicate that existing control files identified by the initialization
parameter CONTROL_FILES can be reused, overwriting any information they may
currently contain. If you omit this clause and any of these control files already exists,
then Oracle Database returns an error.

DATABASE Clause
Specify the name of the database. The value of this parameter must be the existing
database name established by the previous CREATE DATABASE statement or CREATE
CONTROLFILE statement.

SET DATABASE Clause
Use SET DATABASE to change the name of the database. The name of a database can be
as long as eight bytes.
When you specify this clause, you must also specify RESETLOGS. If you want to rename
the database and retain your existing log files, then after issuing this CREATE
CONTROLFILE statement you must complete a full database recovery using an ALTER
DATABASE RECOVER USING BACKUP CONTROLFILE statement.

logfile_clause
Use the logfile_clause to specify the redo log files for your database. You must list
all members of all redo log file groups.
Use the redo_log_file_spec form of file_specification (see file_specification on
page 8-29) to list regular redo log files in an operating system file system or to list
Oracle ASM disk group redo log files. When using a form of ASM_filename, you
cannot specify the autoextend_clause of the redo_log_file_spec.

14-14 Oracle Database SQL Language Reference

CREATE CONTROLFILE

If you specify RESETLOGS in this clause, then you must use one of the file creation
forms of ASM_filename. If you specify NORESETLOGS, then you must specify one of the
reference forms of ASM_filename.
See Also: ASM_filename on page 8-32 for information on the
different forms of syntax and Oracle Automatic Storage Management
Administrator's Guide for general information about using Oracle ASM

Specify the logfile group number. If you specify GROUP values, then
Oracle Database verifies these values with the GROUP values when the database was
last open.

GROUP integer

If you omit this clause, then the database creates logfiles using system default values.
In addition, if either the DB_CREATE_ONLINE_LOG_DEST_n or DB_CREATE_FILE_DEST
initialization parameter has been set, and if you have specified RESETLOGS, then the
database creates two logs in the default logfile destination specified in the DB_CREATE_
ONLINE_LOG_DEST_n parameter, and if it is not set, then in the DB_CREATE_FILE_DEST
parameter.
See Also: file_specification on page 8-29 for a full description of this
clause
RESETLOGS Specify RESETLOGS if you want Oracle Database to ignore the contents
of the files listed in the LOGFILE clause. These files do not have to exist. You must
specify this clause if you have specified the SET DATABASE clause.

Each redo_log_file_spec in the LOGFILE clause must specify the SIZE parameter. The
database assigns all online redo log file groups to thread 1 and enables this thread for
public use by any instance. After using this clause, you must open the database using
the RESETLOGS clause of the ALTER DATABASE statement.
Specify NORESETLOGS if you want Oracle Database to use all files in
the LOGFILE clause as they were when the database was last open. These files must
exist and must be the current online redo log files rather than restored backups. The
database reassigns the redo log file groups to the threads to which they were
previously assigned and reenables the threads as they were previously enabled.
NORESETLOGS

You cannot specify NORESETLOGS if you have specified the SET DATABASE clause to
change the name of the database. Refer to "SET DATABASE Clause" on page 14-14 for
more information.
DATAFILE Clause
Specify the data files of the database. You must list all data files. These files must all
exist, although they may be restored backups that require media recovery.
Do not include in the DATAFILE clause any data files in read-only tablespaces. You can
add these types of files to the database later. Also, do not include in this clause any
temporary data files (temp files).
Use the datafile_tempfile_spec form of file_specification (see file_specification on
page 8-29) to list regular data files and temp files in an operating system file system or
to list Oracle ASM disk group files. When using a form of ASM_filename, you must use
one of the reference forms of ASM_filename. Refer to ASM_filename on page 8-32 for
information on the different forms of syntax.
Oracle Automatic Storage Management Administrator's Guide
for general information about using Oracle ASM

See Also:

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-15

CREATE CONTROLFILE

You cannot specify the autoextend_clause of file_
specification in this DATAFILE clause.

Restriction on DATAFILE

MAXLOGFILES Clause
Specify the maximum number of online redo log file groups that can ever be created
for the database. Oracle Database uses this value to determine how much space to
allocate in the control file for the names of redo log files. The default and maximum
values depend on your operating system. The value that you specify should not be less
than the greatest GROUP value for any redo log file group.
MAXLOGMEMBERS Clause
Specify the maximum number of members, or identical copies, for a redo log file
group. Oracle Database uses this value to determine how much space to allocate in the
control file for the names of redo log files. The minimum value is 1. The maximum and
default values depend on your operating system.
MAXLOGHISTORY Clause
This parameter is useful only if you are using Oracle Database in ARCHIVELOG mode.
Specify your current estimate of the maximum number of archived redo log file
groups needed for automatic media recovery of the database. The database uses this
value to determine how much space to allocate in the control file for the names of
archived redo log files.
The minimum value is 0. The default value is a multiple of the MAXINSTANCES value
and depends on your operating system. The maximum value is limited only by the
maximum size of the control file. The database will continue to add additional space to
the appropriate section of the control file as needed, so that you do not need to
re-create the control file if your your original configuration is no longer adequate. As a
result, the actual value of this parameter can eventually exceed the value you specify.
MAXDATAFILES Clause
Specify the initial sizing of the data files section of the control file at CREATE DATABASE
or CREATE CONTROLFILE time. An attempt to add a file whose number is greater than
MAXDATAFILES, but less than or equal to DB_FILES, causes the control file to expand
automatically so that the data files section can accommodate more files.
The number of data files accessible to your instance is also limited by the initialization
parameter DB_FILES.
MAXINSTANCES Clause
Specify the maximum number of instances that can simultaneously have the database
mounted and open. This value takes precedence over the value of the initialization
parameter INSTANCES. The minimum value is 1. The maximum and default values
depend on your operating system.
ARCHIVELOG | NOARCHIVELOG
Specify ARCHIVELOG to archive the contents of redo log files before reusing them. This
clause prepares for the possibility of media recovery as well as instance or system
failure recovery.
If you omit both the ARCHIVELOG clause and NOARCHIVELOG clause, then Oracle
Database chooses NOARCHIVELOG mode by default. After creating the control file, you
can change between ARCHIVELOG mode and NOARCHIVELOG mode with the ALTER
DATABASE statement.

14-16 Oracle Database SQL Language Reference

CREATE CONTROLFILE

FORCE LOGGING
Use this clause to put the database into FORCE LOGGING mode after control file creation.
When the database is in this mode, Oracle Database logs all changes in the database
except changes to temporary tablespaces and temporary segments. This setting takes
precedence over and is independent of any NOLOGGING or FORCE LOGGING settings you
specify for individual tablespaces and any NOLOGGING settings you specify for
individual database objects. If you omit this clause, then the database will not be in
FORCE LOGGING mode after the control file is created.

Note: FORCE LOGGING mode can have performance effects. Refer to
Oracle Database Administrator's Guide for information on when to
use this setting.

character_set_clause
If you specify a character set, then Oracle Database reconstructs character set
information in the control file. If media recovery of the database is subsequently
required, then this information will be available before the database is open, so that
tablespace names can be correctly interpreted during recovery. This clause is required
only if you are using a character set other than the default, which depends on your
operating system. Oracle Database prints the current database character set to the alert
log in $ORACLE_HOME/log during startup.
If you are re-creating your control file and you are using Recovery Manager for
tablespace recovery, and if you specify a different character set from the one stored in
the data dictionary, then tablespace recovery will not succeed. However, at database
open, the control file character set will be updated with the correct character set from
the data dictionary.
You cannot modify the character set of the database with this clause.
See Also: Oracle Database Backup and Recovery User's Guide for more
information on tablespace recovery

Examples
14

This statement re-creates a control file. In this
statement, database demo was created with the WE8DEC character set. The example
uses the word path where you would normally insert the path on your system to the
appropriate Oracle Database directories.
Creating a Controlfile: Example

STARTUP NOMOUNT
CREATE CONTROLFILE REUSE DATABASE "demo" NORESETLOGS NOARCHIVELOG
MAXLOGFILES 32
MAXLOGMEMBERS 2
MAXDATAFILES 32
MAXINSTANCES 1
MAXLOGHISTORY 449
LOGFILE
GROUP 1 '/path/oracle/dbs/t_log1.f' SIZE 500K,
GROUP 2 '/path/oracle/dbs/t_log2.f' SIZE 500K
# STANDBY LOGFILE
DATAFILE
'/path/oracle/dbs/t_db1.f',
'/path/oracle/dbs/dbu19i.dbf',

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-17

CREATE CONTROLFILE

'/path/oracle/dbs/tbs_11.f',
'/path/oracle/dbs/smundo.dbf',
'/path/oracle/dbs/demo.dbf'
CHARACTER SET WE8DEC
;

14-18 Oracle Database SQL Language Reference

CREATE DATABASE

CREATE DATABASE
Caution: This statement prepares a database for initial use and
erases any data currently in the specified files. Use this statement only
when you understand its ramifications.

Note Regarding Security Enhancements: In this release of Oracle
Database and in subsequent releases, several enhancements are being made
to ensure the security of default database user accounts. You can find a
security checklist for this release in Oracle Database Security Guide. Oracle
recommends that you read this checklist and configure your database
accordingly.

Purpose
14

Use the CREATE DATABASE statement to create a database, making it available for
general use.
This statement erases all data in any specified data files that already exist in order to
prepare them for initial database use. If you use the statement on an existing database,
then all data in the data files is lost.
After creating the database, this statement mounts it in either exclusive or parallel
mode, depending on the value of the CLUSTER_DATABASE initialization parameter and
opens it, making it available for normal use. You can then create tablespaces for the
database.
See Also:
■

■

■

ALTER DATABASE on page 10-8 for information on modifying a
database
Oracle Database Java Developer's Guide for information on creating
an Oracle Java virtual machine
CREATE TABLESPACE on page 16-83 for information on creating
tablespaces

Prerequisites
14

To create a database, you must have the SYSDBA system privilege. An initialization
parameter file with the name of the database to be created must be available, and you
must be in STARTUP NOMOUNT mode.

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-19

CREATE DATABASE

Syntax
14

create_database::=
USER

SYS

IDENTIFIED

USER

SYSTEM

CONTROLFILE

database
CREATE

IDENTIFIED

password
BY

password

REUSE

MAXDATAFILES

integer

MAXINSTANCES

integer

CHARACTER

BY

SET

charset

DATABASE

;
NATIONAL

CHARACTER

SET

charset

BIGFILE
SET

DEFAULT

TABLESPACE
SMALLFILE

database_logging_clauses
tablespace_clauses
set_time_zone_clause

(database_logging_clauses::= on page 14-20, tablespace_clauses::= on page 14-21, set_time_
zone_clause::= on page 14-22)
database_logging_clauses::=
,
GROUP
LOGFILE

integer
file_specification

MAXLOGFILES

integer

MAXLOGMEMBERS
MAXLOGHISTORY

integer
integer

ARCHIVELOG
NOARCHIVELOG
FORCE

LOGGING

(file_specification::= on page 8-29)

14-20 Oracle Database SQL Language Reference

CREATE DATABASE

tablespace_clauses::=
EXTENT

MANAGEMENT

LOCAL

,
DATAFILE

file_specification
,

SYSAUX

DATAFILE

file_specification

default_tablespace
default_temp_tablespace
undo_tablespace

(file_specification::= on page 8-29, default_tablespace::= on page 14-21, default_temp_
tablespace::= on page 14-21, undo_tablespace::= on page 14-21)
default_tablespace::=
DATAFILE
DEFAULT

TABLESPACE

datafile_tempfile_spec

extent_management_clause

tablespace

default_temp_tablespace::=
BIGFILE
SMALLFILE
DEFAULT

TEMPORARY

TABLESPACE

tablespace

,
TEMPFILE

file_specification

extent_management_clause

(file_specification::= on page 8-29)
extent_management_clause::=

AUTOALLOCATE
SIZE

size_clause

UNIFORM
EXTENT

MANAGEMENT

LOCAL

(size_clause::= on page 8-47)
undo_tablespace::=
BIGFILE

,

SMALLFILE

DATAFILE
UNDO

TABLESPACE

file_specification

tablespace

(file_specification::= on page 8-29)
SQL Statements: CREATE CLUSTER to CREATE JAVA

14-21

CREATE DATABASE

set_time_zone_clause::=
+
hh
SET

TIME_ZONE

=

’

:

mi

–

’

time_zone_region

Semantics
14

database
Specify the name of the database to be created. The name must match the value of the
DB_NAME initialization parameter. The name can be up to 8 bytes long and can contain
only ASCII characters. Oracle Database writes this name into the control file. If you
subsequently issue an ALTER DATABASE statement that explicitly specifies a database
name, then Oracle Database verifies that name with the name in the control file.
The database name is case insensitive and is stored in uppercase ASCII characters. If
you specify the database name as a quoted identifier, then the quotation marks are
silently ignored.
You cannot use special characters from European or Asian
character sets in a database name. For example, characters with
umlauts are not allowed.

Note:

If you omit the database name from a CREATE DATABASE statement, then Oracle
Database uses the name specified by the initialization parameter DB_NAME. The DB_NAME
initialization parameter must be set in the database initialization parameter file, and if
you specify a different name from the value of that parameter, then the database
returns an error. Refer to "Database Object Naming Rules" on page 3-111 for additional
rules to which database names should adhere.

USER SYS ..., USER SYSTEM ...
Use these clauses to establish passwords for the SYS and SYSTEM users. These clauses
are not mandatory in this release. However, if you specify either clause, then you must
specify both clauses.
If you do not specify these clauses, then Oracle Database creates default passwords
change_on_install for user SYS and manager for user SYSTEM. You can subsequently
change these passwords using the ALTER USER statement. You can also use ALTER USER
to add password management attributes after database creation.
See Also:

ALTER USER on page 13-6

CONTROLFILE REUSE Clause
Specify CONTROLFILE REUSE to reuse existing control files identified by the initialization
parameter CONTROL_FILES, overwriting any information they currently contain.
Normally you use this clause only when you are re-creating a database, rather than
creating one for the first time. When you create a database for the first time, Oracle
Database creates a control file in the default destination, which is dependent on the
value or several initialization parameters. See CREATE CONTROLFILE, "Semantics" on
page 14-13.

14-22 Oracle Database SQL Language Reference

CREATE DATABASE

You cannot use this clause if you also specify a parameter value that requires that the
control file be larger than the existing files. These parameters are MAXLOGFILES,
MAXLOGMEMBERS, MAXLOGHISTORY, MAXDATAFILES, and MAXINSTANCES.
If you omit this clause and any of the files specified by CONTROL_FILES already exist,
then the database returns an error.

MAXDATAFILES Clause
Specify the initial sizing of the data files section of the control file at CREATE DATABASE
or CREATE CONTROLFILE time. An attempt to add a file whose number is greater than
MAXDATAFILES, but less than or equal to DB_FILES, causes the Oracle Database control
file to expand automatically so that the data files section can accommodate more files.
The number of data files accessible to your instance is also limited by the initialization
parameter DB_FILES.

MAXINSTANCES Clause
Specify the maximum number of instances that can simultaneously have this database
mounted and open. This value takes precedence over the value of initialization
parameter INSTANCES. The minimum value is 1. The maximum value is 1055. The
default depends on your operating system.

CHARACTER SET Clause
Specify the character set the database uses to store data. The supported character sets
and default value of this parameter depend on your operating system.
Restriction on CHARACTER SET

You cannot specify the AL16UTF16 character set as

the database character set.
See Also: Oracle Database Globalization Support Guide for more
information about choosing a character set

NATIONAL CHARACTER SET Clause
Specify the national character set used to store data in columns specifically defined as
NCHAR, NCLOB, or NVARCHAR2. Valid values are AL16UTF16 and UTF8. The default is
AL16UTF16.
See Also: Oracle Database Globalization Support Guide for information
on Unicode data type support

SET DEFAULT TABLESPACE Clause
Use this clause to determine the default type of subsequently created tablespaces and
of the SYSTEM and SYSAUX tablespaces. Specify either BIGFILE or SMALLFILE to set the
default type of subsequently created tablespaces as a bigfile or smallfile tablespace,
respectively.
■

■

A bigfile tablespace contains only one data file or temp file, which can contain up
to approximately 4 billion (232) blocks. The maximum size of the single data file or
temp file is 128 terabytes (TB) for a tablespace with 32K blocks and 32TB for a
tablespace with 8K blocks.
A smallfile tablespace is a traditional Oracle tablespace, which can contain 1022
data files or temp files, each of which can contain up to approximately 4 million
(222) blocks.

If you omit this clause, then Oracle Database creates smallfile tablespaces by default.

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-23

CREATE DATABASE

See Also:
■

■

Oracle Database Administrator's Guide for more information about
bigfile tablespaces
"Setting the Default Type of Tablespaces: Example" on page 10-43
for an example using this syntax

database_logging_clauses
Use the database_logging_clauses to determine how Oracle Database will handle
redo log files for this database.
LOGFILE Clause
Specify one or more files to be used as redo log files. Use the redo_log_file_spec
form of file_specification to create regular redo log files in an operating system file
system or to create Oracle ASM disk group redo log files. When using a form of ASM_
filename, you cannot specify the autoextend_clause of redo_log_file_spec.
The redo_log_file_spec clause specifies a redo log file group containing one or more
redo log file members (copies). All redo log files specified in a CREATE DATABASE
statement are added to redo log thread number 1.
See Also: file_specification on page 8-29 for a full description of this
clause

If you omit the LOGFILE clause, then Oracle Database creates an Oracle-managed log
file member in the default destination, which is one of the following locations (in order
of precedence):
■

■

■

■

If DB_CREATE_ONLINE_LOG_DEST_n is set, then the database creates a log file
member in each directory specified, up to the value of the MAXLOGMEMBERS
initialization parameter.
If the DB_CREATE_ONLINE_LOG_DEST_n parameter is not set, but both the DB_
CREATE_FILE_DEST and DB_RECOVERY_FILE_DEST initialization parameters are set,
then the database creates one Oracle-managed log file member in each of those
locations. The log file in the DB_CREATE_FILE_DEST destination is the first member.
If only the DB_CREATE_FILE_DEST initialization parameter is specified, then Oracle
Database creates a log file member in that location.
If only the DB_RECOVERY_FILE_DEST initialization parameter is specified, then
Oracle Database creates a log file member in that location.

In all these cases, the parameter settings must correctly specify operating system
filenames or creation form Oracle ASM filenames, as appropriate.
If no values are set for any of these parameters, then the database creates a log file in
the default location for the operating system on which the database is running. This
log file is not an Oracle Managed File.
Specify the number that identifies the redo log file group. The value
of integer can range from 1 to the value of the MAXLOGFILES parameter. A database
must have at least two redo log file groups. You cannot specify multiple redo log file
groups having the same GROUP value. If you omit this parameter, then Oracle Database
generates its value automatically. You can examine the GROUP value for a redo log file
group through the dynamic performance view V$LOG.
GROUP integer

14-24 Oracle Database SQL Language Reference

CREATE DATABASE

MAXLOGFILES Clause
Specify the maximum number of redo log file groups that can ever be created for the
database. Oracle Database uses this value to determine how much space to allocate in
the control file for the names of redo log files. The default, minimum, and maximum
values depend on your operating system.
MAXLOGMEMBERS Clause
Specify the maximum number of members, or copies, for a redo log file group. Oracle
Database uses this value to determine how much space to allocate in the control file for
the names of redo log files. The minimum value is 1. The maximum and default values
depend on your operating system.
MAXLOGHISTORY Clause
This parameter is useful only if you are using Oracle Database in ARCHIVELOG mode
with Oracle Real Application Clusters (Oracle RAC). Specify the maximum number of
archived redo log files for automatic media recovery of Oracle RAC. The database uses
this value to determine how much space to allocate in the control file for the names of
archived redo log files. The minimum value is 0. The default value is a multiple of the
MAXINSTANCES value and depends on your operating system. The maximum value is
limited only by the maximum size of the control file.
ARCHIVELOG
Specify ARCHIVELOG if you want the contents of a redo log file group to be archived
before the group can be reused. This clause prepares for the possibility of media
recovery.
NOARCHIVELOG
Specify NOARCHIVELOG if the contents of a redo log file group need not be archived
before the group can be reused. This clause does not allow for the possibility of media
recovery.
The default is NOARCHIVELOG mode. After creating the database, you can change
between ARCHIVELOG mode and NOARCHIVELOG mode with the ALTER DATABASE
statement.
FORCE LOGGING
Use this clause to put the database into FORCE LOGGING mode. Oracle Database will log
all changes in the database except for changes in temporary tablespaces and
temporary segments. This setting takes precedence over and is independent of any
NOLOGGING or FORCE LOGGING settings you specify for individual tablespaces and any
NOLOGGING settings you specify for individual database objects.
FORCE LOGGING mode is persistent across instances of the database. If you shut down
and restart the database, then the database is still in FORCE LOGGING mode. However, if
you re-create the control file, then Oracle Database will take the database out of FORCE
LOGGING mode unless you specify FORCE LOGGING in the CREATE CONTROLFILE statement.

Note: FORCE LOGGING mode can have performance effects. Refer to
Oracle Database Administrator's Guide for information on when to
use this setting.

See Also:

CREATE CONTROLFILE on page 14-12

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-25

CREATE DATABASE

tablespace_clauses
Use the tablespace clauses to configure the SYSTEM and SYSAUX tablespaces and to
specify a default temporary tablespace and an undo tablespace.
extent_management_clause
Use this clause to create a locally managed SYSTEM tablespace. If you omit this clause,
then the SYSTEM tablespace will be dictionary managed.
When you create a locally managed SYSTEM tablespace, you
cannot change it to be dictionary managed, nor can you create any
other dictionary-managed tablespaces in this database.

Caution:

If you specify this clause, then the database must have a default temporary tablespace,
because a locally managed SYSTEM tablespace cannot store temporary segments.
■

■

If you specify EXTENT MANAGEMENT LOCAL but you do not specify the DATAFILE
clause, then you can omit the default_temp_tablespace clause. Oracle Database
will create a default temporary tablespace called TEMP with one data file of size
10M with autoextend disabled.
If you specify both EXTENT MANAGEMENT LOCAL and the DATAFILE clause, then you
must also specify the default_temp_tablespace clause and explicitly specify a
temp file for that temporary tablespace.

If you have opened the instance in automatic undo mode, similar requirements exist
for the database undo tablespace:
■

■

If you specify EXTENT MANAGEMENT LOCAL but you do not specify the DATAFILE
clause, then you can omit the undo_tablespace clause. Oracle Database will create
an undo tablespace named SYS_UNDOTBS.
If you specify both EXTENT MANAGEMENT LOCAL and the DATAFILE clause, then you
must also specify the undo_tablespace clause and explicitly specify a data file for
that tablespace.
See Also: Oracle Database Administrator's Guide for more information
on locally managed and dictionary-managed tablespaces

DATAFILE Clause
Specify one or more files to be used as data files. All these files become part of the
SYSTEM tablespace. Use the data file_tempfile_spec form of file_specification to
create regular data files and temp files in an operating system file system or to create
Oracle ASM disk group files.
This clause is optional, as is the DATAFILE clause of the
undo_tablespace clause. Therefore, to avoid ambiguity, if your
intention is to specify a data file for the SYSTEM tablespace with this
clause, then do not specify it immediately after an undo_tablespace
clause that does not include the optional DATAFILE clause. If you do
so, then Oracle Database will interpret the DATAFILE clause to be part
of the undo_tablespace clause.
Caution:

The syntax for specifying data files for the SYSTEM tablespace is the same as that for
specifying data files during tablespace creation using the CREATE TABLESPACE

14-26 Oracle Database SQL Language Reference

CREATE DATABASE

statement, whether you are storing files using Oracle ASM or in a file system or raw
device.
CREATE TABLESPACE on page 16-83 for information on
specifying data files

See Also:

If you are running the database in automatic undo mode and you specify a data file
name for the SYSTEM tablespace, then Oracle Database expects to generate data files for
all tablespaces. Oracle Database does this automatically if you are using Oracle
Managed Files—you have set a value for the DB_CREATE_FILE_DEST initialization
parameter. However, if you are not using Oracle Managed Files and you specify this
clause, then you must also specify the undo_tablespace clause and the default_temp_
tablespace clause.
If you omit this clause, then:
■

■

If the DB_CREATE_FILE_DEST initialization parameter is set, then Oracle Database
creates a 100 MB Oracle-managed data file with a system-generated name in the
default file destination specified in the parameter.
If the DB_CREATE_FILE_DEST initialization parameter is not set, then Oracle
Database creates one data file whose name and size depend on your operating
system.
See Also:

file_specification on page 8-29 for syntax

SYSAUX Clause
Oracle Database creates both the SYSTEM and SYSAUX tablespaces as part of every
database. Use this clause if you are not using Oracle Managed Files and you want to
specify one or more data files for the SYSAUX tablespace.
You must specify this clause if you have specified one or more data files for the SYSTEM
tablespace using the DATAFILE clause. If you are using Oracle Managed Files and you
omit this clause, then the database creates the SYSAUX data files in the default location
set up for Oracle Managed Files.
If you have enabled Oracle Managed Files and you omit the SYSAUX clause, then the
database creates the SYSAUX tablespace as an online, permanent, locally managed
tablespace with one data file of 100 MB, with logging enabled and automatic
segment-space management.
The syntax for specifying data files for the SYSAUX tablespace is the same as that for
specifying data files during tablespace creation using the CREATE TABLESPACE
statement, whether you are storing files using Oracle ASM or in a file system or raw
device.
See Also:
■

■

CREATE TABLESPACE on page 16-83 for information on creating
the SYSAUX tablespace during database upgrade and for
information on specifying data files in a tablespace
Oracle Database Administrator's Guide for more information on
creating the SYSAUX tablespace

default_tablespace
Specify this clause to create a default permanent tablespace for the database. Oracle
Database creates a smallfile tablespace and subsequently will assign to this tablespace
any non-SYSTEM users for whom you do not specify a different permanent tablespace.
SQL Statements: CREATE CLUSTER to CREATE JAVA

14-27

CREATE DATABASE

If you do not specify this clause, then the SYSTEM tablespace is the default permanent
tablespace for non-SYSTEM users.
The DATAFILE clause and extent_management_clause have the same semantics they
have in a CREATE TABLESPACE statement. Refer to "DATAFILE | TEMPFILE Clause" on
page 16-87 and extent_management_clause on page 16-90 for information on these
clauses.
default_temp_tablespace
Specify this clause to create a default temporary tablespace for the database. Oracle
Database will assign to this temporary tablespace any users for whom you do not
specify a different temporary tablespace. If you do not specify this clause, and if the
database does not create a default temporary tablespace automatically in the process
of creating a locally managed SYSTEM tablespace, then the SYSTEM tablespace is the
default temporary tablespace.
Specify BIGFILE or SMALLFILE to determine whether the default temporary tablespace
is a bigfile or smallfile tablespace. These clauses have the same semantics as in the
"SET DEFAULT TABLESPACE Clause" on page 14-23.
The TEMPFILE clause part of this clause is optional if you have enabled Oracle
Managed Files by setting the DB_CREATE_FILE_DEST initialization parameter. If you
have not specified a value for this parameter, then the TEMPFILE clause is required. If
you have specified BIGFILE, then you can specify only one temp file in this clause.
The syntax for specifying temp files for the default temporary tablespace is the same as
that for specifying temp files during temporary tablespace creation using the CREATE
TABLESPACE statement, whether you are storing files using Oracle ASM or in a file
system or raw device.
See Also: CREATE TABLESPACE on page 16-83 for information on
specifying temp files

On some operating systems, Oracle does not allocate space
for a temp file until the temp file blocks are actually accessed. This
delay in space allocation results in faster creation and resizing of
temp files, but it requires that sufficient disk space is available
when the temp files are later used. To avoid potential problems,
before you create or resize a temp file, ensure that the available disk
space exceeds the size of the new temp file or the increased size of a
resized temp file. The excess space should allow for anticipated
increases in disk space use by unrelated operations as well. Then
proceed with the creation or resizing operation.

Note:

Restrictions on Default Temporary Tablespaces

Default temporary tablespaces are

subject to the following restrictions:
■

You cannot specify the SYSTEM tablespace in this clause.

■

The default temporary tablespace must have a standard block size.

The extent_management_clause clause has the same semantics in CREATE DATABASE
and CREATE TABLESPACE statements. For complete information, refer to the CREATE
TABLESPACE ... extent_management_clause on page 16-90.

14-28 Oracle Database SQL Language Reference

CREATE DATABASE

undo_tablespace
If you have opened the instance in automatic undo mode (the UNDO_MANAGEMENT
initialization parameter is set to AUTO, which is the default), then you can specify the
undo_tablespace to create a tablespace to be used for undo data. Oracle strongly
recommends that you use automatic undo mode. However, if you want undo space
management to be handled by way of rollback segments, then you must omit this
clause. You can also omit this clause if you have set a value for the UNDO_TABLESPACE
initialization parameter. If that parameter has been set, and if you specify this clause,
then tablespace must be the same as that parameter value.
■

■

■

Specify BIGFILE if you want the undo tablespace to be a bigfile tablespace. A
bigfile tablespace contains only one data file, which can be up to 8 exabytes (8
million terabytes) in size.
Specify SMALLFILE if you want the undo tablespace to be a smallfile tablespace. A
smallfile tablespace is a traditional Oracle Database tablespace, which can contain
1022 data files or temp files, each of which can contain up to approximately 4
million (222) blocks.
The DATAFILE clause part of this clause is optional if you have enabled Oracle
Managed Files by setting the DB_CREATE_FILE_DEST initialization parameter. If you
have not specified a value for this parameter, then the DATAFILE clause is required.
If you have specified BIGFILE, then you can specify only one data file in this
clause.

The syntax for specifying data files for the undo tablespace is the same as that for
specifying data files during tablespace creation using the CREATE TABLESPACE
statement, whether you are storing files using Oracle ASM or in a file system or raw
device.
CREATE TABLESPACE on page 16-83 for information on
specifying data files

See Also:

If you specify this clause, then Oracle Database creates an undo tablespace named
tablespace, creates the specified data file(s) as part of the undo tablespace, and
assigns this tablespace as the undo tablespace of the instance. Oracle Database will
manage undo data using this undo tablespace. The DATAFILE clause of this clause has
the same behavior as described in "DATAFILE Clause" on page 14-26.
If you have specified a value for the UNDO_TABLESPACE initialization parameter in your
initialization parameter file before mounting the database, then you must specify the
same name in this clause. If these names differ, then Oracle Database will return an
error when you open the database.
If you omit this clause, then Oracle Database creates a default database with a default
smallfile undo tablespace named SYS_UNDOTBS and assigns this default tablespace as
the undo tablespace of the instance. This undo tablespace allocates disk space from the
default files used by the CREATE DATABASE statement, and it has an initial extent of 10M.
Oracle Database handles the system-generated data file as described in "DATAFILE
Clause" on page 14-26. If Oracle Database is unable to create the undo tablespace, then
the entire CREATE DATABASE operation fails.
See Also:
■

■

Oracle Database Administrator's Guide for information on automatic
undo management and undo tablespaces
CREATE TABLESPACE on page 16-83 for information on creating
an undo tablespace after database creation

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-29

CREATE DATABASE

set_time_zone_clause
Use the SET TIME_ZONE clause to set the time zone of the database. You can specify the
time zone in two ways:
■

■

By specifying a displacement from UTC (Coordinated Universal Time—formerly
Greenwich Mean Time). The valid range of hh:mi is -12:00 to +14:00.
By specifying a time zone region. To see a listing of valid time zone region names,
query the TZNAME column of the V$TIMEZONE_NAMES dynamic performance view.
Oracle recommends that you set the database time zone to
UTC (0:00). Doing so can improve performance, especially across
databases, as no conversion of time zones will be required.

Note:

Oracle Database Reference for information on the dynamic
performance views

See Also:

Oracle Database normalizes all TIMESTAMP WITH LOCAL TIME ZONE data to the time zone
of the database when the data is stored on disk. If you do not specify the SET TIME_
ZONE clause, then the database uses the operating system time zone of the server. If the
operating system time zone is not a valid Oracle Database time zone, then the database
time zone defaults to UTC.

Examples
14

Creating a Database: Example

The following statement creates a database and fully

specifies each argument:
CREATE DATABASE sample
CONTROLFILE REUSE
LOGFILE
GROUP 1 ('diskx:log1.log', 'disky:log1.log') SIZE 50K,
GROUP 2 ('diskx:log2.log', 'disky:log2.log') SIZE 50K
MAXLOGFILES 5
MAXLOGHISTORY 100
MAXDATAFILES 10
MAXINSTANCES 2
ARCHIVELOG
CHARACTER SET AL32UTF8
NATIONAL CHARACTER SET AL16UTF16
DATAFILE
'disk1:df1.dbf' AUTOEXTEND ON,
'disk2:df2.dbf' AUTOEXTEND ON NEXT 10M MAXSIZE UNLIMITED
DEFAULT TEMPORARY TABLESPACE temp_ts
UNDO TABLESPACE undo_ts
SET TIME_ZONE = '+02:00';

This example assumes that you have enabled Oracle Managed Files by specifying a
value for the DB_CREATE_FILE_DEST parameter in your initialization parameter file.
Therefore no file specification is needed for the DEFAULT TEMPORARY TABLESPACE and
UNDO TABLESPACE clauses.

14-30 Oracle Database SQL Language Reference

CREATE DATABASE LINK

CREATE DATABASE LINK
Purpose
14

Use the CREATE DATABASE LINK statement to create a database link. A database link is a
schema object in one database that enables you to access objects on another database.
The other database need not be an Oracle Database system. However, to access
non-Oracle systems you must use Oracle Heterogeneous Services.
After you have created a database link, you can use it in SQL statements to refer to
tables, views, and PL/SQL objects in the other database by appending @dblink to the
table, view, or PL/SQL object name. You can query a table or view in the other
database with the SELECT statement. You can also access remote tables and views using
any INSERT, UPDATE, DELETE, or LOCK TABLE statement.
See Also:
■

■

■

■

■

■

Oracle Database Advanced Application Developer's Guide for
information about accessing remote tables or views with PL/SQL
functions, procedures, packages, and data types
Oracle Database Administrator's Guide for information on
distributed database systems
Oracle Database Reference for descriptions of existing database links
in the ALL_DB_LINKS, DBA_DB_LINKS, and USER_DB_LINKS data
dictionary views and for information on monitoring the
performance of existing links through the V$DBLINK dynamic
performance view
ALTER DATABASE LINK on page 10-46 for information on
altering a database link when the password of a connection or
authentication user changes.
DROP DATABASE LINK on page 17-40 for information on
dropping existing database links
INSERT on page 18-54, UPDATE on page 19-73, DELETE on
page 17-26, and LOCK TABLE on page 18-71 for using links in
DML operations

Prerequisites
14

To create a private database link, you must have the CREATE DATABASE LINK system
privilege. To create a public database link, you must have the CREATE PUBLIC DATABASE
LINK system privilege. Also, you must have the CREATE SESSION system privilege on
the remote Oracle Database.
Oracle Net must be installed on both the local and remote Oracle Databases.

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-31

CREATE DATABASE LINK

Syntax
14

create_database_link::=
SHARED

PUBLIC

CREATE

DATABASE

LINK

dblink

CURRENT_USER
CONNECT

TO

dblink_authentication
user

IDENTIFIED

BY

password

dblink_authentication

USING

connect_string
;

(dblink::= on page 3-118)
dblink_authentication::=
AUTHENTICATED

BY

user

IDENTIFIED

BY

password

Semantics
14

PUBLIC
Specify PUBLIC to create a public database link visible to all users. If you omit this
clause, then the database link is private and is available only to you.
The data accessible on the remote database depends on the identity the database link
uses when connecting to the remote database:
■

■

■

If you specify CONNECT TO user IDENTIFIED BY password, then the database link
connects with the specified user and password.
If you specify CONNECT TO CURRENT_USER, then the database link connects with the
user in effect based on the scope in which the link is used.
If you omit both of those clauses, then the database link connects to the remote
database as the locally connected user.
See Also:

"Defining a Public Database Link: Example" on page 14-34

SHARED
Specify SHARED to create a database link that can be shared by multiple sessions using a
single network connection from the source database to the target database. In a shared
server configuration, shared database links can keep the number of connections into
the remote database from becoming too large. Shared links are typically also public
database links. However, a shared private database link can be useful when many
clients access the same local schema, and therefore use the same private database link.
In a shared database link, multiple sessions in the source database share the same
connection to the target database. Once a session is established on the target database,
that session is disassociated from the connection, to make the connection available to
another session on the source database. To prevent an unauthorized session from
14-32 Oracle Database SQL Language Reference

CREATE DATABASE LINK

attempting to connect through the database link, when you specify SHARED you must
also specify the dblink_authentication clause for the users authorized to use the
database link.
See Also: Oracle Database Administrator's Guide for more information
about shared database links

dblink
Specify the complete or partial name of the database link. If you specify only the
database name, then Oracle Database implicitly appends the database domain of the
local database.
Use only ASCII characters for dblink. Multibyte characters are not supported. The
database link name is case insensitive and is stored in uppercase ASCII characters. If
you specify the database name as a quoted identifier, then the quotation marks are
silently ignored.
If the value of the GLOBAL_NAMES initialization parameter is TRUE, then the database link
must have the same name as the database to which it connects. If the value of GLOBAL_
NAMES is FALSE, and if you have changed the global name of the database, then you can
specify the global name.
The maximum number of database links that can be open in one session or one
instance of an Oracle RAC configuration depends on the value of the OPEN_LINKS and
OPEN_LINKS_PER_INSTANCE initialization parameters.
Restriction on Creating Database Links You cannot create a database link in another
user's schema, and you cannot qualify dblink with the name of a schema. Periods are
permitted in names of database links, so Oracle Database interprets the entire name,
such as ralph.linktosales, as the name of a database link in your schema rather than
as a database link named linktosales in the schema ralph.)
See Also:
■

■

■

"References to Objects in Remote Databases" on page 3-117 for
guidelines for naming database links
Oracle Database Reference for information on the GLOBAL_NAMES,
OPEN_LINKS, and OPEN_LINKS_PER_INSTANCE initialization
parameters
"RENAME GLOBAL_NAME Clause" on page 10-40 (an ALTER
DATABASE clause) for information on changing the database global
name

CONNECT TO Clause
The CONNECT TO clause lets you specify the user and credentials, if any, to be used to
connect to the remote database.
CURRENT_USER Clause
Specify CURRENT_USER to create a current user database link. The current user must be
a global user with a valid account on the remote database.
If the database link is used directly rather than from within a stored object, then the
current user is the same as the connected user.
When executing a stored object (such as a procedure, view, or trigger) that initiates a
database link, CURRENT_USER is the name of the user that owns the stored object, and

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-33

CREATE DATABASE LINK

not the name of the user that called the object. For example, if the database link
appears inside procedure scott.p (created by scott), and user jane calls procedure
scott.p, then the current user is scott.
However, if the stored object is an invoker-rights function, procedure, or package, then
the invoker's authorization ID is used to connect as a remote user. For example, if the
privileged database link appears inside procedure scott.p (an invoker-rights
procedure created by scott), and user Jane calls procedure scott.p, then CURRENT_
USER is jane and the procedure executes with Jane's privileges.
See Also:
■

■

CREATE FUNCTION on page 14-58 for more information on
invoker-rights functions
"Defining a CURRENT_USER Database Link: Example" on
page 14-35

user IDENTIFIED BY password
Specify the user name and password used to connect to the remote database using a
fixed user database link. If you omit this clause, then the database link uses the user
name and password of each user who is connected to the database. This is called a
connected user database link.
See Also:

"Defining a Fixed-User Database Link: Example" on

page 14-35

dblink_authentication
You can specify this clause only if you are creating a shared database link—that is, you
have specified the SHARED clause. Specify the username and password on the target
instance. This clause authenticates the user to the remote server and is required for
security. The specified username and password must be a valid username and
password on the remote instance. The username and password are used only for
authentication. No other operations are performed on behalf of this user.

USING 'connect string'
Specify the service name of a remote database. If you specify only the database name,
then Oracle Database implicitly appends the database domain to the connect string to
create a complete service name. Therefore, if the database domain of the remote
database is different from that of the current database, then you must specify the
complete service name.
See Also: Oracle Database Administrator's Guide for information on
specifying remote databases

Examples
14

The examples that follow assume two databases, one with the database name local
and the other with the database name remote. The examples use the Oracle Database
domain. Your database domain will be different.
The following statement defines a shared
public database link named remote that refers to the database specified by the service
name remote:
Defining a Public Database Link: Example

CREATE PUBLIC DATABASE LINK remote
USING 'remote';

14-34 Oracle Database SQL Language Reference

CREATE DATABASE LINK

This database link allows user hr on the local database to update a table on the
remote database (assuming hr has appropriate privileges):
UPDATE employees@remote
SET salary=salary*1.1
WHERE last_name = 'Baer';

Defining a Fixed-User Database Link: Example In the following statement, user hr

on the remote database defines a fixed-user database link named local to the hr
schema on the local database:
CREATE DATABASE LINK local
CONNECT TO hr IDENTIFIED BY password
USING 'local';

After this database link is created, hr can query tables in the schema hr on the local
database in this manner:
SELECT * FROM employees@local;

User hr can also use DML statements to modify data on the local database:
INSERT INTO employees@local
(employee_id, last_name, email, hire_date, job_id)
VALUES (999, 'Claus', 'sclaus@example.com', SYSDATE, 'SH_CLERK');
UPDATE jobs@local SET min_salary = 3000
WHERE job_id = 'SH_CLERK';
DELETE FROM employees@local
WHERE employee_id = 999;

Using this fixed database link, user hr on the remote database can also access tables
owned by other users on the same database. This statement assumes that user hr has
SELECT privileges on the oe.customers table. The statement connects to the user hr on
the local database and then queries the oe.customers table:
SELECT * FROM oe.customers@local;

The following statement
defines a current-user database link to the remote database, using the entire service
name as the link name:

Defining a CURRENT_USER Database Link: Example

CREATE DATABASE LINK remote.us.example.com
CONNECT TO CURRENT_USER
USING 'remote';

The user who issues this statement must be a global user registered with the LDAP
directory service.
You can create a synonym to hide the fact that a particular table is on the remote
database. The following statement causes all future references to emp_table to access
the employees table owned by hr on the remote database:
CREATE SYNONYM emp_table
FOR oe.employees@remote.us.example.com;

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-35

CREATE DIMENSION

CREATE DIMENSION
Purpose
14

Use the CREATE DIMENSION statement to create a dimension. A dimension defines a
parent-child relationship between pairs of column sets, where all the columns of a
column set must come from the same table. However, columns in one column set
(called a level) can come from a different table than columns in another set. The
optimizer uses these relationships with materialized views to perform query rewrite.
The SQL Access Advisor uses these relationships to recommend creation of specific
materialized views.
Oracle Database does not automatically validate the
relationships you declare when creating a dimension. To validate the
relationships specified in the hierarchy_clause and the dimension_
join_clause of CREATE DIMENSION, you must run the DBMS_
OLAP.VALIDATE_DIMENSION procedure.

Note:

See Also:
■

■

CREATE MATERIALIZED VIEW on page 15-4 for more
information on materialized views
Oracle Database Performance Tuning Guide for more information on
query rewrite, the optimizer and the SQL Access Advisor

Prerequisites
14

To create a dimension in your own schema, you must have the CREATE DIMENSION
system privilege. To create a dimension in another user's schema, you must have the
CREATE ANY DIMENSION system privilege. In either case, you must have the SELECT
object privilege on any objects referenced in the dimension.

Syntax
14

create_dimension::=

schema
CREATE

hierarchy_clause

.

DIMENSION

dimension

level_clause

attribute_clause

;

extended_attribute_clause

level_clause::=
level_table
LEVEL

level

.

level_column

IS

SKIP

,
(

level_table

.

level_column

14-36 Oracle Database SQL Language Reference

)

WHEN

NULL

CREATE DIMENSION

hierarchy_clause::=
dimension_join_clause
HIERARCHY

hierarchy

(

child_level

CHILD

OF

parent_level

)

dimension_join_clause::=
child_key_column
JOIN

KEY

,
(

REFERENCES

child_key_column

parent_level

)

attribute_clause::=
dependent_column
ATTRIBUTE

level

DETERMINES

,
(

dependent_column

)

extended_attribute_clause::=
dependent_column
ATTRIBUTE

attribute

LEVEL

level

DETERMINES

,
(

dependent_column

)

Semantics
14

schema
Specify the schema in which the dimension will be created. If you do not specify
schema, then Oracle Database creates the dimension in your own schema.

dimension
Specify the name of the dimension. The name must be unique within its schema and
satisfy the requirements listed in "Database Object Naming Rules" on page 3-111.

level_clause
The level_clause defines a level in the dimension. A level defines dimension
hierarchies and attributes.
level

Specify the name of the level.

level_table . level_column Specify the columns in the level. You can specify up to 32
columns. The tables you specify in this clause must already exist.

Specify this clause to indicate that if the specified level is NULL,
then the level is to be skipped. This clause lets you preserve the hierarchical chain of
parent-child relationship by an alternative path that skips over the specified level. See
hierarchy_clause on page 14-38.

SKIP WHEN NULL

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-37

CREATE DIMENSION

Restrictions on Dimension Level Columns

Dimension level columns are subject to

the following restrictions:
■
■

All of the columns in a level must come from the same table.
If columns in different levels come from different tables, then you must specify the
dimension_join_clause.

■

The set of columns you specify must be unique to this level.

■

The columns you specify cannot be specified in any other dimension.

■

Each level_column must be non-null unless the level is specified with SKIP WHEN
NULL. The non-null columns need not have NOT NULL constraints. The column for
which you specify SKIP WHEN NULL cannot have a NOT NULL constraint).

hierarchy_clause
The hierarchy_clause defines a linear hierarchy of levels in the dimension. Each
hierarchy forms a chain of parent-child relationships among the levels in the
dimension. Hierarchies in a dimension are independent of each other. They may, but
need not, have columns in common.
Each level in the dimension should be specified at most once in this clause, and each
level must already have been named in the level_clause.
hierarchy Specify the name of the hierarchy. This name must be unique in the
dimension.
child_level Specify the name of a level that has an n:1 relationship with a parent

level. The level_columns of child_level cannot be null, and each child_level value
uniquely determines the value of the next named parent_level.
If the child level_table is different from the parent level_table, then you must
specify a join relationship between them in the dimension_join_clause.
parent_level Specify the name of a level.

dimension_join_clause
The dimension_join_clause lets you specify an inner equijoin relationship for a
dimension whose columns are contained in multiple tables. This clause is required and
permitted only when the columns specified in the hierarchy are not all in the same
table.
child_key_column
Specify one or more columns that are join-compatible with columns in the parent level.
If you do not specify the schema and table of each child_column, then the schema and
table are inferred from the CHILD OF relationship in the hierarchy_clause. If you do
specify the schema and column of a child_key_column, then the schema and table
must match the schema and table of columns in the child of parent_level in the
hierarchy_clause.
parent_level
Specify the name of a level.
Restrictions on Join Dimensions

restrictions:

14-38 Oracle Database SQL Language Reference

Join dimensions are subject to the following

CREATE DIMENSION

■

■

■
■

■
■

■

You can specify only one dimension_join_clause for a given pair of levels in the
same hierarchy.
The child_key_columns must be non-null, and the parent key must be unique and
non-null. You need not define constraints to enforce these conditions, but queries
may return incorrect results if these conditions are not true.
Each child key must join with a key in the parent_level table.
Self-joins are not permitted. The child_key_columns cannot be in the same table as
parent_level.
All of the child key columns must come from the same table.
The number of child key columns must match the number of columns in parent_
level, and the columns must be joinable.
You cannot specify multiple child key columns unless the parent level consists of
multiple columns.

attribute_clause
The attribute_clause lets you specify the columns that are uniquely determined by a
hierarchy level. The columns in level must all come from the same table as the
dependent_columns. The dependent_columns need not have been specified in the
level_clause.
For example, if the hierarchy levels are city, state, and country, then city might
determine mayor, state might determine governor, and country might determine
president.

extended_attribute_clause
This clause lets you specify an attribute name for one or more level-to-column
relations. The type of attribute you create with this clause is not different from the type
of attribute created using the attribute_clause. The only difference is that this clause
lets you assign a name to the attribute that is different from the level name.

Examples
14

This statement was used to create the customers_
dim dimension in the sample schema sh:

Creating a Dimension: Examples

CREATE DIMENSION customers_dim
LEVEL customer
IS (customers.cust_id)
LEVEL city
IS (customers.cust_city)
LEVEL state
IS (customers.cust_state_province)
LEVEL country
IS (countries.country_id)
LEVEL subregion IS (countries.country_subregion)
LEVEL region
IS (countries.country_region)
HIERARCHY geog_rollup (
customer
CHILD OF
city
CHILD OF
state
CHILD OF
country
CHILD OF
subregion
CHILD OF
region
JOIN KEY (customers.country_id) REFERENCES country
)
ATTRIBUTE customer DETERMINES
(cust_first_name, cust_last_name, cust_gender,

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-39

CREATE DIMENSION

cust_marital_status, cust_year_of_birth,
cust_income_level, cust_credit_limit)
ATTRIBUTE country DETERMINES (countries.country_name)
;

Alternatively, the
extended_attribute_clause could have been used instead of the attribute_clause,
as shown in the following example:
Creating a Dimension with Extended Attributes: Example

CREATE DIMENSION customers_dim
LEVEL customer
IS (customers.cust_id)
LEVEL city
IS (customers.cust_city)
LEVEL state
IS (customers.cust_state_province)
LEVEL country
IS (countries.country_id)
LEVEL subregion IS (countries.country_subregion)
LEVEL region
IS (countries.country_region)
HIERARCHY geog_rollup (
customer
CHILD OF
city
CHILD OF
state
CHILD OF
country
CHILD OF
subregion
CHILD OF
region
JOIN KEY (customers.country_id) REFERENCES country
)
ATTRIBUTE customer_info LEVEL customer DETERMINES
(cust_first_name, cust_last_name, cust_gender,
cust_marital_status, cust_year_of_birth,
cust_income_level, cust_credit_limit)
ATTRIBUTE country DETERMINES (countries.country_name);

Creating a Dimension with NULL Column Values: Example The following example
shows how to create the dimension if one of the level columns is null and you want to
preserve the hierarchical chain. The example uses the cust_marital_status column
for simplicity because it is not a NOT NULL column. If it had such a constraint, then you
would have to disable the constraint before using the SKIP WHEN NULL clause.
CREATE DIMENSION customers_dim
LEVEL customer IS (customers.cust_id)
LEVEL status IS (customers.cust_marital_status) SKIP WHEN NULL
LEVEL city IS (customers.cust_city)
LEVEL state IS (customers.cust_state_province)
LEVEL country IS (countries.country_id)
LEVEL subregion IS (countries.country_subregion) SKIP WHEN NULL
LEVEL region IS (countries.country_region)
HIERARCHY geog_rollup (
customer CHILD OF
city CHILD OF
state CHILD OF
country CHILD OF
subregion CHILD OF
region
JOIN KEY (customers.country_id) REFERENCES country
)
ATTRIBUTE customer DETERMINES
(cust_first_name, cust_last_name, cust_gender,
cust_marital_status, cust_year_of_birth,
cust_income_level, cust_credit_limit)
ATTRIBUTE country DETERMINES (countries.country_name)
;

14-40 Oracle Database SQL Language Reference

CREATE DIRECTORY

CREATE DIRECTORY
Purpose
14

Use the CREATE DIRECTORY statement to create a directory object. A directory object
specifies an alias for a directory on the server file system where external binary file
LOBs (BFILEs) and external table data are located. You can use directory names when
referring to BFILEs in your PL/SQL code and OCI calls, rather than hard coding the
operating system path name, for management flexibility.
All directories are created in a single namespace and are not owned by an individual
schema. You can secure access to the BFILEs stored within the directory structure by
granting object privileges on the directories to specific users.
See Also:
■

■

■

"Large Object (LOB) Data Types" on page 3-24 for more
information on BFILE objects
GRANT on page 18-33 for more information on granting object
privileges
external_table_clause::= of CREATE TABLE on page 16-17

Prerequisites
14

You must have the CREATE ANY DIRECTORY system privilege to create directories.
When you create a directory, you are automatically granted the READ, WRITE, and
EXECUTE object privileges on the directory, and you can grant these privileges to other
users and roles. The DBA can also grant these privileges to other users and roles.
WRITE privileges on a directory are useful in connection with external tables. They let
the grantee determine whether the external table agent can write a log file or a bad file
to the directory.
For file storage, you must also create a corresponding operating system directory, an
Oracle Automatic Storage Management (Oracle ASM) disk group, or a directory
within an Oracle ASM disk group. Your system or database administrator must ensure
that the operating system directory has the correct read and write permissions for
Oracle Database processes.
Privileges granted for the directory are created independently of the permissions
defined for the operating system directory, and the two may or may not correspond
exactly. For example, an error occurs if sample user hr is granted READ privilege on the
directory object but the corresponding operating system directory does not have READ
permission defined for Oracle Database processes.

Syntax
14

create_directory::=
OR
CREATE

REPLACE
DIRECTORY

directory

AS

’

path_name

’

;

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-41

CREATE DIRECTORY

Semantics
14

OR REPLACE
Specify OR REPLACE to re-create the directory database object if it already exists. You
can use this clause to change the definition of an existing directory without dropping,
re-creating, and regranting database object privileges previously granted on the
directory.
Users who had previously been granted privileges on a redefined directory can still
access the directory without being regranted the privileges.
DROP DIRECTORY on page 17-42 for information on
removing a directory from the database

See Also:

directory
Specify the name of the directory object to be created. The maximum length of
directory is 30 bytes. You cannot qualify a directory object with a schema name.
Oracle Database does not verify that the directory you specify actually exists.
Therefore, take care that you specify a valid directory in your operating system. In
addition, if your operating system uses case-sensitive path names, then be sure you
specify the directory in the correct format. You need not include a trailing slash at the
end of the path name.
Do not refer to a parent directory in the directory name. For example, the following
syntax is valid:
CREATE DIRECTORY mydir AS '/scratch/data/file_data';

However, the following syntax is not valid:
CREATE DIRECTORY mydir AS '/scratch/../file_data';

path_name
Specify the full path name of the operating system directory of the server where the
files are located. The single quotation marks are required, with the result that the path
name is case sensitive.

Examples
14

The following statement creates a directory database
object that points to a directory on the server:

Creating a Directory: Examples

CREATE DIRECTORY admin AS '/disk1/oracle/admin';

The following statement redefines directory database object bfile_dir to enable
access to BFILEs stored in the operating system directory /usr/bin/bfile_dir:
CREATE OR REPLACE DIRECTORY bfile_dir AS '/usr/bin/bfile_dir';

14-42 Oracle Database SQL Language Reference

CREATE DISKGROUP

CREATE DISKGROUP
This SQL statement is valid only if you are using Oracle ASM
and you have started an Oracle ASM instance. You must issue this
statement from within the Oracle ASM instance, not from a normal
database instance. For information on starting an Oracle ASM
instance, refer to Oracle Automatic Storage Management Administrator's
Guide.

Note:

Purpose
14

Use the CREATE DISKGROUP clause to name a group of disks and specify that Oracle
Database should manage the group for you. Oracle Database manages a disk group as
a logical unit and evenly spreads each file across the disks to balance I/O. Oracle
Database also automatically distributes database files across all available disks in disk
groups and rebalances storage automatically whenever the storage configuration
changes.
This statement creates a disk group, assigns one or more disks to the disk group, and
mounts the disk group for the first time. Note that CREATE DISKGROUP only mounts a
disk group on the local node. If you want Oracle ASM to mount the disk group
automatically in subsequent instances, then you must add the disk group name to the
value of the ASM_DISKGROUPS initialization parameter in the initialization parameter
file. If you use an SPFILE, then the disk group is added to the initialization parameter
automatically.
See Also:
■

■

■

■

■

ALTER DISKGROUP on page 10-51 for information on modifying
disk groups
Oracle Automatic Storage Management Administrator's Guide for
information on Oracle ASM and using disk groups to simplify
database administration
ASM_DISKGROUPS for more information about adding disk group
names to the initialization parameter file
V$ASM_OPERATION for information on monitoring Oracle ASM
operations
DROP DISKGROUP on page 17-43 for information on dropping a
disk group

Prerequisites
14

You must have the SYSASM system privilege to issue this statement.
Before issuing this statement, you must format the disks using an operating system
format utility. Also ensure that the Oracle Database user has read/write permission
and the disks can be discovered using the ASM_DISKSTRING.
When you store your database files in Oracle ASM disk groups, rather than in a file
system or on raw devices, before the database instance can access your files in the disk
groups, you must configure and start up an Oracle ASM instance to manage the disk
groups.

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-43

CREATE DISKGROUP

Each database instance communicates with a single Oracle ASM instance on the same
node as the database. Multiple database instances on the same node can communicate
with a single Oracle ASM instance.

Syntax
14

create_diskgroup::=
HIGH
NORMAL

REDUNDANCY

EXTERNAL
CREATE

DISKGROUP

diskgroup_name

QUORUM
REGULAR

FAILGROUP

,

failgroup_name
DISK

qualified_disk_clause

,
ATTRIBUTE

’

attribute_name

’

=

’

attribute_value

’
;

qualified_disk_clause::=
FORCE
NAME

disk_name

SIZE

size_clause

NOFORCE

search_string

(size_clause::= on page 8-47)

Semantics
14

diskgroup_name
Specify the name of the disk group. Disk groups are subject to the same naming
conventions and restrictions as database schema objects. Refer to "Database Object
Naming Rules" on page 3-111 for information on database object names. However,
disk groups are not schema objects. Disk group names are not case sensitive, even if
you specify them with quotation marks. They are always stored internally as
uppercase.

Oracle does not recommend using quoted identifiers for disk
group names. These quoted identifiers are accepted when issuing the
CREATE DISKGROUP statement in SQL*Plus, but they may not be valid
when using other tools that manage disk groups.
Note:

REDUNDANCY Clause
The REDUNDANCY clause lets you specify the redundancy level of the disk group.

14-44 Oracle Database SQL Language Reference

CREATE DISKGROUP

■

NORMAL REDUNDANCY requires the existence of at least two failure groups (see the
FAILGROUP clause that follows). Oracle ASM provides redundancy for all files in
the disk group according to the attributes specified in the disk group templates.
NORMAL REDUNDANCY disk groups can tolerate the loss of one group. Refer to ALTER
DISKGROUP ... diskgroup_template_clauses on page 10-63 for more information on disk
group templates.
NORMAL REDUNDANCY is the default if you omit the REDUNDANCY clause. Therefore, if
you omit this clause, you must create at least two failure groups, or the create
operation will fail.

■

■

HIGH REDUNDANCY requires the existence of at least three failure groups. Oracle ASM
fixes mirroring at 3-way mirroring, with each extent getting two mirrored copies.
HIGH REDUNDANCY disk groups can tolerate the loss of two failure groups.
EXTERNAL REDUNDANCY indicates that Oracle ASM does not provide any redundancy
for the disk group. The disks within the disk group must provide redundancy (for
example, using a storage array), or you must be willing to tolerate loss of the disk
group if a disk fails (for example, in a test environment). You cannot specify the
FAILGROUP clause if you specify EXTERNAL REDUNDANCY.

You cannot change the redundancy level after the disk group has been created.

QUORUM | REGULAR
Use these keywords to qualify either failure group or disk specifications.
■
■

REGULAR disks, or disks in non-quorum failure groups, can contain any files.
QUORUM disks, or disks in quorum failure groups, cannot contain any database files,
the Oracle Cluster Registry (OCR), or dynamic volumes. However, QUORUM disks
can contain the voting file for Cluster Synchronization Services (CSS). Oracle ASM
uses quorum disks or disks in quorum failure groups for voting files whenever
possible.
Disks in quorum failure groups are not considered when determining redundancy
requirements.

If you specify neither keyword, then REGULAR is the default.
Specify either QUORUM or REGULAR before the keyword FAILGROUP if you are explicitly
specifying the failure group. If you are creating a disk group with implicitly created
failure groups, then specify these keywords before the keyword DISK.
See Also: Oracle Automatic Storage Management Administrator's Guide
for more information about quorum and regular disks and failure
groups

FAILGROUP Clause
Use this clause to specify a name for one or more failure groups. If you omit this
clause, and you have specified NORMAL or HIGH REDUNDANCY, then Oracle Database
automatically adds each disk in the disk group to its own failure group. The implicit
name of the failure group is the same as the operating system independent disk name
(see "NAME Clause" on page 14-46).
You cannot specify this clause if you are creating an EXTERNAL REDUNDANCY disk group.

qualified_disk_clause
Specify DISK qualified_disk_clause to add a disk to a disk group.

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-45

CREATE DISKGROUP

search_string For each disk you are adding to the disk group, specify the operating
system dependent search string that Oracle ASM will use to find the disk. The search_
string must point to a subset of the disks returned by discovery using the strings in
the ASM_DISKSTRING initialization parameter. If search_string does not point to any
disks the Oracle Database user has read/write access to, then Oracle ASM returns an
error. If it points to one or more disks that have already been assigned to a different
disk group, then Oracle Database returns an error unless you also specify FORCE.

For each valid candidate disk, Oracle ASM formats the disk header to indicate that it is
a member of the new disk group.
See Also: The ASM_DISKSTRING initialization parameter for more
information on specifying the search string
NAME Clause The NAME clause is valid only if the search_string points to a single
disk. This clause lets you specify an operating system independent name for the disk.
The name can be up to 30 alphanumeric characters. The first character must be
alphabetic. If you omit this clause and you assigned a label to a disk through ASMLIB,
then that label is used as the disk name. If you omit this clause and you did not assign
a label through ASMLIB, then Oracle ASM creates a default name of the form
diskgroup_name_####, where #### is the disk number. You use this name to refer to the
disk in subsequent Oracle ASM operations.
SIZE Clause Use this clause to specify in bytes the size of the disk. If you specify a

size greater than the capacity of the disk, then Oracle ASM returns an error. If you
specify a size less than the capacity of the disk, then you limit the disk space Oracle
ASM will use. If you omit this clause, then Oracle ASM attempts programmatically to
determine the size of the disk.
Specify FORCE if you want Oracle ASM to add the disk to the disk group even
if the disk is already a member of a different disk group.

FORCE

Caution:

Using FORCE in this way may destroy existing disk groups.

For this clause to be valid, the disk must already be a member of a disk group and the
disk cannot be part of a mounted disk group.
Specify NOFORCE if you want Oracle ASM to return an error if the disk is
already a member of a different disk group. NOFORCE is the default.

NOFORCE

ATTRIBUTE Clause
Use this clause to set attribute values for the disk group. You can view the current
attribute values by querying the V$ASM_ATTRIBUTE view. Table 14–1 lists the attributes
you can set with this clause. All attribute values are strings.

14-46 Oracle Database SQL Language Reference

CREATE DISKGROUP

Table 14–1

Disk Group Attributes

Attribute

Valid Values

Description

ACCESS_
CONTROL.ENABLED

true or false

Specifies whether Oracle ASM File Access Control is enabled for a
disk group. If set to true, accessing Oracle ASM files is subject to
access control. If false, any user can access every file in the disk
group. All other operations behave independently of this attribute.
The default value is false.
If both the compatible.rdbms and compatible.asm attributes are set
to at least 11.2, you can set this attribute in an ALTER DISKGROUP ... SET
ATTRIBUTE statement. You cannot set this attribute when creating a
disk group.
When you set up file access control on an existing disk group, the files
previously created remain accessible by everyone, unless you run the
ALTER DISKGROUP SET PERMISSION statement to restrict the
permissions.
Note: This attribute is used in conjunction with ACCESS_
CONTROL.UMASK to manage Oracle ASM File Access Control. After
setting the ACCESS_CONTROL.ENABLED disk attribute, you must set
permissions with the ACCESS_CONTROL.UMASK attribute.

ACCESS_CONTROL.UMASK

A three-digit
number where
each digit is 0,
2, or 6.

Determines which permissions are masked out on the creation of an
Oracle ASM file for the user that owns the file (first digit), users in the
same user group (second digit), and others not in the user group
(third digit). This attribute applies to all files on a disk group. Setting
to 0 masks out nothing. Setting to 2 masks out write permission.
Setting to 6 masks out both read and write permissions. The default
value is 066.
If both the compatible.rdbms and compatible.asm attributes are set
to at least 11.2, you can set this attribute in an ALTER DISKGROUP ... SET
ATTRIBUTE statement. You cannot set this attribute when creating a
disk group.
When you set up file access control on an existing disk group, the files
previously created remain accessible by everyone, unless you run the
ALTER DISKGROUP SET PERMISSION statement to restrict the
permissions.
Note: This attribute is used in conjunction with ACCESS_
CONTROL.ENABLED to manage Oracle ASM File Access Control. Before
setting ACCESS_CONTROL.UMASK, you must set ACCESS_CONTROL.ENABLED
to true.

AU_SIZE

Size in bytes.
Specifies the allocation unit size. This attribute can be set only during
Valid values are disk group creation; it cannot be modified with an ALTER DISKGROUP
powers of 2
statement.
from 1M to
64M. Examples
'4M', '4194304'.

COMPATIBLE.ADVM

Valid Oracle
Database
version
number1

Determines whether the disk group can contain Oracle ADVM
volumes. The value must be set to 11.2 or higher. Before setting this
attribute, the COMPATIBLE.ASM value must be 11.2 or higher. Also, the
Oracle ADVM volume drivers must be loaded.
By default, the value of the COMPATIBLE.ADVM attribute is empty until
set.

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-47

CREATE DISKGROUP

Table 14–1 (Cont.) Disk Group Attributes
Attribute

Valid Values

Description

COMPATIBLE.ASM

Valid Oracle
Database
version
number1

Determines the minimum software version for an Oracle ASM
instance that can use the disk group. This setting also affects the
format of the data structures for the Oracle ASM metadata on the
disk.
For Oracle ASM in Oracle Database 11g, 10.1 is the default setting for
the COMPATIBLE.ASM attribute when using the SQL CREATE DISKGROUP
statement, the ASMCMD mkdg command, and Oracle Enterprise
Manager Create Disk Group page. When creating a disk group with
ASMCA, the default setting is 11.2.

14-48 Oracle Database SQL Language Reference

CREATE DISKGROUP

Table 14–1 (Cont.) Disk Group Attributes
Attribute

Valid Values

Description

COMPATIBLE.RDBMS

Valid Oracle
Database
version
number1

Determines the minimum COMPATIBLE database initialization
parameter setting for any database instance that is allowed to use the
disk group.
Before advancing the COMPATIBLE.RDBMS attribute, ensure that the
values for the COMPATIBLE initialization parameter for all of the
databases that access the disk group are set to at least the value of the
new setting for COMPATIBLE.RDBMS. For example, if the COMPATIBLE
initialization parameters of the databases are set to either 11.1 or 11.2,
then COMPATIBLE.RDBMS can be set to any value between 10.1 and 11.1
inclusively.
For Oracle ASM in Oracle Database 11g, 10.1 is the default setting for
the COMPATIBLE.RDBMS attribute when using the SQL CREATE
DISKGROUP statement, the ASMCMD mkdg command, ASMCA Create
Disk Group page, and Oracle Enterprise Manager Create Disk
Group page.

DISK_REPAIR_TIME

0 to 136 years

When disks are taken offline, Oracle ASM drops them after a default
period of time. If both the compatible.rdbms and compatible.asm
attributes are set to at least 11.1, you can set the disk_repair_time
attribute in an ALTER DISKGROUP ... SET ATTRIBUTE statement to change
that default period of time so that the disk can be repaired and
brought back online. You cannot set this attribute when creating a
disk group.
The time can be specified in units of minute (M) or hour (H). The
specified time elapses only when the disk group is mounted. If you
omit the unit, then the default is H. If you omit this attribute, and
both compatible.rdbms and compatible.asm are set to at least 11.1,
then the default is 3.6 H. Otherwise the disk is dropped immediately.
You can override this attribute with an ALTER DISKGROUP ... OFFLINE
DISK statement and the DROP AFTER clause.
Note: If a disk is taken offline using the current value of disk_
repair_time, and the value of this attribute is subsequently changed,
then the changed value is used by Oracle ASM in the disk offline
logic.
See Also: The ALTER DISKGROUP ... disk_offline_clause on page 10-61 and
Oracle Automatic Storage Management Administrator's Guide for more
information

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-49

CREATE DISKGROUP

Table 14–1 (Cont.) Disk Group Attributes
Attribute

Valid Values

Description

SECTOR_SIZE

Sector size of
the disks in the
disk group.
Valid values are
'512' '4096', and
'4K'.

All disks in the disk group must have a sector size equal to the
attribute value specified. When processing the CREATE DISKGROUP
statement, Oracle ASM queries the operating system for the sector
size of every disk specified in this statement before it is added to the
disk group, ensuring that a disk group is made up of disks with
identical sector size. If a disk is found to have a different sector size
than is specified for this attribute, then the statement fails. Similar
checks are performed when mounting the disk group. If you omit this
attribute in the CREATE DISKGROUP statement, then Oracle ASM
proceeds with the create operation as long as all specified disks are
found to have identical sector size, and that value is assumed as the
disk group sector size.
When new disks are added to an existing disk group, using the ALTER
DISKGROUP ... ADD DISK statement, the new disks must also have sector
size value identical to the disk group attribute. Oracle ASM verifies
this and the ALTER DISKGROUP statement fails if any of the disks to be
added are found to be of a different sector size.
By setting a value for this attribute, you can establish the sector size
you intend for the disk group, rather than letting Oracle ASM assume
a value for all the disks in the disk group. As a result, users can query
the SECTOR_SIZE column of the V$ASM_ATTRIBUTE view to determine
the intended sector size before attempting to add a new disk to the
disk group.

1

Specify at least the first two digits of a valid Oracle Database release number. Refer to Oracle Database Administrator's Guide for
information on specifying valid version numbers. For example, you can specify compatibility as '10.2' or '11.2'.

See Also: Oracle Automatic Storage Management Administrator's Guide
for more information on managing these attribute settings

Examples
14

The following example assumes that the ASM_DISKSTRING parameter is a superset of
/devices/disks/c*, /devices/disks/c* points to at least one device to be used as an
Oracle ASM disk, and the Oracle Database user has read/write permission to the
disks.
See Also: Oracle Automatic Storage Management Administrator's Guide
for information on Oracle ASM and using disk groups to simplify
database administration
Creating a Diskgroup: Example The following statement creates an Oracle ASM disk
group dgroup_01 where no redundancy for the disk group is provided by Oracle ASM
and includes all disks that match the search_string:
CREATE DISKGROUP dgroup_01
EXTERNAL REDUNDANCY
DISK '/devices/disks/c*';

14-50 Oracle Database SQL Language Reference

CREATE EDITION

CREATE EDITION
Purpose
14

This statement creates a new edition as a child of an existing edition. An edition makes
it possible to have two or more versions of the same editionable objects in the
database. When you create an edition, it immediately inherits all of the editionable
objects of its parent edition. The following object types are editionable:
■

Synonym

■

View

■

Function

■

Procedure

■

Package (specification and body)

■

Type (specification and body)

■

Library

■

Trigger

An editionable object is an object of one of the above editionable object types in an
editions-enabled schema. The ability to have multiple versions of these objects in the
database greatly facilitates online application upgrades.
Note: All database object types not listed above are not editionable.
Changes to object types that are not editionable are immediately
visible across all editions in the database.

Every newly created or upgraded Oracle Database has one default edition named
ORA$BASE, which serves as the parent of the first edition created with a CREATE EDITION
statement. You can subsequently designate a user-defined edition as the database
default edition using an ALTER DATABASE DEFAULT EDITION statement.
See Also:
■

■

Oracle Database Advanced Application Developer's Guide for a more
complete discussion of editionable object types and editions
The ALTER DATABASE "DEFAULT EDITION Clause" on page 10-38
for information on designating an edition as the default edition
for the database

Prerequisites
14

To create an edition, you must have the CREATE ANY EDITION system privilege, granted
either directly or through a role. To create an edition as a child of another edition, you
must have the USE object privilege on the parent edition.

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-51

CREATE EDITION

Syntax
14

create_edition::=
AS
CREATE

EDITION

CHILD

OF

parent_edition

edition

;

Semantics
14

edition
Specify the name of the edition to be created. The name must satisfy the rules listed in
"Database Object Naming Rules" on page 3-111. To learn the editions that have been
created for the database, query the EDITION_NAME column of the DBA_OBJECTS or ALL_
OBJECTS data dictionary view.
When you create an edition, the system automatically grants you the USE object
privilege WITH GRANT OPTION on the edition you create.
Oracle strongly recommends that you do not name editions
with the prefixes ORA, ORACLE, SYS, DBA, and DBMS, as these prefixes are
reserved for internal use.

Note:

AS CHILD OF Clause
If you use this clause, then the new edition is created as a child of parent_edition. If
you omit this clause, then the new edition is created as a child of the leaf edition. At
the time of its creation, the new edition inherits all editioned objects from its parent
edition.
An edition can have only one child edition. If you specify for
parent_edition an edition that already has a child edition, an error is returned.
Restriction on Editions

Examples
14

The following very simple examples are intended to show the syntax for creating and
working with an edition. For realistic examples of using editions refer to Oracle
Database Advanced Application Developer's Guide.
In the following statements, the user HR is given the privileges needed to create and
use an edition:
GRANT CREATE ANY EDITION, DROP ANY EDITION to HR;
Grant succeeded.
ALTER USER hr ENABLE EDITIONS;
User altered.

HR creates a new edition TEST_ED for testing purposes:
CREATE EDITION test_ed;

HR then creates an editioning view ed_view in the default edition ORA$BASE for testing
purposes, first verifying that the current edition is the default edition:
SELECT SYS_CONTEXT('userenv', 'current_edition_name') FROM DUAL;
SYS_CONTEXT('USERENV','CURRENT_EDITION_NAME')
--------------------------------------------------------------------------------

14-52 Oracle Database SQL Language Reference

CREATE EDITION

ORA$BASE
1 row selected.
CREATE EDITIONING VIEW e_view AS
SELECT last_name, first_name, email FROM employees;
View created.
DESCRIBE e_view
Name
----------------------------------------LAST_NAME
FIRST_NAME
EMAIL

Null?
Type
-------- ---------------------------NOT NULL VARCHAR2(25)
VARCHAR2(20)
NOT NULL VARCHAR2(25)

The view is then actualized in the TEST_ED edition when HR uses the TEST_ED edition
and re-creates the view in a different form:
ALTER SESSION SET EDITION = TEST_ED;
Session altered.
CREATE OR REPLACE EDITIONING VIEW e_view AS
SELECT last_name, first_name, email, salary FROM employees;
View created.

The view in the TEST_ED edition has an additional column:
DESCRIBE e_view
Name
----------------------------------------LAST_NAME
FIRST_NAME
EMAIL
SALARY

Null?
Type
-------- ---------------------------NOT NULL VARCHAR2(25)
VARCHAR2(20)
NOT NULL VARCHAR2(25)
NUMBER(8,2)

The view in the ORA$BASE edition remains isolated from the test environment:
ALTER SESSION SET EDITION = ora$base;
Session altered.
DESCRIBE e_view;
Name
----------------------------------------LAST_NAME
FIRST_NAME
EMAIL

Null?
Type
-------- ---------------------------NOT NULL VARCHAR2(25)
VARCHAR2(20)
NOT NULL VARCHAR2(25)

Even if the view is dropped in the test environment, it remains in the ORA$BASE edition:
ALTER SESSION SET EDITION = TEST_ED;
Session altered.
DROP VIEW e_view;
View dropped.
ALTER SESSION SET EDITION = ORA$BASE;
Session altered.
DESCRIBE e_view;
Name
Null?
Type
----------------------------------------- -------- ---------------------------SQL Statements: CREATE CLUSTER to CREATE JAVA

14-53

CREATE EDITION

LAST_NAME
FIRST_NAME
EMAIL

NOT NULL VARCHAR2(25)
VARCHAR2(20)
NOT NULL VARCHAR2(25)

When the testing of upgrade that necessitated the TEST_ED edition is complete, the
edition can be dropped:
DROP EDITION TEST_ED;

14-54 Oracle Database SQL Language Reference

CREATE FLASHBACK ARCHIVE

CREATE FLASHBACK ARCHIVE
Purpose
14

Use the CREATE FLASHBACK ARCHIVE statement to create a flashback data archive, which
provides the ability to automatically track and archive transactional data changes to
specified database objects. A flashback data archive consists of multiple tablespaces
and stores historic data from all transactions against tracked tables. The data is stored
in internal history tables.
Flashback data archives retain historical data for the time duration specified using the
RETENTION parameter. Historical data can be queried using the Flashback Query AS OF
clause. Archived historic data that has aged beyond the specified retention period is
automatically purged.
Flashback data archives retain historical data across data definition language (DDL)
changes to tables enabled for flashback data archive. Flashback data archives supports
many common DDL statements, including some DDL statements that alter table
definitions or incur data movement. DDL statements that are not supported result in
error ORA-55610.
See Also:
■

■

■

Oracle Database Advanced Application Developer's Guide for general
information on using flashback data archives
The CREATE TABLE flashback_archive_clause on page 16-66 for
information on designating a table as a tracked table
ALTER FLASHBACK ARCHIVE on page 10-74 for information on
changing the quota and retention attributes of the flashback data
archive, as well as adding or changing tablespace storage for the
flashback data archive

Prerequisites
14

You must have the FLASHBACK ARCHIVE ADMINISTER system privilege to create a
flashback data archive. In addition, you must have the CREATE TABLESPACE system
privilege to create a flashback data archive, as well as sufficient quota on the
tablespace in which the historical information will reside. To designate a flashback
data archive as the system default flashback data archive, you must be logged in as
SYSDBA.

Syntax
14

create_flashback_archive::=
DEFAULT
CREATE

FLASHBACK

ARCHIVE

flashback_archive

TABLESPACE

tablespace

NO
flashback_archive_quota

OPTIMIZE

DATA
flashback_archive_retention

;

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-55

CREATE FLASHBACK ARCHIVE

flashback_archive_quota::=
M
G
QUOTA

integer

T
P
E

flashback_archive_retention::=
YEAR
RETENTION

integer

MONTH
DAY

Semantics
14

DEFAULT
You must be logged in as SYSDBA to specify DEFAULT. Use this clause to designate this
flashback data archive as the default flashback data archive for the database. When a
CREATE TABLE or ALTER TABLE statement specifies the flashback_archive_clause
without specifying a flashback data archive name, the database uses the default
flashback data archive to store data from that table.
You cannot specify this clause if a default flashback data archive already exists.
However, you can replace an existing default flashback data archive using the ALTER
FLASHBACK ARCHIVE ... SET DEFAULT clause.
See Also: The CREATE TABLE flashback_archive_clause on page 16-66
for more information

flashback_archive
Specify the name of the flashback data archive. The name must satisfy the
requirements specified in "Database Object Naming Rules" on page 3-111.

TABLESPACE Clause
Specify the tablespace where the archived data for this flashback data archive is to be
stored. You can specify only one tablespace with this clause. However, you can
subsequently add tablespaces to the flashback data archive with an ALTER FLASHBACK
ARCHIVE statement.

flashback_archive_quota
Specify the amount of space in the initial tablespace to be reserved for the archived
data. If the space for archiving in a flashback data archive becomes full, then DML
operations on tracked tables that use this flashback data archive will fail. The database
issues an out-of-space alert when the content of the flashback data archive is 90% of
the specified quota, to allow time to purge old data or add additional quota. If you
omit this clause, then the flashback data archive has unlimited quota on the specified
tablespace.

14-56 Oracle Database SQL Language Reference

CREATE FLASHBACK ARCHIVE

[NO] OPTIMIZE DATA
Specify OPTIMIZE DATA to instruct the database to optimize the storage of data in
history tables using any of the following features: Advanced Row Compression,
Advanced LOB Compression, Advanced LOB Deduplication, segment-level
compression tiering, and row-level compression tiering.
Specify NO OPTIMIZE DATA to instruct the database not to optimize the storage of data in
history tables. This is the default.
The [NO] OPTIMIZE DATA clause is available starting with
Oracle Database 11g Release 2 (11.2.0.4).

Note:

flashback_archive_retention
Specify the length of time in months, days, or years that the archived data should be
retained in the flashback data archive. If the length of time causes the flashback data
archive to become full, then the database responds as described in flashback_archive_
quota on page 14-56.

Examples
14

The following statement creates two flashback data archives for testing purposes. The
first is designated as the default for the database. For both of them, the space quota is 1
megabyte, and the archive retention is one day.
CREATE FLASHBACK ARCHIVE DEFAULT test_archive1
TABLESPACE example
QUOTA 1 M
RETENTION 1 DAY;
CREATE FLASHBACK ARCHIVE test_archive2
TABLESPACE example
QUOTA 1 M
RETENTION 1 DAY;

The next statement alters the default flashback data archive to extend the retention
period to 1 month:
ALTER FLASHBACK ARCHIVE test_archive1
MODIFY RETENTION 1 MONTH;

The next statement specifies tracking for the oe.customers table. The flashback data
archive is not specified, so data will be archived in the default flashback data archive,
test_archive1:
ALTER TABLE oe.customers
FLASHBACK ARCHIVE;

The next statement specifies tracking for the oe.orders table. In this case, data will be
archived in the specified flashback data archive, test_archive2:
ALTER TABLE oe.orders
FLASHBACK ARCHIVE test_archive2;

The next statement drops test_archive2 flashback data archive:
DROP FLASHBACK ARCHIVE test_archive2;

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-57

CREATE FUNCTION

CREATE FUNCTION
Purpose
14

Functions are defined using PL/SQL. Therefore, this section provides some general
information but refers to Oracle Database PL/SQL Language Reference for details of
syntax and semantics.
Use the CREATE FUNCTION statement to create a standalone stored function or a call
specification.
■

■

A stored function (also called a user function or user-defined function) is a set of
PL/SQL statements you can call by name. Stored functions are very similar to
procedures, except that a function returns a value to the environment in which it is
called. User functions can be used as part of a SQL expression.
A call specification declares a Java method or a third-generation language (3GL)
routine so that it can be called from PL/SQL. You can also use the CALL SQL
statement to call such a method or routine. The call specification tells Oracle
Database which Java method, or which named function in which shared library, to
invoke when a call is made. It also tells the database what type conversions to
make for the arguments and return value.
You can also create a function as part of a package using the
CREATE PACKAGE statement.
Note:

See Also:
■

■

■

CREATE PROCEDURE on page 15-48 for a general discussion of
procedures and functions, CREATE PACKAGE on page 15-42 for
information on creating packages, ALTER FUNCTION on
page 10-77 and DROP FUNCTION on page 17-48 for information
on modifying and dropping a function
CREATE LIBRARY on page 15-2 for information on shared
libraries
Oracle Database Advanced Application Developer's Guide for more
information about registering external functions

Prerequisites
14

To create or replace a function in your own schema, you must have the CREATE
PROCEDURE system privilege. To create or replace a function in another user's schema,
you must have the CREATE ANY PROCEDURE system privilege.

Syntax
14

Functions are defined using PL/SQL. Therefore, the syntax diagram in this book
shows only the SQL keywords. Refer to Oracle Database PL/SQL Language Reference for
the PL/SQL syntax, semantics, and examples.
create_function::=
OR
CREATE

REPLACE
FUNCTION

plsql_source

14-58 Oracle Database SQL Language Reference

CREATE FUNCTION

(plsql_source: See Oracle Database PL/SQL Language Reference.)

Semantics
14

OR REPLACE
Specify OR REPLACE to re-create the function if it already exists. Use this clause to
change the definition of an existing function without dropping, re-creating, and
regranting object privileges previously granted on the function. If you redefine a
function, then Oracle Database recompiles it.
Users who had previously been granted privileges on a redefined function can still
access the function without being regranted the privileges.
If any function-based indexes depend on the function, then Oracle Database marks the
indexes DISABLED.
See Also: ALTER FUNCTION for information on recompiling functions
using SQL

plsql_source
See Oracle Database PL/SQL Language Reference for the syntax and semantics of the
plsql_source, including examples.

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-59

CREATE INDEX

CREATE INDEX
Purpose
14

Use the CREATE INDEX statement to create an index on:
■

One or more columns of a table, a partitioned table, an index-organized table, or a
cluster

■

One or more scalar typed object attributes of a table or a cluster

■

A nested table storage table for indexing a nested table column

An index is a schema object that contains an entry for each value that appears in the
indexed column(s) of the table or cluster and provides direct, fast access to rows.
Oracle Database supports several types of index:
■

Normal indexes. (By default, Oracle Database creates B-tree indexes.)

■

Bitmap indexes, which store rowids associated with a key value as a bitmap.

■

■

■

Partitioned indexes, which consist of partitions containing an entry for each value
that appears in the indexed column(s) of the table.
Function-based indexes, which are based on expressions. They enable you to
construct queries that evaluate the value returned by an expression, which in turn
may include built-in or user-defined functions.
Domain indexes, which are instances of an application-specific index of type
indextype.
See Also:
■

Oracle Database Concepts for a discussion of indexes

■

ALTER INDEX on page 10-78 and DROP INDEX on page 17-50

Prerequisites
14

To create an index in your own schema, one of the following conditions must be true:
■

The table or cluster to be indexed must be in your own schema.

■

You must have the INDEX object privilege on the table to be indexed.

■

You must have the CREATE ANY INDEX system privilege.

To create an index in another schema, you must have the CREATE ANY INDEX system
privilege. Also, the owner of the schema to contain the index must have either the
UNLIMITED TABLESPACE system privilege or space quota on the tablespaces to contain
the index or index partitions.
To create a domain index in your own schema, in addition to the prerequisites for
creating a conventional index, you must also have the EXECUTE object privilege on the
indextype. If you are creating a domain index in another user's schema, then the index
owner also must have the EXECUTE object privilege on the indextype and its underlying
implementation type. Before creating a domain index, you should first define the
indextype.
To create a function-based index, in addition to the prerequisites for creating a
conventional index, if the index is based on user-defined functions, then those
functions must be marked DETERMINISTIC. Also, you must have the EXECUTE object

14-60 Oracle Database SQL Language Reference

CREATE INDEX

privilege on any user-defined function(s) used in the function-based index if those
functions are owned by another user.
CREATE INDEXTYPE on page 14-87

See Also:

Syntax
14

create_index::=
UNIQUE
BITMAP

schema

CREATE

cluster_index_clause
ON

.

INDEX

index

UNUSABLE

table_index_clause

;

bitmap_join_index_clause

cluster_index_clause::=
schema

.

CLUSTER

cluster

index_attributes

(index_attributes::= on page 14-62)
table_index_clause::=
,
ASC
schema

.

t_alias

DESC

table

(

index_properties

index_expr

)

(index_properties::= on page 14-62)
bitmap_join_index_clause::=
,
schema

.
table

schema

.

t_alias
table

.

ASC

.

DESC

(

column

)

,
schema
FROM

.

t_alias
table

local_partitioned_index
WHERE

condition

index_attributes

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-61

CREATE INDEX

(local_partitioned_index::= on page 14-65, index_attributes::= on page 14-62)
index_expr::=
column
column_expression

index_properties::=
global_partitioned_index
local_partitioned_index
index_attributes
domain_index_clause
INDEXTYPE

IS
XMLIndex_clause

(global_partitioned_index::= on page 14-64, local_partitioned_index::= on page 14-65,
index_attributes::= on page 14-62, domain_index_clause::= on page 14-63, XMLIndex_
clause::= on page 14-63)
index_attributes::=
physical_attributes_clause
logging_clause
ONLINE
tablespace
TABLESPACE
DEFAULT
key_compression
SORT
NOSORT
REVERSE
VISIBLE
INVISIBLE
parallel_clause

(physical_attributes_clause::= on page 14-63, logging_clause::= on page 14-63, key_
compression::= on page 14-63, parallel_clause::= on page 14-66)

14-62 Oracle Database SQL Language Reference

CREATE INDEX

physical_attributes_clause::=
PCTFREE

integer

PCTUSED

integer

INITRANS

integer

storage_clause

(storage_clause::= on page 8-50)
logging_clause::=
LOGGING
NOLOGGING
FILESYSTEM_LIKE_LOGGING

key_compression::=
integer
COMPRESS
NOCOMPRESS

domain_index_clause::=
local_domain_index_clause

parallel_clause

PARAMETERS

(

’

ODCI_parameters

’

)

indextype

(parallel_clause::= on page 14-66)
local_domain_index_clause::=
,
PARAMETERS
(

PARTITION

(

’

ODCI_parameters

partition

’

)
)

LOCAL

XMLIndex_clause::=
XDB

.

local_XMLIndex_clause

parallel_clause

XMLIndex_parameters_clause

XMLINDEX

( The XMLIndex_parameters_clause is documented in Oracle XML DB Developer's
Guide.)

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-63

CREATE INDEX

local_XMLIndex_clause::=
,
XMLIndex_parameters_clause
(

PARTITION

partition

)

LOCAL

global_partitioned_index::=
RANGE
GLOBAL

PARTITION

(

column_list

)

(

BY

index_partitioning_clause

)

individual_hash_partitions
HASH

(

column_list

)
hash_partitions_by_quantity

(index_partitioning_clause::= on page 14-65, individual_hash_partitions::= on page 14-64,
hash_partitions_by_quantity::= on page 14-65)
individual_hash_partitions::=
,
partition
(

partitioning_storage_clause

PARTITION

)

(partitioning_storage_clause::= on page 14-64)
partitioning_storage_clause::=
TABLESPACE

tablespace
TABLESPACE

tablespace

OVERFLOW
table_compression
key_compression
LOB_partitioning_storage
SECUREFILE
BASICFILE
VARRAY

varray_item

STORE

AS

LOB

LOB_segname

LOB_partitioning_storage::=
LOB

(

LOB_item

)

(

STORE

BASICFILE

LOB_segname

SECUREFILE

(

AS

14-64 Oracle Database SQL Language Reference

TABLESPACE

TABLESPACE

tablespace

)

tablespace

)

CREATE INDEX

hash_partitions_by_quantity::=
,
STORE
PARTITIONS

IN

(

tablespace

)

hash_partition_quantity

table_compression

,

key_compression

OVERFLOW

STORE

IN

(

tablespace

)

index_partitioning_clause::=
,

partition
PARTITION

VALUES

LESS

THAN

(

literal

segment_attributes_clause
)

(segment_attributes_clause::= on page 14-66)
local_partitioned_index::=
on_range_partitioned_table
on_list_partitioned_table
on_hash_partitioned_table
on_comp_partitioned_table
LOCAL

(on_range_partitioned_table::= on page 14-65, on_list_partitioned_table::= on page 14-65,
on_hash_partitioned_table::= on page 14-66, on_comp_partitioned_table::= on page 14-66)
on_range_partitioned_table::=
,
segment_attributes_clause
partition
(

key_compression

UNUSABLE

PARTITION

)

(segment_attributes_clause::= on page 14-66)
on_list_partitioned_table::=
,
segment_attributes_clause
partition
(

key_compression

PARTITION

UNUSABLE
)

(segment_attributes_clause::= on page 14-66)

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-65

CREATE INDEX

segment_attributes_clause::=
physical_attributes_clause
TABLESPACE

tablespace

logging_clause

(physical_attributes_clause::= on page 14-63, logging_clause::= on page 14-63
on_hash_partitioned_table::=
,
STORE

IN

(

tablespace

)
,

partition
(

TABLESPACE

tablespace

key_compression

UNUSABLE

PARTITION

)

on_comp_partitioned_table::=
,
STORE

IN

(

tablespace

)

,
segment_attributes_clause
partition
(

key_compression

UNUSABLE

index_subpartition_clause

PARTITION

)

(segment_attributes_clause::= on page 14-66, index_subpartition_clause::= on page 14-66)
index_subpartition_clause::=
,
STORE

IN

(

tablespace

)
,

subpartition
(

TABLESPACE

SUBPARTITION

parallel_clause::=
NOPARALLEL
integer
PARALLEL

(storage_clause::= on page 8-50)

14-66 Oracle Database SQL Language Reference

tablespace

key_compression

UNUSABLE
)

CREATE INDEX

Semantics
14

UNIQUE
Specify UNIQUE to indicate that the value of the column (or columns) upon which the
index is based must be unique.
Restrictions on Unique Indexes Unique indexes are subject to the following
restrictions:
■

You cannot specify both UNIQUE and BITMAP.

■

You cannot specify UNIQUE for a domain index.
See Also: "Unique Constraints" on page 8-9 for information on the
conditions that satisfy a unique constraint

BITMAP
Specify BITMAP to indicate that index is to be created with a bitmap for each distinct
key, rather than indexing each row separately. Bitmap indexes store the rowids
associated with a key value as a bitmap. Each bit in the bitmap corresponds to a
possible rowid. If the bit is set, then it means that the row with the corresponding
rowid contains the key value. The internal representation of bitmaps is best suited for
applications with low levels of concurrent transactions, such as data warehousing.
Oracle does not index table rows in which all key columns
are null except in the case of bitmap indexes. Therefore, if you want
an index on all rows of a table, then you must either specify NOT
NULL constraints for the index key columns or create a bitmap
index.
Note:

Restrictions on Bitmap Indexes Bitmap indexes are subject to the following
restrictions:
■
■

You cannot specify BITMAP when creating a global partitioned index.
You cannot create a bitmap secondary index on an index-organized table unless
the index-organized table has a mapping table associated with it.

■

You cannot specify both UNIQUE and BITMAP.

■

You cannot specify BITMAP for a domain index.

■

A bitmap index can have a maximum of 30 columns.
See Also:
■

Oracle Database Concepts and Oracle Database Performance Tuning
Guide for more information about using bitmap indexes

■

CREATE TABLE on page 16-6 for information on mapping tables

■

"Bitmap Index Examples" on page 14-84

schema
Specify the schema to contain the index. If you omit schema, then Oracle Database
creates the index in your own schema.

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-67

CREATE INDEX

index
Specify the name of the index to be created. The name must satisfy the requirements
listed in "Database Object Naming Rules" on page 3-111.
See Also: "Creating an Index: Example" on page 14-80 and "Creating
an Index on an XMLType Table: Example" on page 14-81

cluster_index_clause
Use the cluster_index_clause to identify the cluster for which a cluster index is to be
created. If you do not qualify cluster with schema, then Oracle Database assumes the
cluster is in your current schema. You cannot create a cluster index for a hash cluster.
See Also: CREATE CLUSTER on page 14-2 and "Creating a Cluster
Index: Example" on page 14-81

table_index_clause
Specify the table on which you are defining the index. If you do not qualify table with
schema, then Oracle Database assumes the table is contained in your own schema.
You create an index on a nested table column by creating the index on the nested table
storage table. Include the NESTED_TABLE_ID pseudocolumn of the storage table to
create a UNIQUE index, which effectively ensures that the rows of a nested table value
are distinct.
See Also:

"Indexes on Nested Tables: Example" on page 14-85

You can perform DDL operations (such as ALTER TABLE, DROP TABLE, CREATE INDEX) on
a temporary table only when no session is bound to it. A session becomes bound to a
temporary table by performing an INSERT operation on the table. A session becomes
unbound to the temporary table by issuing a TRUNCATE statement or at session
termination, or, for a transaction-specific temporary table, by issuing a COMMIT or
ROLLBACK statement.
Restrictions on the table_index_clause This clause is subject to the following
restrictions:
■
■

■

■

If index is locally partitioned, then table must be partitioned.
If table is index-organized, then this statement creates a secondary index. The
index contains the index key and the logical rowid of the index-organized table.
The logical rowid excludes columns that are also part of the index key. You cannot
specify REVERSE for this secondary index, and the combined size of the index key
and the logical rowid should be less than the block size.
If table is a temporary table, then index will also be temporary with the same
scope (session or transaction) as table. The following restrictions apply to indexes
on temporary tables:
–

The only part of index_properties you can specify is index_attributes.

–

Within index_attributes, you cannot specify the physical_attributes_
clause, the parallel_clause, the logging_clause, or TABLESPACE.

–

You cannot create a domain index or a partitioned index on a temporary table.

You cannot create an index on an external table.
See Also: CREATE TABLE on page 16-6 and Oracle Database Concepts
for more information on temporary tables

14-68 Oracle Database SQL Language Reference

CREATE INDEX

t_alias
Specify a correlation name (alias) for the table upon which you are building the index.
Note: This alias is required if the index_expr references any object
type attributes or object type methods. See "Creating a Function-based
Index on a Type Method: Example" on page 14-82 and "Indexing on
Substitutable Columns: Examples" on page 14-85.

index_expr
For index_expr, specify the column or column expression upon which the index is
based.
Specify the name of one or more columns in the table. A bitmap index can
have a maximum of 30 columns. Other indexes can have as many as 32 columns. These
columns define the index key.

column

If a unique index is local nonprefixed (see local_partitioned_index), then the index key
must contain the partitioning key.
See Also: Oracle Database VLDB and Partitioning Guide for
information on prefixed and nonprefixed indexes

You can create an index on a scalar object attribute column or on the system-defined
NESTED_TABLE_ID column of the nested table storage table. If you specify an object
attribute column, then the column name must be qualified with the table name. If you
specify a nested table column attribute, then it must be qualified with the outermost
table name, the containing column name, and all intermediate attribute names leading
to the nested table column attribute.
Restrictions on Index Columns
■

■

The following restrictions apply to index columns:

You cannot create an index on columns or attributes whose type is user-defined,
LONG, LONG RAW, LOB, or REF, except that Oracle Database supports an index on REF
type columns or attributes that have been defined with a SCOPE clause.
Only normal (B-tree) indexes can be created on encrypted columns, and they can
only be used for equality searches.

column_expression Specify an expression built from columns of table, constants,

SQL functions, and user-defined functions. When you specify column_expression,
you create a function-based index.
"Column Expressions" on page 6-6, "Notes on
Function-based Indexes" on page 14-70, "Restrictions on
Function-based Indexes" on page 14-71, and "Function-Based Index
Examples" on page 14-81

See Also:

Name resolution of the function is based on the schema of the index creator.
User-defined functions used in column_expression are fully name resolved during the
CREATE INDEX operation.
After creating a function-based index, collect statistics on both the index and its base
table using the DBMS_STATS package. Such statistics will enable Oracle Database to
correctly decide when to use the index.

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-69

CREATE INDEX

Function-based unique indexes can be useful in defining a conditional unique
constraint on a column or combination of columns. Refer to "Using a Function-based
Index to Define Conditional Uniqueness: Example" on page 14-82 for an example.
Oracle Database PL/SQL Packages and Types Reference for
more information on the DBMS_STATS package

See Also:

Notes on Function-based Indexes

The following notes apply to function-based

indexes:
■

When you subsequently query a table that uses a function-based index, Oracle
Database will not use the index unless the query filters out nulls. However, Oracle
Database will use a function-based index in a query even if the columns specified
in the WHERE clause are in a different order than their order in the column_
expression that defined the function-based index.
See Also:

■

■

■

■

■

"Function-Based Index Examples" on page 14-81

If the function on which the index is based becomes invalid or is dropped, then
Oracle Database marks the index DISABLED. Queries on a DISABLED index fail if the
optimizer chooses to use the index. DML operations on a DISABLED index fail
unless the index is also marked UNUSABLE and the parameter SKIP_UNUSABLE_
INDEXES is set to true. Refer to ALTER SESSION on page 11-45 for more
information on this parameter.
If a public synonym for a function, package, or type is used in column_expression,
and later an actual object with the same name is created in the table owner's
schema, then Oracle Database disables the function-based index. When you
subsequently enable the function-based index using ALTER INDEX ... ENABLE or
ALTER INDEX ... REBUILD, the function, package, or type used in the column_
expression continues to resolve to the function, package, or type to which the
public synonym originally pointed. It will not resolve to the new function,
package, or type.
If the definition of a function-based index generates internal conversion to
character data, then use caution when changing NLS parameter settings.
Function-based indexes use the current database settings for NLS parameters. If
you reset these parameters at the session level, then queries using the
function-based index may return incorrect results. Two exceptions are the collation
parameters (NLS_SORT and NLS_COMP). Oracle Database handles the conversions
correctly even if these have been reset at the session level.
Oracle Database cannot convert data in all cases, even when conversion is
explicitly requested. For example, an attempt to convert the string '105 lbs' from
VARCHAR2 to NUMBER using the TO_NUMBER function fails with an error. Therefore, if
column_expression contains a data conversion function such as TO_NUMBER or TO_
DATE, and if a subsequent INSERT or UPDATE statement includes data that the
conversion function cannot convert, then the index will cause the INSERT or
UPDATE statement to fail.
If column_expression contains a datetime format model, then the function-based
index expression defining the column may contain format elements that are
different from those specified. For example, define a function-based index using
the yyyy datetime format element:
CREATE INDEX cust_eff_ix ON customers
(NVL(cust_eff_to, TO_DATE('9000-01-01 00:00:00', 'yyyy-mm-dd hh24:mi:ss')));

14-70 Oracle Database SQL Language Reference

CREATE INDEX

Query the ALL_IND_EXPRESSIONS view to see that the function-based index
expression defining the column uses the syyyy datetime format element:
SELECT column_expression
FROM all_ind_expressions
WHERE index_name='CUST_EFF_IX';
COLUMN_EXPRESSION
-----------------------------------------------------------------------------NVL("CUST_EFF_TO",TO_DATE(' 9000-01-01 00:00:00', 'syyyy-mm-dd hh24:mi:ss'))

Restrictions on Function-based Indexes Function-based indexes are subject to the

following restrictions:
■

■

■

■

■

■

The value returned by the function referenced in column_expression is subject to
the same restrictions as are the index columns of a B-tree index. Refer to
"Restrictions on Index Columns" on page 14-69.
Any user-defined function referenced in column_expression must be declared as
DETERMINISTIC.
For a function-based globally partitioned index, the column_expression cannot be
the partitioning key.
The column_expression can be any of the forms of expression described in
Column Expressions on page 6-6.
All functions must be specified with parentheses, even if they have no parameters.
Otherwise Oracle Database interprets them as column names.
Any function you specify in column_expression must return a repeatable value.
For example, you cannot specify the SYSDATE or USER function or the ROWNUM
pseudocolumn.
CREATE FUNCTION on page 14-58 and Oracle Database
PL/SQL Language Reference

See Also:

ASC | DESC
Use ASC or DESC to indicate whether the index should be created in ascending or
descending order. Indexes on character data are created in ascending or descending
order of the character values in the database character set.
Oracle Database treats descending indexes as if they were function-based indexes. As
with other function-based indexes, the database does not use descending indexes until
you first analyze the index and the table on which the index is defined. See the
column_expression clause of this statement.
Ascending unique indexes allow multiple NULL values. However, in descending
unique indexes, multiple NULL values are treated as duplicate values and therefore are
not permitted.
You cannot specify either of
these clauses for a domain index. You cannot specify DESC for a reverse index. Oracle
Database ignores DESC if index is bitmapped or if the COMPATIBLE initialization
parameter is set to a value less than 8.1.0.
Restriction on Ascending and Descending Indexes

index_attributes
Specify the optional index attributes.

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-71

CREATE INDEX

Use the physical_attributes_clause to establish
values for physical and storage characteristics for the index.

physical_attributes_clause

If you omit this clause, then Oracle Database sets PCTFREE to 10 and INITRANS to 2.
Restriction on Index Physical Attributes You cannot specify the PCTUSED parameter

for an index.
See Also: physical_attributes_clause on page 8-44 and storage_clause on
page 8-46 for a complete description of these clauses
TABLESPACE For tablespace, specify the name of the tablespace to hold the index,
index partition, or index subpartition. If you omit this clause, then Oracle Database
creates the index in the default tablespace of the owner of the schema containing the
index.

For a local index, you can specify the keyword DEFAULT in place of tablespace. New
partitions or subpartitions added to the local index will be created in the same
tablespace(s) as the corresponding partitions or subpartitions of the underlying table.
key_compression Specify COMPRESS to enable key compression, which eliminates
repeated occurrence of key column values and may substantially reduce storage. Use
integer to specify the prefix length (number of prefix columns to compress).

Oracle Database compresses indexes that are nonunique or unique indexes of at least
two columns. If you want to use compression for a partitioned index, then you must
create the index with compression enabled at the index level. You can subsequently
enable and disable the compression setting for individual partitions of such a
partitioned index. You can also enable and disable compression when rebuilding
individual partitions. You can modify an existing non-partitioned index to enable or
disable compression only when rebuilding the index.
■

■

For unique indexes, the valid range of prefix length values is from 1 to the number
of key columns minus 1. The default prefix length is the number of key columns
minus 1.
For nonunique indexes, the valid range of prefix length values is from 1 to the
number of key columns. The default prefix length is the number of key columns.

Restriction on Key Compression You cannot specify COMPRESS for a bitmap index.
See Also:
NOCOMPRESS

"Compressing an Index: Example" on page 14-80
Specify NOCOMPRESS to disable key compression. This is the default.

SORT | NOSORT By default, Oracle Database sorts indexes in ascending order when
it creates the index. You can specify NOSORT to indicate to the database that the rows
are already stored in the database in ascending order, so that Oracle Database does not
have to sort the rows when creating the index. If the rows of the indexed column or
columns are not stored in ascending order, then the database returns an error. For
greatest savings of sort time and space, use this clause immediately after the initial
load of rows into a table. If you specify neither of these keywords, then SORT is the
default.
Restrictions on NOSORT

This parameter is subject to the following restrictions:

■

You cannot specify REVERSE with this clause.

■

You cannot use this clause to create a cluster index partitioned or bitmap index.

14-72 Oracle Database SQL Language Reference

CREATE INDEX

■

You cannot specify this clause for a secondary index on an index-organized table.

Specify REVERSE to store the bytes of the index block in reverse order,
excluding the rowid.

REVERSE

Restrictions on Reverse Indexes Reverse indexes are subject to the following
restrictions:
■

You cannot specify NOSORT with this clause.

■

You cannot reverse a bitmap index or an index on an index-organized table.

Use this clause to specify whether the index is visible or
invisible to the optimizer. An invisible index is maintained by DML operations, but it
is not be used by the optimizer during queries unless you explicitly set the parameter
OPTIMIZER_USE_INVISIBLE_INDEXES to TRUE at the session or system level.
VISIBLE | INVISIBLE

To determine whether an existing index is visible or invisible to the optimizer, you can
query the VISIBILITY column of the USER_, DBA_, ALL_INDEXES data dictionary views.
See Also: Oracle Database Administrator's Guide for more information
on this feature
logging_clause Specify whether the creation of the index will be logged (LOGGING) or
not logged (NOLOGGING) in the redo log file. This setting also determines whether
subsequent Direct Loader (SQL*Loader) and direct-path INSERT operations against the
index are logged or not logged. LOGGING is the default.

If index is nonpartitioned, then this clause specifies the logging attribute of the index.
If index is partitioned, then this clause determines:
■

■
■

The default value of all partitions specified in the CREATE statement, unless you
specify the logging_clause in the PARTITION description clause
The default value for the segments associated with the index partitions
The default value for local index partitions or subpartitions added implicitly
during subsequent ALTER TABLE ... ADD PARTITION operations

The logging attribute of the index is independent of that of its base table.
If you omit this clause, then the logging attribute is that of the tablespace in which it
resides.
See Also:
■
■

■

logging_clause on page 8-38 for a full description of this clause
Oracle Database VLDB and Partitioning Guide for more information
about logging and parallel DML
"Creating an Index in NOLOGGING Mode: Example" on
page 14-80

Specify ONLINE to indicate that DML operations on the table will be allowed
during creation of the index.

ONLINE

Restrictions on Online Index Building

Online index building is subject to the

following restrictions:

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-73

CREATE INDEX

■

Parallel DML is not supported during online index building. If you specify ONLINE
and then issue parallel DML statements, then Oracle Database returns an error.

■

You cannot specify ONLINE for a bitmap index or a cluster index.

■

You cannot specify ONLINE for a conventional index on a UROWID column.

■

For a nonunique secondary index on an index-organized table, the number of
index key columns plus the number of primary key columns that are included in
the logical rowid in the index-organized table cannot exceed 32. The logical rowid
excludes columns that are part of the index key.
Oracle Database Concepts for a description of online index
building and rebuilding

See Also:

parallel_clause
Specify the parallel_clause if you want creation of the index to be parallelized.
For complete information on this clause, refer to parallel_clause on page 16-63 in the
documentation on CREATE TABLE.

Index Partitioning Clauses
Use the global_partitioned_index clause and the local_partitioned_index clauses
to partition index.
The storage of partitioned database entities in tablespaces of different block sizes is
subject to several restrictions. Refer to Oracle Database VLDB and Partitioning Guide for
a discussion of these restrictions.
See Also:

"Partitioned Index Examples" on page 14-83

global_partitioned_index
The global_partitioned_index clause lets you specify that the partitioning of the
index is user defined and is not equipartitioned with the underlying table. By default,
nonpartitioned indexes are global indexes.
You can partition a global index by range or by hash. In both cases, you can specify up
to 32 columns as partitioning key columns. The partitioning column list must specify a
left prefix of the index column list. If the index is defined on columns a, b, and c, then
for the columns you can specify (a, b, c), or (a, b), or (a, c), but you cannot specify (b,
c) or (c) or (b, a). If you omit the partition names, then Oracle Database assigns
names of the form SYS_Pn.
GLOBAL PARTITION BY RANGE Use this clause to create a range-partitioned global
index. Oracle Database will partition the global index on the ranges of values from the
table columns you specify in the column list.
See Also:

"Creating a Range-Partitioned Global Index: Example" on

page 14-83
GLOBAL PARTITION BY HASH Use this clause to create a hash-partitioned global
index. Oracle Database assigns rows to the partitions using a hash function on values
in the partitioning key columns.
See Also: The CREATE TABLE clause hash_partitions on page 16-55 for
information on the two methods of hash partitioning and "Creating a
Hash-Partitioned Global Index: Example" on page 14-83

14-74 Oracle Database SQL Language Reference

CREATE INDEX

Restrictions on Global Partitioned Indexes Global partitioned indexes are subject to
the following restrictions:
■

■

■

■

The partitioning key column list cannot contain the ROWID pseudocolumn or a
column of type ROWID.
The only property you can specify for hash partitions is tablespace storage.
Therefore, you cannot specify LOB or varray storage clauses in the partitioning_
storage_clause of individual_hash_partitions.
You cannot specify the OVERFLOW clause of hash_partitions_by_quantity, as that
clause is valid only for index-organized table partitions.
In the partitioning_storage_clause, you cannot specify table_compression, but
you can specify key_compression.
If your enterprise has or will have databases using different
character sets, then use caution when partitioning on character
columns. The sort sequence of characters is not identical in all
character sets.

Note:

See Also: Oracle Database Globalization Support Guide for more
information on character set support
index_partitioning_clause Use this clause to describe the individual index partitions.
The number of repetitions of this clause determines the number of partitions. If you
omit partition, then Oracle Database generates a name with the form SYS_Pn.

For VALUES LESS THAN (value_list), specify the noninclusive upper bound for the
current partition in a global index. The value list is a comma-delimited, ordered list of
literal values corresponding to the column list in the global_partitioned_index
clause. Always specify MAXVALUE as the value of the last partition.
If the index is partitioned on a DATE column, and if the date
format does not specify the first two digits of the year, then you must
use the TO_DATE function with a 4-character format mask for the year.
The date format is determined implicitly by NLS_TERRITORY or
explicitly by NLS_DATE_FORMAT. Refer to Oracle Database Globalization
Support Guide for more information on these initialization parameters.
Note:

See Also:

"Range Partitioning Example" on page 16-76

local_partitioned_index
The local_partitioned_index clauses let you specify that the index is partitioned on
the same columns, with the same number of partitions and the same partition bounds
as table. For composite-partitioned tables, this clause lets you specify that the index is
subpartitioned on the same columns, with the same number of subpartitions and the
same subpartition bounds as table. Oracle Database automatically maintains local
index partitioning as the underlying table is repartitioned.
If you specify only the keyword LOCAL and do not specify a subclause, then Oracle
Database creates each index partition in the same tablespace as its corresponding table
partition and assigns it the same name as its corresponding table partition. If table is a
composite-partitioned table, then Oracle Database creates each index subpartition in

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-75

CREATE INDEX

the same tablespace as its corresponding table subpartition and assigns it the same
name as its corresponding table subpartition.
This clause lets you specify the names and attributes of
index partitions on a range-partitioned table. If you specify this clause, then the
number of PARTITION clauses must be equal to the number of table partitions, and in
the same order. If you omit partition, then Oracle Database generates a name that is
consistent with the corresponding table partition. If the name conflicts with an existing
index partition name, then the database uses the form SYS_Pn.
on_range_partitioned_table

You cannot specify key compression for an index partition unless you have specified
key compression for the index.
For more information on the UNUSABLE clause, refer to UNUSABLE on page 14-79.
The on_list_partitioned_table clause is identical to on_
range_partitioned_table on page 14-76.

on_list_partitioned_table

This clause lets you specify names and tablespace
storage for index partitions on a hash-partitioned table.

on_hash_partitioned_table

If you specify any PARTITION clauses, then the number of these clauses must be equal
to the number of table partitions. If you omit partition, then Oracle Database
generates a name that is consistent with the corresponding table partition. If the name
conflicts with an existing index partition name, then the database uses the form SYS_
Pn. You can optionally specify tablespace storage for one or more individual partitions.
If you do not specify tablespace storage either here or in the STORE IN clause, then the
database stores each index partition in the same tablespace as the corresponding table
partition.
The STORE IN clause lets you specify one or more tablespaces across which Oracle
Database will distribute all the index hash partitions. The number of tablespaces need
not equal the number of index partitions. If the number of index partitions is greater
than the number of tablespaces, then the database cycles through the names of the
tablespaces.
For more information on the UNUSABLE clause, refer to UNUSABLE on page 14-79.
on_comp_partitioned_table This clause lets you specify the name and attributes of
index partitions on a composite-partitioned table.

The STORE IN clause is valid only for range-hash or list-hash composite-partitioned
tables. It lets you specify one or more default tablespaces across which Oracle
Database will distribute all index hash subpartitions for all partitions. You can
override this storage by specifying different default tablespace storage for the
subpartitions of an individual partition in the second STORE IN clause in the index_
subpartition_clause.
For range-range, range-list, and list-list composite-partitioned tables, you can specify
default attributes for the range or list subpartitions in the PARTITION clause. You can
override this storage by specifying different attributes for the range or list
subpartitions of an individual partition in the SUBPARTITION clause of the index_
subpartition_clause.
You cannot specify key compression for an index partition unless you have specified
key compression for the index.
For more information on the UNUSABLE clause, refer to UNUSABLE on page 14-79.

14-76 Oracle Database SQL Language Reference

CREATE INDEX

index_subpartition_clause This clause lets you specify names and tablespace storage

for index subpartitions in a composite-partitioned table.
The STORE IN clause is valid only for hash subpartitions of a range-hash and list-hash
composite-partitioned table. It lets you specify one or more tablespaces across which
Oracle Database will distribute all the index hash subpartitions. The SUBPARTITION
clause is valid for all subpartition types.
If you specify any SUBPARTITION clauses, then the number of those clauses must be
equal to the number of table subpartitions. If you omit subpartition, then the
database generates a name that is consistent with the corresponding table subpartition.
If the name conflicts with an existing index subpartition name, then the database uses
the form SYS_SUBPn.
The number of tablespaces need not equal the number of index subpartitions. If the
number of index subpartitions is greater than the number of tablespaces, then the
database cycles through the names of the tablespaces.
If you do not specify tablespace storage for subpartitions either in the on_comp_
partitioned_table clause or in the index_subpartition_clause, then Oracle
Database uses the tablespace specified for index. If you also do not specify tablespace
storage for index, then the database stores the subpartition in the same tablespace as
the corresponding table subpartition.
For more information on the UNUSABLE clause, refer to CREATE INDEX ... UNUSABLE on
page 14-79.
domain_index_clause
Use the domain_index_clause to indicate that index is a domain index, which is an
instance of an application-specific index of type indextype.
Creating a domain index requires a number of preceding operations. You must first
create an implementation type for an indextype. You must also create a functional
implementation and then create an operator that uses the function. Next you create an
indextype, which associates the implementation type with the operator. Finally, you
create the domain index using this clause. Refer to Appendix F, "Extended Examples",
which contains an example of creating a simple domain index, including all of these
operations.
In the index_expr (in table_index_clause), specify the table columns or
object attributes on which the index is defined. You can define multiple domain
indexes on a single column only if the underlying indextypes are different and the
indextypes support a disjoint set of user-defined operators.

index_expr

Restrictions on Domain Indexes Domain indexes are subject to the following
restrictions:
■

The index_expr (in table_index_clause) can specify only a single column, and
the column cannot be of data type REF, varray, nested table, LONG, or LONG RAW.

■

You cannot create a bitmap or unique domain index.

■

You cannot create a domain index on a temporary table.

■

You can create a local domain index only on a range-, list-, hash-, or
interval-partitioned table.

indextype For indextype, specify the name of the indextype. This name should be a
valid schema object that has already been created.

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-77

CREATE INDEX

If you have installed Oracle Text, then you can use various built-in indextypes to
create Oracle Text domain indexes. For more information on Oracle Text and the
indexes it uses, refer to Oracle Text Reference.
See Also:

CREATE INDEXTYPE on page 14-87

local_domain_index_clause Use this clause to specify that the index is a local index
on a partitioned table.
■

■

The PARTITIONS clause lets you specify names for the index partitions. The
number of partitions you specify must match the number of partitions in the base
table. If you omit this clause, then the database creates the partitions with
system-generated names of the form SYS_Pn.
The PARAMETERS clause lets you specify the parameter string specific to an
individual partition. If you omit this clause, then the parameter string associated
with the index is also associated with the partition.

parallel_clause Use the parallel_clause to parallelize creation of the domain
index. For a nonpartitioned domain index, Oracle Database passes the explicit or
default degree of parallelism to the ODCIIndexCreate cartridge routine, which in turn
establishes parallelism for the index. For local domain indexes, this clause causes the
index partitions to be created in parallel.
See Also: Oracle Database Data Cartridge Developer's Guide for
complete information on the Oracle Data Cartridge Interface (ODCI)
routines
PARAMETERS In the PARAMETERS clause, specify the parameter string that is passed
uninterpreted to the appropriate ODCI indextype routine. The maximum length of the
parameter string is 1000 characters.

When you specify this clause at the top level of the syntax, the parameters become the
default parameters for the index partitions. If you specify this clause as part of the
local_domain_index_clause, then you override any default parameters with
parameters for the individual partition.
After the domain index is created, Oracle Database invokes the appropriate ODCI
routine. If the routine does not return successfully, then the domain index is marked
FAILED. The only operations supported on an failed domain index are DROP INDEX and
(for non-local indexes) REBUILD INDEX.
See Also: Oracle Database Data Cartridge Developer's Guide for
information on the Oracle Data Cartridge Interface (ODCI) routines

XMLIndex_clause
The XMLIndex_clause lets you define an XMLIndex index, typically on a column
contain XML data. An XMLIndex index is a type of domain index designed specifically
for the domain of XML data.
XMLIndex_parameters_clause This clause lets you specify information about the
path table and about the secondary indexes corresponding to the components of
XMLIndex. This clause also lets you specify information about the structured
component of the index. The maximum length of the parameter string is 1000
characters.

When you specify this clause at the top level of the syntax, the parameters become the
parameters of the index and the default parameters for the index partitions. If you
14-78 Oracle Database SQL Language Reference

CREATE INDEX

specify this clause as part of the local_xmlindex_clause clause, then you override any
default parameters with parameters for the individual partition.
See Also: Oracle XML DB Developer's Guide for the syntax and
semantics of the XMLIndex_parameters_clause clause, as well as
detailed information about the use of XMLIndex

bitmap_join_index_clause
Use the bitmap_join_index_clause to define a bitmap join index. A bitmap join
index is defined on a single table. For an index key made up of dimension table
columns, it stores the fact table rowids corresponding to that key. In a data
warehousing environment, the table on which the index is defined is commonly
referred to as a fact table, and the tables with which this table is joined are commonly
referred to as dimension tables. However, a star schema is not a requirement for
creating a join index.
ON In the ON clause, first specify the fact table, and then inside the parentheses

specify the columns of the dimension tables on which the index is defined.
FROM
WHERE

In the FROM clause, specify the joined tables.
In the WHERE clause, specify the join condition.

If the underlying fact table is partitioned, then you must also specify one of the local_
partitioned_index clauses (see local_partitioned_index on page 14-75).
Restrictions on Bitmap Join Indexes In addition to the restrictions on bitmap
indexes in general (see BITMAP on page 14-67), the following restrictions apply to
bitmap join indexes:
■

You cannot create a bitmap join index on a temporary table.

■

No table may appear twice in the FROM clause.

■

You cannot create a function-based join index.

■

■

■

The dimension table columns must be either primary key columns or have unique
constraints.
If a dimension table has a composite primary key, then each column in the primary
key must be part of the join.
You cannot specify the local_partitioned_index clause unless the fact table is
partitioned.
See Also: Oracle Database Data Warehousing Guide for information on
fact and dimension tables and on using bitmap indexes in a data
warehousing environment

UNUSABLE
Specify UNUSABLE to create an index in an UNUSABLE state. An unusable index must be
rebuilt, or dropped and re-created, before it can be used.
If the index is partitioned, then all index partitions are marked UNUSABLE. You can then
subsequently choose to rebuild only some of the index partitions to make them
USABLE. Doing so can be useful if you want to maintain indexes only on some index
partitions—for example, if you want to enable index access for new partitions but not
for old partitions.

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-79

CREATE INDEX

When an index, or some partitions or subpartitions of an index, are created UNUSABLE,
no segment is allocated for the unusable object. The unusable index or index partition
consumes no space in the database.
If an index, or some partitions or subpartitions of the index, are marked UNUSABLE,
then the index will be considered as an access path by the optimizer only under the
following circumstances: the optimizer must know at compile time which partitions
are to be accessed, and all of those partitions to be accessed must be marked USABLE.
Therefore, the query cannot contain any bind variables.
Restrictions on UNUSABLE The following restrictions apply when marking an index

UNUSABLE:
■
■

You cannot specify this clause for an index on a temporary table.
Unusable indexes or index partitions will still have a segment under the following
conditions:
–

The index (or index partition) is owned by SYS, SYSTEM, PUBLIC, OUTLN, or XDB

–

The index (or index partition) is stored in dictionary-managed tablespaces

–

The global partitioned or nonpartitioned index on a partitioned table becomes
unusable due to a partition maintenance operation

Examples
14

General Index Examples
The following statement shows how the sample index
ord_customer_ix on the customer_id column of the sample table oe.orders was
created:
Creating an Index: Example

CREATE INDEX ord_customer_ix
ON orders (customer_id);

To create the ord_customer_ix_demo index with the
COMPRESS clause, you might issue the following statement:
Compressing an Index: Example

CREATE INDEX ord_customer_ix_demo
ON orders (customer_id, sales_rep_id)
COMPRESS 1;

The index will compress repeated occurrences of customer_id column values.
If the sample table orders had
been created using a fast parallel load (so all rows were already sorted), then you
could issue the following statement to quickly create an index.

Creating an Index in NOLOGGING Mode: Example

/* Unless you first sort the table oe.orders, this example fails
because you cannot specify NOSORT unless the base table is
already sorted.
*/
CREATE INDEX ord_customer_ix_demo
ON orders (order_mode)
NOSORT
NOLOGGING;

14-80 Oracle Database SQL Language Reference

CREATE INDEX

Creating a Cluster Index: Example To create an index for the personnel cluster,
which was created in "Creating a Cluster: Example" on page 14-7, issue the following
statement:
CREATE INDEX idx_personnel ON CLUSTER personnel;

No index columns are specified, because cluster indexes are automatically built on all
the columns of the cluster key. For cluster indexes, all rows are indexed.
Creating an Index on an XMLType Table: Example The following example creates an
index on the area element of the xwarehouses table (created in "XMLType Table
Examples" on page 16-75):
CREATE INDEX area_index ON xwarehouses e
(EXTRACTVALUE(VALUE(e),'/Warehouse/Area'));

Such an index would greatly improve the performance of queries that select from the
table based on, for example, the square footage of a warehouse, as shown in this
statement:
SELECT e.getClobVal() AS warehouse
FROM xwarehouses e
WHERE EXISTSNODE(VALUE(e),'/Warehouse[Area>50000]') = 1;

See Also:

EXISTSNODE on page 5-89 and VALUE on page 5-340

Function-Based Index Examples
The following examples show how to create and use function-based indexes.
Creating a Function-Based Index: Example The following statement creates a
function-based index on the employees table based on an uppercase evaluation of the
last_name column:
CREATE INDEX upper_ix ON employees (UPPER(last_name));

See the "Prerequisites" on page 14-60 for the privileges and parameter settings required
when creating function-based indexes.
To increase the likelihood that Oracle Database will use the index rather than
performing a full table scan, be sure that the value returned by the function is not null
in subsequent queries. For example, this statement will use the index, unless some
other condition exists that prevents the optimizer from doing so:
SELECT first_name, last_name
FROM employees WHERE UPPER(last_name) IS NOT NULL
ORDER BY UPPER(last_name);

Without the WHERE clause, Oracle Database may perform a full table scan.
In the next statements showing index creation and subsequent query, Oracle Database
will use index income_ix even though the columns are in reverse order in the query:
CREATE INDEX income_ix
ON employees(salary + (salary*commission_pct));
SELECT first_name||' '||last_name "Name"
FROM employees
WHERE (salary*commission_pct) + salary > 15000
ORDER BY employee_id;

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-81

CREATE INDEX

Creating a Function-Based Index on a LOB Column: Example The following

statement uses the text_length function to create a function-based index on a LOB
column in the sample pm schema. See Oracle Database PL/SQL Language Reference for the
example that creates this function. The example selects rows from the sample table
print_media where that CLOB column has fewer than 1000 characters.
CREATE INDEX src_idx ON print_media(text_length(ad_sourcetext));
SELECT product_id FROM print_media
WHERE text_length(ad_sourcetext) < 1000
ORDER BY product_id;
PRODUCT_ID
---------2056
2268
3060
3106

Creating a Function-based Index on a Type Method: Example This example entails
an object type rectangle containing two number attributes: length and width. The
area() method computes the area of the rectangle.
CREATE TYPE rectangle AS OBJECT
( length
NUMBER,
width
NUMBER,
MEMBER FUNCTION area RETURN NUMBER DETERMINISTIC
);
CREATE OR REPLACE TYPE BODY rectangle AS
MEMBER FUNCTION area RETURN NUMBER IS
BEGIN
RETURN (length*width);
END;
END;

Now, if you create a table rect_tab of type rectangle, you can create a function-based
index on the area() method as follows:
CREATE TABLE rect_tab OF rectangle;
CREATE INDEX area_idx ON rect_tab x (x.area());

You can use this index efficiently to evaluate a query of the form:
SELECT * FROM rect_tab x WHERE x.area() > 100;

The
following statement creates a unique function-based index on the oe.orders table that
prevents a customer from taking advantage of promotion ID 2 ("blowout sale") more
than once:
Using a Function-based Index to Define Conditional Uniqueness: Example

CREATE UNIQUE INDEX promo_ix ON orders
(CASE WHEN promotion_id =2 THEN customer_id ELSE NULL END,
CASE WHEN promotion_id = 2 THEN promotion_id ELSE NULL END);
INSERT INTO orders (order_id, order_date, customer_id, order_total, promotion_id)
VALUES (2459, systimestamp, 106, 251, 2);
1 row created.
INSERT INTO orders (order_id, order_date, customer_id, order_total, promotion_id)
VALUES (2460, systimestamp+1, 106, 110, 2);

14-82 Oracle Database SQL Language Reference

CREATE INDEX

insert into orders (order_id, order_date, customer_id, order_total, promotion_id)
*
ERROR at line 1:
ORA-00001: unique constraint (OE.PROMO_IX) violated

The objective is to remove from the index any rows where the promotion_id is not
equal to 2. Oracle Database does not store in the index any rows where all the keys are
NULL. Therefore, in this example, both customer_id and promotion_id are mapped to
NULL unless promotion_id is equal to 2. The result is that the index constraint is
violated only if promotion_id is equal to 2 for two rows with the same customer_id
value.
Partitioned Index Examples
The following statement
creates a global prefixed index cost_ix on the sample table sh.sales with three
partitions that divide the range of costs into three groups:
Creating a Range-Partitioned Global Index: Example

CREATE INDEX cost_ix ON sales (amount_sold)
GLOBAL PARTITION BY RANGE (amount_sold)
(PARTITION p1 VALUES LESS THAN (1000),
PARTITION p2 VALUES LESS THAN (2500),
PARTITION p3 VALUES LESS THAN (MAXVALUE));

Creating a Hash-Partitioned Global Index: Example The following statement creates
a hash-partitioned global index cust_last_name_ix on the sample table sh.customers
with four partitions:
CREATE INDEX cust_last_name_ix ON customers (cust_last_name)
GLOBAL PARTITION BY HASH (cust_last_name)
PARTITIONS 4;

The following statement
creates a local index on the category_id column of the hash_products partitioned
table (which was created in "Hash Partitioning Example" on page 16-79). The STORE IN
clause immediately following LOCAL indicates that hash_products is hash partitioned.
Oracle Database will distribute the hash partitions between the tbs1 and tbs2
tablespaces:
Creating an Index on a Hash-Partitioned Table: Example

CREATE INDEX prod_idx ON hash_products(category_id) LOCAL
STORE IN (tbs_01, tbs_02);

The creator of the index must have quota on the tablespaces specified. See CREATE
TABLESPACE on page 16-83 for examples that create tablespaces tbs_01 and tbs_02.
Creating an Index on a Composite-Partitioned Table: Example The following

statement creates a local index on the composite_sales table, which was created in
"Composite-Partitioned Table Examples" on page 16-79. The STORAGE clause specifies
default storage attributes for the index. However, this default is overridden for the five
subpartitions of partitions q3_2000 and q4_2000, because separate TABLESPACE storage
is specified.
The creator of the index must have quota on the tablespaces specified. See CREATE
TABLESPACE on page 16-83 for examples that create tablespaces tbs_02 and tbs_03.
CREATE INDEX sales_ix ON composite_sales(time_id, prod_id)
STORAGE (INITIAL 1M)
LOCAL
(PARTITION q1_1998,

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-83

CREATE INDEX

PARTITION q2_1998,
PARTITION q3_1998,
PARTITION q4_1998,
PARTITION q1_1999,
PARTITION q2_1999,
PARTITION q3_1999,
PARTITION q4_1999,
PARTITION q1_2000,
PARTITION q2_2000
(SUBPARTITION pq2001, SUBPARTITION pq2002,
SUBPARTITION pq2003, SUBPARTITION pq2004,
SUBPARTITION pq2005, SUBPARTITION pq2006,
SUBPARTITION pq2007, SUBPARTITION pq2008),
PARTITION q3_2000
(SUBPARTITION c1 TABLESPACE tbs_02,
SUBPARTITION c2 TABLESPACE tbs_02,
SUBPARTITION c3 TABLESPACE tbs_02,
SUBPARTITION c4 TABLESPACE tbs_02,
SUBPARTITION c5 TABLESPACE tbs_02),
PARTITION q4_2000
(SUBPARTITION pq4001 TABLESPACE tbs_03,
SUBPARTITION pq4002 TABLESPACE tbs_03,
SUBPARTITION pq4003 TABLESPACE tbs_03,
SUBPARTITION pq4004 TABLESPACE tbs_03)
);

Bitmap Index Examples
The following creates a bitmap index on the table oe.hash_products, which was
created in "Hash Partitioning Example" on page 16-79:
CREATE BITMAP INDEX product_bm_ix
ON hash_products(list_price)
LOCAL(PARTITION ix_p1 TABLESPACE tbs_01,
PARTITION ix_p2,
PARTITION ix_p3 TABLESPACE tbs_02,
PARTITION ix_p4,
PARTITION ix_p5 TABLESPACE tbs_03)
TABLESPACE tbs_04;

Because hash_products is a partitioned table, the bitmap join index must be locally
partitioned. In this example, the user must have quota on tablespaces specified. See
CREATE TABLESPACE on page 16-83 for examples that create tablespaces tbs_01,
tbs_02, tbs_03, and tbs_04.
The next series of statements shows how one might create a bitmap join index on a fact
table using a join with a dimension table.
CREATE TABLE hash_products
( product_id
NUMBER(6)
, product_name
VARCHAR2(50)
, product_description VARCHAR2(2000)
, category_id
NUMBER(2)
, weight_class
NUMBER(1)
, warranty_period
INTERVAL YEAR TO MONTH
, supplier_id
NUMBER(6)
, product_status
VARCHAR2(20)
, list_price
NUMBER(8,2)
, min_price
NUMBER(8,2)
, catalog_url
VARCHAR2(50)
, CONSTRAINT
pk_product_id PRIMARY KEY (product_id)

14-84 Oracle Database SQL Language Reference

CREATE INDEX

, CONSTRAINT

product_status_lov_demo
CHECK (product_status in ('orderable'
,'planned'
,'under development'
,'obsolete')

) )
PARTITION BY HASH (product_id)
PARTITIONS 5
STORE IN (example);
CREATE TABLE sales_quota
( product_id
NUMBER(6)
, customer_name
VARCHAR2(50)
, order_qty
NUMBER(6)
,CONSTRAINT u_product_id UNIQUE(product_id)
);
CREATE BITMAP INDEX product_bm_ix
ON hash_products(list_price)
FROM hash_products h, sales_quota s
WHERE h.product_id = s.product_id
LOCAL(PARTITION ix_p1 TABLESPACE example,
PARTITION ix_p2,
PARTITION ix_p3 TABLESPACE example,
PARTITION ix_p4,
PARTITION ix_p5 TABLESPACE example)
TABLESPACE example;

Indexes on Nested Tables: Example
The sample table pm.print_media contains a nested table column ad_textdocs_ntab,
which is stored in storage table textdocs_nestedtab. The following example creates a
unique index on storage table textdocs_nestedtab:
CREATE UNIQUE INDEX nested_tab_ix
ON textdocs_nestedtab(NESTED_TABLE_ID, document_typ);

Including pseudocolumn NESTED_TABLE_ID ensures distinct rows in nested table
column ad_textdocs_ntab.
Indexing on Substitutable Columns: Examples
You can build an index on attributes of the declared type of a substitutable column. In
addition, you can reference the subtype attributes by using the appropriate TREAT
function. The following example uses the table books, which is created in
"Substitutable Table and Column Examples" on page 16-71. The statement creates an
index on the salary attribute of all employee authors in the books table:
CREATE INDEX salary_i
ON books (TREAT(author AS employee_t).salary);

The target type in the argument of the TREAT function must be the type that added the
attribute being referenced. In the example, the target of TREAT is employee_t, which is
the type that added the salary attribute.
If this condition is not satisfied, then Oracle Database interprets the TREAT function as
any functional expression and creates the index as a function-based index. For
example, the following statement creates a function-based index on the salary
attribute of part-time employees, assigning nulls to instances of all other types in the
type hierarchy.

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-85

CREATE INDEX

CREATE INDEX salary_func_i ON persons p
(TREAT(VALUE(p) AS part_time_emp_t).salary);

You can also build an index on the type-discriminant column underlying a
substitutable column by using the SYS_TYPEID function.
Oracle Database uses the type-discriminant column to
evaluate queries that involve the IS OF type condition. The cardinality
of the typeid column is normally low, so Oracle recommends that you
build a bitmap index in this situation.

Note:

The following statement creates a bitmap index on the typeid of the author column of
the books table:
CREATE BITMAP INDEX typeid_i ON books (SYS_TYPEID(author));

See Also:
■

■

Oracle Database PL/SQL Language Reference to see the creation of
the type hierarchy underlying the books table
the functions TREAT on page 5-327 and SYS_TYPEID on
page 5-289 and the condition "IS OF type Condition" on page 7-25

14-86 Oracle Database SQL Language Reference

CREATE INDEXTYPE

CREATE INDEXTYPE
Purpose
14

Use the CREATE INDEXTYPE statement to create an indextype, which is an object that
specifies the routines that manage a domain (application-specific) index. Indextypes
reside in the same namespace as tables, views, and other schema objects. This
statement binds the indextype name to an implementation type, which in turn
specifies and refers to user-defined index functions and procedures that implement the
indextype.
See Also: Oracle Database Data Cartridge Developer's Guide for more
information on implementing indextypes

Prerequisites
14

To create an indextype in your own schema, you must have the CREATE INDEXTYPE
system privilege. To create an indextype in another schema, you must have the CREATE
ANY INDEXTYPE system privilege. In either case, you must have the EXECUTE object
privilege on the implementation type and the supported operators.
An indextype supports one or more operators, so before creating an indextype, you
must first design the operator or operators to be supported and provide functional
implementation for those operators.
CREATE OPERATOR on page 15-35

See Also:

Syntax
14

create_indextype::=
OR

REPLACE

schema

CREATE

.

INDEXTYPE

indextype

,
schema

,

.

FOR

operator

(

paramater_type

)

using_type_clause

RANGE
WITH

LOCAL

PARTITION

storage_table_clause
;

using_type_clause::=
schema
USING

.

array_DML_clause
implementation_type

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-87

CREATE INDEXTYPE

array_DML_clause::=
WITH
WITHOUT
ARRAY

DML

,
schema
schema

.

,

(

.
varray_type

type

)

storage_table_clause::=
SYSTEM
WITH

MANAGED

STORAGE

TABLES

USER

Semantics
14

schema
Specify the name of the schema in which the indextype resides. If you omit schema,
then Oracle Database creates the indextype in your own schema.

indextype
Specify the name of the indextype to be created.

FOR Clause
Use the FOR clause to specify the list of operators supported by the indextype.
■

■

For schema, specify the schema containing the operator. If you omit schema, then
Oracle assumes the operator is in your own schema.
For operator, specify the name of the operator supported by the indextype.
All the operators listed in this clause must be valid operators.

■

For parameter_type, list the types of parameters to the operator.

using_type_clause
The USING clause lets you specify the type that provides the implementation for the
new indextype.
For implementation_type, specify the name of the type that implements the
appropriate Oracle Data Cartridge Interface (ODCI).
■

You must specify a valid type that implements the routines in the ODCI.

■

The implementation type must reside in the same schema as the indextype.
See Also: Oracle Database Data Cartridge Developer's Guide for
additional information on this interface

14-88 Oracle Database SQL Language Reference

CREATE INDEXTYPE

WITH LOCAL PARTITION
Use this clause to indicate that the indextype can be used to create local domain
indexes on range- and list-partitioned tables. You use this clause in combination with
the storage_table in several ways (see storage_table_clause on page 14-89).
■

■

The recommended method is to specify WITH LOCAL [RANGE] PARTITION WITH
SYSTEM MANAGED STORAGE TABLES. This combination uses system-managed storage
tables, which are the preferred storage management, and lets you create local
domain indexes on both range- and list-partitioned tables. In this case the RANGE
keyword is optional and ignored, because it is no longer needed if you specify
WITH LOCAL PARTITION WITH SYSTEM MANAGED STORAGE TABLES.
You can specify WITH LOCAL RANGE PARTITION (including the RANGE keyword) and
omit the storage_table clause. Local domain indexes on range-partitioned tables
are supported with user-managed storage tables for backward compatibility. Oracle
does not recommend this combination because it uses the less efficient
user-managed storage tables.

If you omit this clause entirely, then you cannot subsequently use this indextype to
create a local domain index on a range- or list-partitioned table.

storage_table_clause
Use this clause to specify how storage tables and partition maintenance operations for
indexes built on this indextype are managed:
■

■

Specify WITH SYSTEM MANAGED STORAGE TABLES to indicate that the storage of
statistics data is to be managed by the system. The type you specify in
statistics_type should be storing the statistics related information in tables that
are maintained by the system. Also, the indextype you specify must already have
been created or altered to support the WITH SYSTEM MANAGED STORAGE TABLES clause.
Specify WITH USER MANAGED STORAGE TABLES to indicate that the tables that store the
user-defined statistics will be managed by the user. This is the default behavior.
See Also: Oracle Database Data Cartridge Developer's Guide for more
information about storage tables for domain indexes

array_DML_clause
Use this clause to let the indextype support the array interface for the
ODCIIndexInsert method.
type and varray_type If the data type of the column to be indexed is a user-defined
object type, then you must specify this clause to identify the varray varray_type that
Oracle should use to hold column values of type. If the indextype supports a list of
types, then you can specify a corresponding list of varray types. If you omit schema for
either type or varray_type, then Oracle assumes the type is in your own schema.

If the data type of the column to be indexed is a built-in system type, then any varray
type specified for the indextype takes precedence over the ODCI types defined by the
system.
See Also: Oracle Database Data Cartridge Developer's Guide for more
information on the ODCI array interface

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-89

CREATE INDEXTYPE

Examples
14

Creating an Indextype: Example The following statement creates an indextype
named position_indextype and specifies the position_between operator that is
supported by the indextype and the position_im type that implements the index
interface. Refer to "Using Extensible Indexing" on page F-1 for an extensible indexing
scenario that uses this indextype:
CREATE INDEXTYPE position_indextype
FOR position_between(NUMBER, NUMBER, NUMBER)
USING position_im;

14-90 Oracle Database SQL Language Reference

CREATE JAVA

CREATE JAVA
Purpose
14

Use the CREATE JAVA statement to create a schema object containing a Java source,
class, or resource.
See Also:

Oracle Database Java Developer's Guide for Java concepts and
information about Java stored procedures

■

Oracle Database JDBC Developer's Guide for information on JDBC

■

Prerequisites
14

To create or replace a schema object containing a Java source, class, or resource in your
own schema, you must have CREATE PROCEDURE system privilege. To create or replace
such a schema object in another user's schema, you must have CREATE ANY PROCEDURE
system privilege.

Syntax
14

create_java::=
RESOLVE
AND
OR

REPLACE

COMPILE

NOFORCE

CREATE

schema

SOURCE

.

NAMED

primary_name
invoker_rights_clause

RESOURCE
JAVA
SCHEMA

schema

CLASS

,
RESOLVER

(

(

schema_name

match_string

)

)

–

BFILE

(

directory_object_name

,

server_file_name

)

CLOB
USING

BLOB

subquery

BFILE
;
’
AS

key_for_BLOB

’

source_char

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-91

CREATE JAVA

invoker_rights_clause::=
CURRENT_USER
AUTHID
DEFINER

Semantics
14

OR REPLACE
Specify OR REPLACE to re-create the schema object containing the Java class, source, or
resource if it already exists. Use this clause to change the definition of an existing
object without dropping, re-creating, and regranting object privileges previously
granted.
If you redefine a Java schema object and specify RESOLVE or COMPILE, then Oracle
Database recompiles or resolves the object. Whether or not the resolution or
compilation is successful, the database invalidates classes that reference the Java
schema object.
Users who had previously been granted privileges on a redefined function can still
access the function without being regranted the privileges.
See Also:

ALTER JAVA on page 10-100 for additional information

RESOLVE | COMPILE
RESOLVE and COMPILE are synonymous keywords. They specify that Oracle Database
should attempt to resolve the Java schema object that is created if this statement
succeeds.
■

■

When applied to a class, resolution of referenced names to other class schema
objects occurs.
When applied to a source, source compilation occurs.

Restriction on RESOLVE and COMPILE You cannot specify these keywords for a
Java resource.

NOFORCE
Specify NOFORCE to roll back the results of this CREATE command if you have specified
either RESOLVE or COMPILE and the resolution or compilation fails. If you do not specify
this option, then Oracle Database takes no action if the resolution or compilation fails,
and the created schema object remains.

JAVA SOURCE Clause
Specify JAVA SOURCE to load a Java source file.

JAVA CLASS Clause
Specify JAVA CLASS to load a Java class file.

JAVA RESOURCE Clause
Specify JAVA RESOURCE to load a Java resource file.

14-92 Oracle Database SQL Language Reference

CREATE JAVA

NAMED Clause
The NAMED clause is required for a Java source or resource. The primary_name must be
enclosed in double quotation marks and its length must not exceed 4000 bytes in the
database character set.
■

■

For a Java source, this clause specifies the name of the schema object in which the
source code is held. A successful CREATE JAVA SOURCE statement will also create
additional schema objects to hold each of the Java classes defined by the source.
For a Java resource, this clause specifies the name of the schema object to hold the
Java resource.

Use double quotation marks to preserve a lower- or mixed-case primary_name.
If you do not specify schema, then Oracle Database creates the object in your own
schema.
Restrictions on NAMED Java Classes

The NAMED clause is subject to the following

restrictions:
■

You cannot specify NAMED for a Java class.

■

The primary_name cannot contain a database link.

SCHEMA Clause
The SCHEMA clause applies only to a Java class. This optional clause specifies the
schema in which the object containing the Java file will reside. If you do not specify
this clause, then Oracle Database creates the object in your own schema.

invoker_rights_clause
Use the invoker_rights_clause to indicate whether the methods of the class execute
with the privileges and in the schema of the user who owns the class or with the
privileges and in the schema of CURRENT_USER.
This clause also determines how Oracle Database resolves external names in queries,
DML operations, and dynamic SQL statements in the member functions and
procedures of the type.
AUTHID CURRENT_USER
CURRENT_USER indicates that the methods of the class execute with the privileges of
CURRENT_USER. This clause is the default and creates an invoker-rights class.
This clause also specifies that external names in queries, DML operations, and
dynamic SQL statements resolve in the schema of CURRENT_USER. External names in all
other statements resolve in the schema in which the methods reside.
AUTHID DEFINER
DEFINER indicates that the methods of the class execute with the privileges of the
owner of the schema in which the class resides, and that external names resolve in the
schema where the class resides. This clause creates a definer-rights class.
See Also:
■
■

Oracle Database Java Developer's Guide
Oracle Database PL/SQL Language Reference for information on how
CURRENT_USER is determined

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-93

CREATE JAVA

RESOLVER Clause
The RESOLVER clause lets you specify a mapping of the fully qualified Java name to a
Java schema object, where:
■

■

■

match_string is either a fully qualified Java name, a wildcard that can match such
a Java name, or a wildcard that can match any name.
schema_name designates a schema to be searched for the corresponding Java
schema object.
A dash (-) as an alternative to schema_name indicates that if match_string matches
a valid Java name, Oracle Database can leave the name unresolved. The resolution
succeeds, but the name cannot be used at run time by the class.

This mapping is stored with the definition of the schema objects created in this
command for use in later resolutions (either implicit or in explicit ALTER JAVA ...
RESOLVE statements).

USING Clause
The USING clause determines a sequence of character data (CLOB or BFILE) or binary
data (BLOB or BFILE) for the Java class or resource. Oracle Database uses the sequence
of characters to define one file for a Java class or resource, or one source file and one or
more derived classes for a Java source.
BFILE Clause
Specify the directory and filename of a previously created file on the operating system
(directory_object_name) and server file (server_file_name) containing the sequence.
BFILE is usually interpreted as a character sequence by CREATE JAVA SOURCE and as a
binary sequence by CREATE JAVA CLASS or CREATE JAVA RESOURCE.
CLOB | BLOB | BFILE subquery
Specify a subquery that selects a single row and column of the type specified (CLOB,
BLOB, or BFILE). The value of the column makes up the sequence of characters.
Note: In earlier releases, the USING clause implicitly supplied the
keyword SELECT. This is no longer the case. However, the subquery
without the keyword SELECT is still supported for backward
compatibility.

key_for_BLOB
The key_for_BLOB clause supplies the following implicit query:
SELECT LOB FROM CREATE$JAVA$LOB$TABLE
WHERE NAME = 'key_for_BLOB';

For you to use this case, the table
CREATE$JAVA$LOB$TABLE must exist in the current schema and must have a column
LOB of type BLOB and a column NAME of type VARCHAR2.
Restriction on the key_for_BLOB Clause

AS source_char
Specify a sequence of characters for a Java source.

14-94 Oracle Database SQL Language Reference

CREATE JAVA

Examples
14

Creating a Java Class Object: Example The following statement creates a schema
object containing a Java class using the name found in a Java binary file:
CREATE JAVA CLASS USING BFILE (java_dir, 'Agent.class')
/

This example assumes the directory object java_dir, which points to the operating
system directory containing the Java class Agent.class, already exists. In this
example, the name of the class determines the name of the Java class schema object.
Creating a Java Source Object: Example The following statement creates a Java
source schema object:
CREATE JAVA SOURCE NAMED "Welcome" AS
public class Welcome {
public static String welcome() {
return "Welcome World";
} }
/

The following statement creates a Java
resource schema object named apptext from a bfile:

Creating a Java Resource Object: Example

CREATE JAVA RESOURCE NAMED "appText"
USING BFILE (java_dir, 'textBundle.dat')
/

SQL Statements: CREATE CLUSTER to CREATE JAVA

14-95

CREATE JAVA

14-96 Oracle Database SQL Language Reference

15
15

SQL Statements: CREATE LIBRARY to
CREATE SPFILE

This chapter contains the following SQL statements:
■

CREATE LIBRARY

■

CREATE MATERIALIZED VIEW

■

CREATE MATERIALIZED VIEW LOG

■

CREATE OPERATOR

■

CREATE OUTLINE

■

CREATE PACKAGE

■

CREATE PACKAGE BODY

■

CREATE PFILE

■

CREATE PROCEDURE

■

CREATE PROFILE

■

CREATE RESTORE POINT

■

CREATE ROLE

■

CREATE ROLLBACK SEGMENT

■

CREATE SCHEMA

■

CREATE SEQUENCE

■

CREATE SPFILE

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-1

CREATE LIBRARY

CREATE LIBRARY
Purpose
15

Use the CREATE LIBRARY statement to create a schema object associated with an
operating-system shared library. The name of this schema object can then be used in
the call_spec of CREATE FUNCTION or CREATE PROCEDURE statements, or when declaring
a function or procedure in a package or type, so that SQL and PL/SQL can call to
third-generation-language (3GL) functions and procedures.
See Also: CREATE FUNCTION on page 14-58 and Oracle Database
PL/SQL Language Reference for more information on functions and
procedures

Prerequisites
15

The CREATE LIBRARY statement is valid only on platforms that support shared libraries
and dynamic linking.
To create a library in your own schema, you must have the CREATE LIBRARY system
privilege. To create a library in another user's schema, you must have the CREATE ANY
LIBRARY system privilege.
To use the library in the call_spec of a CREATE FUNCTION statement, or when declaring
a function in a package or type, you must have the EXECUTE object privilege on the
library and the CREATE FUNCTION system privilege. Refer to Oracle Database PL/SQL
Language Reference for information on the call_spec of a CREATE FUNCTION statement.
To use the library in the call_spec of a CREATE PROCEDURE statement, or when
declaring a procedure in a package or type, you must have the EXECUTE object
privilege on the library and the CREATE PROCEDURE system privilege. Refer to Oracle
Database PL/SQL Language Reference for information on the call_spec of a CREATE
PROCEDURE statement.
To execute a procedure or function defined with the call_spec (including a procedure
or function defined within a package or type), you must have the EXECUTE object
privilege on the procedure or function (but you do not need the EXECUTE object
privilege on the library).

Syntax
15

Libraries are defined using PL/SQL. Therefore, the syntax diagram in this book shows
only the SQL keywords. Refer to Oracle Database PL/SQL Language Reference for the
PL/SQL syntax, semantics, and examples.
create_library::=
OR
CREATE

REPLACE
LIBRARY

plsql_source

(plsql_source: See Oracle Database PL/SQL Language Reference.)

15-2 Oracle Database SQL Language Reference

CREATE LIBRARY

Semantics
15

OR REPLACE
Specify OR REPLACE to re-create the library if it already exists. Use this clause to change
the definition of an existing library without dropping, re-creating, and regranting
object privileges granted on it.
Users who had previously been granted privileges on a redefined library can still
access the library without being regranted the privileges.

plsql_source
See Oracle Database PL/SQL Language Reference for the syntax and semantics of the
plsql_source.

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-3

CREATE MATERIALIZED VIEW

CREATE MATERIALIZED VIEW
Purpose
15

Use the CREATE MATERIALIZED VIEW statement to create a materialized view. A
materialized view is a database object that contains the results of a query. The FROM
clause of the query can name tables, views, and other materialized views. Collectively
these objects are called master tables (a replication term) or detail tables (a data
warehousing term). This reference uses "master tables" for consistency. The databases
containing the master tables are called the master databases.
The keyword SNAPSHOT is supported in place of MATERIALIZED
VIEW for backward compatibility.
Note:

For replication purposes, materialized views allow you to maintain copies of remote
data on your local node. The copies can be updatable with the Advanced Replication
feature and are read-only without this feature. You can select data from a materialized
view as you would from a table or view. In replication environments, the materialized
views commonly created are primary key, rowid, object, and subquery materialized
views.
See Also: Oracle Database Advanced Replication for information on the
types of materialized views used to support replication

For data warehousing purposes, the materialized views commonly created are
materialized aggregate views, single-table materialized aggregate views, and
materialized join views. All three types of materialized views can be used by query
rewrite, an optimization technique that transforms a user request written in terms of
master tables into a semantically equivalent request that includes one or more
materialized views.
See Also:
■
■

ALTER MATERIALIZED VIEW on page 11-3
Oracle Database Data Warehousing Guide for information on the
types of materialized views used to support data warehousing

Prerequisites
15

The privileges required to create a materialized view should be granted directly rather
than through a role.
To create a materialized view in your own schema:
■

■

You must have been granted the CREATE MATERIALIZED VIEW system privilege and
either the CREATE TABLE or CREATE ANY TABLE system privilege.
You must also have access to any master tables of the materialized view that you
do not own, either through a SELECT object privilege on each of the tables or
through the SELECT ANY TABLE system privilege.

To create a materialized view in another user's schema:
■
■

You must have the CREATE ANY MATERIALIZED VIEW system privilege.
The owner of the materialized view must have the CREATE TABLE system privilege.
The owner must also have access to any master tables of the materialized view

15-4 Oracle Database SQL Language Reference

CREATE MATERIALIZED VIEW

that the schema owner does not own (for example, if the master tables are on a
remote database) and to any materialized view logs defined on those master
tables, either through a SELECT object privilege on each of the tables or through the
SELECT ANY TABLE system privilege.
To create a refresh-on-commit materialized view (ON COMMIT REFRESH clause), in
addition to the preceding privileges, you must have the ON COMMIT REFRESH object
privilege on any master tables that you do not own or you must have the ON COMMIT
REFRESH system privilege.
To create the materialized view with query rewrite enabled, in addition to the
preceding privileges:
■

■

If the schema owner does not own the master tables, then the schema owner must
have the GLOBAL QUERY REWRITE privilege or the QUERY REWRITE object privilege on
each table outside the schema.
If you are defining the materialized view on a prebuilt container (ON PREBUILT
TABLE clause), then you must have the SELECT privilege WITH GRANT OPTION on the
container table.

The user whose schema contains the materialized view must have sufficient quota in
the target tablespace to store the master table and index of the materialized view or
must have the UNLIMITED TABLESPACE system privilege.
When you create a materialized view, Oracle Database creates one internal table and at
least one index, and may create one view, all in the schema of the materialized view.
Oracle Database uses these objects to maintain the materialized view data. You must
have the privileges necessary to create these objects.
You can create the following types of local materialized views (including both ON
COMMIT and ON DEMAND) on master tables with commit SCN-based materialized view
logs:
■

Materialized aggregate views, including materialized aggregate views on a single
table

■

Materialized join views

■

Primary-key-based and rowid-based single table materialized views

■

UNION ALL materialized views, where each UNION ALL branch is one of the above
materialized view types

You cannot create remote materialized views on master tables with commit SCN-based
materialized view logs.
Creating a materialized view on master tables with different types of materialized
view logs (that is, a master table with timestamp-based materialized view logs and a
master table with commit SCN-based materialized view logs) is not supported and
causes ORA-32414.
See Also:
■

■

■

CREATE TABLE on page 16-6, CREATE VIEW on page 17-14, and
CREATE INDEX on page 14-60 for information on these privileges
Oracle Database Advanced Replication for information about the
prerequisites that apply to creating replication materialized views
Oracle Database Data Warehousing Guide for information about the
prerequisites that apply to creating data warehousing
materialized views

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-5

CREATE MATERIALIZED VIEW

Syntax
15

create_materialized_view::=
schema
CREATE

MATERIALIZED

.

VIEW

materialized_view

,
scoped_table_ref_constraint
encryption_spec

(
schema

.

ENCRYPT

OF

object_type

column_alias

WITH
REDUCED

PRECISION

WITHOUT
ON

PREBUILT

TABLE

physical_properties

materialized_view_props

physical_attributes_clause
TABLESPACE
USING

INDEX

USING

NO

tablespace

INDEX

create_mv_refresh

DISABLE
QUERY
FOR

UPDATE

REWRITE

ENABLE
AS

subquery

;

(scoped_table_ref_constraint::= on page 15-6, physical_properties::= on page 15-7,
materialized_view_props::= on page 15-7, physical_attributes_clause::= on page 15-8,
create_mv_refresh::= on page 15-8, subquery::= on page 19-5)
scoped_table_ref_constraint::=
schema

ref_column
SCOPE

FOR

(

)

.

scope_table_name

IS

ref_attribute

15-6 Oracle Database SQL Language Reference

c_alias

)

CREATE MATERIALIZED VIEW

physical_properties::=
deferred_segment_creation

table_compression
segment_attributes_clause
segment_attributes_clause

table_compression

HEAP
deferred_segment_creation
ORGANIZATION

segment_attributes_clause
INDEX

index_org_table_clause

EXTERNAL

external_table_clause

,
CLUSTER

cluster

(

column

)

(deferred_segment_creation::= on page 15-8, segment_attributes_clause::= on page 15-8,
table_compression::= on page 15-9, index_org_table_clause::= on page 15-7)
materialized_view_props::=
CACHE
column_properties

table_partitioning_clauses

NOCACHE

parallel_clause

build_clause

(column_properties::= on page 15-9, table_partitioning_clauses::= on page 16-17—part of
CREATE TABLE syntax, parallel_clause::= on page 15-12, build_clause::= on page 15-12)
index_org_table_clause::=
mapping_table_clause
PCTTHRESHOLD

integer

key_compression

index_org_overflow_clause

(mapping_table_clause: not supported with materialized views, key_compression::= on
page 15-7, index_org_overflow_clause::= on page 15-7)
key_compression::=
integer
COMPRESS
NOCOMPRESS

index_org_overflow_clause::=
INCLUDING

column_name

segment_attributes_clause
OVERFLOW

(segment_attributes_clause::= on page 15-8)

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-7

CREATE MATERIALIZED VIEW

create_mv_refresh::=
FAST
COMPLETE
FORCE
ON

DEMAND

ON

COMMIT

START

WITH
date

NEXT
PRIMARY

KEY

WITH

REFRESH

ROWID
MASTER
LOCAL
DEFAULT

ROLLBACK

SEGMENT

MASTER

USING

LOCAL
ROLLBACK

SEGMENT

rollback_segment

ENFORCED
USING

CONSTRAINTS
TRUSTED

NEVER

REFRESH

deferred_segment_creation::=
IMMEDIATE
SEGMENT

CREATION
DEFERRED

segment_attributes_clause::=
physical_attributes_clause
TABLESPACE

tablespace

logging_clause

(physical_attributes_clause::= on page 15-8, logging_clause::= on page 15-9)
physical_attributes_clause::=
PCTFREE

integer

PCTUSED

integer

INITRANS

integer

storage_clause

15-8 Oracle Database SQL Language Reference

CREATE MATERIALIZED VIEW

(logging_clause::= on page 8-38)
logging_clause::=
LOGGING
NOLOGGING
FILESYSTEM_LIKE_LOGGING

table_compression::=
BASIC
OLTP
LOW

FOR

HIGH

QUERY
ARCHIVE
COMPRESS
NOCOMPRESS

column_properties::=
object_type_col_properties
nested_table_col_properties
,
(

varray_col_properties

LOB_partition_storage

)

LOB_storage_clause
XMLType_column_properties

(object_type_col_properties::= on page 15-9, nested_table_col_properties::= on page 15-10,
varray_col_properties::= on page 15-10, LOB_partition_storage::= on page 15-11, LOB_
storage_clause::= on page 15-10, XMLType_column_properties: not supported for
materialized views)
object_type_col_properties::=
COLUMN

column

substitutable_column_clause

(substitutable_column_clause::= on page 15-9)
substitutable_column_clause::=
ELEMENT

TYPE
IS

OF

(

ONLY

type

)

NOT
SUBSTITUTABLE

AT

ALL

LEVELS

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-9

CREATE MATERIALIZED VIEW

nested_table_col_properties::=
LOCAL
substitutable_column_clause

nested_item
NESTED

GLOBAL

TABLE
COLUMN_VALUE

(
(

object_properties

)

physical_properties

)

column_properties
STORE

AS

storage_table

AS

LOCATOR

RETURN
VALUE

(substitutable_column_clause::= on page 15-9, object_properties::= on page 16-9, physical_
properties::= on page 16-10—part of CREATE TABLE syntax, column_properties::= on
page 15-9)
varray_col_properties::=
substitutable_column_clause
varray_storage_clause
VARRAY

varray_item
substitutable_column_clause

(substitutable_column_clause::= on page 15-9, varray_storage_clause::= on page 15-10)
varray_storage_clause::=
SECUREFILE
LOB_segname
BASICFILE
STORE

(

AS

LOB_storage_parameters

LOB
LOB_segname

(LOB_parameters::= on page 15-11)
LOB_storage_clause::=
SECUREFILE

,
(

LOB_item

)

STORE

BASICFILE

AS
(

LOB

LOB_storage_parameters

SECUREFILE
BASICFILE
(

LOB_item

)

STORE

AS

LOB_segname
(

15-10 Oracle Database SQL Language Reference

LOB_storage_parameters

)

)

)

CREATE MATERIALIZED VIEW

(LOB_storage_parameters::= on page 15-11)
LOB_storage_parameters::=
TABLESPACE

tablespace
storage_clause

LOB_parameters
storage_clause

(LOB_parameters::= on page 15-11, storage_clause::= on page 8-50)
LOB_parameters::=
ENABLE
STORAGE

IN

ROW

DISABLE
CHUNK

integer

PCTVERSION
FREEPOOLS

integer
integer

LOB_retention_clause
LOB_deduplicate_clause
LOB_compression_clause
ENCRYPT

encryption_spec

DECRYPT
CACHE

logging_clause

NOCACHE
CACHE

READS

(storage_clause::= on page 8-50, logging_clause::= on page 15-9)
LOB_partition_storage::=
LOB_storage_clause
PARTITION

partition
varray_col_properties

LOB_partitioning_storage
(

SUBPARTITION

subpartition

)
varray_col_properties

(LOB_storage_clause::= on page 15-10, varray_col_properties::= on page 15-10)

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-11

CREATE MATERIALIZED VIEW

parallel_clause::=
NOPARALLEL
integer
PARALLEL

build_clause::=
IMMEDIATE
BUILD
DEFERRED

Semantics
15

schema
Specify the schema to contain the materialized view. If you omit schema, then Oracle
Database creates the materialized view in your schema.

materialized_view
Specify the name of the materialized view to be created. The name must satisfy the
requirements listed in "Database Object Naming Rules" on page 3-111. Oracle Database
generates names for the table and indexes used to maintain the materialized view by
adding a prefix or suffix to the materialized view name.

OF object_type
The OF object_type clause lets you explicitly create an object materialized view of
type object_type.
See Also: See CREATE TABLE ... object_table on page 16-7 for more
information on the OF type_name clause

scoped_table_ref_constraint
Use the SCOPE FOR clause to restrict the scope of references to a single object table. You
can refer either to the table name with scope_table_name or to a column alias. The
values in the REF column or attribute point to objects in scope_table_name or c_alias,
in which object instances of the same type as the REF column are stored. If you specify
aliases, then they must have a one-to-one correspondence with the columns in the
SELECT list of the defining query of the materialized view.
See Also: "SCOPE REF Constraints" on page 8-13 for more
information

column_alias
You can specify a column alias for each column of the materialized view. The column
alias list explicitly resolves any column name conflict, eliminating the need to specify
aliases in the SELECT clause of the materialized view. If you specify any column alias in
this clause, then you must specify an alias for each data source referenced in the
SELECT clause.
Use this clause to encrypt this column of the materialized view.
Refer to the CREATE TABLE clause encryption_spec on page 16-27 for more information on
column encryption.
ENCRYPT clause

15-12 Oracle Database SQL Language Reference

CREATE MATERIALIZED VIEW

ON PREBUILT TABLE Clause
The ON PREBUILT TABLE clause lets you register an existing table as a preinitialized
materialized view. This clause is particularly useful for registering large materialized
views in a data warehousing environment. The table must have the same name and be
in the same schema as the resulting materialized view.
If the materialized view is dropped, then the preexisting table reverts to its identity as
a table.
Caution: This clause assumes that the table object reflects the
materialization of a subquery. Oracle strongly recommends that you
ensure that this assumption is true in order to ensure that the
materialized view correctly reflects the data in its master tables.

The ON PREBUILT TABLE clause could be useful in the following scenarios:
■

■

You have a table representing the result of a query. Creating the table was an
expensive operation that possibly took a long time. You want to create a
materialized view on the query. You can use the ON PREBUILT TABLE clause to avoid
the expense of executing the query and populating the container for the
materialized view.
You temporarily discontinue having a materialized view, but keep its container
table, using the DROP MATERIALIZED VIEW ... PRESERVE TABLE statement. You then
decide to recreate the materialized view and you know that the master tables of
the view have not changed. You can create the materialized view using the ON
PREBUILT TABLE clause. This avoids the expense and time of creating and
populating the container table for the materialized view.

Specify WITH REDUCED PRECISION to authorize the loss of
precision that will result if the precision of the table or materialized view columns do
not exactly match the precision returned by subquery.

WITH REDUCED PRECISION

WITHOUT REDUCED PRECISION Specify WITHOUT REDUCED PRECISION to require that
the precision of the table or materialized view columns match exactly the precision
returned by subquery, or the create operation will fail. This is the default.
Restrictions on Using Prebuilt Tables Prebuilt tables are subject to the following
restrictions:
■

■

■

Each column alias in subquery must correspond to a column in the prebuilt table,
and corresponding columns must have matching data types.
If you specify this clause, then you cannot specify a NOT NULL constraint for any
column that is not referenced in subquery unless you also specify a default value
for that column.
You cannot specify the ON PREBUILT TABLE clause when creating a rowid
materialized view.
See Also: "Creating Prebuilt Materialized Views: Example" on
page 15-24

physical_properties_clause
The components of the physical_properties_clause have the same semantics for
materialized views that they have for tables, with exceptions and additions described
in the sections that follow.
SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-13

CREATE MATERIALIZED VIEW

Restriction on the physical_properties_clause

You cannot specify ORGANIZATION

EXTERNAL for a materialized view.
deferred_segment_creation
Use this clause to determine when the segment for this materialized view should be
created. See the CREATE TABLE clause deferred_segment_creation on page 16-32 for more
information.
segment_attributes_clause
Use the segment_attributes_clause to establish values for the PCTFREE, PCTUSED, and
INITRANS parameters, the storage characteristics for the materialized view, to assign a
tablespace, and to specify whether logging is to occur. In the USING INDEX clause, you
cannot specify PCTFREE or PCTUSED.
TABLESPACE Clause Specify the tablespace in which the materialized view is to be
created. If you omit this clause, then Oracle Database creates the materialized view in
the default tablespace of the schema containing the materialized view.
See Also: physical_attributes_clause on page 8-44 and storage_clause on
page 8-46 for a complete description of these clauses, including
default values
logging_clause Specify LOGGING or NOLOGGING to establish the logging characteristics

for the materialized view. The logging characteristic affects the creation of the
materialized view and any nonatomic refresh that is initiated by way of the DBMS_
REFRESH package. The default is the logging characteristic of the tablespace in which
the materialized view resides.
logging_clause on page 8-38 for a full description of this
clause and Oracle Database PL/SQL Packages and Types Reference for
more information on atomic and nonatomic refresh

See Also:

table_compression
Use the table_compression clause to instruct the database whether to compress data
segments to reduce disk and memory use.
See Also: Refer to the CREATE TABLE table_compression on page 16-34
for the full semantics of this clause

index_org_table_clause
The ORGANIZATION INDEX clause lets you create an index-organized materialized view.
In such a materialized view, data rows are stored in an index defined on the primary
key of the materialized view. You can specify index organization for the following
types of materialized views:
■

Read-only and updatable object materialized views. You must ensure that the
master table has a primary key.

■

Read-only and updatable primary key materialized views.

■

Read-only rowid materialized views.

The keywords and parameters of the index_org_table_clause have the same
semantics as described in CREATE TABLE, with the restrictions that follow.
See Also:

The index_org_table_clause of CREATE TABLE on page 16-37

15-14 Oracle Database SQL Language Reference

CREATE MATERIALIZED VIEW

Restrictions on Index-Organized Materialized Views

Index-organized materialized

views are subject to the following restrictions:
■

■

You cannot specify the following CREATE MATERIALIZED VIEW clauses: CACHE or
NOCACHE, CLUSTER, or ON PREBUILT TABLE.
In the index_org_table_clause:
–

You cannot specify the mapping_table_clause.

–

You can specify COMPRESS only for a materialized view based on a composite
primary key. You can specify NOCOMPRESS for a materialized view based on
either a simple or composite primary key.

CLUSTER Clause
The CLUSTER clause lets you create the materialized view as part of the specified
cluster. A cluster materialized view uses the space allocation of the cluster. Therefore,
you do not specify physical attributes or the TABLESPACE clause with the CLUSTER
clause.
Restriction on Cluster Materialized Views If you specify CLUSTER, then you cannot
specify the table_partitioning_clauses in materialized_view_props.

materialized_view_props
Use these property clauses to describe a materialized view that is not based on an
existing table. To create a materialized view that is based on an existing table, use the
ON PREBUILT TABLE clause.
column_properties
The column_properties clause lets you specify the storage characteristics of a LOB,
nested table, varray, or XMLType column. The object_type_col_properties are not
relevant for a materialized view.
CREATE TABLE on page 16-6 for detailed information
about specifying the parameters of this clause

See Also:

table_partitioning_clauses
The table_partitioning_clauses let you specify that the materialized view is
partitioned on specified ranges of values or on a hash function. Partitioning of
materialized views is the same as partitioning of tables.
table_partitioning_clauses on page 16-51 in the CREATE
TABLE documentation

See Also:

CACHE | NOCACHE
For data that will be accessed frequently, CACHE specifies that the blocks retrieved for
this table are placed at the most recently used end of the least recently used (LRU) list
in the buffer cache when a full table scan is performed. This attribute is useful for
small lookup tables. NOCACHE specifies that the blocks are placed at the least recently
used end of the LRU list.
NOCACHE has no effect on materialized views for which you
specify KEEP in the storage_clause.

Note:

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-15

CREATE MATERIALIZED VIEW

CREATE TABLE on page 16-6 for information about
specifying CACHE or NOCACHE

See Also:

parallel_clause
The parallel_clause lets you indicate whether parallel operations will be supported
for the materialized view and sets the default degree of parallelism for queries and
DML on the materialized view after creation.
For complete information on this clause, refer to parallel_clause on page 16-63 in the
documentation on CREATE TABLE.
build_clause
The build_clause lets you specify when to populate the materialized view.
IMMEDIATE Specify IMMEDIATE to indicate that the materialized view is to be

populated immediately. This is the default.
DEFERRED Specify DEFERRED to indicate that the materialized view is to be

populated by the next REFRESH operation. The first (deferred) refresh must always be a
complete refresh. Until then, the materialized view has a staleness value of UNUSABLE,
so it cannot be used for query rewrite.

USING INDEX Clause
The USING INDEX clause lets you establish the value of the INITRANS and STORAGE
parameters for the default index Oracle Database uses to maintain the materialized
view data. If USING INDEX is not specified, then default values are used for the index.
Oracle Database uses the default index to speed up incremental (FAST) refresh of the
materialized view.
Restriction on USING INDEX clause You cannot specify the PCTUSED parameter in

this clause.

USING NO INDEX Clause
Specify USING NO INDEX to suppress the creation of the default index. You can create an
alternative index explicitly by using the CREATE INDEX statement. You should create
such an index if you specify USING NO INDEX and you are creating the materialized view
with the incremental refresh method (REFRESH FAST).

create_mv_refresh
Use the create_mv_refresh clause to specify the default methods, modes, and times
for the database to refresh the materialized view. If the master tables of a materialized
view are modified, then the data in the materialized view must be updated to make
the materialized view accurately reflect the data currently in its master tables. This
clause lets you schedule the times and specify the method and mode for the database
to refresh the materialized view.
Note: This clause only sets the default refresh options. For
instructions on actually implementing the refresh, refer to Oracle
Database Advanced Replication and Oracle Database Data Warehousing
Guide.

15-16 Oracle Database SQL Language Reference

CREATE MATERIALIZED VIEW

See Also:
■

■

"Periodic Refresh of Materialized Views: Example" on page 15-24
and "Automatic Refresh Times for Materialized Views: Example"
on page 15-25
Oracle Database PL/SQL Packages and Types Reference for more
information on refresh methods

FAST Clause
Specify FAST to indicate the incremental refresh method, which performs the refresh
according to the changes that have occurred to the master tables. The changes for
conventional DML changes are stored in the materialized view log associated with the
master table. The changes for direct-path INSERT operations are stored in the direct
loader log.
If you specify REFRESH FAST, then the CREATE statement will fail unless materialized
view logs already exist for the materialized view master tables. Oracle Database
creates the direct loader log automatically when a direct-path INSERT takes place. No
user intervention is needed.
For both conventional DML changes and for direct-path INSERT operations, other
conditions may restrict the eligibility of a materialized view for fast refresh.
Materialized views are not eligible for fast refresh if the defining query contains an
analytic function or the XMLTable function.
See Also:
■

■

■

■
■

Oracle Database Advanced Replication for restrictions on fast refresh
in replication environments
Oracle Database Data Warehousing Guide for restrictions on fast
refresh in data warehousing environments
The EXPLAIN_MVIEW procedure of the DBMS_MVIEW package for help
diagnosing problems with fast refresh and the TUNE_MVIEW
procedure of the DBMS_MVIEW package correction of fast refresh
problems
"Analytic Functions" on page 5-11
"Creating a Fast Refreshable Materialized View: Example" on
page 15-25

COMPLETE Clause
Specify COMPLETE to indicate the complete refresh method, which is implemented by
executing the defining query of the materialized view. If you request a complete
refresh, then Oracle Database performs a complete refresh even if a fast refresh is
possible.
FORCE Clause
Specify FORCE to indicate that when a refresh occurs, Oracle Database will perform a
fast refresh if one is possible or a complete refresh if fast refresh is not possible. If you
do not specify a refresh method (FAST, COMPLETE, or FORCE), then FORCE is the default.
ON COMMIT Clause
Specify ON COMMIT to indicate that a refresh is to occur whenever the database commits
a transaction that operates on a master table of the materialized view. This clause may
SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-17

CREATE MATERIALIZED VIEW

increase the time taken to complete the commit, because the database performs the
refresh operation as part of the commit process.
You cannot specify both ON COMMIT and ON DEMAND. If you specify ON COMMIT, then you
cannot also specify START WITH or NEXT.
Restrictions on Refreshing ON COMMIT
■
This clause is not supported for materialized views containing object types or
Oracle-supplied types.
■
■

This clause is not supported for materialized views with remote tables.
If you specify this clause, then you cannot subsequently execute a distributed
transaction on any master table of this materialized view. For example, you cannot
insert into the master by selecting from a remote table. The ON DEMAND clause does
not impose this restriction on subsequent distributed transactions on master
tables.

ON DEMAND Clause
Specify ON DEMAND to indicate that database will not refresh the materialized view
unless the user manually launches a refresh through one of the three DBMS_MVIEW
refresh procedures.
You cannot specify both ON COMMIT and ON DEMAND. If you omit both ON COMMIT and ON
DEMAND, then ON DEMAND is the default. You can override this default setting by
specifying the START WITH or NEXT clauses, either in the same CREATE MATERIALIZED
VIEW statement or a subsequent ALTER MATERIALIZED VIEW statement.
START WITH and NEXT take precedence over ON DEMAND. Therefore, in most circumstances
it is not meaningful to specify ON DEMAND when you have specified START WITH or NEXT.
See Also:
■

■

Oracle Database PL/SQL Packages and Types Reference for
information on these procedures
Oracle Database Data Warehousing Guide on the types of
materialized views you can create by specifying REFRESH ON
DEMAND

START WITH Clause
Specify a datetime expression for the first automatic refresh time.
NEXT Clause
Specify a datetime expression for calculating the interval between automatic refreshes.
Both the START WITH and NEXT values must evaluate to a time in the future. If you omit
the START WITH value, then the database determines the first automatic refresh time by
evaluating the NEXT expression with respect to the creation time of the materialized
view. If you specify a START WITH value but omit the NEXT value, then the database
refreshes the materialized view only once. If you omit both the START WITH and NEXT
values, or if you omit the create_mv_refresh entirely, then the database does not
automatically refresh the materialized view.
WITH PRIMARY KEY Clause
Specify WITH PRIMARY KEY to create a primary key materialized view. This is the
default and should be used in all cases except those described for WITH ROWID. Primary
key materialized views allow materialized view master tables to be reorganized
15-18 Oracle Database SQL Language Reference

CREATE MATERIALIZED VIEW

without affecting the eligibility of the materialized view for fast refresh. The master
table must contain an enabled primary key constraint, and the defining query of the
materialized view must specify all of the primary key columns directly. In the defining
query, the primary key columns cannot be specified as the argument to a function such
as UPPER.
Restriction on Primary Key Materialized Views You cannot specify this clause for an
object materialized view. Oracle Database implicitly refreshes objects materialized
WITH OBJECT ID.

Oracle Database Advanced Replication for detailed
information about primary key materialized views and "Creating
Primary Key Materialized Views: Example" on page 15-24

See Also:

WITH ROWID Clause
Specify WITH ROWID to create a rowid materialized view. Rowid materialized views are
useful if the materialized view does not include all primary key columns of the master
tables. Rowid materialized views must be based on a single table and cannot contain
any of the following:
■

Distinct or aggregate functions

■

GROUP BY or CONNECT BY clauses

■

Subqueries

■

Joins

■

Set operations

The WITH ROWID clause has no effect if there are multiple master tables in the defining
query.
Rowid materialized views are not eligible for fast refresh after a master table
reorganization until a complete refresh has been performed.
Restriction on Rowid Materialized Views You cannot specify this clause for an object

materialized view. Oracle Database implicitly refreshes objects materialized WITH
OBJECT ID.
See Also: "Creating Materialized Aggregate Views: Example" on
page 15-23 and "Creating Rowid Materialized Views: Example" on
page 15-24

USING ROLLBACK SEGMENT Clause
This clause is not valid if your database is in automatic undo mode, because in that
mode Oracle Database uses undo tablespaces instead of rollback segments. Oracle
strongly recommends that you use automatic undo mode. This clause is supported for
backward compatibility with replication environments containing older versions of
Oracle Database that still use rollback segments.
For rollback_segment, specify the remote rollback segment to be used during
materialized view refresh.
DEFAULT DEFAULT specifies that Oracle Database will choose automatically which
rollback segment to use. If you specify DEFAULT, then you cannot specify rollback_
segment. DEFAULT is most useful when modifying, rather than creating, a materialized
view.

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-19

CREATE MATERIALIZED VIEW

See Also:

ALTER MATERIALIZED VIEW on page 11-3

MASTER MASTER specifies the remote rollback segment to be used at the remote

master site for the individual materialized view.
LOCAL LOCAL specifies the remote rollback segment to be used for the local refresh
group that contains the materialized view. This is the default.

Oracle Database Advanced Replication for information on
specifying the local materialized view rollback segment using the
DBMS_REFRESH package
See Also:

If you omit rollback_segment, then the database automatically chooses the rollback
segment to be used. One master rollback segment is stored for each materialized view
and is validated during materialized view creation and refresh. If the materialized
view is complex, then the database ignores any master rollback segment you specify.
USING ... CONSTRAINTS Clause
The USING ... CONSTRAINTS clause lets Oracle Database choose more rewrite options
during the refresh operation, resulting in more efficient refresh execution. The clause
lets Oracle Database use unenforced constraints, such as dimension relationships or
constraints in the RELY state, rather than relying only on enforced constraints during
the refresh operation.
The USING TRUSTED CONSTRAINTS clause enables you to create a materialized view on
top of a table that has a non-NULL Virtual Private Database (VPD) policy on it. In this
case, ensure that the materialized view behaves correctly. Materialized view results are
computed based on the rows and columns filtered by VPD policy. Therefore, you must
coordinate the materialized view definition with the VPD policy to ensure the correct
results. Without the USING TRUSTED CONSTRAINTS clause, any VPD policy on a master
table will prevent a materialized view from being created.
The USING TRUSTED CONSTRAINTS clause lets Oracle
Database use dimension and constraint information that has been
declared trustworthy by the database administrator but that has not
been validated by the database. If the dimension and constraint
information is valid, then performance may improve. However, if this
information is invalid, then the refresh procedure may corrupt the
materialized view even though it returns a success status.

Caution:

If you omit this clause, then the default is USING ENFORCED CONSTRAINTS.
NEVER REFRESH Clause
Specify NEVER REFRESH to prevent the materialized view from being refreshed with any
Oracle Database refresh mechanism or packaged procedure. Oracle Database will
ignore any REFRESH statement on the materialized view issued from such a procedure.
If you specify this clause, then you can perform DML operations on the materialized
view. To reverse this clause, you must issue an ALTER MATERIALIZED VIEW ... REFRESH
statement.

15-20 Oracle Database SQL Language Reference

CREATE MATERIALIZED VIEW

FOR UPDATE Clause
Specify FOR UPDATE to allow a subquery, primary key, object, or rowid materialized
view to be updated. When used in conjunction with Advanced Replication, these
updates will be propagated to the master.

QUERY REWRITE Clause
The QUERY REWRITE clause lets you specify whether the materialized view is eligible to
be used for query rewrite.
ENABLE Clause

Specify ENABLE to enable the materialized view for query rewrite.

Restrictions on Enabling Query Rewrite Enabling of query rewrite is subject to the
following restrictions:
■

■

You can enable query rewrite only if all user-defined functions in the materialized
view are DETERMINISTIC.
You can enable query rewrite only if expressions in the statement are repeatable.
For example, you cannot include CURRENT_TIME or USER, sequence values (such as
the CURRVAL or NEXTVAL pseudocolumns), or the SAMPLE clause (which may sample
different rows as the contents of the materialized view change).
Notes:
■

■

Query rewrite is disabled by default, so you must specify this
clause to make materialized views eligible for query rewrite.
After you create the materialized view, you must collect statistics
on it using the DBMS_STATS package. Oracle Database needs the
statistics generated by this package to optimize query rewrite.

See Also:
■

■

■

■

Oracle Database Data Warehousing Guide for more information on
query rewrite
Oracle Database PL/SQL Packages and Types Reference for
information about the DBMS_STATS package
The EXPLAIN_MVIEW procedure of the DBMS_MVIEW package for help
diagnosing problems with query rewrite and the TUNE_MVIEW
procedure of the DBMS_MVIEW package correction of query rewrite
problems
CREATE FUNCTION on page 14-58

DISABLE Clause Specify DISABLE to indicate that the materialized view is not
eligible for use by query rewrite. A disabled materialized view can be refreshed.

AS subquery
Specify the defining query of the materialized view. When you create the materialized
view, Oracle Database executes this subquery and places the results in the materialized
view. This subquery is any valid SQL subquery. However, not all subqueries are fast
refreshable, nor are all subqueries eligible for query rewrite.
Notes on the Defining Query of a Materialized View The following notes apply to
materialized views:
SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-21

CREATE MATERIALIZED VIEW

■

■

■

Oracle Database does not execute the defining query immediately if you specify
BUILD DEFERRED.
Oracle recommends that you qualify each table and view in the FROM clause of the
defining query of the materialized view with the schema containing it.
In order to create a materialized view whose defining query selects from a master
table that has a Virtual Private Database (VPD) policy, you must specify the
REFRESH USING TRUSTED CONSTRAINTS clause.

Restrictions on the Defining Query of a Materialized View

The materialized view

query is subject to the following restrictions:
■

■

■

■

The defining query of a materialized view can select from tables, views, or
materialized views owned by the user SYS, but you cannot enable QUERY REWRITE
on such a materialized view.
You cannot define a materialized view with a subquery in the select list of the
defining query. You can, however, include subqueries elsewhere in the defining
query, such as in the WHERE clause.
You cannot use the AS OF clause of the flashback_query_clause in the defining
query of a materialized view.
Materialized join views and materialized aggregate views with a GROUP BY clause
cannot select from an index-organized table.

■

Materialized views cannot contain columns of data type LONG or LONG RAW.

■

Materialized views cannot contain virtual columns.

■

■

■

You cannot create a materialized view log on a temporary table. Therefore, if the
defining query references a temporary table, then this materialized view will not
be eligible for FAST refresh, nor can you specify the QUERY REWRITE clause in this
statement.
If the FROM clause of the defining query references another materialized view, then
you must always refresh the materialized view referenced in the defining query
before refreshing the materialized view you are creating in this statement.
Materialized views with join expressions in the defining query cannot have XML
data type columns. The XML data types include XMLType and URI data type
columns.

If you are creating a materialized view enabled for query rewrite, then:
■

■

The defining query cannot contain, either directly or through a view, references to
ROWNUM, USER, SYSDATE, remote tables, sequences, or PL/SQL functions that write
or read database or package state.
Neither the materialized view nor the master tables of the materialized view can
be remote.

If you want the materialized view to be eligible for fast refresh using a materialized
view log, then some additional restrictions may apply.

15-22 Oracle Database SQL Language Reference

CREATE MATERIALIZED VIEW

See Also:
■

■

■

Oracle Database Data Warehousing Guide for more information on
restrictions relating to data warehousing
Oracle Database Advanced Replication for more information on
restrictions relating to replication
"Creating Materialized Join Views: Example" on page 15-24,
"Creating Subquery Materialized Views: Example" on page 15-23,
and "Creating a Nested Materialized View: Example" on
page 15-26

Examples
15

The following examples require the materialized logs that are created in the
"Examples" section of CREATE MATERIALIZED VIEW LOG on page 15-27.
The following statement creates a
very simple materialized view based on the employees and table in the hr schema:

Creating a Simple Materialized View: Example

CREATE MATERIALIZED VIEW mv1 AS SELECT * FROM hr.employees;

By default, Oracle Database creates a primary key materialized view with refresh on
demand only. If a materialized view log exists on employees, then mv1 can be altered to
be capable of fast refresh. If no such log exists, then only full refresh of mv1 is possible.
Oracle Database uses default storage properties for mv1. The only privileges required
for this operation are the CREATE MATERIALIZED VIEW system privilege, and the SELECT
object privilege on hr.employees.
Creating Subquery Materialized Views: Example The following statement creates a
subquery materialized view based on the customers and countries tables in the sh
schema at the remote database:
CREATE MATERIALIZED VIEW foreign_customers FOR UPDATE
AS SELECT * FROM sh.customers@remote cu
WHERE EXISTS
(SELECT * FROM sh.countries@remote co
WHERE co.country_id = cu.country_id);

Creating Materialized Aggregate Views: Example The following statement creates
and populates a materialized aggregate view on the sample sh.sales table and
specifies the default refresh method, mode, and time. It uses the materialized view log
created in "Creating a Materialized View Log: Examples" on page 15-33, as well as the
two additional logs shown here:
CREATE MATERIALIZED VIEW LOG ON times
WITH ROWID, SEQUENCE (time_id, calendar_year)
INCLUDING NEW VALUES;
CREATE MATERIALIZED VIEW LOG ON products
WITH ROWID, SEQUENCE (prod_id)
INCLUDING NEW VALUES;
CREATE MATERIALIZED VIEW sales_mv
BUILD IMMEDIATE
REFRESH FAST ON COMMIT
AS SELECT t.calendar_year, p.prod_id,
SUM(s.amount_sold) AS sum_sales
FROM times t, products p, sales s

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-23

CREATE MATERIALIZED VIEW

WHERE t.time_id = s.time_id AND p.prod_id = s.prod_id
GROUP BY t.calendar_year, p.prod_id;

Creating Materialized Join Views: Example The following statement creates and
populates the materialized aggregate view sales_by_month_by_state using tables in
the sample sh schema. The materialized view will be populated with data as soon as
the statement executes successfully. By default, subsequent refreshes will be
accomplished by reexecuting the defining query of the materialized view:
CREATE MATERIALIZED VIEW sales_by_month_by_state
TABLESPACE example
PARALLEL 4
BUILD IMMEDIATE
REFRESH COMPLETE
ENABLE QUERY REWRITE
AS SELECT t.calendar_month_desc, c.cust_state_province,
SUM(s.amount_sold) AS sum_sales
FROM times t, sales s, customers c
WHERE s.time_id = t.time_id AND s.cust_id = c.cust_id
GROUP BY t.calendar_month_desc, c.cust_state_province;

Creating Prebuilt Materialized Views: Example The following statement creates a
materialized aggregate view for the preexisting summary table, sales_sum_table:
CREATE TABLE sales_sum_table
(month VARCHAR2(8), state VARCHAR2(40), sales NUMBER(10,2));
CREATE MATERIALIZED VIEW sales_sum_table
ON PREBUILT TABLE WITH REDUCED PRECISION
ENABLE QUERY REWRITE
AS SELECT t.calendar_month_desc AS month,
c.cust_state_province AS state,
SUM(s.amount_sold) AS sales
FROM times t, customers c, sales s
WHERE s.time_id = t.time_id AND s.cust_id = c.cust_id
GROUP BY t.calendar_month_desc, c.cust_state_province;

In the preceding example, the materialized view has the same name and also has the
same number of columns with the same data types as the prebuilt table. The WITH
REDUCED PRECISION clause allows for differences between the precision of the
materialized view columns and the precision of the values returned by the subquery.
The following statement creates
the primary key materialized view catalog on the sample table oe.product_
information:

Creating Primary Key Materialized Views: Example

CREATE MATERIALIZED VIEW catalog
REFRESH FAST START WITH SYSDATE NEXT SYSDATE + 1/4096
WITH PRIMARY KEY
AS SELECT * FROM product_information;

The following statement creates a
rowid materialized view on the sample table oe.orders:

Creating Rowid Materialized Views: Example

CREATE MATERIALIZED VIEW order_data REFRESH WITH ROWID
AS SELECT * FROM orders;

Periodic Refresh of Materialized Views: Example The following statement creates
the primary key materialized view emp_data and populates it with data from the
sample table hr.employees:

15-24 Oracle Database SQL Language Reference

CREATE MATERIALIZED VIEW

CREATE MATERIALIZED VIEW LOG ON employees
WITH PRIMARY KEY
INCLUDING NEW VALUES;
CREATE MATERIALIZED VIEW emp_data
PCTFREE 5 PCTUSED 60
TABLESPACE example
STORAGE (INITIAL 50K)
REFRESH FAST NEXT sysdate + 7
AS SELECT * FROM employees;

The preceding statement does not include a START WITH parameter, so Oracle Database
determines the first automatic refresh time by evaluating the NEXT value using the
current SYSDATE. A materialized view log was created for the employee table, so Oracle
Database performs a fast refresh of the materialized view every 7 days, beginning 7
days after the materialized view is created.
Because the materialized view conforms to the conditions for fast refresh, the database
will perform a fast refresh. The preceding statement also establishes storage
characteristics that the database uses to maintain the materialized view.
Automatic Refresh Times for Materialized Views: Example The following statement
creates the complex materialized view all_customers that queries the employee tables
on the remote and local databases:
CREATE MATERIALIZED VIEW all_customers
PCTFREE 5 PCTUSED 60
TABLESPACE example
STORAGE (INITIAL 50K)
USING INDEX STORAGE (INITIAL 25K)
REFRESH START WITH ROUND(SYSDATE + 1) + 11/24
NEXT NEXT_DAY(TRUNC(SYSDATE), 'MONDAY') + 15/24
AS SELECT * FROM sh.customers@remote
UNION
SELECT * FROM sh.customers@local;

Oracle Database automatically refreshes this materialized view tomorrow at 11:00 a.m.
and subsequently every Monday at 3:00 p.m. The default refresh method is FORCE. The
defining query contains a UNION operator, which is not supported for fast refresh, so
the database will automatically perform a complete refresh.
The preceding statement also establishes storage characteristics for both the
materialized view and the index that the database uses to maintain it:
■

■

The first STORAGE clause establishes the sizes of the first and second extents of the
materialized view as 50 kilobytes each.
The second STORAGE clause, appearing with the USING INDEX clause, establishes the
sizes of the first and second extents of the index as 25 kilobytes each.

The following statement
creates a fast-refreshable materialized view that selects columns from the order_items
table in the sample oe schema, using the UNION set operator to restrict the rows
returned from the product_information and inventories tables using WHERE
conditions. The materialized view logs for order_items and product_information
were created in the "Examples" section of CREATE MATERIALIZED VIEW LOG on
page 15-33. This example also requires a materialized view log on oe.inventories.
Creating a Fast Refreshable Materialized View: Example

CREATE MATERIALIZED VIEW LOG ON inventories
WITH (quantity_on_hand);

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-25

CREATE MATERIALIZED VIEW

CREATE MATERIALIZED VIEW warranty_orders REFRESH FAST AS
SELECT order_id, line_item_id, product_id FROM order_items o
WHERE EXISTS
(SELECT * FROM inventories i WHERE o.product_id = i.product_id
AND i.quantity_on_hand IS NOT NULL)
UNION
SELECT order_id, line_item_id, product_id FROM order_items
WHERE quantity > 5;

The materialized view warranty_orders requires that materialized view logs be
defined on order_items (with product_id as a join column) and on inventories (with
quantity_on_hand as a filter column). See "Specifying Filter Columns for Materialized
View Logs: Example" and "Specifying Join Columns for Materialized View Logs:
Example" on page 15-34.
The following example uses the
materialized view from the preceding example as a master table to create a
materialized view tailored for a particular sales representative in the sample oe
schema:
Creating a Nested Materialized View: Example

CREATE MATERIALIZED VIEW my_warranty_orders
AS SELECT w.order_id, w.line_item_id, o.order_date
FROM warranty_orders w, orders o
WHERE o.order_id = o.order_id
AND o.sales_rep_id = 165;

15-26 Oracle Database SQL Language Reference

CREATE MATERIALIZED VIEW LOG

CREATE MATERIALIZED VIEW LOG
Purpose
15

Use the CREATE MATERIALIZED VIEW LOG statement to create a materialized view log,
which is a table associated with the master table of a materialized view.
The keyword SNAPSHOT is supported in place of MATERIALIZED
VIEW for backward compatibility.
Note:

When DML changes are made to master table data, Oracle Database stores rows
describing those changes in the materialized view log and then uses the materialized
view log to refresh materialized views based on the master table. This process is called
incremental or fast refresh. Without a materialized view log, Oracle Database must
reexecute the materialized view query to refresh the materialized view. This process is
called a complete refresh. Usually, a fast refresh takes less time than a complete
refresh.
A materialized view log is located in the master database in the same schema as the
master table. A master table can have only one materialized view log defined on it.
Oracle Database can use this materialized view log to perform fast refreshes for all
fast-refreshable materialized views based on the master table.
To fast refresh a materialized join view, you must create a materialized view log for
each of the tables referenced by the materialized view.
There are two types of materialized view logs: timestamp-based materialized view
logs and commit SCN-based materialized view logs. Timestamp-based materialized
view logs use timestamps and require some setup operations when preparing to
refresh the materialized view. Commit SCN-based materialized view logs use commit
SCN data rather than timestamps, which removes the need for the setup operations
and thus can improve the speed of the materialized view refresh. If you specify the
COMMIT SCN clause, then a commit SCN-based materialized view log is created.
Otherwise, a time-stamp based materialized view log is created.
Note that only new materialized view logs can take advantage of COMMIT SCN. Existing
materialized view logs cannot be altered to add COMMIT SCN unless they are dropped
and recreated. Refer to Oracle Database Data Warehousing Guide for more information.
See Also:
■

■

■

■

CREATE MATERIALIZED VIEW on page 15-4, ALTER
MATERIALIZED VIEW on page 11-3, Oracle Database Concepts,
Oracle Database Data Warehousing Guide, and Oracle Database
Advanced Replication for information on materialized views in
general
ALTER MATERIALIZED VIEW LOG on page 11-18 for
information on modifying a materialized view log
DROP MATERIALIZED VIEW LOG on page 17-57 for
information on dropping a materialized view log
Oracle Database Utilities for information on using direct loader logs

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-27

CREATE MATERIALIZED VIEW LOG

Prerequisites
15

The privileges required to create a materialized view log directly relate to the
privileges necessary to create the underlying objects associated with a materialized
view log.
■

■

If you own the master table, then you can create an associated materialized view
log if you have the CREATE TABLE privilege.
If you are creating a materialized view log for a table in another user's schema,
then you must have the CREATE ANY TABLE and COMMENT ANY TABLE system
privileges, as well as either the SELECT object privilege on the master table or the
SELECT ANY TABLE system privilege.

In either case, the owner of the materialized view log must have sufficient quota in the
tablespace intended to hold the materialized view log or must have the UNLIMITED
TABLESPACE system privilege.
See Also: Oracle Database Data Warehousing Guide for more
information about the prerequisites for creating a materialized view
log

Syntax
15

create_materialized_vw_log::=
schema
CREATE

MATERIALIZED

VIEW

LOG

.

ON

table

physical_attributes_clause
TABLESPACE

tablespace

logging_clause
CACHE
NOCACHE

parallel_clause

table_partitioning_clauses

,
OBJECT
PRIMARY

ID
KEY

ROWID
SEQUENCE
COMMIT
WITH

,

SCN
(

column

new_values_clause
)

mv_log_purge_clause
;

(physical_attributes_clause::= on page 15-8, logging_clause::= on page 15-29, parallel_
clause::= on page 15-29, table_partitioning_clauses::= on page 16-17 (in CREATE TABLE),
new_values_clause::= on page 15-29, mv_log_purge_clause::= on page 15-29.)

15-28 Oracle Database SQL Language Reference

CREATE MATERIALIZED VIEW LOG

physical_attributes_clause::=
PCTFREE

integer

PCTUSED

integer

INITRANS

integer

storage_clause

(storage_clause::= on page 8-50)
logging_clause::=
LOGGING
NOLOGGING
FILESYSTEM_LIKE_LOGGING

parallel_clause::=
NOPARALLEL
integer
PARALLEL

new_values_clause::=
INCLUDING
NEW

VALUES

EXCLUDING

mv_log_purge_clause::=
SYNCHRONOUS
ASYNCHRONOUS
IMMEDIATE
NEXT

datetime_expr

REPEAT

PURGE
START
START

WITH
WITH

INTERVAL

interval_expr

datetime_expr
datetime_expr

NEXT
REPEAT

datetime_expr
INTERVAL

interval_expr

Semantics
15

schema
Specify the schema containing the materialized view log master table. If you omit
schema, then Oracle Database assumes the master table is contained in your own
schema. Oracle Database creates the materialized view log in the schema of its master
table. You cannot create a materialized view log for a table in the schema of the user
SYS.

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-29

CREATE MATERIALIZED VIEW LOG

table
Specify the name of the master table for which the materialized view log is to be
created. Oracle Database encrypts any columns in the materialized view log that are
encrypted in the master table, using the same encryption algorithm.
Restrictions on Master Tables of Materialized View Logs

The following restrictions

apply to master tables of materialized view logs:
■

You cannot create a materialized view log for a temporary table or for a view.

■

You cannot create a materialized view log for a master table with a virtual column.
See Also: "Creating a Materialized View Log: Examples" on
page 15-33

physical_attributes_clause
Use the physical_attributes_clause to define physical and storage characteristics
for the materialized view log.
See Also: physical_attributes_clause on page 8-44 and storage_clause on
page 8-46 for a complete description these clauses, including default
values

TABLESPACE Clause
Specify the tablespace in which the materialized view log is to be created. If you omit
this clause, then the database creates the materialized view log in the default
tablespace of the schema of the materialized view log.

logging_clause
Specify either LOGGING or NOLOGGING to establish the logging characteristics for the
materialized view log. The default is the logging characteristic of the tablespace in
which the materialized view log resides.
See Also:

logging_clause on page 8-38 for a full description of this

clause

CACHE | NOCACHE
For data that will be accessed frequently, CACHE specifies that the blocks retrieved for
this log are placed at the most recently used end of the least recently used (LRU) list in
the buffer cache when a full table scan is performed. This attribute is useful for small
lookup tables.
NOCACHE specifies that the blocks are placed at the least recently used end of the LRU
list. The default is NOCACHE.
NOCACHE has no effect on materialized view logs for which you
specify KEEP in the storage_clause.

Note:

See Also: CREATE TABLE on page 16-6 for information about
specifying CACHE or NOCACHE

parallel_clause
The parallel_clause lets you indicate whether parallel operations will be supported
for the materialized view log.
15-30 Oracle Database SQL Language Reference

CREATE MATERIALIZED VIEW LOG

For complete information on this clause, refer to parallel_clause on page 16-63 in the
documentation on CREATE TABLE.

table_partitioning_clauses
Use the table_partitioning_clauses to indicate that the materialized view log is
partitioned on specified ranges of values or on a hash function. Partitioning of
materialized view logs is the same as partitioning of tables.
table_partitioning_clauses on page 16-51 in the CREATE
TABLE documentation

See Also:

WITH Clause
Use the WITH clause to indicate whether the materialized view log should record the
primary key, rowid, object ID, or a combination of these row identifiers when rows in
the master are changed. You can also use this clause to add a sequence to the
materialized view log to provide additional ordering information for its records.
This clause also specifies whether the materialized view log records additional
columns that might be referenced as filter columns, which are non-primary-key
columns referenced by subquery materialized views, or join columns, which are
non-primary-key columns that define a join in the subquery WHERE clause.
If you omit this clause, or if you specify the clause without PRIMARY KEY, ROWID, or
OBJECT ID, then the database stores primary key values by default. However, the
database does not store primary key values implicitly if you specify only OBJECT ID or
ROWID at create time. A primary key log, created either explicitly or by default,
performs additional checking on the primary key constraint.
OBJECT ID Specify OBJECT ID to indicate that the system-generated or user-defined
object identifier of every modified row should be recorded in the materialized view
log.
Restriction on OBJECT ID You can specify OBJECT ID only when creating a log on an

object table, and you cannot specify it for storage tables.
PRIMARY KEY Specify PRIMARY KEY to indicate that the primary key of all rows
changed should be recorded in the materialized view log.
ROWID Specify ROWID to indicate that the rowid of all rows changed should be
recorded in the materialized view log.
SEQUENCE Specify SEQUENCE to indicate that a sequence value providing additional
ordering information should be recorded in the materialized view log. Sequence
numbers are necessary to support fast refresh after some update scenarios.
See Also: Oracle Database Data Warehousing Guide for more
information on the use of sequence numbers in materialized view logs
and for examples that use this clause
COMMIT SCN Without the COMMIT SCN clause, the materialized view log is based on

timestamps and requires some setup operations when preparing to refresh the
materialized view. Specify COMMIT SCN to instruct the database to use commit SCN data
rather than timestamps. This setting removes the need for the setup operations and
thus can improve the speed of the materialized view refresh.

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-31

CREATE MATERIALIZED VIEW LOG

You can create the following types of local materialized views (including both ON
COMMIT and ON DEMAND) on master tables with commit SCN-based materialized view
logs:
■

Materialized aggregate views, including materialized aggregate views on a single
table

■

Materialized join views

■

Primary-key-based and rowid-based single table materialized views

■

UNION ALL materialized views, where each UNION ALL branch is one of the above
materialized view types

You cannot create remote materialized views on master tables with commit SCN-based
materialized view logs.
Restrictions on COMMIT SCN
■

■

The following restrictions apply to COMMIT SCN:

Use of COMMIT SCN on a table with one or more LOB columns is not supported and
causes ORA-32421.
Creating a materialized view on master tables with different types of materialized
view logs (that is, a master table with timestamp-based materialized view logs and
a master table with commit SCN-based materialized view logs) is not supported
and causes ORA-32414.

column Specify the columns whose values you want to be recorded in the
materialized view log for all rows that are changed. Typically these columns are filter
columns and join columns.
Restrictions on the WITH Clause
■

■

This clause is subject to the following restrictions:

You can specify only one PRIMARY KEY, one ROWID, one OBJECT ID, one SEQUENCE,
and one column list for each materialized view log.
Primary key columns are implicitly recorded in the materialized view log.
Therefore, you cannot specify any of the following combinations if column
contains one of the primary key columns:
WITH ... PRIMARY KEY ... (column)
WITH ... (column) ... PRIMARY KEY
WITH (column)

See Also:
■

■

■

CREATE MATERIALIZED VIEW on page 15-4 for information on
explicit and implicit inclusion of materialized view log values
Oracle Database Advanced Replication for more information about
filter columns and join columns
"Specifying Filter Columns for Materialized View Logs: Example"
on page 15-34 and "Specifying Join Columns for Materialized
View Logs: Example" on page 15-34

NEW VALUES Clause
The NEW VALUES clause lets you determine whether Oracle Database saves both old and
new values for update DML operations in the materialized view log.
See Also: "Including New Values in Materialized View Logs:
Example" on page 15-34

15-32 Oracle Database SQL Language Reference

CREATE MATERIALIZED VIEW LOG

INCLUDING Specify INCLUDING to save both new and old values in the log. If this log
is for a table on which you have a single-table materialized aggregate view, and if you
want the materialized view to be eligible for fast refresh, then you must specify
INCLUDING.

Specify EXCLUDING to disable the recording of new values in the log.
This is the default. You can use this clause to avoid the overhead of recording new
values. Do not use this clause if you have a fast-refreshable single-table materialized
aggregate view defined on the master table.

EXCLUDING

mv_log_purge_clause
Use this clause to specify the purge time for the materialized view log.
■

■

■

IMMEDIATE SYNCHRONOUS: the materialized view log is purged immediately after
refresh. This is the default.
IMMEDIATE ASYNCHRONOUS: the materialized view log is purged in a separate Oracle
Scheduler job after the refresh operation.
START WITH, NEXT, and REPEAT INTERVAL set up a scheduled purge that is
independent of the materialized view refresh and is initiated during CREATE or
ALTER MATERIALIZED VIEW LOG statement. This is very similar to scheduled refresh
syntax in a CREATE or ALTER MATERIALIZED VIEW statement:
–

The START WITH datetime expression specifies when the purge starts.

–

The NEXT datetime expression computes the next run time for the purge.

If you specify REPEAT INTERVAL, then the next run time will be: SYSDATE +
interval_expr.
A CREATE MATERIALIZED VIEW LOG statement with a scheduled purge creates an
Oracle Scheduler job to perform log purge. The job calls the DBMS_SNAPSHOT.PURGE_
LOG procedure to purge the materialized view logs. This process allows you to
amortize the purging costs over several materialized view refreshes.
Restriction on mv_log_purge_clause

This clause is not valid for materialized view

logs on temporary tables.
See Also: Oracle Database Data Warehousing Guide for more
information on purging materialized view logs

Examples
15

The following statement creates a
materialized view log on the oe.customers table that specifies physical and storage
characteristics:

Creating a Materialized View Log: Examples

CREATE MATERIALIZED VIEW LOG ON customers
PCTFREE 5
TABLESPACE example
STORAGE (INITIAL 10K);

The materialized view log on customers supports fast refresh for primary key
materialized views only.
The following statement creates another version of the materialized view log with the
ROWID clause, which enables fast refresh for more types of materialized views:
CREATE MATERIALIZED VIEW LOG ON customers WITH PRIMARY KEY, ROWID;

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-33

CREATE MATERIALIZED VIEW LOG

This materialized view log on customers makes fast refresh possible for rowid
materialized views and for materialized join views. To provide for fast refresh of
materialized aggregate views, you must also specify the SEQUENCE and INCLUDING NEW
VALUES clauses, as shown in the example that follows.
Specify a Purge Repeat Interval for a Materialized View Log: Example The following
statement creates a materialized view log on the oe.orders table. The contents of the
log will be purged once every five days, beginning five days after the creation date of
the materialized view log:
CREATE MATERIALIZED VIEW LOG ON orders
PCTFREE 5
TABLESPACE example
STORAGE (INITIAL 10K)
PURGE REPEAT INTERVAL '5' DAY;

Specifying Filter Columns for Materialized View Logs: Example The following
statement creates a materialized view log on the sh.sales table and is used in
"Creating Materialized Aggregate Views: Example" on page 15-23. It specifies as filter
columns all of the columns of the table referenced in that materialized view.
CREATE MATERIALIZED VIEW LOG ON sales
WITH ROWID, SEQUENCE(amount_sold, time_id, prod_id)
INCLUDING NEW VALUES;

Specifying Join Columns for Materialized View Logs: Example The following
statement creates a materialized view log on the order_items table of the sample oe
schema. The log records primary keys and product_id, which is used as a join column
in "Creating a Fast Refreshable Materialized View: Example" on page 15-25.
CREATE MATERIALIZED VIEW LOG ON order_items WITH (product_id);

Including New Values in Materialized View Logs: Example The following example
creates a materialized view log on the oe.product_information table that specifies
INCLUDING NEW VALUES:
CREATE MATERIALIZED VIEW LOG ON product_information
WITH ROWID, SEQUENCE (list_price, min_price, category_id), PRIMARY KEY
INCLUDING NEW VALUES;

You could create the following materialized aggregate view to use the product_
information log:
CREATE MATERIALIZED VIEW products_mv
REFRESH FAST ON COMMIT
AS SELECT SUM(list_price - min_price), category_id
FROM product_information
GROUP BY category_id;

This materialized view is eligible for fast refresh because the log defined on its master
table includes both old and new values.

15-34 Oracle Database SQL Language Reference

CREATE OPERATOR

CREATE OPERATOR
Purpose
15

Use the CREATE OPERATOR statement to create a new operator and define its bindings.
Operators can be referenced by indextypes and by SQL queries and DML statements.
The operators, in turn, reference functions, packages, types, and other user-defined
objects.
See Also: Oracle Database Data Cartridge Developer's Guide and Oracle
Database Concepts for a discussion of these dependencies and of
operators in general

Prerequisites
15

To create an operator in your own schema, you must have the CREATE OPERATOR system
privilege. To create an operator in another schema, you must have the CREATE ANY
OPERATOR system privilege. In either case, you must also have the EXECUTE object
privilege on the functions and operators referenced.

Syntax
15

create_operator::=
OR

REPLACE

schema

CREATE

.

OPERAT0R

operator

binding_clause

;

binding_clause::=
,
,
BINDING

(

implementation_clause

parameter_type

)

RETURN

return_type

using_function_clause

implementation_clause::=
,
,
ANCILLARY

TO

primary_operator

(

parameter_type

)

context_clause

context_clause::=
COMPUTE
WITH

INDEX

WITH

COLUMN

CONTEXT

,

SCAN

CONTEXT

ANCILLARY

DATA

implementation_type

CONTEXT

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-35

CREATE OPERATOR

using_function_clause::=
package
schema

.

type

.
.

USING

function_name

Semantics
15

OR REPLACE
Specify OR REPLACE to replace the definition of the operator schema object.
Restriction on Replacing an Operator You can replace the definition only if the
operator has no dependent objects, such as indextypes supporting the operator.

schema
Specify the schema containing the operator. If you omit schema, then the database
creates the operator in your own schema.

operator
Specify the name of the operator to be created. The name must satisfy the
requirements listed in "Database Object Naming Rules" on page 3-111.

binding_clause
Use the binding_clause to specify one or more parameter data types (parameter_
type) for binding the operator to a function. The signature of each binding—the
sequence of the data types of the arguments to the corresponding function—must be
unique according to the rules of overloading.
The parameter_type can itself be an object type. If it is, then you can optionally qualify
it with its schema.
Restriction on Binding Operators You cannot specify a parameter_type of REF, LONG,

or LONG RAW.
Oracle Database PL/SQL Language Reference for more
information about overloading

See Also:

RETURN Clause
Specify the return data type for the binding.
The return_type can itself be an object type. If so, then you can optionally qualify it
with its schema.
Restriction on Binding Return Data Type You cannot specify a return_type of REF,

LONG, or LONG RAW.

implementation_clause
Use this clause to describe the implementation of the binding.

15-36 Oracle Database SQL Language Reference

CREATE OPERATOR

ANCILLARY TO Clause
Use the ANCILLARY TO clause to indicate that the operator binding is ancillary to the
specified primary operator binding (primary_operator). If you specify this clause,
then do not specify a previous binding with just one number parameter.
context_clause
Use the context_clause to describe the functional implementation of a binding that is
not ancillary to a primary operator binding.
Use this clause to indicate that the
functional evaluation of the operator uses the index and a scan context that is specified
by the implementation type.

WITH INDEX CONTEXT, SCAN CONTEXT

COMPUTE ANCILLARY DATA Specify COMPUTE ANCILLARY DATA to indicate that the

operator binding computes ancillary data.
WITH COLUMN CONTEXT Specify WITH COLUMN CONTEXT to indicate that Oracle
Database should pass the column information to the functional implementation for the
operator.

If you specify this clause, then the signature of the function implemented must include
one extra ODCIFuncCallInfo structure.
See Also: Oracle Database Data Cartridge Developer's Guide for
instructions on using the ODCIFuncCallInfo routine

using_function_clause
The using_function_clause lets you specify the function that provides the
implementation for the binding. The function_name can be a standalone function,
packaged function, type method, or a synonym for any of these.
If the function is subsequently dropped, then the database marks all dependent objects
INVALID, including the operator. However, if you then subsequently issue an ALTER
OPERATOR ... DROP BINDING statement to drop the binding, then subsequent queries and
DML will revalidate the dependent objects.

Examples
15

This example creates a very simple
functional implementation of equality and then creates an operator that uses the
function. For a more complete set of examples, see Oracle Database Data Cartridge
Developer's Guide.

Creating User-Defined Operators: Example

CREATE FUNCTION eq_f(a VARCHAR2, b VARCHAR2) RETURN NUMBER AS
BEGIN
IF a = b THEN RETURN 1;
ELSE RETURN 0;
END IF;
END;
/
CREATE OPERATOR eq_op
BINDING (VARCHAR2, VARCHAR2)
RETURN NUMBER
USING eq_f;

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-37

CREATE OUTLINE

CREATE OUTLINE
Purpose
15

Stored outlines will be desupported in a future release in favor
of SQL plan management. In Oracle Database 11g Release 2 (11.2),
stored outlines continue to function as in past releases. Oracle strongly
recommends that you use SQL plan management for new
applications. SQL plan management creates SQL plan baselines,
which offer superior SQL performance stability compared with stored
outlines.

Note:

You can migrate existing stored outlines to SQL plan baselines by
using the MIGRATE_STORED_OUTLINE function of the DBMS_SPM package
or Enterprise Manager DB Control. When the migration is complete,
the stored outlines are marked as migrated and can be removed. You
can drop all migrated stored outlines on your system by using the
DROP_MIGRATED_STORED_OUTLINE function of the DBMS_SPM package.
See Also: Oracle Database Performance Tuning Guide for more
information about SQL plan management and Oracle Database PL/SQL
Packages and Types Reference for information about the DBMS_SPM
package
Use the CREATE OUTLINE statement to create a stored outline, which is a set of
attributes used by the optimizer to generate an execution plan. You can then instruct
the optimizer to use a set of outlines to influence the generation of execution plans
whenever a particular SQL statement is issued, regardless of changes in factors that
can affect optimization. You can also modify an outline so that it takes into account
changes in these factors.
The SQL statement you want to affect must be an exact string
match of the statement specified when creating the outline.

Note:

See Also:
■

■

■

Oracle Database Performance Tuning Guide for information on
execution plans
ALTER OUTLINE on page 11-28 for information on modifying an
outline
ALTER SESSION on page 11-45 and ALTER SYSTEM on
page 11-58 for information on the USE_STORED_OUTLINES and USE_
PRIVATE_OUTLINES parameters

Prerequisites
15

To create a public or private outline, you must have the CREATE ANY OUTLINE system
privilege.
If you are creating a clone outline from a source outline, then you must also have the
SELECT_CATALOG_ROLE role.

15-38 Oracle Database SQL Language Reference

CREATE OUTLINE

You can enable or disable the use of stored outlines dynamically for an individual
session or for the system:
■

Enable the USE_STORED_OUTLINES parameter to use public outlines.

■

Enable the USE_PRIVATE_OUTLINES parameter to use private stored outlines.
See Also: Oracle Database Performance Tuning Guide for information
on using outlines for performance tuning

Syntax
15

create_outline::=
PUBLIC
OR

REPLACE

PRIVATE

CREATE

outline
OUTLINE

PUBLIC
PRIVATE
FROM

FOR

source_outline

CATEGORY

category

ON

statement
;

Note: None of the clauses after outline are required. However, you
must specify at least one clause after outline, and it must be either
the FROM clause or the ON clause.

Semantics
15

OR REPLACE
Specify OR REPLACE to replace an existing outline with a new outline of the same name.

PUBLIC | PRIVATE
Specify PUBLIC if you are creating an outline for use by PUBLIC. This is the default.
Specify PRIVATE to create an outline for private use by the current session only. The
data of this outline is stored in the current schema.

outline
Specify the unique name to be assigned to the stored outline. If you do not specify
outline, then the database generates an outline name.
See Also:

"Creating an Outline: Example" on page 15-40

FROM source_outline Clause
Use the FROM clause to create a new outline by copying an existing one. By default,
Oracle Database looks for source_category in the public area. If you specify PRIVATE,
then the database looks for the outline in the current schema.

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-39

CREATE OUTLINE

Restriction on Copying an Outline If you specify the FROM clause, then you cannot

specify the ON clause.
See Also: "Creating a Private Clone Outline: Example" on page 15-40
and "Publicizing a Private Outline to the Public Area: Example" on
page 15-40

FOR CATEGORY Clause
Specify an optional name used to group stored outlines. For example, you could
specify a category of outlines for end-of-week use and another for end-of-quarter use.
If you do not specify category, then the outline is stored in the DEFAULT category.

ON Clause
Specify the SQL statement for which the database will create an outline when the
statement is compiled. This clause is optional only if you are creating a copy of an
existing outline using the FROM clause.
You can specify any one of the following statements: SELECT, DELETE, UPDATE, INSERT ...
SELECT, CREATE TABLE ... AS SELECT.
Restrictions on the ON Clause

This clause is subject to the following restrictions:

■

If you specify the ON clause, then you cannot specify the FROM clause.

■

You cannot create an outline on a multitable INSERT statement.

■

The SQL statement in the ON clause cannot include any DML operation on a
remote object.
In subsequent statements, you can specify additional outlines
for the same SQL statement, but each outline for the same statement
must specify a different category in the CATEGORY clause.

Note:

Examples
15

The following statement creates a stored outline by
compiling the ON statement. The outline is called salaries and is stored in the
category special.
Creating an Outline: Example

CREATE OUTLINE salaries FOR CATEGORY special
ON SELECT last_name, salary FROM employees;

When this same SELECT statement is subsequently compiled, if the USE_STORED_
OUTLINES parameter is set to special, the database generates the same execution plan
as was generated when the outline salaries was created.
Creating a Private Clone Outline: Example The following statement creates a stored
private outline my_salaries based on the public category salaries created in the
preceding example.
CREATE OR REPLACE PRIVATE OUTLINE my_salaries
FROM salaries;

Publicizing a Private Outline to the Public Area: Example The following statement
copies back (publicizes) a private outline to the public area after private editing:
CREATE OR REPLACE OUTLINE public_salaries

15-40 Oracle Database SQL Language Reference

CREATE OUTLINE

FROM PRIVATE my_salaries;

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-41

CREATE PACKAGE

CREATE PACKAGE
Purpose
15

Packages are defined using PL/SQL. Therefore, this section provides some general
information but refers to Oracle Database PL/SQL Language Reference for details of
syntax and semantics.
Use the CREATE PACKAGE statement to create the specification for a stored package,
which is an encapsulated collection of related procedures, functions, and other
program objects stored together in the database. The package specification declares
these objects. The package body, specified subsequently, defines these objects.
See Also:
■

■

■

■

CREATE PACKAGE BODY on page 15-44 for information on
specifying the implementation of the package
CREATE FUNCTION on page 14-58 and CREATE PROCEDURE
on page 15-48 for information on creating standalone functions
and procedures
ALTER PACKAGE on page 11-30 and DROP PACKAGE on
page 17-62 for information on modifying and dropping a package
Oracle Database Advanced Application Developer's Guide and Oracle
Database PL/SQL Packages and Types Reference for detailed
discussions of packages and how to use them

Prerequisites
15

To create or replace a package in your own schema, you must have the CREATE
PROCEDURE system privilege. To create or replace a package in another user's schema,
you must have the CREATE ANY PROCEDURE system privilege.
To embed a CREATE PACKAGE statement inside an Oracle Database precompiler
program, you must terminate the statement with the keyword END-EXEC followed by
the embedded SQL statement terminator for the specific language.
See Also:

Oracle Database PL/SQL Language Reference for more

information

Syntax
15

Packages are defined using PL/SQL. Therefore, the syntax diagram in this book shows
only the SQL keywords. Refer to Oracle Database PL/SQL Language Reference for the
PL/SQL syntax, semantics, and examples.
create_package::=
OR
CREATE

REPLACE
PACKAGE

plsql_source

(plsql_source: See Oracle Database PL/SQL Language Reference.)

15-42 Oracle Database SQL Language Reference

CREATE PACKAGE

Semantics
15

OR REPLACE
Specify OR REPLACE to re-create the package specification if it already exists. Use this
clause to change the specification of an existing package without dropping,
re-creating, and regranting object privileges previously granted on the package. If you
change a package specification, then Oracle Database recompiles it.
Users who had previously been granted privileges on a redefined package can still
access the package without being regranted the privileges.
If any function-based indexes depend on the package, then the database marks the
indexes DISABLED.
ALTER PACKAGE for information on recompiling package
specifications

See Also:

plsql_source
See Oracle Database PL/SQL Language Reference for the syntax and semantics of the
plsql_source, including examples.

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-43

CREATE PACKAGE BODY

CREATE PACKAGE BODY
Purpose
15

Package bodies are defined using PL/SQL. Therefore, this section provides some
general information but refers to Oracle Database PL/SQL Language Reference for details
of syntax and semantics.
Use the CREATE PACKAGE BODY statement to create the body of a stored package, which
is an encapsulated collection of related procedures, stored functions, and other
program objects stored together in the database. The package body defines these
objects. The package specification, defined in an earlier CREATE PACKAGE statement,
declares these objects.
Packages are an alternative to creating procedures and functions as standalone schema
objects.
See Also:
■

■

■

■

CREATE FUNCTION on page 14-58 and CREATE PROCEDURE
on page 15-48 for information on creating standalone functions
and procedures
CREATE PACKAGE on page 15-42 for a discussion of packages,
including how to create packages
ALTER PACKAGE on page 11-30 for information on modifying a
package
DROP PACKAGE on page 17-62 for information on removing a
package from the database

Prerequisites
15

To create or replace a package in your own schema, you must have the CREATE
PROCEDURE system privilege. To create or replace a package in another user's schema,
you must have the CREATE ANY PROCEDURE system privilege. In both cases, the package
body must be created in the same schema as the package.
To embed a CREATE PACKAGE BODY statement inside an Oracle Database precompiler
program, you must terminate the statement with the keyword END-EXEC followed by
the embedded SQL statement terminator for the specific language.
See Also:

Oracle Database PL/SQL Language Reference

Syntax
15

Package bodies are defined using PL/SQL. Therefore, the syntax diagram in this book
shows only the SQL keywords. Refer to Oracle Database PL/SQL Language Reference for
the PL/SQL syntax, semantics, and examples.
create_package_body::=
OR
CREATE

REPLACE
PACKAGE

BODY

plsql_source

(plsql_source: See Oracle Database PL/SQL Language Reference.)

15-44 Oracle Database SQL Language Reference

CREATE PACKAGE BODY

Semantics
15

OR REPLACE
Specify OR REPLACE to re-create the package body if it already exists. Use this clause to
change the body of an existing package without dropping, re-creating, and regranting
object privileges previously granted on it. If you change a package body, then Oracle
Database recompiles it.
Users who had previously been granted privileges on a redefined package can still
access the package without being regranted the privileges.
See Also: ALTER PACKAGE on page 11-30 for information on
recompiling package bodies

plsql_source
See Oracle Database PL/SQL Language Reference for the syntax and semantics of the
plsql_source.

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-45

CREATE PFILE

CREATE PFILE
Purpose
15

Use the CREATE PFILE statement to export either a binary server parameter file or the
current in-memory parameter settings into a text initialization parameter file. Creating
a text parameter file is a convenient way to get a listing of the current parameter
settings being used by the database, and it lets you edit the file easily in a text editor
and then convert it back into a server parameter file using the CREATE SPFILE
statement.
Upon successful execution of this statement, Oracle Database creates a text parameter
file on the server. In an Oracle Real Application Clusters environment, it will contain
all parameter settings of all instances. It will also contain any comments that appeared
on the same line with a parameter setting in the server parameter file.
See Also:

CREATE SPFILE on page 15-71 for information on server
parameter files

■

Oracle Database Administrator's Guide for additional information on
text initialization parameter files and binary server parameter files

■

Oracle Real Application Clusters Administration and Deployment
Guide for information on using server parameter files in an Oracle
Real Application Clusters environment

■

Prerequisites
15

You must have the SYSDBA or the SYSOPER role to execute this statement. You can
execute this statement either before or after instance startup.

Syntax
15

create_pfile::=
=
=
CREATE

’

pfile_name

’

PFILE

’

spfile_name

’

SPFILE
FROM

;
MEMORY

Semantics
15

pfile_name
Specify the name of the text parameter file you want to create. If you do not specify
pfile_name, then Oracle Database uses the platform-specific default initialization
parameter file name. pfile_name can include a path prefix. If you do not specify such a
path prefix, then the database adds the path prefix for the default storage location,
which is platform dependent.

spfile_name
Specify the name of the binary server parameter from which you want to create a text
file.

15-46 Oracle Database SQL Language Reference

CREATE PFILE

■

■

If you specify spfile_name, then the file must exist on the server. If the file does
not reside in the default directory for server parameter files on your operating
system, then you must specify the full path.
If you do not specify spfile_name, then the database uses the spfile that is
currently associated with the instance, usually the one that was used a startup. If
no spfile is associated with the instance, then the database looks for the
platform-specific default server parameter file name. If that file does not exist, then
the database returns an error.
See Also: the appropriate operating-system-specific documentation
for default parameter file names

MEMORY
Specify MEMORY to create a pfile using the current system-wide parameter settings. In
an Oracle RAC environment, the created file will contain the parameter settings from
each instance.

Examples
15

Creating a Parameter File: Example The following example creates a text parameter
file my_init.ora from a binary server parameter file s_params.ora:
CREATE PFILE = 'my_init.ora' FROM SPFILE = 's_params.ora';

Typically you will need to specify the full path and filename
for parameter files on your operating system. Refer to your Oracle
operating system documentation for path information.

Note:

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-47

CREATE PROCEDURE

CREATE PROCEDURE
Purpose
15

Procedures are defined using PL/SQL. Therefore, this section provides some general
information but refers to Oracle Database PL/SQL Language Reference for details of
syntax and semantics.
Use the CREATE PROCEDURE statement to create a standalone stored procedure or a call
specification.
A procedure is a group of PL/SQL statements that you can call by name. A call
specification (sometimes called call spec) declares a Java method or a third-generation
language (3GL) routine so that it can be called from SQL and PL/SQL. The call spec
tells Oracle Database which Java method to invoke when a call is made. It also tells the
database what type conversions to make for the arguments and return value.
Stored procedures offer advantages in the areas of development, integrity, security,
performance, and memory allocation.
See Also:
■

■

■

■

■

Oracle Database Advanced Application Developer's Guide for more
information on stored procedures, including how to call stored
procedures and for information about registering external
procedures.
CREATE FUNCTION on page 14-58 for information specific to
functions, which are similar to procedures in many ways.
CREATE PACKAGE on page 15-42 for information on creating
packages. The CREATE PROCEDURE statement creates a procedure as
a standalone schema object. You can also create a procedure as
part of a package.
ALTER PROCEDURE on page 11-31 and DROP PROCEDURE on
page 17-64 for information on modifying and dropping a
standalone procedure.
CREATE LIBRARY on page 15-2 for more information about
shared libraries.

Prerequisites
15

To create or replace a procedure in your own schema, you must have the CREATE
PROCEDURE system privilege. To create or replace a procedure in another user's schema,
you must have the CREATE ANY PROCEDURE system privilege.
To invoke a call spec, you may need additional privileges, for example, the EXECUTE
object privilege on the C library for a C call spec.
To embed a CREATE PROCEDURE statement inside an Oracle precompiler program, you
must terminate the statement with the keyword END-EXEC followed by the embedded
SQL statement terminator for the specific language.
Oracle Database PL/SQL Language Reference or Oracle
Database Java Developer's Guide for more information

See Also:

15-48 Oracle Database SQL Language Reference

CREATE PROCEDURE

Syntax
15

Procedures are defined using PL/SQL. Therefore, the syntax diagram in this book
shows only the SQL keywords. Refer to Oracle Database PL/SQL Language Reference for
the PL/SQL syntax, semantics, and examples.
create_procedure::=
OR

REPLACE

CREATE

PROCEDURE

plsql_source

(plsql_source: See Oracle Database PL/SQL Language Reference.)

Semantics
15

OR REPLACE
Specify OR REPLACE to re-create the procedure if it already exists. Use this clause to
change the definition of an existing procedure without dropping, re-creating, and
regranting object privileges previously granted on it. If you redefine a procedure, then
Oracle Database recompiles it.
Users who had previously been granted privileges on a redefined procedure can still
access the procedure without being regranted the privileges.
If any function-based indexes depend on the procedure, then Oracle Database marks
the indexes DISABLED.
See Also: ALTER PROCEDURE on page 11-31 for information on
recompiling procedures

plsql_source
See Oracle Database PL/SQL Language Reference for the syntax and semantics of the
plsql_source.

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-49

CREATE PROFILE

CREATE PROFILE
Oracle recommends that you use the Database Resource
Manager rather than this SQL statement to establish resource limits.
The Database Resource Manager offers a more flexible means of
managing and tracking resource use. For more information on the
Database Resource Manager, refer to Oracle Database Administrator's
Guide.

Note:

Purpose
15

Use the CREATE PROFILE statement to create a profile, which is a set of limits on
database resources. If you assign the profile to a user, then that user cannot exceed
these limits.
See Also: Oracle Database Security Guide for a detailed description
and explanation of how to use password management and protection

Prerequisites
15

To create a profile, you must have the CREATE PROFILE system privilege.
To specify resource limits for a user, you must:
■

Enable resource limits dynamically with the ALTER SYSTEM statement or with the
initialization parameter RESOURCE_LIMIT. This parameter does not apply to
password resources. Password resources are always enabled.

■

Create a profile that defines the limits using the CREATE PROFILE statement

■

Assign the profile to the user using the CREATE USER or ALTER USER statement
See Also:
■

■

■

ALTER SYSTEM on page 11-58 for information on enabling
resource limits dynamically
Oracle Database Reference for information on the RESOURCE_LIMIT
parameter
CREATE USER on page 17-7 and ALTER USER on page 13-6 for
information on profiles

Syntax
15

create_profile::=
resource_parameters
CREATE

PROFILE

profile

LIMIT

;
password_parameters

15-50 Oracle Database SQL Language Reference

CREATE PROFILE

resource_parameters::=
SESSIONS_PER_USER
CPU_PER_SESSION
CPU_PER_CALL
integer
CONNECT_TIME
UNLIMITED
IDLE_TIME
DEFAULT
LOGICAL_READS_PER_SESSION
LOGICAL_READS_PER_CALL
COMPOSITE_LIMIT
size_clause
PRIVATE_SGA

UNLIMITED
DEFAULT

(size_clause::= on page 8-47
password_parameters::=
FAILED_LOGIN_ATTEMPTS
PASSWORD_LIFE_TIME
expr
PASSWORD_REUSE_TIME
UNLIMITED
PASSWORD_REUSE_MAX
DEFAULT
PASSWORD_LOCK_TIME
PASSWORD_GRACE_TIME
function
PASSWORD_VERIFY_FUNCTION

NULL
DEFAULT

Semantics
15

profile
Specify the name of the profile to be created. The name must satisfy the requirements
listed in "Database Object Naming Rules" on page 3-111. Use profiles to limit the
database resources available to a user for a single call or a single session.
Oracle Database enforces resource limits in the following ways:
■

■

If a user exceeds the CONNECT_TIME or IDLE_TIME session resource limit, then the
database rolls back the current transaction and ends the session. When the user
process next issues a call, the database returns an error.
If a user attempts to perform an operation that exceeds the limit for other session
resources, then the database aborts the operation, rolls back the current statement,
and immediately returns an error. The user can then commit or roll back the
current transaction, and must then end the session.

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-51

CREATE PROFILE

■

If a user attempts to perform an operation that exceeds the limit for a single call,
then the database aborts the operation, rolls back the current statement, and
returns an error, leaving the current transaction intact.
Notes:
■

■

You can use fractions of days for all parameters that limit time,
with days as units. For example, 1 hour is 1/24 and 1 minute is
1/1440.
You can specify resource limits for users regardless of whether the
resource limits are enabled. However, Oracle Database does not
enforce the limits until you enable them.

See Also:

"Creating a Profile: Example" on page 15-54

UNLIMITED
When specified with a resource parameter, UNLIMITED indicates that a user assigned
this profile can use an unlimited amount of this resource. When specified with a
password parameter, UNLIMITED indicates that no limit has been set for the parameter.

DEFAULT
Specify DEFAULT if you want to omit a limit for this resource in this profile. A user
assigned this profile is subject to the limit for this resource specified in the DEFAULT
profile. The DEFAULT profile initially defines unlimited resources. You can change those
limits with the ALTER PROFILE statement.
Any user who is not explicitly assigned a profile is subject to the limits defined in the
DEFAULT profile. Also, if the profile that is explicitly assigned to a user omits limits for
some resources or specifies DEFAULT for some limits, then the user is subject to the
limits on those resources defined by the DEFAULT profile.

resource_parameters
SESSIONS_PER_USER

Specify the number of concurrent sessions to which you

want to limit the user.
CPU_PER_SESSION

Specify the CPU time limit for a session, expressed in

hundredth of seconds.
CPU_PER_CALL Specify the CPU time limit for a call (a parse, execute, or fetch),
expressed in hundredths of seconds.
CONNECT_TIME Specify the total elapsed time limit for a session, expressed in
minutes.
IDLE_TIME Specify the permitted periods of continuous inactive time during a

session, expressed in minutes. Long-running queries and other operations are not
subject to this limit.
Specify the permitted number of data blocks read
in a session, including blocks read from memory and disk.

LOGICAL_READS_PER_SESSION

15-52 Oracle Database SQL Language Reference

CREATE PROFILE

Specify the permitted number of data blocks read for
a call to process a SQL statement (a parse, execute, or fetch).

LOGICAL_READS_PER_CALL

Specify the amount of private space a session can allocate in the
shared pool of the system global area (SGA). Refer to size_clause on page 8-47 for
information on that clause.

PRIVATE_SGA

This limit applies only if you are using shared server
architecture. The private space for a session in the SGA includes
private SQL and PL/SQL areas, but not shared SQL and PL/SQL
areas.

Note:

Specify the total resource cost for a session, expressed in service
units. Oracle Database calculates the total service units as a weighted sum of CPU_PER_
SESSION, CONNECT_TIME, LOGICAL_READS_PER_SESSION, and PRIVATE_SGA.

COMPOSITE_LIMIT

See Also:
■

■

ALTER RESOURCE COST on page 11-35 for information on how
to specify the weight for each session resource
"Setting Profile Resource Limits: Example" on page 15-54

password_parameters
Use the following clauses to set password parameters. Parameters that set lengths of
time—that is, all the password parameters except FAILED_LOGIN_ATTEMPTS and
PASSWORD_REUSE_MAX—are interpreted in number of days. For testing purposes you
can specify minutes (n/1440) or even seconds (n/86400) for these parameters. You can
also use a decimal value for this purpose (for example .0833 for approximately one
hour). The minimum value is 1 second. The maximum value is 24855 days. For
FAILED_LOGIN_ATTEMPTS and PASSWORD_REUSE_MAX, you must specify an integer.
Specify the number of consecutive failed attempts to
log in to the user account before the account is locked. If you omit this clause, then the
default is 10 times.

FAILED_LOGIN_ATTEMPTS

PASSWORD_LIFE_TIME Specify the number of days the same password can be used
for authentication. If you also set a value for PASSWORD_GRACE_TIME, then the password
expires if it is not changed within the grace period, and further connections are
rejected. If you omit this clause, then the default is 180 days.

Oracle Database Security Guide for information on setting
PASSWORD_LIFE_TIME to a low value
See Also:

PASSWORD_REUSE_TIME and PASSWORD_REUSE_MAX These two parameters
must be set in conjunction with each other. PASSWORD_REUSE_TIME specifies the number
of days before which a password cannot be reused. PASSWORD_REUSE_MAX specifies the
number of password changes required before the current password can be reused. For
these parameter to have any effect, you must specify a value for both of them.
■

If you specify a value for both of these parameters, then the user cannot reuse a
password until the password has been changed the number of times specified for
PASSWORD_REUSE_MAX during the number of days specified for PASSWORD_REUSE_
TIME.

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-53

CREATE PROFILE

For example, if you specify PASSWORD_REUSE_TIME to 30 and PASSWORD_REUSE_MAX
to 10, then the user can reuse the password after 30 days if the password has
already been changed 10 times.
■

■

■

If you specify a value for either of these parameters and specify UNLIMITED for the
other, then the user can never reuse a password.
If you specify DEFAULT for either parameter, then Oracle Database uses the value
defined in the DEFAULT profile. By default, all parameters are set to UNLIMITED in
the DEFAULT profile. If you have not changed the default setting of UNLIMITED in
the DEFAULT profile, then the database treats the value for that parameter as
UNLIMITED.
If you set both of these parameters to UNLIMITED, then the database ignores both of
them. This is the default if you omit both parameters.

Specify the number of days an account will be locked
after the specified number of consecutive failed login attempts. If you omit this clause,
then the default is 1 day.

PASSWORD_LOCK_TIME

Specify the number of days after the grace period begins
during which a warning is issued and login is allowed. If you omit this clause, then
the default is 7 days.

PASSWORD_GRACE_TIME

PASSWORD_VERIFY_FUNCTION The PASSWORD_VERIFY_FUNCTION clause lets a
PL/SQL password complexity verification script be passed as an argument to the
CREATE PROFILE statement. Oracle Database provides a default script, but you can
create your own routine or use third-party software instead.
■

■

For function, specify the name of the password complexity verification routine.
The function must exist in the SYS schema and you must have EXECUTE privilege
on the function.
Specify NULL to indicate that no password verification is performed.

If you specify expr for any of the password parameters, then the expression can be of
any form except scalar subquery expression.
Restriction on Password Parameters When you assign a profile to an external user

or a global user, the password parameters do not take effect for that user.
See Also:

"Setting Profile Password Limits: Example" on page 15-55

Examples
15

Creating a Profile: Example The following statement creates the profile new_profile:
CREATE PROFILE new_profile
LIMIT PASSWORD_REUSE_MAX 10
PASSWORD_REUSE_TIME 30;

Setting Profile Resource Limits: Example

app_user:
CREATE PROFILE app_user LIMIT
SESSIONS_PER_USER
CPU_PER_SESSION
CPU_PER_CALL
CONNECT_TIME
LOGICAL_READS_PER_SESSION

15-54 Oracle Database SQL Language Reference

UNLIMITED
UNLIMITED
3000
45
DEFAULT

The following statement creates the profile

CREATE PROFILE

LOGICAL_READS_PER_CALL
PRIVATE_SGA
COMPOSITE_LIMIT

1000
15K
5000000;

If you assign the app_user profile to a user, then the user is subject to the following
limits in subsequent sessions:
■

The user can have any number of concurrent sessions.

■

In a single session, the user can consume an unlimited amount of CPU time.

■

A single call made by the user cannot consume more than 30 seconds of CPU time.

■

A single session cannot last for more than 45 minutes.

■

■

■
■

■

In a single session, the number of data blocks read from memory and disk is
subject to the limit specified in the DEFAULT profile.
A single call made by the user cannot read more than 1000 data blocks from
memory and disk.
A single session cannot allocate more than 15 kilobytes of memory in the SGA.
In a single session, the total resource cost cannot exceed 5 million service units.
The formula for calculating the total resource cost is specified by the ALTER
RESOURCE COST statement.
Since the app_user profile omits a limit for IDLE_TIME and for password limits, the
user is subject to the limits on these resources specified in the DEFAULT profile.

The following statement creates the app_
user2 profile with password limits values set:

Setting Profile Password Limits: Example

CREATE PROFILE app_user2 LIMIT
FAILED_LOGIN_ATTEMPTS 5
PASSWORD_LIFE_TIME 60
PASSWORD_REUSE_TIME 60
PASSWORD_REUSE_MAX 5
PASSWORD_VERIFY_FUNCTION verify_function
PASSWORD_LOCK_TIME 1/24
PASSWORD_GRACE_TIME 10;

This example uses the default Oracle Database password verification function,
verify_function. Refer to Oracle Database Security Guide for information on using this
verification function provided or designing your own verification function.

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-55

CREATE RESTORE POINT

CREATE RESTORE POINT
Purpose
15

Use the CREATE RESTORE POINT statement to create a restore point, which is a name
associated with a timestamp or an SCN of the database. A restore point can be used to
flash back a table or the database to the time specified by the restore point without the
need to determine the SCN or timestamp. Restore points are also useful in various
RMAN operations, including backups and database duplication. You can use RMAN
to create restore points in the process of implementing an archival backup.
See Also:
■

■

Oracle Database Backup and Recovery User's Guide for more
information on creating and using restore points and guaranteed
restore points, for information on database duplication, and for
information on archival backups
FLASHBACK DATABASE on page 18-24, FLASHBACK TABLE
on page 18-27, and DROP RESTORE POINT on page 17-66 for
information on using and dropping restore points

Prerequisites
15

To create a normal restore point, you must have either SELECT ANY DICTIONARY or
FLASHBACK ANY TABLE privilege. To create a guaranteed restore point, you must have
the SYSDBA system privileges.
To view or use a restore point, you must have the SELECT ANY DICTIONARY or FLASHBACK
ANY TABLE system privilege or the SELECT_CATALOG_ROLE role.
You can create a restore point on a primary or standby database. The database can be
open, or mounted but not open. If the database is mounted, then it must have been
shut down consistently before being mounted unless it is a physical standby database.
You must have created a fast recovery area before creating a guaranteed restore point.
You need not enable flashback database before you create the restore point. The
database must be in ARCHIVELOG mode if you are creating a guaranteed restore point.

Syntax
15

create_restore_point::=
TIMESTAMP
AS

OF

expr
SCN

CREATE

RESTORE

POINT

restore_point

PRESERVE
GUARANTEE

FLASHBACK

DATABASE
;

15-56 Oracle Database SQL Language Reference

CREATE RESTORE POINT

Semantics
15

restore_point
Specify the name of the restore point. The name is a character value of up to 128
characters.
The database can retain at least 2048 normal restore points. Normal restore points are
retained in the database for at least the number of days specified for the CONTROL_
FILE_RECORD_KEEP_TIME initialization parameter. The default value of that parameter
is 7 days. Guaranteed restore points are retained in the database until explicitly
dropped by the user.
If you specify neither PRESERVE nor GUARANTEE FLASHBACK DATABASE, then the
resulting restore point enables you to flash the database back to a restore point within
the time period determined by the DB_FLASHBACK_RETENTION_TARGET initialization
parameter. The database automatically manages such restore points. When the
maximum number of restore points is reached, according to the rules described in
restore_point above, the database automatically drops the oldest restore point.
Under some circumstances the restore points will be retained in the RMAN recovery
catalog for use in restoring long-term backups. You can explicitly drop a restore point
using the DROP RESTORE POINT statement.

AS OF Clause
Use this clause to create a restore point at a specified datetime or SCN in the past. If
you specify TIMESTAMP, then expr must be a valid datetime expression resolving to a
time in the past. If you specify SCN, then expr must be a valid SCN in the database in
the past. In either case, expr must refer to a datetime or SCN in the current incarnation
of the database.

PRESERVE
Specify PRESERVE to indicate that the restore point must be explicitly deleted. Such
restore points are useful for preserving a flashback database.

GUARANTEE FLASHBACK DATABASE
A guaranteed restore point enables you to flash the database back deterministically to
the restore point regardless of the DB_FLASHBACK_RETENTION_TARGET initialization
parameter setting. The guaranteed ability to flash back depends on sufficient space
being available in the fast recovery area.
Guaranteed restore points guarantee only that the database will maintain enough
flashback logs to flashback the database to the guaranteed restore point. It does not
guarantee that the database will have enough undo to flashback any table to the same
restore point.
Guaranteed restore points are always preserved. They must be dropped explicitly by
the user using the DROP RESTORE POINT statement. They do not age out. Guaranteed
restore points can use considerable space in the fast recovery area. Therefore, Oracle
recommends that you create guaranteed restore points only after careful consideration.

Examples
15

The following example creates a
normal restore point, updates a table, and then flashes back the altered table to the
restore point. The example assumes the user hr has the appropriate system privileges
to use each of the statements.

Creating and Using a Restore Point: Example

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-57

CREATE RESTORE POINT

CREATE RESTORE POINT good_data;
SELECT salary FROM employees WHERE employee_id = 108;
SALARY
---------12000
UPDATE employees SET salary = salary*10
WHERE employee_id = 108;
SELECT salary FROM employees
WHERE employee_id = 108;
SALARY
---------120000
COMMIT;
FLASHBACK TABLE employees TO RESTORE POINT good_data;
SELECT salary FROM employees
WHERE employee_id = 108;
SALARY
---------12000

15-58 Oracle Database SQL Language Reference

CREATE ROLE

CREATE ROLE
Purpose
15

Use the CREATE ROLE statement to create a role, which is a set of privileges that can be
granted to users or to other roles. You can use roles to administer database privileges.
You can add privileges to a role and then grant the role to a user. The user can then
enable the role and exercise the privileges granted by the role.
A role contains all privileges granted to the role and all privileges of other roles
granted to it. A new role is initially empty. You add privileges to a role with the GRANT
statement.
If you create a role that is NOT IDENTIFIED or is IDENTIFIED EXTERNALLY or BY password,
then Oracle Database grants you the role with ADMIN OPTION. However, if you create a
role IDENTIFIED GLOBALLY, then the database does not grant you the role. A global role
cannot be granted to a user or role directly. Global roles can be granted only through
enterprise roles.
See Also:
■

GRANT on page 18-33 for information on granting roles

■

ALTER USER on page 13-6 for information on enabling roles

■

■

■
■

ALTER ROLE on page 11-38 and DROP ROLE on page 17-67 for
information on modifying or removing a role from the database
SET ROLE on page 19-61 for information on enabling and
disabling roles for the current session
Oracle Database Security Guide for general information about roles
Oracle Database Enterprise User Security Administrator's Guide for
details on enterprise roles

Prerequisites
15

You must have the CREATE ROLE system privilege.

Syntax
15

create_role::=
NOT

IDENTIFIED
BY

password
schema

IDENTIFIED

USING

.
package

EXTERNALLY
GLOBALLY
CREATE

ROLE

role

;

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-59

CREATE ROLE

Semantics
15

role
Specify the name of the role to be created. Oracle recommends that the role contain at
least one single-byte character regardless of whether the database character set also
contains multibyte characters. The maximum length of the role name is 30 bytes. The
maximum number of user-defined roles that can be enabled for a single user at one
time is 148.
Some roles are defined by SQL scripts provided on your distribution media.
See Also: GRANT on page 18-33 for a list of these predefined roles
and SET ROLE on page 19-61 for information on enabling and
disabling roles for a user

NOT IDENTIFIED Clause
Specify NOT IDENTIFIED to indicate that this role is authorized by the database and that
no password is required to enable the role.

IDENTIFIED Clause
Use the IDENTIFIED clause to indicate that a user must be authorized by the specified
method before the role is enabled with the SET ROLE statement.
BY password The BY password clause lets you create a local role and indicates that
the user must specify the password to the database when enabling the role. The
password can contain only single-byte characters from your database character set
regardless of whether this character set also contains multibyte characters.
USING package The USING package clause lets you create a secure application role,
which is a role that can be enabled only by applications using an authorized package.
If you do not specify schema, then the database assumes the package is in your own
schema.
See Also: Oracle Database Security Guide for information on creating
a secure application role

Specify EXTERNALLY to create an external role. An external user must
be authorized by an external service, such as an operating system or third-party
service, before enabling the role.

EXTERNALLY

Depending on the operating system, the user may have to specify a password to the
operating system before the role is enabled.
GLOBALLY Specify GLOBALLY to create a global role. A global user must be

authorized to use the role by the enterprise directory service before the role is enabled
at login.
If you omit both the NOT IDENTIFIED clause and the IDENTIFIED clause, then the role
defaults to NOT IDENTIFIED.

Examples
15

Creating a Role: Example

The following statement creates the role dw_manager:

CREATE ROLE dw_manager;

15-60 Oracle Database SQL Language Reference

CREATE ROLE

Users who are subsequently granted the dw_manager role will inherit all of the
privileges that have been granted to this role.
You can add a layer of security to roles by specifying a password, as in the following
example:
CREATE ROLE dw_manager
IDENTIFIED BY warehouse;

Users who are subsequently granted the dw_manager role must specify the password
warehouse to enable the role with the SET ROLE statement.
The following statement creates global role warehouse_user:
CREATE ROLE warehouse_user IDENTIFIED GLOBALLY;

The following statement creates the same role as an external role:
CREATE ROLE warehouse_user IDENTIFIED EXTERNALLY;

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-61

CREATE ROLLBACK SEGMENT

CREATE ROLLBACK SEGMENT
Oracle strongly recommends that you run your database in
automatic undo management mode instead of using rollback
segments. Do not use rollback segments unless you must do so for
compatibility with earlier versions of Oracle Database. Refer to Oracle
Database Administrator's Guide for information on automatic undo
management.

Note:

Purpose
15

Use the CREATE ROLLBACK SEGMENT statement to create a rollback segment, which is an
object that Oracle Database uses to store data necessary to reverse, or undo, changes
made by transactions.
The information in this section assumes that your database is not running in automatic
undo mode (the UNDO_MANAGEMENT initialization parameter is set to MANUAL or not set at
all). If your database is running in automatic undo mode (the UNDO_MANAGEMENT
initialization parameter is set to AUTO, which is the default), then rollback segments are
not permitted. However, errors generated in rollback segment operations are
suppressed.
Further, if your database has a locally managed SYSTEM tablespace, then you cannot
create rollback segments in any dictionary-managed tablespace. Instead, you must
either use the automatic undo management feature or create locally managed
tablespaces to hold the rollback segments.
A tablespace can have multiple rollback segments. Generally,
multiple rollback segments improve performance.

Note:
■

■

The tablespace must be online for you to add a rollback segment
to it.
When you create a rollback segment, it is initially offline. To make
it available for transactions by your Oracle Database instance,
bring it online using the ALTER ROLLBACK SEGMENT statement. To
bring it online automatically whenever you start up the database,
add the segment name to the value of the ROLLBACK_SEGMENT
initialization parameter.

To use objects in a tablespace other than the SYSTEM tablespace:
■

■

If you are using rollback segments for undo, then at least one rollback segment
(other than the SYSTEM rollback segment) must be online.
If you are running the database in automatic undo mode, then at least one UNDO
tablespace must be online.

15-62 Oracle Database SQL Language Reference

CREATE ROLLBACK SEGMENT

See Also:
■

■

■

■

ALTER ROLLBACK SEGMENT on page 11-40 for information on
altering a rollback segment
DROP ROLLBACK SEGMENT on page 17-68 for information on
removing a rollback segment
Oracle Database Reference for information on the UNDO_MANAGEMENT
parameter
Oracle Database Administrator's Guide for information on automatic
undo mode

Prerequisites
15

To create a rollback segment, you must have the CREATE ROLLBACK SEGMENT system
privilege.

Syntax
15

create_rollback_segment::=
TABLESPACE
PUBLIC
CREATE

tablespace

storage_clause
ROLLBACK

SEGMENT

rollback_segment

;

(storage_clause on page 8-48)

Semantics
15

PUBLIC
Specify PUBLIC to indicate that the rollback segment is public and is available to any
instance. If you omit this clause, then the rollback segment is private and is available
only to the instance naming it in its initialization parameter ROLLBACK_SEGMENTS.
rollback_segment
Specify the name of the rollback segment to be created.
TABLESPACE
Use the TABLESPACE clause to identify the tablespace in which the rollback segment is
created. If you omit this clause, then the database creates the rollback segment in the
SYSTEM tablespace.

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-63

CREATE ROLLBACK SEGMENT

Oracle Database must access rollback segments frequently.
Therefore, Oracle strongly recommends that you do not create
rollback segments in the SYSTEM tablespace, either explicitly or
implicitly by omitting this clause. In addition, to avoid high
contention for the tablespace containing the rollback segment, it
should not contain other objects such as tables and indexes, and it
should require minimal extent allocation and deallocation.
Note:

To achieve these goals, create rollback segments in locally managed
tablespaces with autoallocation disabled—in tablespaces created with
the EXTENT MANAGEMENT LOCAL clause with the UNIFORM setting. The
AUTOALLOCATE setting is not supported.

See Also:

CREATE TABLESPACE on page 16-83

storage_clause
The storage_clause lets you specify storage characteristics for the rollback segment.
■

■

The OPTIMAL parameter of the storage_clause is of particular interest, because it
applies only to rollback segments.
You cannot specify the PCTINCREASE parameter of the storage_clause with CREATE
ROLLBACK SEGMENT.
See Also:

storage_clause on page 8-48

Examples
15

Creating a Rollback Segment: Example The following statement creates a rollback
segment with default storage values in an appropriately configured tablespace:
CREATE TABLESPACE rbs_ts
DATAFILE 'rbs01.dbf' SIZE 10M
EXTENT MANAGEMENT LOCAL UNIFORM SIZE 100K;
/* This example and the next will fail if your database is in
automatic undo mode.
*/
CREATE ROLLBACK SEGMENT rbs_one
TABLESPACE rbs_ts;

The preceding statement is equivalent to the following:
CREATE ROLLBACK SEGMENT rbs_one
TABLESPACE rbs_ts
STORAGE
( INITIAL 10K );

15-64 Oracle Database SQL Language Reference

CREATE SCHEMA

CREATE SCHEMA
Purpose
15

Use the CREATE SCHEMA statement to create multiple tables and views and perform
multiple grants in your own schema in a single transaction.
To execute a CREATE SCHEMA statement, Oracle Database executes each included
statement. If all statements execute successfully, then the database commits the
transaction. If any statement results in an error, then the database rolls back all the
statements.
This statement does not actually create a schema. Oracle
Database automatically creates a schema when you create a user (see
CREATE USER on page 17-7). This statement lets you populate your
schema with tables and views and grant privileges on those objects
without having to issue multiple SQL statements in multiple
transactions.

Note:

Prerequisites
15

The CREATE SCHEMA statement can include CREATE TABLE, CREATE VIEW, and GRANT
statements. To issue a CREATE SCHEMA statement, you must have the privileges
necessary to issue the included statements.

Syntax
15

create_schema::=
create_table_statement
CREATE

SCHEMA

AUTHORIZATION

schema

create_view_statement

;

grant_statement

Semantics
15

schema
Specify the name of the schema. The schema name must be the same as your Oracle
Database username.

create_table_statement
Specify a CREATE TABLE statement to be issued as part of this CREATE SCHEMA statement.
Do not end this statement with a semicolon (or other terminator character).
See Also:

CREATE TABLE on page 16-6

create_view_statement
Specify a CREATE VIEW statement to be issued as part of this CREATE SCHEMA statement.
Do not end this statement with a semicolon (or other terminator character).
See Also:

CREATE VIEW on page 17-14
SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-65

CREATE SCHEMA

grant_statement
Specify a GRANT statement to be issued as part of this CREATE SCHEMA statement. Do not
end this statement with a semicolon (or other terminator character). You can use this
clause to grant object privileges on objects you own to other users. You can also grant
system privileges to other users if you were granted those privileges WITH ADMIN
OPTION.
See Also:

GRANT on page 18-33

The CREATE SCHEMA statement supports the syntax of these statements only as defined
by standard SQL, rather than the complete syntax supported by Oracle Database.
The order in which you list the CREATE TABLE, CREATE VIEW, and GRANT statements is
unimportant. The statements within a CREATE SCHEMA statement can reference existing
objects or objects you create in other statements within the same CREATE SCHEMA
statement.
Restriction on Granting Privileges to a Schema The syntax of the parallel_clause
is allowed for a CREATE TABLE statement in CREATE SCHEMA, but parallelism is not used
when creating the objects.
See Also: The parallel_clause on page 16-63 in the CREATE TABLE
documentation

Examples
15

Creating a Schema: Example The following statement creates a schema named oe
for the sample order entry user oe, creates the table new_product, creates the view new_
product_view, and grants the SELECT object privilege on new_product_view to the
sample human resources user hr.
CREATE SCHEMA AUTHORIZATION oe
CREATE TABLE new_product
(color VARCHAR2(10) PRIMARY KEY, quantity NUMBER)
CREATE VIEW new_product_view
AS SELECT color, quantity FROM new_product WHERE color = 'RED'
GRANT select ON new_product_view TO hr;

15-66 Oracle Database SQL Language Reference

CREATE SEQUENCE

CREATE SEQUENCE
Purpose
15

Use the CREATE SEQUENCE statement to create a sequence, which is a database object
from which multiple users may generate unique integers. You can use sequences to
automatically generate primary key values.
When a sequence number is generated, the sequence is incremented, independent of
the transaction committing or rolling back. If two users concurrently increment the
same sequence, then the sequence numbers each user acquires may have gaps, because
sequence numbers are being generated by the other user. One user can never acquire
the sequence number generated by another user. After a sequence value is generated
by one user, that user can continue to access that value regardless of whether the
sequence is incremented by another user.
Sequence numbers are generated independently of tables, so the same sequence can be
used for one or for multiple tables. It is possible that individual sequence numbers will
appear to be skipped, because they were generated and used in a transaction that
ultimately rolled back. Additionally, a single user may not realize that other users are
drawing from the same sequence.
After a sequence is created, you can access its values in SQL statements with the
CURRVAL pseudocolumn, which returns the current value of the sequence, or the
NEXTVAL pseudocolumn, which increments the sequence and returns the new value.
Note on Using Sequences with Deferred Segments If you attempt to insert a
sequence value into a table that uses deferred segment creation, the first value that the
sequence returns will be skipped.
See Also:
■

■

■

Chapter 2, "Pseudocolumns" for more information on using the
CURRVAL and NEXTVAL
"How to Use Sequence Values" on page 2-4 for information on
using sequences
ALTER SEQUENCE on page 11-43 or DROP SEQUENCE on
page 18-2 for information on modifying or dropping a sequence

Prerequisites
15

To create a sequence in your own schema, you must have the CREATE SEQUENCE system
privilege.
To create a sequence in another user's schema, you must have the CREATE ANY SEQUENCE
system privilege.

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-67

CREATE SEQUENCE

Syntax
15

create_sequence::=
INCREMENT

BY
integer

START

WITH

MAXVALUE

integer

NOMAXVALUE
MINVALUE

integer

NOMINVALUE
CYCLE
NOCYCLE
CACHE

integer

NOCACHE
ORDER
schema
CREATE

SEQUENCE

.

NOORDER
sequence

;

Semantics
15

schema
Specify the schema to contain the sequence. If you omit schema, then Oracle Database
creates the sequence in your own schema.

sequence
Specify the name of the sequence to be created. The name must satisfy the
requirements listed in "Database Object Naming Rules" on page 3-111.
If you specify none of the following clauses, then you create an ascending sequence
that starts with 1 and increases by 1 with no upper limit. Specifying only INCREMENT BY
-1 creates a descending sequence that starts with -1 and decreases with no lower limit.
■

■

■

To create a sequence that increments without bound, for ascending sequences,
omit the MAXVALUE parameter or specify NOMAXVALUE. For descending sequences,
omit the MINVALUE parameter or specify the NOMINVALUE.
To create a sequence that stops at a predefined limit, for an ascending sequence,
specify a value for the MAXVALUE parameter. For a descending sequence, specify a
value for the MINVALUE parameter. Also specify NOCYCLE. Any attempt to generate a
sequence number once the sequence has reached its limit results in an error.
To create a sequence that restarts after reaching a predefined limit, specify values
for both the MAXVALUE and MINVALUE parameters. Also specify CYCLE.

Specify the interval between sequence numbers. This integer value
can be any positive or negative integer, but it cannot be 0. This value can have 28 or
fewer digits for an ascending sequence and 27 or fewer digits for a descending
sequence. The absolute of this value must be less than the difference of MAXVALUE and

INCREMENT BY

15-68 Oracle Database SQL Language Reference

CREATE SEQUENCE

MINVALUE. If this value is negative, then the sequence descends. If the value is positive,
then the sequence ascends. If you omit this clause, then the interval defaults to 1.
START WITH Specify the first sequence number to be generated. Use this clause to
start an ascending sequence at a value greater than its minimum or to start a
descending sequence at a value less than its maximum. For ascending sequences, the
default value is the minimum value of the sequence. For descending sequences, the
default value is the maximum value of the sequence. This integer value can have 28 or
fewer digits for positive values and 27 or fewer digits for negative values.

This value is not necessarily the value to which an ascending
or descending cycling sequence cycles after reaching its maximum or
minimum value, respectively.

Note:

MAXVALUE Specify the maximum value the sequence can generate. This integer
value can have 28 or fewer digits for positive values and 27 or fewer digits for negative
values. MAXVALUE must be equal to or greater than START WITH and must be greater than
MINVALUE.

Specify NOMAXVALUE to indicate a maximum value of 1028-1 for an
ascending sequence or -1 for a descending sequence. This is the default.

NOMAXVALUE

Specify the minimum value of the sequence. This integer value can have
28 or fewer digits for positive values and 27 or fewer digits for negative values.
MINVALUE must be less than or equal to START WITH and must be less than MAXVALUE.
MINVALUE

Specify NOMINVALUE to indicate a minimum value of 1 for an
ascending sequence or -(1027 -1) for a descending sequence. This is the default.

NOMINVALUE

Specify CYCLE to indicate that the sequence continues to generate values after
reaching either its maximum or minimum value. After an ascending sequence reaches
its maximum value, it generates its minimum value. After a descending sequence
reaches its minimum, it generates its maximum value.

CYCLE

NOCYCLE Specify NOCYCLE to indicate that the sequence cannot generate more
values after reaching its maximum or minimum value. This is the default.

Specify how many values of the sequence the database preallocates and
keeps in memory for faster access. This integer value can have 28 or fewer digits. The
minimum value for this parameter is 2. For sequences that cycle, this value must be
less than the number of values in the cycle. You cannot cache more values than will fit
in a given cycle of sequence numbers. Therefore, the maximum value allowed for
CACHE must be less than the value determined by the following formula:
CACHE

(CEIL (MAXVALUE - MINVALUE)) / ABS (INCREMENT)

If a system failure occurs, then all cached sequence values that have not been used in
committed DML statements are lost. The potential number of lost values is equal to the
value of the CACHE parameter.
Note: Oracle recommends using the CACHE setting to enhance
performance if you are using sequences in an Oracle Real Application
Clusters environment.

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-69

CREATE SEQUENCE

Specify NOCACHE to indicate that values of the sequence are not
preallocated. If you omit both CACHE and NOCACHE, then the database caches 20
sequence numbers by default.

NOCACHE

ORDER Specify ORDER to guarantee that sequence numbers are generated in order of
request. This clause is useful if you are using the sequence numbers as timestamps.
Guaranteeing order is usually not important for sequences used to generate primary
keys.

ORDER is necessary only to guarantee ordered generation if you are using Oracle Real
Application Clusters. If you are using exclusive mode, then sequence numbers are
always generated in order.
NOORDER Specify NOORDER if you do not want to guarantee sequence numbers are
generated in order of request. This is the default.

Example
15

The following statement creates the sequence
customers_seq in the sample schema oe. This sequence could be used to provide
customer ID numbers when rows are added to the customers table.
Creating a Sequence: Example

CREATE SEQUENCE customers_seq
START WITH
1000
INCREMENT BY
1
NOCACHE
NOCYCLE;

The first reference to customers_seq.nextval returns 1000. The second returns 1001.
Each subsequent reference will return a value 1 greater than the previous reference.

15-70 Oracle Database SQL Language Reference

CREATE SPFILE

CREATE SPFILE
Purpose
15

Use the CREATE SPFILE statement to create a server parameter file either from a
traditional plain-text initialization parameter file or from the current system-wide
settings. Server parameter files are binary files that exist only on the server and are
called from client locations to start up the database.
Server parameter files let you make persistent changes to individual parameters.
When you use a server parameter file, you can specify in an ALTER SYSTEM SET
parameter statement that the new parameter value should be persistent. This means
that the new value applies not only in the current instance, but also to any instances
that are started up subsequently. Traditional plain-text parameter files do not let you
make persistent changes to parameter values.
Server parameter files are located on the server, so they allow for automatic database
tuning by Oracle Database and for backup by Recovery Manager (RMAN).
To use a server parameter file when starting up the database, you must create it using
the CREATE SPFILE statement.
All instances in an Oracle Real Application Clusters environment must use the same
server parameter file. However, when otherwise permitted, individual instances can
have different settings of the same parameter within this one file. Instance-specific
parameter definitions are specified as SID.parameter = value, where SID is the
instance identifier.
The method of starting up the database with a server parameter file depends on
whether you create a default or nondefault server parameter file. Refer to "Creating a
Server Parameter File: Examples" on page 15-73 for examples of how to use server
parameter files.
See Also:
■

■

■

CREATE PFILE on page 15-46 for information on creating a
regular text parameter file from a binary server parameter file
Oracle Database Administrator's Guide for information on
traditional plain-text initialization parameter files and server
parameter files
Oracle Real Application Clusters Administration and Deployment
Guide for information on using server parameter files in an Oracle
Real Application Clusters environment

Prerequisites
15

You must have the SYSDBA or the SYSOPER system privilege to execute this statement.
You can execute this statement before or after instance startup. However, if you have
already started an instance using spfile_name, you cannot specify the same spfile_
name in this statement.

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-71

CREATE SPFILE

Syntax
15

create_spfile::=
=
=
CREATE

’

spfile_name

’

SPFILE

’

pfile_name

’

PFILE
FROM

;
MEMORY

Semantics
15

spfile_name
This clause lets you specify a name for the server parameter file you are creating.
■

■

■

If you do not specify spfile_name, then Oracle Database uses the platform-specific
default server parameter filename. If spfile_name already exists on the server,
then this statement will overwrite it. When using a default server parameter file,
you start up the database without referring to the file by name.
If you do specify spfile_name, then you are creating a nondefault server
parameter file. In this case, to start up the database, you must first create a
single-line traditional parameter file that points to the server parameter file, and
then name the single-line file in your STARTUP command.
The spfile_name can be either a traditional filename or an Oracle Automatic
Storage Management (Oracle ASM) filename. By using the Oracle ASM filename
syntax, you can create the spfile in an Oracle ASM disk group. For an Oracle ASM
instance, creation of the single-line traditional parameter file to start up the
database is not required.

spfile_name can include a path prefix. If you do not specify such a path prefix, then
the database adds the path prefix for the default storage location, which is platform
dependent.
See Also:
■

■

■

"Creating a Server Parameter File: Examples" on page 15-73 for
information on starting up the database with default and
nondefault server parameter files
file_specification on page 8-29 for the syntax of traditional and
Oracle ASM filenames and ALTER DISKGROUP on page 10-51 for
information on modifying the characteristics of an Oracle ASM
file
The appropriate operating-system-specific documentation for
default parameter file names

pfile_name
Specify the traditional plain-text initialization parameter file from which you want to
create a server parameter file. The traditional parameter file must reside on the server.
■

■

If you specify pfile_name and the traditional parameter file does not reside in the
default directory for parameter files on your operating system, then you must
specify the full path.
If you do not specify pfile_name, then Oracle Database looks in the default
directory for parameter files on your operating system for the default parameter

15-72 Oracle Database SQL Language Reference

CREATE SPFILE

filename and uses that file. If that file does not exist in the expected directory, then
the database returns an error.
In an Oracle Real Application Clusters environment, you must
first combine all instance parameter files into one file before specifying
that filename in this statement to create a server parameter file. For
information on accomplishing this step, see Oracle Real Application
Clusters Administration and Deployment Guide.

Note:

MEMORY
Specify MEMORY to create an spfile using the current system-wide parameter settings. In
an Oracle RAC environment, the created file will contain the parameter settings from
each instance.

Examples
15

The following example creates a default
server parameter file from a traditional plain-text parameter file named t_init1.ora:

Creating a Server Parameter File: Examples

CREATE SPFILE
FROM PFILE = '$ORACLE_HOME/work/t_init1.ora';

Typically you will need to specify the full path and filename
for parameter files on your operating system.

Note:

When you create a default server parameter file, you subsequently start up the
database using that server parameter file by using the SQL*Plus command STARTUP
without the PFILE parameter, as follows:
STARTUP

The following example creates a nondefault server parameter file s_params.ora from a
traditional plain-text parameter file named t_init1.ora:
CREATE SPFILE = 's_params.ora'
FROM PFILE = '$ORACLE_HOME/work/t_init1.ora';

When you create a nondefault server parameter file, you subsequently start up the
database by first creating a traditional parameter file containing the following single
line:
spfile = 's_params.ora'

The name of this parameter file must comply with the naming conventions of your
operating system. You then use the single-line parameter file in the STARTUP command.
The following example shows how to start up the database, assuming that the
single-line parameter file is named new_param.ora:
STARTUP PFILE=new_param.ora

SQL Statements: CREATE LIBRARY to CREATE SPFILE 15-73

CREATE SPFILE

15-74 Oracle Database SQL Language Reference

16

16
SQL Statements: CREATE SYNONYM to
CREATE TRIGGER
This chapter contains the following SQL statements:
■

CREATE SYNONYM

■

CREATE TABLE

■

CREATE TABLESPACE

■

CREATE TRIGGER

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-1

CREATE SYNONYM

CREATE SYNONYM
Purpose
16

Use the CREATE SYNONYM statement to create a synonym, which is an alternative name
for a table, view, sequence, operator, procedure, stored function, package, materialized
view, Java class schema object, user-defined object type, or another synonym. A
synonym places a dependency on its target object and becomes invalid if the target
object is changed or dropped.
Synonyms provide both data independence and location transparency. Synonyms
permit applications to function without modification regardless of which user owns
the table or view and regardless of which database holds the table or view. However,
synonyms are not a substitute for privileges on database objects. Appropriate
privileges must be granted to a user before the user can use the synonym.
You can refer to synonyms in the following DML statements: SELECT, INSERT, UPDATE,
DELETE, FLASHBACK TABLE, EXPLAIN PLAN, and LOCK TABLE.
You can refer to synonyms in the following DDL statements: AUDIT, NOAUDIT, GRANT,
REVOKE, and COMMENT.
See Also:

Oracle Database Concepts for general information on

synonyms

Prerequisites
16

To create a private synonym in your own schema, you must have the CREATE SYNONYM
system privilege.
To create a private synonym in another user's schema, you must have the CREATE ANY
SYNONYM system privilege.
To create a PUBLIC synonym, you must have the CREATE PUBLIC SYNONYM system
privilege.

Syntax
16

create_synonym::=
OR

REPLACE

PUBLIC

schema

CREATE

SYNONYM

schema
FOR

.

@

.
synonym

dblink

object

;

Semantics
16

OR REPLACE
Specify OR REPLACE to re-create the synonym if it already exists. Use this clause to
change the definition of an existing synonym without first dropping it.
Restriction on Replacing a Synonym You cannot use the OR REPLACE clause for a type
synonym that has any dependent tables or dependent valid user-defined object types.

16-2 Oracle Database SQL Language Reference

CREATE SYNONYM

PUBLIC
Specify PUBLIC to create a public synonym. Public synonyms are accessible to all users.
However each user must have appropriate privileges on the underlying object in order
to use the synonym.
When resolving references to an object, Oracle Database uses a public synonym only if
the object is not prefaced by a schema and is not followed by a database link.
If you omit this clause, then the synonym is private. A private synonym name must be
unique in its schema. A private synonym is accessible to users other than the owner
only if those users have appropriate privileges on the underlying database object and
specify the schema along with the synonym name.
Notes on Public Synonyms
■

■

The following notes apply to public synonyms:

If you create a public synonym and it subsequently has dependent tables or
dependent valid user-defined object types, then you cannot create another
database object of the same name as the synonym in the same schema as the
dependent objects.
Take care not to create a public synonym with the same name as an existing
schema. If you do so, then all PL/SQL units that use that name will be invalidated.

schema
Specify the schema to contain the synonym. If you omit schema, then Oracle Database
creates the synonym in your own schema. You cannot specify a schema for the
synonym if you have specified PUBLIC.

synonym
Specify the name of the synonym to be created. The name must satisfy the
requirements listed in "Database Object Naming Rules" on page 3-111.
Synonyms longer than 30 bytes can be created and dropped.
However, unless they represent a Java name they will not work in any
other SQL command. Names longer than 30 bytes are transformed
into an obscure shorter string for storage in the data dictionary.

Note:

See Also: "CREATE SYNONYM: Examples" on page 16-4 and
"Oracle Database Resolution of Synonyms: Example" on page 16-4

FOR Clause
Specify the object for which the synonym is created. The schema object for which you
are creating the synonym can be of the following types:
■

Table or object table

■

View or object view

■

Sequence

■

Stored procedure, function, or package

■

Materialized view

■

Java class schema object

■

User-defined object type

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-3

CREATE SYNONYM

■

Synonym

The schema object need not currently exist and you need not have privileges to access
the object.
Restriction on the FOR Clause The schema object cannot be contained in a package.
schema Specify the schema in which the object resides. If you do not qualify object
with schema, then the database assumes that the schema object is in your own schema.

If you are creating a synonym for a procedure or function on a remote database, then
you must specify schema in this CREATE statement. Alternatively, you can create a local
public synonym on the database where the object resides. However, the database link
must then be included in all subsequent calls to the procedure or function.
You can specify a complete or partial database link to create a synonym for a
schema object on a remote database where the object is located. If you specify dblink
and omit schema, then the synonym refers to an object in the schema specified by the
database link. Oracle recommends that you specify the schema containing the object in
the remote database.
dblink

If you omit dblink, then Oracle Database assumes the object is located on the local
database.
Restriction on Database Links You cannot specify dblink for a Java class synonym.
See Also:
■

■

"References to Objects in Remote Databases" on page 3-117 for
more information on referring to database links
CREATE DATABASE LINK on page 14-31 for more information
on creating database links

Examples
16

CREATE SYNONYM: Examples To define the synonym offices for the table

locations in the schema hr, issue the following statement:
CREATE SYNONYM offices
FOR hr.locations;

To create a PUBLIC synonym for the employees table in the schema hr on the remote
database, you could issue the following statement:
CREATE PUBLIC SYNONYM emp_table
FOR hr.employees@remote.us.example.com;

A synonym may have the same name as the underlying object, provided the
underlying object is contained in another schema.
Oracle Database Resolution of Synonyms: Example Oracle Database attempts to
resolve references to objects at the schema level before resolving them at the PUBLIC
synonym level. For example, the schemas oe and sh both contain tables named
customers. In the next example, user SYSTEM creates a PUBLIC synonym named
customers for oe.customers:
CREATE PUBLIC SYNONYM customers FOR oe.customers;

16-4 Oracle Database SQL Language Reference

CREATE SYNONYM

If the user sh then issues the following statement, then the database returns the count
of rows from sh.customers:
SELECT COUNT(*) FROM customers;

To retrieve the count of rows from oe.customers, the user sh must preface customers
with the schema name. (The user sh must have select permission on oe.customers as
well.)
SELECT COUNT(*) FROM oe.customers;

If the user hr's schema does not contain an object named customers, and if hr has
select permission on oe.customers, then hr can access the customers table in oe's
schema by using the public synonym customers:
SELECT COUNT(*) FROM customers;

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-5

CREATE TABLE

CREATE TABLE
Purpose
16

Use the CREATE TABLE statement to create one of the following types of tables:
■
■

A relational table, which is the basic structure to hold user data.
An object table, which is a table that uses an object type for a column definition.
An object table is explicitly defined to hold object instances of a particular type.

You can also create an object type and then use it in a column when creating a
relational table.
Tables are created with no data unless a subquery is specified. You can add rows to a
table with the INSERT statement. After creating a table, you can define additional
columns, partitions, and integrity constraints with the ADD clause of the ALTER TABLE
statement. You can change the definition of an existing column or partition with the
MODIFY clause of the ALTER TABLE statement.
See Also:
■

■

Oracle Database Administrator's Guide and CREATE TYPE on
page 17-3 for more information about creating objects
ALTER TABLE on page 12-2 and DROP TABLE on page 18-5 for
information on modifying and dropping tables

Prerequisites
16

To create a relational table in your own schema, you must have the CREATE TABLE
system privilege. To create a table in another user's schema, you must have the CREATE
ANY TABLE system privilege. Also, the owner of the schema to contain the table must
have either space quota on the tablespace to contain the table or the UNLIMITED
TABLESPACE system privilege.
In addition to these table privileges, to create an object table or a relational table with
an object type column, the owner of the table must have the EXECUTE object privilege in
order to access all types referenced by the table, or you must have the EXECUTE ANY
TYPE system privilege. These privileges must be granted explicitly and not acquired
through a role.
Additionally, if the table owner intends to grant access to the table to other users, then
the owner must have been granted the EXECUTE object privilege on the referenced
types WITH GRANT OPTION, or have the EXECUTE ANY TYPE system privilege WITH ADMIN
OPTION. Without these privileges, the table owner has insufficient privileges to grant
access to the table to other users.
To enable a unique or primary key constraint, you must have the privileges necessary
to create an index on the table. You need these privileges because Oracle Database
creates an index on the columns of the unique or primary key in the schema
containing the table.
To create an external table, you must have the required read and write operating
system privileges on the appropriate operating system directories. You must have the
READ object privilege on the database directory object corresponding to the operating
system directory in which the external data resides. You must also have the WRITE
object privilege on the database directory in which the files will reside if you specify a
log file or bad file in the opaque_format_spec or if you unload data into an external
table from a database table by specifying the AS subquery clause.
16-6 Oracle Database SQL Language Reference

CREATE TABLE

To create an XMLType table in a different database schema from your own, you must
have not only privilege CREATE ANY TABLE but also privilege CREATE ANY INDEX. This is
because a unique index is created on column OBJECT_ID when you create the table.
Column OBJECT_ID stores a system-generated object identifier.
See Also:

CREATE INDEX on page 14-60

■

Oracle Database Administrator's Guide for more information about
the privileges required to create tables using types

■

Syntax
16

create_table::=
GLOBAL

TEMPORARY

relational_table

schema

CREATE

TABLE

.

table

object_table

;

XMLType_table

(relational_table::= on page 16-7, object_table::= on page 16-7, XMLType_table::= on
page 16-8)
relational_table::=
DELETE
ON
(

relational_properties

COMMIT

)

physical_properties

ROWS
PRESERVE

table_properties

Each of the clauses following the table name is optional for
any given relational table. However, for every table you must at least
specify either column names and data types using the relational_
properties clause or an AS subquery clause using the table_
properties clause.

Note:

(relational_properties::= on page 16-8, physical_properties::= on page 16-10, table_
properties::= on page 16-11)
object_table::=
schema

.

object_table_substitution

OF

object_type

DELETE
ON
(

object_properties

OID_clause

)

OID_index_clause

COMMIT

ROWS
PRESERVE

physical_properties

table_properties

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-7

CREATE TABLE

(object_table_substitution::= on page 16-9, object_properties::= on page 16-9, oid_clause::=
on page 16-9, oid_index_clause::= on page 16-9, physical_properties::= on page 16-10,
table_properties::= on page 16-11)
XMLType_table::=
(
OF

object_properties

)

XMLTYPE

XMLType_storage

XMLSchema_spec

XMLTYPE

DELETE
ON

COMMIT

ROWS

XMLType_virtual_columns

OID_index_clause

PRESERVE

physical_properties

OID_clause

table_properties

(XMLType_storage::= on page 16-15, XMLSchema_spec::= on page 16-15, XMLType_
virtual_columns::= on page 16-15, oid_clause::= on page 16-9, oid_index_clause::= on
page 16-9, physical_properties::= on page 16-10, table_properties::= on page 16-11)
relational_properties::=
,
column_definition
virtual_column_definition
out_of_line_constraint
out_of_line_ref_constraint
supplemental_logging_props

(column_definition::= on page 16-8, virtual_column_definition::= on page 16-9,
constraint::= on page 8-5, supplemental_logging_props::= on page 16-16)
column_definition::=
SORT
column

DEFAULT

expr

ENCRYPT

encryption_spec

datatype

inline_constraint
inline_ref_constraint

(encryption_spec::= on page 16-9, constraint::= on page 8-5)

16-8 Oracle Database SQL Language Reference

CREATE TABLE

virtual_column_definition::=
datatype

GENERATED

ALWAYS

column

VIRTUAL

AS

(

column_expression

BY

password

)

inline_constraint

(constraint::= on page 8-5)
encryption_spec::=
USING

’

encrypt_algorithm

’

IDENTIFIED

NO
’

integrity_algorithm

’

SALT

object_table_substitution::=
NOT
SUBSTITUTABLE

AT

ALL

LEVELS

object_properties::=
inline_constraint
DEFAULT

column

expr

inline_ref_constraint

attribute
out_of_line_constraint
out_of_line_ref_constraint
supplemental_logging_props

(constraint::= on page 8-5, supplemental_logging_props::= on page 16-16)
oid_clause::=
SYSTEM
OBJECT

IDENTIFIER

GENERATED

IS
PRIMARY

KEY

oid_index_clause::=
index
OIDINDEX

physical_attributes_clause
(

)
TABLESPACE

tablespace

(physical_attributes_clause::= on page 16-10)

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-9

CREATE TABLE

physical_properties::=
deferred_segment_creation

table_compression
segment_attributes_clause
segment_attributes_clause

table_compression

HEAP
deferred_segment_creation
ORGANIZATION

segment_attributes_clause
INDEX
EXTERNAL

index_org_table_clause
external_table_clause

,
CLUSTER

cluster

(

column

)

(deferred_segment_creation::= on page 16-10, segment_attributes_clause::= on page 16-10,
table_compression::= on page 16-11, index_org_table_clause::= on page 16-16, external_
table_clause::= on page 16-17)
deferred_segment_creation::=
IMMEDIATE
SEGMENT

CREATION
DEFERRED

segment_attributes_clause::=
physical_attributes_clause
TABLESPACE

tablespace

logging_clause

(physical_attributes_clause::= on page 16-10, logging_clause::= on page 16-14)
physical_attributes_clause::=
PCTFREE

integer

PCTUSED

integer

INITRANS

integer

storage_clause

(storage_clause::= on page 8-50)

16-10 Oracle Database SQL Language Reference

CREATE TABLE

table_compression::=
BASIC
OLTP
LOW

FOR

HIGH

QUERY
ARCHIVE
COMPRESS
NOCOMPRESS

table_properties::=
CACHE
column_properties

table_partitioning_clauses

NOCACHE

DEFAULT
RESULT_CACHE

(

ROWDEPENDENCIES

MODE

)
FORCE

enable_disable_clause

parallel_clause

row_movement_clause

flashback_archive_clause

NOROWDEPENDENCIES

AS

subquery

(column_properties::= on page 16-11, table_partitioning_clauses::= on page 16-17, parallel_
clause::= on page 16-24, enable_disable_clause::= on page 16-24, row_movement_clause::=
on page 16-16, flashback_archive_clause::= on page 16-16, subquery::= on page 19-5)
column_properties::=
object_type_col_properties
nested_table_col_properties
,
varray_col_properties

(

LOB_partition_storage

)

LOB_storage_clause
XMLType_column_properties

(object_type_col_properties::= on page 16-11, nested_table_col_properties::= on page 16-12,
varray_col_properties::= on page 16-12, LOB_storage_clause::= on page 16-13, LOB_
partition_storage::= on page 16-14, XMLType_column_properties::= on page 16-15)
object_type_col_properties::=
COLUMN

column

substitutable_column_clause

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-11

CREATE TABLE

substitutable_column_clause::=
ELEMENT

TYPE
IS

OF

(

ONLY

type

)

NOT
SUBSTITUTABLE

AT

ALL

LEVELS

nested_table_col_properties::=
LOCAL
substitutable_column_clause

nested_item
NESTED

GLOBAL

TABLE
COLUMN_VALUE

(
(

object_properties

)

physical_properties

)

column_properties
STORE

AS

storage_table

AS

LOCATOR

RETURN
VALUE

(substitutable_column_clause::= on page 16-12, object_properties::= on page 16-9, physical_
properties::= on page 16-10, column_properties::= on page 16-11)
varray_col_properties::=
substitutable_column_clause
varray_storage_clause
VARRAY

varray_item
substitutable_column_clause

(substitutable_column_clause::= on page 16-12, varray_storage_clause::= on page 16-12)
varray_storage_clause::=
SECUREFILE
LOB_segname
BASICFILE
STORE

AS

(
LOB
LOB_segname

(LOB_parameters::= on page 16-13)

16-12 Oracle Database SQL Language Reference

LOB_storage_parameters

)

CREATE TABLE

LOB_storage_clause::=
SECUREFILE

,
(

LOB_item

)

STORE

BASICFILE

AS
(

LOB

LOB_storage_parameters

)

SECUREFILE
BASICFILE
(

LOB_item

)

STORE

AS

LOB_segname
(

LOB_storage_parameters

)

(LOB_storage_parameters::= on page 16-13)
LOB_storage_parameters::=
TABLESPACE

tablespace
storage_clause

LOB_parameters
storage_clause

(LOB_parameters::= on page 16-13, storage_clause::= on page 8-50)
LOB_parameters::=
ENABLE
STORAGE

IN

ROW

DISABLE
CHUNK

integer

PCTVERSION
FREEPOOLS

integer
integer

LOB_retention_clause
LOB_deduplicate_clause
LOB_compression_clause
ENCRYPT

encryption_spec

DECRYPT
CACHE

logging_clause

NOCACHE
CACHE

READS

(LOB_deduplicate_clause::= on page 16-14, LOB_compression_clause::= on page 16-14,
encryption_spec::= on page 16-9, logging_clause::= on page 8-38)

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-13

CREATE TABLE

Several of the LOB parameters are no longer needed if you are
using SecureFiles for LOB storage. Refer to LOB_storage_parameters on
page 16-43 for more information.

Note:

LOB_retention_clause::=
MAX
MIN

integer

AUTO
NONE
RETENTION

LOB_deduplicate_clause::=
DEDUPLICATE
KEEP_DUPLICATES

LOB_compression_clause::=
HIGH
MEDIUM
LOW
COMPRESS
NOCOMPRESS

logging_clause::=
LOGGING
NOLOGGING
FILESYSTEM_LIKE_LOGGING

LOB_partition_storage::=
LOB_storage_clause
PARTITION

partition
varray_col_properties

LOB_partitioning_storage
(

SUBPARTITION

subpartition

)
varray_col_properties

(LOB_storage_clause::= on page 16-13, varray_col_properties::= on page 16-12, LOB_
partitioning_storage::= on page 16-15)

16-14 Oracle Database SQL Language Reference

CREATE TABLE

LOB_partitioning_storage::=
LOB

(

LOB_item

)

(

STORE

BASICFILE

LOB_segname

SECUREFILE

(

TABLESPACE

TABLESPACE

tablespace

tablespace

)

)

AS

XMLType_column_properties::=
COLUMN

XMLType_storage

XMLTYPE

XMLSchema_spec

column

(XMLType_storage::= on page 16-15, XMLSchema_spec::= on page 16-15)
XMLType_storage::=
STORE

OBJECT

RELATIONAL
(

AS

SECUREFILE

LOB_parameters

)

LOB_segname

BASICFILE

(

CLOB
BINARY

LOB_parameters

)

XML

LOBS
ALL

VARRAYS

AS
TABLES

(LOB_parameters::= on page 16-13)
XMLSchema_spec::=
XMLSCHEMA

XMLSchema_URL

element
ELEMENT
XMLSchema_URL

ALLOW

#

element

ALLOW
NONSCHEMA

ANYSCHEMA

DISALLOW

DISALLOW

XMLType_virtual_columns::=
,
VIRTUAL

COLUMNS

(

column

AS

(

expr

)

)

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-15

CREATE TABLE

row_movement_clause::=
ENABLE
ROW

MOVEMENT

DISABLE

flashback_archive_clause::=
flashback_archive
FLASHBACK
NO

ARCHIVE

FLASHBACK

ARCHIVE

index_org_table_clause::=
mapping_table_clause
PCTTHRESHOLD

integer

key_compression

index_org_overflow_clause

(mapping_table_clauses::= on page 16-16, key_compression::= on page 16-16, index_org_
overflow_clause::= on page 16-16)
mapping_table_clauses::=
MAPPING

TABLE

NOMAPPING

key_compression::=
integer
COMPRESS
NOCOMPRESS

index_org_overflow_clause::=
INCLUDING

column_name

segment_attributes_clause
OVERFLOW

(segment_attributes_clause::= on page 16-10)
supplemental_logging_props::=
supplemental_log_grp_clause
SUPPLEMENTAL

LOG
supplemental_id_key_clause

supplemental_log_grp_clause::=
,
NO
GROUP

log_group

(

LOG

column

16-16 Oracle Database SQL Language Reference

ALWAYS
)

CREATE TABLE

supplemental_id_key_clause::=
,
ALL
PRIMARY
DATA

KEY

(

)

COLUMNS

UNIQUE
FOREIGN

KEY

external_table_clause::=
integer
REJECT
TYPE

LIMIT

access_driver_type

UNLIMITED

(

external_data_properties

)

(external_data_properties::= on page 16-17)
external_data_properties::=
(
ACCESS

opaque_format_spec

USING
DEFAULT

DIRECTORY

)

PARAMETERS
CLOB

subquery

directory

,
directory
LOCATION

(

:
’

location_specifier

’

)

(opaque_format_spec: This clause specifies all access parameters for the ORACLE_
LOADER and ORACLE_DATAPUMP access drivers. See Oracle Database Utilities for
descriptions of these parameters.)
table_partitioning_clauses::=
range_partitions
list_partitions
hash_partitions
composite_range_partitions
composite_list_partitions
composite_hash_partitions
reference_partitioning
system_partitioning

(range_partitions::= on page 16-18, list_partitions::= on page 16-18, hash_partitions::= on
page 16-18, composite_range_partitions::= on page 16-19, composite_list_partitions::= on
page 16-19, composite_hash_partitions::= on page 16-19, reference_partitioning::= on
page 16-20, system_partitioning::= on page 16-20)

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-17

CREATE TABLE

range_partitions::=
,
PARTITION

BY

RANGE

(

column

)

,
STORE
INTERVAL

(

expr

IN

(

tablespace

)

)

,
partition
(

PARTITION

range_values_clause

table_partition_description

)

(range_values_clause::= on page 16-22, table_partition_description::= on page 16-23)
list_partitions::=
PARTITION

BY

LIST

(

column

)

,
partition
(

PARTITION

list_values_clause

table_partition_description

)

(list_values_clause::= on page 16-23, table_partition_description::= on page 16-23)
hash_partitions::=
,
individual_hash_partitions
PARTITION

BY

HASH

(

column

)
hash_partitions_by_quantity

(individual_hash_partitions::= on page 16-18, hash_partitions_by_quantity::= on
page 16-19)
individual_hash_partitions::=
,
partition
(

partitioning_storage_clause

PARTITION

)

(partitioning_storage_clause::= on page 16-23)

16-18 Oracle Database SQL Language Reference

CREATE TABLE

hash_partitions_by_quantity::=
,
STORE
PARTITIONS

IN

(

tablespace

)

hash_partition_quantity

table_compression

,

key_compression

OVERFLOW

STORE

IN

(

tablespace

)

(table_compression::= on page 16-11, key_compression::= on page 16-16)
composite_range_partitions::=
,
PARTITION

BY

RANGE

(

column

)

,
STORE
INTERVAL

(

expr

subpartition_by_range
subpartition_by_list

IN

(

tablespace

)

)

,
(

range_partition_desc

)

subpartition_by_hash

(subpartition_by_range::= on page 16-21. subpartition_by_list::= on page 16-21,
subpartition_by_hash::= on page 16-22, range_partition_desc::= on page 16-20)
composite_list_partitions::=
subpartition_by_range
PARTITION

BY

LIST

(

column

)

subpartition_by_list

,
(

list_partition_desc

)

subpartition_by_hash

(subpartition_by_range::= on page 16-21. subpartition_by_list::= on page 16-21,
subpartition_by_hash::= on page 16-22, list_partition_desc::= on page 16-21)
composite_hash_partitions::=
,

subpartition_by_range
individual_hash_partitions

PARTITION

BY

HASH

(

column

)

subpartition_by_list
hash_partitions_by_quantity
subpartition_by_hash

(subpartition_by_range::= on page 16-21, subpartition_by_list::= on page 16-21,
subpartition_by_hash::= on page 16-22, individual_hash_partitions::= on page 16-18, hash_
partitions_by_quantity::= on page 16-19)

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-19

CREATE TABLE

reference_partitioning::=
,
(
PARTITION

BY

REFERENCE

(

constraint

reference_partition_desc

)

)

(constraint::= on page 8-5, reference_partition_desc::= on page 16-20)
reference_partition_desc::=
partition

table_partition_description

PARTITION

(table_partition_description::= on page 16-23)
system_partitioning::=
PARTITIONS

integer
,

reference_partition_desc
PARTITION

BY

SYSTEM

(reference_partition_desc::= on page 16-20)
range_partition_desc::=
,
range_subpartition_desc
,
(

list_subpartition_desc
,
individual_hash_subparts

partition
PARTITION

hash_subparts_by_quantity
range_values_clause

table_partition_description

(range_values_clause::= on page 16-22, table_partition_description::= on page 16-23, range_
subpartition_desc::= on page 16-22, list_subpartition_desc::= on page 16-22, individual_
hash_subparts::= on page 16-22, hash_subparts_by_quantity::= on page 16-22)

16-20 Oracle Database SQL Language Reference

)

CREATE TABLE

list_partition_desc::=
,
range_subpartition_desc
,
(

list_subpartition_desc

)

,
individual_hash_subparts
partition

hash_subparts_by_quantity

PARTITION

list_values_clause

table_partition_description

(list_values_clause::= on page 16-23, table_partition_description::= on page 16-23, range_
subpartition_desc::= on page 16-22, list_subpartition_desc::= on page 16-22, individual_
hash_subparts::= on page 16-22, hash_subparts_by_quantity::= on page 16-22)
subpartition_template::=
,
range_subpartition_desc
,
(

list_subpartition_desc

)

,
SUBPARTITION

TEMPLATE
individual_hash_subparts
hash_subpartition_quantity

(range_subpartition_desc::= on page 16-22, list_subpartition_desc::= on page 16-22,
individual_hash_subparts::= on page 16-22)
subpartition_by_range::=
,
SUBPARTITION

BY

RANGE

(

subpartition_template

column

)

(subpartition_template::= on page 16-21)
subpartition_by_list::=
subpartition_template
SUBPARTITION

BY

LIST

(

column

)

(subpartition_template::= on page 16-21)

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-21

CREATE TABLE

subpartition_by_hash::=
,
SUBPARTITION

BY

HASH

(

column

)

,
STORE
SUBPARTITIONS

IN

(

tablespace

)

integer

subpartition_template

(subpartition_template::= on page 16-21)
range_subpartition_desc::=
subpartition

partitioning_storage_clause

SUBPARTITION

range_values_clause

(range_values_clause::= on page 16-22, partitioning_storage_clause::= on page 16-23)
list_subpartition_desc::=
subpartition

partitioning_storage_clause

SUBPARTITION

list_values_clause

(list_values_clause::= on page 16-23, partitioning_storage_clause::= on page 16-23)
individual_hash_subparts::=
subpartition

partitioning_storage_clause

SUBPARTITION

(partitioning_storage_clause::= on page 16-23)
hash_subparts_by_quantity::=
,
STORE
SUBPARTITIONS

IN

(

tablespace

integer

range_values_clause::=
,
literal
VALUES

LESS

THAN

(

)
MAXVALUE

16-22 Oracle Database SQL Language Reference

)

CREATE TABLE

list_values_clause::=
,
literal
VALUES

NULL

(

)

DEFAULT

table_partition_description::=
table_compression
deferred_segment_creation

segment_attributes_clause

key_compression

LOB_storage_clause
varray_col_properties

segment_attributes_clause
OVERFLOW

nested_table_col_properties

(deferred_segment_creation::= on page 16-10, segment_attributes_clause::= on page 16-10,
table_compression::= on page 16-11, key_compression::= on page 16-16, segment_attributes_
clause::= on page 16-10, LOB_storage_clause::= on page 16-13, varray_col_properties::= on
page 16-12, nested_table_col_properties::= on page 16-12)
partitioning_storage_clause::=
TABLESPACE

tablespace
TABLESPACE

tablespace

OVERFLOW
table_compression
key_compression
LOB_partitioning_storage
SECUREFILE
BASICFILE
VARRAY

varray_item

STORE

AS

LOB

LOB_segname

(table_compression::= on page 16-11, key_compression::= on page 16-16, LOB_partitioning_
storage::= on page 16-15)

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-23

CREATE TABLE

LOB_partitioning_storage::=
LOB

(

LOB_item

)

(

STORE

BASICFILE

LOB_segname

SECUREFILE

(

TABLESPACE

TABLESPACE

tablespace

tablespace

)

)

AS

parallel_clause::=
NOPARALLEL
integer
PARALLEL

enable_disable_clause::=
,

VALIDATE
UNIQUE

NOVALIDATE

ENABLE

(

PRIMARY

column

)

KEY

DISABLE
CONSTRAINT

constraint_name

KEEP
INDEX
using_index_clause

exceptions_clause

CASCADE

DROP

(using_index_clause::= on page 16-24, exceptions_clause not supported in CREATE
TABLE statements)
using_index_clause::=
schema

.
index

USING

INDEX

(

create_index_statement

)

index_properties

(create_index::= on page 14-61, index_properties::= on page 16-24)
index_properties::=
global_partitioned_index
local_partitioned_index
index_attributes
domain_index_clause
INDEXTYPE

IS
XMLIndex_clause

16-24 Oracle Database SQL Language Reference

CREATE TABLE

(global_partitioned_index::= on page 14-64, local_partitioned_index::= on page 14-65—part
of CREATE INDEX, index_attributes::= on page 14-62, domain_index_clause and
XMLIndex_clause: not supported in using_index_clause)
index_attributes::=
physical_attributes_clause
logging_clause
ONLINE
tablespace
TABLESPACE
DEFAULT
key_compression
SORT
NOSORT
REVERSE
VISIBLE
INVISIBLE
parallel_clause

(physical_attributes_clause::= on page 16-10, logging_clause::= on page 8-38, key_
compression::= on page 16-16, parallel_clause: not supported in using_index_clause)

Semantics
16

relational_table

GLOBAL TEMPORARY
Specify GLOBAL TEMPORARY to indicate that the table is temporary and that its definition
is visible to all sessions with appropriate privileges. The data in a temporary table is
visible only to the session that inserts the data into the table.
When you first create a temporary table, its table metadata is stored in the data
dictionary, but no space is allocated for table data. Space is allocated for the table
segment at the time of the first DML operation on the table. The temporary table
definition persists in the same way as the definitions of regular tables, but the table
segment and any data the table contains are either session-specific or
transaction-specific data. You specify whether the table segment and data are sessionor transaction-specific with the ON COMMIT keywords.
You can perform DDL operations (such as ALTER TABLE, DROP TABLE, CREATE INDEX) on
a temporary table only when no session is bound to it. A session becomes bound to a
temporary table by performing an INSERT operation on the table. A session becomes
unbound to the temporary table by issuing a TRUNCATE statement or at session
termination, or, for a transaction-specific temporary table, by issuing a COMMIT or
ROLLBACK statement.
Oracle Database Concepts for information on temporary
tables and "Creating a Table: Temporary Table Example" on page 16-71

See Also:

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-25

CREATE TABLE

Restrictions on Temporary Tables

Temporary tables are subject to the following

restrictions:
■

Temporary tables cannot be partitioned, clustered, or index organized.

■

You cannot specify any foreign key constraints on temporary tables.

■

Temporary tables cannot contain columns of nested table.

■

■
■

■

You cannot specify the following clauses of the LOB_storage_clause: TABLESPACE,
storage_clause, or logging_clause.
Parallel UPDATE, DELETE and MERGE are not supported for temporary tables.
The only part of the segment_attributes_clause you can specify for a temporary
table is TABLESPACE, which allows you to specify a single temporary tablespace.
Distributed transactions are not supported for temporary tables.

schema
Specify the schema to contain the table. If you omit schema, then the database creates
the table in your own schema.

table
Specify the name of the table or object table to be created. The name must satisfy the
requirements listed in "Database Object Naming Rules" on page 3-111.
See Also:

"Creating Tables: General Examples" on page 16-70

relational_properties
The relational properties describe the components of a relational table.

column_definition
The column_definition lets you define the characteristics of the column.
column
Specify the name of a column of the table.
If you also specify AS subquery, then you can omit column and datatype unless you
are creating an index-organized table. If you specify AS subquery when creating an
index-organized table, then you must specify column, and you must omit datatype.
The absolute maximum number of columns in a table is 1000. When you create an
object table or a relational table with columns of object, nested table, varray, or REF
type, Oracle Database maps the columns of the user-defined types to relational
columns, in effect creating hidden columns that count toward the 1000-column limit.
datatype
Specify the data type of a column.
Notes on Table Column Data Types The following notes apply to the data types of

table columns:
■

■

If you specify AS subquery, then you can omit datatype. If you are creating an
index-organized table and you specify AS subquery, then you must omit the data
type.
You can also omit datatype if the statement designates the column as part of a
foreign key in a referential integrity constraint. Oracle Database automatically

16-26 Oracle Database SQL Language Reference

CREATE TABLE

assigns to the column the data type of the corresponding column of the referenced
key of the referential integrity constraint.
■

Do not create a table with LONG columns. Use LOB columns (CLOB, NCLOB, BLOB)
instead. LONG columns are supported only for backward compatibility.

Restriction on Table Column Data Types You can specify a column of type ROWID,
but Oracle Database does not guarantee that the values in such columns are valid
rowids.

"Data Types" on page 3-1 for information on LONG columns
and on Oracle-supplied data types

See Also:

SORT
The SORT keyword is valid only if you are creating this table as part of a hash cluster
and only for columns that are also cluster columns.
This clause instructs the database to sort the rows of the cluster on this column after
applying the hash function when performing a DML operation. Doing so may
improve response time during subsequent operations on the clustered data.
"CLUSTER Clause" on page 16-41 for information on
creating a cluster table

See Also:

DEFAULT
The DEFAULT clause lets you specify a value to be assigned to the column if a
subsequent INSERT statement omits a value for the column. The data type of the
expression must match the data type of the column. The column must also be long
enough to hold this expression.
The DEFAULT expression can include any SQL function as long as the function does not
return a literal argument, a column reference, or a nested function invocation.
A DEFAULT expression cannot contain
references to PL/SQL functions or to other columns, the pseudocolumns CURRVAL,
NEXTVAL, LEVEL, PRIOR, and ROWNUM, or date constants that are not fully specified.
Restriction on Default Column Values

See Also:

"About SQL Expressions" on page 6-1 for the syntax of

expr
encryption_spec
The ENCRYPT clause lets you use the Transparent Data Encryption (TDE) feature to
encrypt the column you are defining. You can encrypt columns of type CHAR, NCHAR,
VARCHAR2, NVARCHAR2, NUMBER, DATE, LOB, and RAW. The data does not appear in its
encrypted form to authorized users, such as the user who encrypts the column.
Column encryption requires that a system administrator with
appropriate privileges has initialized the security module, opened a
wallet, and set an encryption key. Refer to Oracle Database Advanced
Security Administrator's Guide for general information on encryption
and to security_clauses on page 11-68 for related ALTER SYSTEM
statements.

Note:

Use this clause to specify the name of the algorithm to be
used. Valid algorithms are AES256, AES192, AES128 and 3DES168. If you omit this

USING 'encrypt_algorithm'

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-27

CREATE TABLE

clause, then the database uses AES192. If you encrypt more than one column in the
same table, and if you specify the USING clause for one of the columns, then you must
specify the same encryption algorithm for all the encrypted columns.
IDENTIFIED BY password If you specify this clause, then the database derives the
column key from the specified password.

Use this clause to specify the integrity algorithm to be used.
Valid integrity algorithms are SHA-1 and NOMAC.

'integrity_algorithm'
■

■

If you specify SHA-1, then TDE uses the Secure Hash Algorithm (SHA-1) and adds
a 20-byte Message Authentication Code (MAC) to each encrypted value for
integrity checking. This is the default.
If you specify NOMAC, then TDE does not add a MAC and does not perform the
integrity check. This saves 20 bytes of disk space per encrypted value. Refer to
Oracle Database Advanced Security Administrator's Guide for more information on
using NOMAC to save disk space and improve performance.

All encrypted columns in a table must use the same integrity algorithm. If you already
have a table column using the SHA-1 algorithm, then you cannot use the NOMAC
parameter to encrypt another column in the same table. Refer to the REKEY
encryption_spec clause of ALTER TABLE on page 12-39 to learn how to change the
integrity algorithm used by all encrypted columns in a table.
Specify SALT to instruct the database to append a random string,
called "salt," to the clear text of the column before encrypting it. This is the default.

SALT | NO SALT

Specify NO SALT to prevent the database from appending salt to the clear text of the
column before encrypting it.
The following considerations apply when specifying SALT or NO SALT for encrypted
columns:
■

■

If you want to use the column as an index key, then you must specify NO SALT.
Refer to Oracle Database Advanced Security Administrator's Guide for a description of
"salt" in this context.
If you specify table compression for the table, then the database does not compress
the data in encrypted columns with SALT.

You cannot specify SALT or NO SALT for LOB encryption.
Restrictions on encryption_spec:

The following restrictions apply to column

encryption:
■

■

Transparent Data Encryption is not supported by the traditional import and export
utilities or by transportable-tablespace-based export. Use the Data Pump import
and export utilities with encrypted columns instead.
To encrypt a column in an external table, the table must use ORACLE_DATAPUMP as
its access type.

■

You cannot encrypt a column in tables owned by SYS.

■

You cannot encrypt a foreign key column.
See Also: Oracle Database Advanced Security Administrator's Guide for
more information about Transparent Data Encryption

16-28 Oracle Database SQL Language Reference

CREATE TABLE

virtual_column_definition
The virtual_column_definition clause lets you create a virtual column. A virtual
column is not stored on disk. Rather, the database derives the values in a virtual
column on demand by computing a set of expressions or functions. Virtual columns
can be used in queries, DML, and DDL statements. They can be indexed, and you can
collect statistics on them. Thus, they can be treated much as other columns. Exceptions
and restrictions are listed below in "Notes on Virtual Columns" on page 16-29 and
"Restrictions on Virtual Columns" on page 16-30.
■
■

■

■

■

For column, specify the name of the virtual column.
You can optionally specify the data type of the virtual column. If you omit
datatype, then the database determines the data type of the column based on the
data type of the underlying expressions. All Oracle scalar data types and XMLType
are supported.
The keywords GENERATED ALWAYS are provided for syntactic clarity. They indicate
that the column is not stored on disk, but is evaluated on demand.
The AS column_expression clause determines the content of the column. Refer to
"Column Expressions" on page 6-6 for more information on column_expression.
The keyword VIRTUAL is optional and for syntactic clarity.

Notes on Virtual Columns
■
If column_expression refers to a column on which column-level security is
implemented, then the virtual column does not inherit the security rules of the
base column. In such a case, you must ensure that data in the virtual column is
protected, either by duplicating a column-level security policy on the virtual
column or by applying a function that implicitly masks the data. For example, it is
common for credit card numbers to be protected by a column-level security policy,
while still allowing call center employees to view the last four digits of the credit
card number for validation purposes. In such a case, you could define the virtual
column to take a substring of the last four digits of the credit card number.
■

■

■

■

A table index defined on a virtual column is equivalent to a function-based index
on the table.
You cannot directly update a virtual column. Thus, you cannot specify a virtual
column in the SET clause of an UPDATE statement. However, you can specify a
virtual column in the WHERE clause of an UPDATE statement. Likewise, you can
specify a virtual column in the WHERE clause of a DELETE statement to delete rows
from a table based on the derived value of the virtual column.
A query that specifies in its FROM clause a table containing a virtual column is
eligible for result caching. Refer to "RESULT_CACHE Hint" on page 3-105 for
more information on result caching.
The column_expression can refer to a PL/SQL function if the function is explicitly
designated DETERMINISTIC during its creation. However, if the function is
subsequently replaced, definitions dependent on the virtual column are not
invalidated. In such a case, if the table contains data, queries that reference the
virtual column may return incorrect results if the virtual column is used in the
definition of constraints, indexes, or materialized views or for result caching.
Therefore, in order to replace the deterministic PL/SQL function for a virtual
column.
–

Disable and re-enable any constraints on the virtual column.

–

Rebuild any indexes on the virtual column.

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-29

CREATE TABLE

–

Fully refresh materialized views accessing the virtual column.

–

Flush the result cache if cached queries have accessed the virtual column.

–

Regather statistics on the table.

Restrictions on Virtual Columns
■
You can create virtual columns only in relational heap tables. Virtual columns are
not supported for index-organized, external, object, cluster, or temporary tables.
■

The column_expression in the AS clause has the following restrictions:
–

It cannot refer to another virtual column by name.

–

Any columns referenced in column_expression must be defined on the same
table.

–

It can refer to a deterministic user-defined function, but if it does, then you
cannot use the virtual column as a partitioning key column.

–

The output of column_expression must be a scalar value.
"Column Expressions" on page 6-6 for additional
information and restrictions on column_expression

See Also:

■

■

The virtual column cannot be an Oracle supplied data type, a user-defined type, or
LOB or LONG RAW.
You cannot specify a call to a PL/SQL function in the defining expression for a
virtual column that you want to use as a partitioning column.
See Also: "Adding a Virtual Table Column: Example" on page 12-85
and Oracle Database Administrator's Guide for examples of creating
tables with virtual columns

Constraint Clauses
Use these clauses to create constraints on the table columns. You must specify a
PRIMARY KEY constraint for an index-organized table, and it cannot be DEFERRABLE.
Refer to constraint on page 8-4 for syntax and description of these constraints as well as
examples.
These clauses let you describe
a column of type REF. The only difference between these clauses is that you specify
out_of_line_ref_constraint from the table level, so you must identify the REF
column or attribute you are defining. Specify inline_ref_constraint as part of the
definition of the REF column or attribute.
inline_ref_constraint and out_of_line_ref_constraint

See Also:

"REF Constraint Examples" on page 8-25

inline_constraint Use the inline_constraint to define an integrity constraint as
part of the column definition.

You can create UNIQUE, PRIMARY KEY, and REFERENCES constraints on scalar attributes of
object type columns. You can also create NOT NULL constraints on object type columns
and CHECK constraints that reference object type columns or any attribute of an object
type column.
Use the out_of_line_constraint syntax to define an
integrity constraint as part of the table definition.

out_of_line_constraint

16-30 Oracle Database SQL Language Reference

CREATE TABLE

supplemental_logging_props
The supplemental_logging_props clause lets you instruct the database to put
additional data into the log stream to support log-based tools.
supplemental_log_grp_clause
■

■

Use this clause to create a named log group.

The NO LOG clause lets you omit from the redo log one or more columns that would
otherwise be included in the redo for the named log group. You must specify at
least one fixed-length column without NO LOG in the named log group.
If you specify ALWAYS, then during an update, the database includes in the redo all
columns in the log group. This is called an unconditional log group (sometimes
called an "always log group"), because Oracle Database supplementally logs all the
columns in the log group when the associated row is modified. If you omit ALWAYS,
then the database supplementally logs all the columns in the log group only if any
column in the log group is modified. This is called a conditional log group.

You can query the appropriate USER_, ALL_, or DBA_LOG_GROUP_COLUMNS data dictionary
view to determine whether any supplemental logging has already been specified.
supplemental_id_key_clause Use this clause to specify that all or a combination of

the primary key, unique key, and foreign key columns should be supplementally
logged. Oracle Database will generate either an unconditional log group or a
conditional log group. With an unconditional log group, the database supplementally
logs all the columns in the log group when the associated row is modified. With a
conditional log group, the database supplementally logs all the columns in the log
group only if any column in the log group is modified.
■

■

■

■

If you specify ALL COLUMNS, then the database includes in the redo log all the
fixed-length maximum size columns of that row. Such a redo log is a
system-generated unconditional log group.
If you specify PRIMARY KEY COLUMNS, then for all tables with a primary key, the
database places into the redo log all columns of the primary key whenever an
update is performed. Oracle Database evaluates which columns to supplementally
log as follows:
–

First the database chooses columns of the primary key constraint, if the
constraint is validated or marked RELY and is not marked as DISABLED or
INITIALLY DEFERRED.

–

If no primary key columns exist, then the database looks for the smallest
UNIQUE index with at least one NOT NULL column and uses the columns in that
index.

–

If no such index exists, then the database supplementally logs all scalar
columns of the table.

If you specify UNIQUE COLUMNS, then for all tables with a unique key or a bitmap
index, if any of the unique key or bitmap index columns are modified, the
database places into the redo log all other columns belonging to the unique key or
bitmap index. Such a log group is a system-generated conditional log group.
If you specify FOREIGN KEY COLUMNS, then for all tables with a foreign key, if any
foreign key columns are modified, the database places into the redo log all other
columns belonging to the foreign key. Such a redo log is a system-generated
conditional log group.

If you specify this clause multiple times, then the database creates a separate log group
for each specification. You can query the appropriate USER_, ALL_, or DBA_LOG_GROUPS

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-31

CREATE TABLE

data dictionary view to determine whether any supplemental logging data has already
been specified.
ON COMMIT
The ON COMMIT clause is relevant only if you are creating a temporary table. This clause
specifies whether the data in the temporary table persists for the duration of a
transaction or a session.
DELETE ROWS Specify DELETE ROWS for a transaction-specific temporary table. This
is the default. Oracle Database will truncate the table (delete all its rows) after each
commit.

Specify PRESERVE ROWS for a session-specific temporary table.
Oracle Database will truncate the table (delete all its rows) when you terminate the
session.

PRESERVE ROWS

physical_properties
The physical properties relate to the treatment of extents and segments and to the
storage characteristics of the table.
deferred_segment_creation
Use this clause to determine when the database should create the segment(s) for this
table:
■

SEGMENT CREATION DEFERRED: This clause defers creation of the table segment — as
well as segments for any LOB columns of the table, any indexes created implicitly
as part of table creation, and any indexes subsequently explicitly created on the
table — until the first row of data is inserted into the table. At that time, the
segments for the table, LOB columns and indexes, and explicitly created indexes
are all materialized and inherit any storage properties specified in this CREATE
TABLE statement or, in the case of explicitly created indexes, the CREATE INDEX
statement. These segments are created regardless whether the initial insert
operation is uncommitted or rolled back. This is the default value.
Caution: When creating many tables with deferred segment creation,
ensure that you allocate enough space for your database so that when
the first rows are inserted, there is enough space for all the new
segments.

■

SEGMENT CREATION IMMEDIATE: The table segment is created as part of this CREATE
TABLE statement.

Immediate segment creation is useful, for example, if your application depends upon
the object appearing in the DBA_, USER_, and ALL_SEGMENTS data dictionary views,
because the object will not appear in those views until the segment is created. This
clause overrides the setting of the DEFERRED_SEGMENT_CREATION initialization
parameter.
To determine whether a segment has been created for an existing table or its LOB
columns or indexes, query the SEGMENT_CREATED column of USER_TABLES, USER_
INDEXES, or USER_LOBS.
Notes on Tables Without Segments The following rules apply to a table whose
segment has not yet been materialized:

16-32 Oracle Database SQL Language Reference

CREATE TABLE

■

■

■

■

■

If you create this table with CREATE TABLE ... AS subquery, then if the source table
has no rows, segment creation of the new table is deferred. If the source table has
rows, then segment creation of the new table is not deferred.
If you specify ALTER TABLE ... ALLOCATE EXTENT before the segment is materialized,
then the segment is materialized and then an extent is allocated. However the
ALLOCATE EXTENT clause in a DDL statement on any indexes of the table will return
an error.
In a DDL statement on the table or its LOB columns or indexes, any specification
of DEALLOCATE UNUSED is silently ignored.
ONLINE operations on indexes of a table or table partition without a segment will
silently be disabled; that is, they will proceed OFFLINE.
If any of the following DDL statements are executed on a table with one or more
LOB columns, then the resulting partition(s) or subpartition(s) will be
materialized:
–

ALTER TABLE SPLIT [SUB]PARTITION

–

ALTER TABLE MERGE [SUB]PARTITIONS

–

ALTER TABLE ADD [SUB]PARTITION (hash partition only)

–

ALTER TABLE COALESCE [SUB]PARTITION (hash partition only)

Restrictions on Deferred Segment Creation

This clause is subject to the following

restrictions:
■

■

■

You cannot defer segment creation for the following types of tables:
index-organized tables, clustered tables, global temporary tables, session-specific
temporary tables, internal tables, object tables, XMLType tables, AQ tables, external
tables, and tables owned by SYS, SYSTEM, PUBLIC, OUTLN, or XDB.
Deferred segment creation is supported on partitions and subpartitions beginning
with Oracle Database 11g Release 2 (11.2.0.2).
Deferred segment creation is not supported for bitmap join indexes and domain
indexes.

■

Deferred segment creation is not supported in dictionary-managed tablespaces.

■

Deferred segment creation is not supported in the SYSTEM tablespace.

■

Serializable transactions do not work with deferred segment creation. Trying to
insert data into an empty table with no segment created causes an error.
Oracle Database Concepts for general information on
segment allocation and Oracle Database Reference for more information
about the DEFERRED_SEGMENT_CREATION initialization parameter

See Also:

segment_attributes_clause
The segment_attributes_clause lets you specify physical attributes and tablespace
storage for the table.
physical_attributes_clause The physical_attributes_clause lets you specify the
value of the PCTFREE, PCTUSED, and INITRANS parameters and the storage
characteristics of the table.
■

For a nonpartitioned table, each parameter and storage characteristic you specify
determines the actual physical attribute of the segment associated with the table.

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-33

CREATE TABLE

■

For partitioned tables, the value you specify for the parameter or storage
characteristic is the default physical attribute of the segments associated with all
partitions specified in this CREATE statement (and in subsequent ALTER TABLE ... ADD
PARTITION statements), unless you explicitly override that value in the PARTITION
clause of the statement that creates the partition.

If you omit this clause, then Oracle Database sets PCTFREE to 10, PCTUSED to 40, and
INITRANS to 1.
See Also:
■

■

physical_attributes_clause on page 8-44 and storage_clause on
page 8-46 for a description of these clauses
"Creating a Table: Storage Example" on page 16-71

Specify the tablespace in which Oracle Database creates the table,
object table OIDINDEX, partition, LOB data segment, LOB index segment, or
index-organized table overflow data segment. If you omit TABLESPACE, then the
database creates that item in the default tablespace of the owner of the schema
containing the table.
TABLESPACE

For a heap-organized table with one or more LOB columns, if you omit the TABLESPACE
clause for LOB storage, then the database creates the LOB data and index segments in
the tablespace where the table is created.
For an index-organized table with one or more LOB columns, if you omit TABLESPACE,
then the LOB data and index segments are created in the tablespace in which the
primary key index segment of the index-organized table is created.
For nonpartitioned tables, the value specified for TABLESPACE is the actual physical
attribute of the segment associated with the table. For partitioned tables, the value
specified for TABLESPACE is the default physical attribute of the segments associated
with all partitions specified in the CREATE statement and on subsequent ALTER TABLE ...
ADD PARTITION statements, unless you specify TABLESPACE in the PARTITION
description.
See Also: CREATE TABLESPACE on page 16-83 for more
information on tablespaces

logging_clause
Specify whether the creation of the table and of any indexes required because of
constraints, partition, or LOB storage characteristics will be logged in the redo log file
(LOGGING) or not (NOLOGGING).The logging attribute of the table is independent of that
of its indexes.
This attribute also specifies whether subsequent direct loader (SQL*Loader) and
direct-path INSERT operations against the table, partition, or LOB storage are logged
(LOGGING) or not logged (NOLOGGING).
Refer to logging_clause on page 8-38 for a full description of this clause.
table_compression
The table_compression clause is valid only for heap-organized tables. Use this clause
to instruct the database whether to compress data segments to reduce disk use. The
COMPRESS keyword enables table compression. The NOCOMPRESS keyword disables table
compression. NOCOMPRESS is the default.

16-34 Oracle Database SQL Language Reference

CREATE TABLE

■

When you enable table compression by specifying either COMPRESS or COMPRESS
BASIC, you enable basic table compression. Oracle Database attempts to compress
data during direct-path INSERT operations when it is productive to do so. The
original import utility (imp) does not support direct-path INSERT, and therefore
cannot import data in a compressed format.
Tables with COMPRESS or COMPRESS BASIC use a PCTFREE value of 0 to maximize
compression, unless you explicitly set a value for PCTFREE in the physical_
attributes_clause.
In earlier releases, this type of compression was called DSS table compression and
was enabled using COMPRESS FOR DIRECT_LOAD OPERATIONS. This syntax has been
deprecated.
"Conventional and Direct-Path INSERT" on page 18-54 for
information on direct-path INSERT operations, including restrictions

See Also:

■

When you enable table compression by specifying COMPRESS FOR OLTP, you enable
OLTP table compression. Oracle Database compresses data during all DML
operations on the table. This form of compression is recommended for OLTP
environments.
Tables with COMPRESS FOR OLTP or NOCOMPRESS use the PCTFREE default value of 10,
to maximize compress while still allowing for some future DML changes to the
data, unless you override this default explicitly.
In earlier releases, OLTP table compression was enabled using COMPRESS FOR ALL
OPERATIONS. This syntax has been deprecated.

■

When you specify COMPRESS FOR QUERY or COMPRESS FOR ARCHIVE, you enable
Hybrid Columnar Compression. With Hybrid Columnar Compression, data can
be compressed during bulk load operations. During the load process, data is
transformed into a column-oriented format and then compressed. Oracle Database
uses a compression algorithm appropriate for the level you specify. In general, the
higher the level, the greater the compression ratio. Hybrid Columnar Compression
can result in higher compression ratios, at a greater CPU cost. Therefore, this form
of compression is recommended for data that is not frequently updated.
COMPRESS FOR QUERY is useful in data warehousing environments. Valid values are
LOW and HIGH, with HIGH providing a higher compression ratio. The default is HIGH.
COMPRESS FOR ARCHIVE uses higher compression ratios than COMPRESS FOR QUERY,
and is useful for compressing data that will be stored for long periods of time.
Valid values are LOW and HIGH, with HIGH providing the highest possible
compression ratio. The default is LOW.
Tables with COMPRESS FOR QUERY or COMPRESS FOR ARCHIVE use a PCTFREE value of 0
to maximize compression, unless you explicitly set a value for PCTFREE in the
physical_attributes_clause. For these tables, PCTFREE has no effect for blocks
loaded using direct-path INSERT. PCTFREE is honored for blocks loaded using
conventional INSERT, and for blocks created as a result of DML operations on
blocks originally loaded using direct-path INSERT.
Oracle Database Concepts for more information on Hybrid
Columnar Compression, which is a feature of certain Oracle storage
systems

See Also:

You can specify table compression for the following portions of a heap-organized
table:
SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-35

CREATE TABLE

■

■

■

■

■

■

■

For an entire table, in the physical_properties clause of relational_table or
object_table
For a range partition, in the table_partition_description of the range_
partitions clause
For a composite range partition, in the table_partition_description of the
range_partition_desc
For a composite list partition, in the table_partition_description of the list_
partition_desc
For a list partition, in the table_partition_description of the list_partitions
clause
For a system or reference partition, in the table_partition_description of the
reference_partition_desc
For the storage table of a nested table, in the nested_table_col_properties clause
See Also: Oracle Database PL/SQL Packages and Types Reference for
information about the DBMS_COMPRESSION package, which helps you
choose the correct compression level for an application, and Oracle
Database Administrator's Guide for more information about table
compression, including examples

Restrictions on Table Compression

Table compression is subject to the following

restrictions:
■

■

■

■

■

■

■

■

■

COMPRESS FOR OLTP and COMPRESS BASIC are not supported for tables with more
than 255 columns.
Data segments of BasicFiles LOBs are not compressed. For information on
compression of SecureFiles LOBs, see LOB_compression_clause on page 16-46.
You cannot drop a column from a table that uses COMPRESS BASIC, although you
can set such a column as unused. All of the operations of the ALTER TABLE ... drop_
column_clause are valid for tables that use COMPRESS FOR OLTP, COMPRESS FOR
QUERY, and COMPRESS FOR ARCHIVE.
If you specify COMPRESS FOR OLTP, then chained rows are not compressed unless
the header for the row remains in the original block and all row columns are
moved to another block. If the row chaining results in leaving just the row header
in the block and moving all of the row's columns to the next block, and they all fit
in the next block, then the columns can be compressed.
You cannot specify any type of table compression for an index-organized table,
any overflow segment or partition of an overflow segment, or any mapping table
segment of an index-organized table.
You cannot specify any type of table compression for external tables or for tables
that are part of a cluster.
You cannot specify any type of table compression for tables with LONG or LONG RAW
columns, tables that are owned by the SYS schema and reside in the SYSTEM
tablespace, or tables with ROWDEPENDENCIES enabled.
You cannot specify Hybrid Columnar Compression on tables that are enabled for
flashback archiving.
You cannot specify Hybrid Columnar Compression on the following
object-relational features: object tables, XMLType tables, columns with abstract data

16-36 Oracle Database SQL Language Reference

CREATE TABLE

types, collections stored as tables, or OPAQUE types, including XMLType columns
stored as objects.
■

■

■

When you update a row in a table compressed with Hybrid Columnar
Compression, the ROWID of the row may change.
In tables compressed with Hybrid Columnar Compression, updates to a single
row may result in locks on multiple rows. Concurrency for write transactions may
therefore be affected.
If a table compressed with Hybrid Columnar Compression has a foreign key
constraint, and you insert data using INSERT with the APPEND hint, then the data
will be compressed to a lesser level than is typical with Hybrid Columnar
Compression. To compress the data with Hybrid Columnar Compression, disable
the foreign key constraint, insert the data using INSERT with the APPEND hint, and
then reenable the foreign key constraint.

RECOVERABLE | UNRECOVERABLE
These keywords are deprecated and have been replaced with LOGGING and NOLOGGING,
respectively. Although RECOVERABLE and UNRECOVERABLE are supported for backward
compatibility, Oracle strongly recommends that you use the LOGGING and NOLOGGING
keywords.
Restrictions on [UN]RECOVERABLE This clause is subject to the following
restrictions:
■

You cannot specify RECOVERABLE for partitioned tables or LOB storage
characteristics.

■

You cannot specify UNRECOVERABLE for partitioned or index-organized tables.

■

You can specify UNRECOVERABLE only with AS subquery.

ORGANIZATION
The ORGANIZATION clause lets you specify the order in which the data rows of the table
are stored.
HEAP HEAP indicates that the data rows of table are stored in no particular order.
This is the default.
INDEX INDEX indicates that table is created as an index-organized table. In an
index-organized table, the data rows are held in an index defined on the primary key
for the table.
EXTERNAL

EXTERNAL indicates that table is a read-only table located outside the

database.
See Also:

"External Table Example" on page 16-74

index_org_table_clause
Use the index_org_table_clause to create an index-organized table. Oracle Database
maintains the table rows, both primary key column values and nonkey column values,
in an index built on the primary key. Index-organized tables are therefore best suited
for primary key-based access and manipulation. An index-organized table is an
alternative to:
■

A noncluster table indexed on the primary key by using the CREATE INDEX
statement

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-37

CREATE TABLE

■

A cluster table stored in an indexed cluster that has been created using the CREATE
CLUSTER statement that maps the primary key for the table to the cluster key

You must specify a primary key for an index-organized table, because the primary key
uniquely identifies a row. The primary key cannot be DEFERRABLE. Use the primary key
instead of the rowid for directly accessing index-organized rows.
If an index-organized table is partitioned and contains LOB columns, then you should
specify the index_org_table_clause first, then the LOB_storage_clause, and then the
appropriate table_partitioning_clauses.
You cannot use the TO_LOB function to convert a LONG column to a LOB column in the
subquery of a CREATE TABLE ... AS SELECT statement if you are creating an
index-organized table. Instead, create the index-organized table without the LONG
column, and then use the TO_LOB function in an INSERT ... AS SELECT statement.
The ROWID pseudocolumn of an index-organized table returns logical rowids instead of
physical rowids. A column that you create with the data type ROWID cannot store the
logical rowids of the IOT. The only data you can store in a column of type ROWID is
rowids from heap-organized tables. If you want to store the logical rowids of an IOT,
then create a column of type UROWID instead. A column of type UROWID can store both
physical and logical rowids.
See Also:

"Index-Organized Table Example" on page 16-74

Restrictions on Index-Organized Tables

Index-organized tables are subject to the

following restrictions:
■

■
■

■

The ROWID pseudocolumn of an index-organized table returns logical rowids
instead of physical rowids. A column that you create of type ROWID cannot store
the logical rowids of the IOT. The only data you can store in a ROWID column is
rowids from heap-organized tables. If you want to store the logical rowids of an
IOT, then create a column of type UROWID instead. A column of type UROWID can
store both physical and logical rowids.
You cannot define a virtual column for an index-organized table.
You cannot specify the composite_range_partitions, composite_list_
partitions, or composite_hash_partitions clauses for an index-organized table.
If the index-organized table is a nested table or varray, then you cannot specify
table_partitioning_clauses.

PCTTHRESHOLD integer Specify the percentage of space reserved in the index block
for an index-organized table row. PCTTHRESHOLD must be large enough to hold the
primary key. All trailing columns of a row, starting with the column that causes the
specified threshold to be exceeded, are stored in the overflow segment. PCTTHRESHOLD
must be a value from 1 to 50. If you do not specify PCTTHRESHOLD, then the default is
50.
Restriction on PCTTHRESHOLD You cannot specify PCTTHRESHOLD for individual

partitions of an index-organized table.
mapping_table_clauses Specify MAPPING TABLE to instruct the database to create a

mapping of local to physical ROWIDs and store them in a heap-organized table. This
mapping is needed in order to create a bitmap index on the index-organized table. If
the index-organized table is partitioned, then the mapping table is also partitioned and
its partitions have the same name and physical attributes as the base table partitions.

16-38 Oracle Database SQL Language Reference

CREATE TABLE

Oracle Database creates the mapping table or mapping table partition in the same
tablespace as its parent index-organized table or partition. You cannot query, perform
DML operations on, or modify the storage characteristics of the mapping table or its
partitions.
key_compression The key_compression clauses let you enable or disable key
compression for index-organized tables.
■

Specify COMPRESS to enable key compression, which eliminates repeated
occurrence of primary key column values in index-organized tables. Use integer
to specify the prefix length, which is the number of prefix columns to compress.
The valid range of prefix length values is from 1 to the number of primary key
columns minus 1. The default prefix length is the number of primary key columns
minus 1.

■

Specify NOCOMPRESS to disable key compression in index-organized tables. This is
the default.

Restriction on Key Compression of Index-organized Tables At the partition level,
you can specify COMPRESS, but you cannot specify the prefix length with integer.

The index_org_overflow_clause lets you instruct the
database that index-organized table data rows exceeding the specified threshold are
placed in the data segment specified in this clause.

index_org_overflow_clause

■

■

■

■

When you create an index-organized table, Oracle Database evaluates the
maximum size of each column to estimate the largest possible row. If an overflow
segment is needed but you have not specified OVERFLOW, then the database raises
an error and does not execute the CREATE TABLE statement. This checking function
guarantees that subsequent DML operations on the index-organized table will not
fail because an overflow segment is lacking.
All physical attributes and storage characteristics you specify in this clause after
the OVERFLOW keyword apply only to the overflow segment of the table. Physical
attributes and storage characteristics for the index-organized table itself, default
values for all its partitions, and values for individual partitions must be specified
before this keyword.
If the index-organized table contains one or more LOB columns, then the LOBs
will be stored out-of-line unless you specify OVERFLOW, even if they would
otherwise be small enough be to stored inline.
If table is partitioned, then the database equipartitions the overflow data
segments with the primary key index segments.

INCLUDING column_name Specify a column at which to divide an index-organized
table row into index and overflow portions. The primary key columns are always
stored in the index. column_name can be either the last primary key column or any non
primary key column. All non primary key columns that follow column_name are stored
in the overflow data segment.

If an attempt to divide a row at column_name causes the size of the index portion of the
row to exceed the specified or default PCTTHRESHOLD value, then the database breaks
up the row based on the PCTTHRESHOLD value.
Restriction on the INCLUDING Clause You cannot specify this clause for individual
partitions of an index-organized table.

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-39

CREATE TABLE

external_table_clause
Use the external_table_clause to create an external table, which is a read-only table
whose metadata is stored in the database but whose data in stored outside the
database. Among other capabilities, external tables let you query data without first
loading it into the database.
See Also: Oracle Database Data Warehousing Guide, Oracle Database
Administrator's Guide, and Oracle Database Utilities for information on
the uses for external tables

Because external tables have no data in the database, you define them with a small
subset of the clauses normally available when creating tables.
■

■

■

■

Within the relational_properties clause, you can specify only column and
datatype.
Within the physical_properties_clause, you can specify only the organization of
the table (ORGANIZATION EXTERNAL external_table_clause).
Within the table_properties clause, you can specify only the parallel_clause.
The parallel_clause lets you parallelize subsequent queries on the external data
and subsequent operations that populate the external table.
You can populate the external table at create time by using the AS subquery clause.

No other clauses are permitted in the same CREATE TABLE statement.
See Also:
■
■

"External Table Example" on page 16-74
ALTER TABLE ... "PROJECT COLUMN Clause" on page 12-56 for
information on the effect of changing the default property of the
column projection

Restrictions on External Tables External tables are subject to the following
restrictions:
■

An external table cannot be a temporary table.

■

You cannot specify constraints on an external table.

■

An external table cannot contain virtual columns.

■

You cannot create an index on an external table.

■

An external table cannot have object type, varray, or LONG columns. However, you
can populate LOB columns of an external table with varray or LONG data from an
internal database table.

TYPE access_driver_type indicates the access driver of the external table. The
access driver is the API that interprets the external data for the database. Oracle
Database provides two access drivers: ORACLE_LOADER and ORACLE_DATAPUMP. If you do
not specify TYPE, then the database uses ORACLE_LOADER as the default access driver.
You must specify the ORACLE_DATAPUMP access driver if you specify the AS subquery
clause to unload data from one Oracle Database and reload it into the same or a
different Oracle Database.
TYPE

Oracle Database Utilities for information about the ORACLE_
LOADER and ORACLE_DATAPUMP access drivers

See Also:

16-40 Oracle Database SQL Language Reference

CREATE TABLE

DEFAULT DIRECTORY lets you specify a default directory object
corresponding to a directory on the file system where the external data sources may
reside. The default directory can also be used by the access driver to store auxiliary
files such as error logs.

DEFAULT DIRECTORY

ACCESS PARAMETERS The optional ACCESS PARAMETERS clause lets you assign

values to the parameters of the specific access driver for this external table.
■

The opaque_format_spec specifies all access parameters for the ORACLE_LOADER
and ORACLE_DATAPUMP access drivers. See Oracle Database Utilities for descriptions
of these parameters.
Field names specified in the opaque_format_spec must match columns in the table
definition. Oracle Database ignores any field in the opaque_format_spec that is
not matched by a column in the table definition.

■

USING CLOB subquery lets you derive the parameters and their values through a
subquery. The subquery cannot contain any set operators or an ORDER BY clause. It
must return one row containing a single item of data type CLOB.

Whether you specify the parameters in an opaque_format_spec or derive them using a
subquery, the database does not interpret anything in this clause. It is up to the access
driver to interpret this information in the context of the external data.
LOCATION The LOCATION clause lets you specify one or more external data sources.
Usually the location_specifier is a file, but it need not be. Oracle Database does not
interpret this clause. It is up to the access driver to interpret this information in the
context of the external data. You cannot use wildcards in the location_specifier to
specify multiple files.

The REJECT LIMIT clause lets you specify how many conversion errors
can occur during a query of the external data before an Oracle Database error is
returned and the query is aborted. The default value is 0.

REJECT LIMIT

CLUSTER Clause
The CLUSTER clause indicates that the table is to be part of cluster. The columns listed
in this clause are the table columns that correspond to the cluster columns. Generally,
the cluster columns of a table are the column or columns that make up its primary key
or a portion of its primary key. Refer to CREATE CLUSTER on page 14-2 for more
information.
Specify one column from the table for each column in the cluster key. The columns are
matched by position, not by name.
A cluster table uses the space allocation of the cluster. Therefore, do not use the
PCTFREE, PCTUSED, or INITRANS parameters, the TABLESPACE clause, or the storage_
clause with the CLUSTER clause.
Restrictions on Cluster Tables
■

■

■

Cluster tables are subject to the following restrictions:

Object tables and tables containing LOB columns or columns of the Any*
Oracle-supplied types cannot be part of a cluster.
You cannot specify the parallel_clause or CACHE or NOCACHE for a table that is
part of a cluster.
You cannot specify CLUSTER with either ROWDEPENDENCIES or NOROWDEPENDENCIES
unless the cluster has been created with the same ROWDEPENDENCIES or
NOROWDEPENDENCIES setting.

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-41

CREATE TABLE

table_properties
The table_properties further define the characteristics of the table.

column_properties
Use the column_properties clauses to specify the storage attributes of a column.
object_type_col_properties
The object_type_col_properties determine storage characteristics of an object
column or attribute or of an element of a collection column or attribute.
column

For column, specify an object column or attribute.

substitutable_column_clause The substitutable_column_clause indicates whether
object columns or attributes in the same hierarchy are substitutable for each other. You
can specify that a column is of a particular type, or whether it can contain instances of
its subtypes, or both.
■

■

■

If you specify ELEMENT, then you constrain the element type of a collection column
or attribute to a subtype of its declared type.
The IS OF [TYPE] (ONLY type) clause constrains the type of the object column to a
subtype of its declared type.
NOT SUBSTITUTABLE AT ALL LEVELS indicates that the object column cannot hold
instances corresponding to any of its subtypes. Also, substitution is disabled for
any embedded object attributes and elements of embedded nested tables and
varrays. The default is SUBSTITUTABLE AT ALL LEVELS.

Restrictions on the substitutable_column_clause

This clause is subject to the

following restrictions:
■

■

You cannot specify this clause for an attribute of an object column. However, you
can specify this clause for a object type column of a relational table and for an
object column of an object table if the substitutability of the object table itself has
not been set.
For a collection type column, the only part of this clause you can specify is [NOT]
SUBSTITUTABLE AT ALL LEVELS.

LOB_storage_clause
The LOB_storage_clause lets you specify the storage attributes of LOB data segments.
You must specify at least one clause after the STORE AS keywords. If you specify more
than one clause, then you must specify them in the order shown in the syntax
diagram, from top to bottom.
For a nonpartitioned table, this clause specifies the storage attributes of LOB data
segments of the table.
For a partitioned table, Oracle Database implements this clause depending on where it
is specified:
■

For a partitioned table specified at the table level—when specified in the
physical_properties clause along with one of the partitioning clauses—this
clause specifies the default storage attributes for LOB data segments associated
with each partition or subpartition. These storage attributes apply to all partitions
or subpartitions unless overridden by a LOB_storage_clause at the partition or
subpartition level.

16-42 Oracle Database SQL Language Reference

CREATE TABLE

■

■

For an individual partition of a partitioned table—when specified as part of a
table_partition_description—this clause specifies the storage attributes of the
data segments of the partition or the default storage attributes of any subpartitions
of the partition. A partition-level LOB_storage_clause overrides a table-level LOB_
storage_clause.
For an individual subpartition of a partitioned table—when specified as part of
subpartition_by_hash or subpartition_by_list—this clause specifies the
storage attributes of the data segments of the subpartition. A subpartition-level
LOB_storage_clause overrides both partition-level and table-level LOB_storage_
clauses.

Only the TABLESPACE clause is allowed
when specifying the LOB_storage_clause in a subpartition.

Restriction on the LOB_storage_clause:

See Also:
■

■

Oracle Database SecureFiles and Large Objects Developer's Guide for
detailed information about LOBs, including guidelines for
creating gigabyte LOBs
"Creating a Table: LOB Column Example" on page 16-74

LOB_item
Specify the LOB column name or LOB object attribute for which you are explicitly
defining tablespace and storage characteristics that are different from those of the
table. Oracle Database automatically creates a system-managed index for each LOB_
item you create.
SECUREFILE | BASICFILE
Use this clause to specify the type of LOB storage, either high-performance LOB
(SecureFiles), or the traditional LOB (BasicFiles).
See Also: Oracle Database SecureFiles and Large Objects Developer's
Guide for more information about SecureFiles LOBs

You cannot convert a LOB from one type of storage to the
other. Instead you must migrate to SecureFiles or BasicFiles by using
online redefinition or partition exchange.

Note:

LOB_segname
Specify the name of the LOB data segment. You cannot use LOB_segname if you specify
more than one LOB_item.
LOB_storage_parameters
The LOB_storage_parameters clause lets you specify various elements of LOB storage.
TABLESPACE Clause
to be stored.

Use this clause to specify the tablespace in which LOB data is

storage_clause Use the storage_clause to specify various aspects of LOB segment
storage. Of particular interest in the context of LOB storage is the MAXSIZE clause of the
storage_clause, which can be used in combination with the LOB_retention_clause of
LOB_parameters. Refer to storage_clause on page 8-48 for more information.

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-43

CREATE TABLE

LOB_parameters
Several of the LOB_parameters are no longer needed if you are using SecureFiles for
LOB storage. The PCTVERSION and FREEPOOLS parameters are valid and useful only if
you are using BasicFiles LOB storage.
ENABLE STORAGE IN ROW If you enable storage in row, then the LOB value is
stored in the row (inline) if its length is less than approximately 4000 bytes minus
system control information. This is the default.

For an index-organized table, you cannot
specify this parameter unless you have specified an OVERFLOW segment in the index_
org_table_clause.

Restriction on Enabling Storage in Row

If you disable storage in row, then the LOB value is
stored outside of the row out of line regardless of the length of the LOB value.

DISABLE STORAGE IN ROW

The LOB locator is always stored inline regardless of where the LOB value is stored.
You cannot change the value of STORAGE IN ROW once it is set except by moving the
table. See the move_table_clause on page 12-76 in the ALTER TABLE documentation for
more information.
Specify the number of bytes to be allocated for LOB manipulation. If
integer is not a multiple of the database block size, then the database rounds up in
bytes to the next multiple. For example, if the database block size is 2048 and integer
is 2050, then the database allocates 4096 bytes (2 blocks). The maximum value is 32768
(32K), which is the largest Oracle Database block size allowed. The default CHUNK size
is one Oracle Database block.
CHUNK integer

The value of CHUNK must be less than or equal to the value of NEXT, either the default
value or that specified in the storage_clause. If CHUNK exceeds the value of NEXT, then
the database returns an error. You cannot change the value of CHUNK once it is set.
Specify the maximum percentage of overall LOB storage
space used for maintaining old versions of the LOB. If the database is running in
manual undo mode, then the default value is 10, meaning that older versions of the
LOB data are not overwritten until they consume 10% of the overall LOB storage
space.

PCTVERSION integer

You can specify the PCTVERSION parameter whether the database is running in manual
or automatic undo mode. PCTVERSION is the default in manual undo mode. RETENTION
is the default in automatic undo mode. You cannot specify both PCTVERSION and
RETENTION.
This clause is not valid if you have specified SECUREFILE. If you specify both
SECUREFILE and PCTVERSION, then the database silently ignores the PCTVERSION
parameter.
LOB_retention_clause Use this clause to specify whether you want the LOB
segment retained for flashback purposes, consistent-read purposes, both, or neither.

You can specify the RETENTION parameter only if the database is running in automatic
undo mode. Oracle Database uses the value of the UNDO_RETENTION initialization
parameter to determine the amount of committed undo data to retain in the database.
In automatic undo mode, RETENTION is the default value unless you specify
PCTVERSION. You cannot specify both PCTVERSION and RETENTION.
You can specify the optional settings after RETENTION only if you are using SecureFiles.
The SECUREFILE parameter of the LOB_storage_clause indicates that the database will

16-44 Oracle Database SQL Language Reference

CREATE TABLE

use SecureFiles to manage storage dynamically, taking into account factors such as the
undo mode of the database.
■

■

■

■

Specify MAX to signify that the undo should be retained until the LOB segment has
reached MAXSIZE. If you specify MAX, then you must also specify the MAXSIZE clause
in the storage_clause.
Specify MIN if the database is in flashback mode to limit the undo retention
duration for the specific LOB segment to n seconds.
Specify AUTO if you want to retain undo sufficient for consistent read purposes
only.
Specify NONE if no undo is required for either consistent read or flashback
purposes.

If you do not specify the RETENTION parameter, or you specify RETENTION with no
optional settings, then RETENTION is set to DEFAULT, which is functionally equivalent to
AUTO.
See Also:
■

■

■

■

CREATE TABLE clause LOB_storage_parameters on page 16-43 for
more information on simplified LOB storage using SecureFiles
Oracle Database SecureFiles and Large Objects Developer's Guide for
more information on using SecureFiles
flashback_mode_clause on page 10-41 of ALTER DATABASE for
information on putting a database in flashback mode
"Creating an Undo Tablespace: Example" on page 16-95

FREEPOOLS integer Specify the number of groups of free lists for the LOB segment.
Normally integer will be the number of instances in an Oracle Real Application
Clusters environment or 1 for a single-instance database.

You can specify this parameter only if the database is running in automatic undo
mode. In this mode, FREEPOOLS is the default unless you specify the FREELIST GROUPS
parameter of the storage_clause. If you specify neither FREEPOOLS nor FREELIST
GROUPS, then the database uses a default of FREEPOOLS 1 if the database is in automatic
undo management mode and a default of FREELIST GROUPS 1 if the database is in
manual undo management mode.
This clause is not valid if you have specified SECUREFILE. If you specify both
SECUREFILE and FREEPOOLS, then the database silently ignores the FREEPOOLS
parameter.
Restriction on FREEPOOLS You cannot specify both FREEPOOLS and the FREELIST

GROUPS parameter of the storage_clause.
This clause is valid only for SecureFiles LOBs. Use the
LOB_deduplicate_clause to enable or disable LOB deduplication, which is the
elimination of duplicate LOB data.
LOB_deduplicate_clause

The DEDUPLICATE keyword instructs the database to eliminate duplicate copies of
LOBs. Using a secure hash index to detect duplication, the database coalesces LOBs
with identical content into a single copy, reducing storage consumption and
simplifying storage management.
If you omit this clause, then LOB deduplication is disabled by default.

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-45

CREATE TABLE

This clause implements LOB deduplication for the entire LOB segment. To enable or
disable deduplication for an individual LOB, use the DBMS_LOB.SETOPTIONS procedure.
See Also: Oracle Database SecureFiles and Large Objects Developer's
Guide for more information about LOB deduplication and Oracle
Database PL/SQL Packages and Types Reference for information about
about the DBMS_LOB package
LOB_compression_clause This clause is valid only for SecureFiles LOBs, not for
BasicFiles LOBs. Use the LOB_compression_clause to instruct the database to enable or
disable server-side LOB compression. Random read/write access is possible on
server-side compressed LOB segments. LOB compression is independent from table
compression or index compression. If you omit this clause, then NOCOMPRESS is the
default.

You can specify HIGH, MEDIUM, or LOW to vary the degree of compression. The HIGH
degree of compression incurs higher latency than MEDIUM but provides better
compression. The LOW degree results in significantly higher decompression and
compression speeds, at the cost of slightly lower compression ratio than either HIGH or
MEDIUM. If you omit this optional parameter, then the default is MEDIUM.
This clause implements server-side LOB compression for the entire LOB segment. To
enable or disable compression on an individual LOB, use the DBMS_LOB.SETOPTIONS
procedure.
See Also: Oracle Database SecureFiles and Large Objects Developer's
Guide for more information on server-side LOB storage and Oracle
Database PL/SQL Packages and Types Reference for information about
client-side LOB compression using the UTL_COMPRESS supplied
package and for information about the DBMS_LOB package
ENCRYPT | DECRYPT These clauses are valid only for LOBs that are using
SecureFiles for LOB storage. Specify ENCRYPT to encrypt all LOBs in the column.
Specify DECRYPT to keep the LOB in cleartext. If you omit this clause, then DECRYPT is
the default.

Refer to encryption_spec on page 16-27 for general information on that clause. When
applied to a LOB column, encryption_spec is specific to the individual LOB column,
so the encryption algorithm can differ from that of other LOB columns and other
non-LOB columns. Use the encryption_spec as part of the column_definition to
encrypt the entire LOB column. Use the encryption_spec as part of the LOB_storage_
clause in the table_partition_description to encrypt a LOB partition.
You cannot specify the SALT or NO SALT
clauses of encryption_spec for LOB encryption.

Restriction on encryption_spec for LOBs

See Also: Oracle Database SecureFiles and Large Objects Developer's
Guide for more information on LOB encryption and Oracle Database
PL/SQL Packages and Types Reference for information the DBMS_LOB
package
CACHE | NOCACHE | CACHE READS Refer to CACHE | NOCACHE | CACHE
READS on page 16-62 for information on that clause.

LOB_partition_storage
The LOB_partition_storage clause lets you specify a separate LOB_storage_clause or
varray_col_properties clause for each partition. You must specify the partitions in
16-46 Oracle Database SQL Language Reference

CREATE TABLE

the order of partition position. You can find the order of the partitions by querying the
PARTITION_NAME and PARTITION_POSITION columns of the USER_IND_PARTITIONS view.
If you do not specify a LOB_storage_clause or varray_col_properties clause for a
particular partition, then the storage characteristics are those specified for the LOB
item at the table level. If you also did not specify any storage characteristics for the
LOB item at the table level, then Oracle Database stores the LOB data partition in the
same tablespace as the table partition to which it corresponds.
Restrictions on LOB_partition_storage:

LOB_partition_storage is subject to the

following restrictions:
■

■

In the LOB_parameters of the LOB_storage_clause, you cannot specify
encryption_spec, because it is invalid to specify an encryption algorithm for
partitions and subpartitions.
You can only specify the TABLESPACE clause for hash partitions and all types of
subpartitions.

varray_col_properties
The varray_col_properties let you specify separate storage characteristics for the
LOB in which a varray will be stored. If varray_item is a multilevel collection, then
the database stores all collection items nested within varray_item in the same LOB in
which varray_item is stored.
■

■

■

■

For a nonpartitioned table—when specified in the physical_properties clause
without any of the partitioning clauses—this clause specifies the storage attributes
of the LOB data segments of the varray.
For a partitioned table specified at the table level—when specified in the
physical_properties clause along with one of the partitioning clauses—this
clause specifies the default storage attributes for the varray LOB data segments
associated with each partition (or its subpartitions, if any).
For an individual partition of a partitioned table—when specified as part of a
table_partition_description—this clause specifies the storage attributes of the
varray LOB data segments of that partition or the default storage attributes of the
varray LOB data segments of any subpartitions of this partition. A partition-level
varray_col_properties overrides a table-level varray_col_properties.
For an individual subpartition of a partitioned table—when specified as part of
subpartition_by_hash or subpartition_by_list—this clause specifies the
storage attributes of the varray data segments of this subpartition. A
subpartition-level varray_col_properties overrides both partition-level and
table-level varray_col_properties.

STORE AS [SECUREFILE | BASICFILE] LOB Clause

If you specify STORE AS LOB,

then:
■

■

If the maximum varray size is less than approximately 4000 bytes, then the
database stores the varray as an inline LOB unless you have disabled storage in
row.
If the maximum varray size is greater than approximately 4000 bytes or if you
have disabled storage in row, then the database stores in the varray as an
out-of-line LOB.

If you do not specify STORE AS LOB, then storage is based on the maximum possible size
of the varray rather than on the actual size of a varray column. The maximum size of

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-47

CREATE TABLE

the varray is the number of elements times the element size, plus a small amount for
system control information. If you omit this clause, then:
■

■

If the maximum size of the varray is less than approximately 4000 bytes, then the
database does not store the varray as a LOB, but as inline data.
If the maximum size is greater than approximately 4000 bytes, then the database
always stores the varray as a LOB.
–

If the actual size is less than approximately 4000 bytes, then it is stored as an
inline LOB

–

If the actual size is greater than approximately 4000 bytes, then it is stored as
an out-of-line LOB, as is true for other LOB columns.

substitutable_column_clause The substitutable_column_clause has the same
behavior as described for object_type_col_properties on page 16-42.
See Also:

"Substitutable Table and Column Examples" on page 16-71

Restriction on Varray Column Properties You cannot specify this clause on an
interval partitioned table or a composite-partitioned table.

nested_table_col_properties
The nested_table_col_properties let you specify separate storage characteristics for
a nested table, which in turn enables you to define the nested table as an
index-organized table. Unless you explicitly specify otherwise in this clause:
■

■

■

For a nonpartitioned table, the storage table is created in the same schema and the
same tablespace as the parent table.
For a partitioned table, the storage table is created in the default tablespace of the
schema. By default, nested tables are equipartitioned with the partitioned base
table.
In either case, the storage table uses default storage characteristics, and stores the
nested table values of the column for which it was created.

You must include this clause when creating a table with columns or column attributes
whose type is a nested table. Clauses within nested_table_col_properties that
function the same way they function for the parent table are not repeated here.
Specify the name of a column, or of a top-level attribute of the object
type of the tables, whose type is a nested table.

nested_item

COLUMN_VALUE If the nested table is a multilevel collection, then the inner nested
table or varray may not have a name. In this case, specify COLUMN_VALUE in place of the
nested_item name.
See Also: "Creating a Table: Multilevel Collection Example" on
page 16-73 for examples using nested_item and COLUMN_VALUE

Specify LOCAL to equipartition the nested table with the base table.
This is the default. Oracle Database automatically creates a local partitioned index for
the partitioned nested table.

LOCAL | GLOBAL

Specify GLOBAL to indicate that the nested table is a nonpartitioned nested table of a
partitioned base table.
storage_table

Specify the name of the table where the rows of nested_item reside.

16-48 Oracle Database SQL Language Reference

CREATE TABLE

You cannot query or perform DML statements on storage_table directly, but you can
modify its storage characteristics by specifying its name in an ALTER TABLE statement.
See Also: ALTER TABLE on page 12-2 for information about
modifying nested table column storage characteristics
RETURN [AS]

Specify what Oracle Database returns as the result of a query.

■

VALUE returns a copy of the nested table itself.

■

LOCATOR returns a collection locator to the copy of the nested table.
The locator is scoped to the session and cannot be used across sessions. Unlike a
LOB locator, the collection locator cannot be used to modify the collection
instance.

If you do not specify the segment_attributes_clause or the LOB_storage_clause,
then the nested table is heap organized and is created with default storage
characteristics.
Restrictions on Nested Table Column Properties Nested table column properties are
subject to the following restrictions:
■
■

■
■

You cannot specify this clause for a temporary table.
You cannot specify this clause on an interval partitioned table or a
composite-partitioned table.
You cannot specify the oid_clause.
At create time, you cannot use object_properties to specify an out_of_line_
ref_constraint, inline_ref_constraint, or foreign key constraint for the
attributes of a nested table. However, you can modify a nested table to add such
constraints using ALTER TABLE.
See Also:
■

■

ALTER TABLE on page 12-2 for information about modifying
nested table column storage characteristics
"Nested Table Example" on page 16-73 and "Creating a Table:
Multilevel Collection Example" on page 16-73

XMLType_column_properties
The XMLType_column_properties let you specify storage attributes for an XMLTYPE
column.
XMLType_storage XMLType data can be stored in binary XML, CLOB, or
object-relational columns.
■

Specify BINARY XML to store the XML data in compact binary XML format.
Any LOB parameters you specify are applied to the underlying BLOB column
created for storing the binary XML encoded value.
In earlier releases, binary XML data is stored by default in a BasicFiles LOB.
Beginning with Oracle Database 11g Release 2 (11.2.0.2), if the COMPATIBLE
initialization parameter is 11.2 or higher and you do not specify BASICFILE or
SECUREFILE, then binary XML data is stored in a SecureFiles LOB whenever
possible. If SecureFiles LOB storage is not possible then the binary XML data is
stored in a BasicFiles LOB. This can occur if either of the following is true:

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-49

CREATE TABLE

■

–

The tablespace for the XMLType table does not use automatic segment space
management.

–

A setting in file init.ora prevents SecureFiles LOB storage. For example, see
parameter DB_SECUREFILE in Oracle Database Reference.

Specify CLOB if you want the database to store the XMLType data in a CLOB column.
Storing data in a CLOB column preserves the original content and enhances
retrieval time.
If you specify LOB storage, then you can specify either LOB parameters or the
XMLSchema_spec clause, but not both. Specify the XMLSchema_spec clause if you
want to restrict the table or column to particular schema-based XML instances.
If you do not specify BASICFILE or SECUREFILE with this clause, then the CLOB
column is stored in a BasicFiles LOB.
Oracle recommends against storing XMLType data in a CLOB
column. Use of the CLOB clause in the XMLType_storage clause may be
deprecated in a future release.

Note:

■

Specify OBJECT RELATIONAL if you want the database to store the XMLType data in
object-relational columns. Storing data objects relationally lets you define indexes
on the relational columns and enhances query performance.
If you specify object-relational storage, then you must also specify the XMLSchema_
spec clause.

Use the ALL VARRAYS AS clause if you want the database to store all varrays in an
XMLType column.
In earlier releases, XMLType data is stored in a CLOB column in a BasicFiles LOB by
default. Beginning with Oracle Database 11g Release 2 (11.2.0.2), if the COMPATIBLE
initialization parameter is 11.2 or higher and you do not specify the XMLType_storage
clause, then XMLType data is stored in a binary XML column in a SecureFiles LOB. If
SecureFiles LOB storage is not possible, then it is stored in a binary XML column in a
BasicFiles LOB.
See Also: Oracle Database SecureFiles and Large Objects Developer's
Guide for more information on SecureFiles LOBs
XMLSchema_spec

Refer to the XMLSchema_spec on page 16-70 for the full semantics

of this clause.
See Also:
■

■

■

■

LOB_storage_clause on page 12-45 for information on the LOB_
segname and LOB_parameters clauses
"XMLType Column Examples" on page 16-76 for examples of
XMLType columns in object-relational tables and "Using XML in
SQL Statements" on page F-8 for an example of creating an
XMLSchema
Oracle XML DB Developer's Guide for more information on XMLType
columns and tables and on creating XMLSchemas
Oracle Database PL/SQL Packages and Types Reference for
information on the DBMS_XMLSCHEMA package

16-50 Oracle Database SQL Language Reference

CREATE TABLE

XMLType_virtual_columns
This clause is valid only for XMLType tables with binary XML storage, which you
designate in the XMLType_storage clause. Specify the VIRTUAL COLUMNS clause to define
virtual columns, which can be used as in a function-based index or in the definition of
a constraint. You cannot define a constraint on such a virtual column during creation
of the table, but you can use a subsequent ALTER TABLE statement to add a constraint to
the column.
See Also: Oracle XML DB Developer's Guide for examples of how to
use this clause in an XML environment

table_partitioning_clauses
Use the table_partitioning_clauses to create a partitioned table.
Notes on Partitioning in General

The following notes pertain to all types of

partitioning:
■
■

■

■

You can specify up to a total of 1024K-1 partitions and subpartitions.
You can create a partitioned table with just one partition. A table with one
partition is different from a nonpartitioned table. For example, you cannot add a
partition to a nonpartitioned table.
You can specify a name for every table and LOB partition and for every table and
LOB subpartition, but you need not do so. If you omit the name, then the database
generates names as follows:
–

If you omit a partition name, then the database generates a name of the form
SYS_Pn. System-generated names for LOB data and LOB index partitions take
the form SYS_LOB_Pn and SYS_IL_Pn, respectively.

–

If you specify a subpartition name in subpartition_template, then for each
subpartition created with that template, the database generates a name by
concatenating the partition name with the template subpartition name. For
LOB subpartitions, the generated LOB subpartition name is a concatenation of
the partition name and the template LOB segment name. In either case, if the
concatenation exceeds 30 characters, then the database returns an error and
the statement fails.

–

If you omit a subpartition name when specifying an individual subpartition,
and you have not specified subpartition_template, then the database
generates a name of the form SYS_SUBPn. The corresponding system-generated
names for LOB data and index subpartitions are SYS_LOB_SUBPn and SYS_IL_
SUBPn, respectively.

Tablespace storage can be specified at various levels in the CREATE TABLE statement
for both table segments and LOB segments. The number of tablespaces does not
have to equal the number of partitions or subpartitions. If the number of partitions
or subpartitions is greater than the number of tablespaces, then the database cycles
through the names of the tablespaces.
The database evaluates tablespace storage in the following order of descending
priority:
–

Tablespace storage specified at the individual table subpartition or LOB
subpartition level has the highest priority, followed by storage specified for
the partition or LOB in the subpartition_template.

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-51

CREATE TABLE

■

–

Tablespace storage specified at the individual table partition or LOB partition
level. Storage parameters specified here take precedence over the
subpartition_template.

–

Tablespace storage specified for the table

–

Default tablespace storage specified for the user

By default, nested tables are equipartitioned with the partitioned base table.

Restrictions on Partitioning in General All partitioning is subject to the following
restrictions:
■
■

■

You cannot partition a table that is part of a cluster.
You cannot partition a nested table or varray that is defined as an index-organized
table.
You cannot partition a table containing any LONG or LONG RAW columns.

The storage of partitioned database entities in tablespaces of different block sizes is
subject to several restrictions. Refer to Oracle Database VLDB and Partitioning Guide for
a discussion of these restrictions.
See Also:

"Partitioning Examples" on page 16-76

range_partitions
Use the range_partitions clause to partition the table on ranges of values from the
column list. For an index-organized table, the column list must be a subset of the
primary key columns of the table.
column
Specify an ordered list of columns used to determine into which partition a row
belongs. These columns are the partitioning key. You can specify virtual columns as
partitioning key columns.
INTERVAL Clause
Use this clause to establish interval partitioning for the table. Interval partitions are
partitions based on a numeric range or datetime interval. They extend range
partitioning by instructing the database to create partitions of the specified range or
interval automatically when data inserted into the table exceeds all of the range
partitions.
■
■

■

For expr, specify a valid number or interval expression.
The optional STORE IN clause lets you specify one or more tablespaces into which
the database will store interval partition data.
You must also specify at least one range partition using the PARTITION clause of
range_partitions. The range partition key value determines the high value of the
range partitions, which is called the transition point, and the database creates
interval partitions for data beyond that transition point.

Restrictions on Interval Partitioning The INTERVAL clause is subject to the restrictions
listed in "Restrictions on Partitioning in General" on page 16-52 and "Restrictions on
Range Partitioning" on page 16-54. The following additional restrictions apply:
■

You can specify only one partitioning key column, and it must be of NUMBER, DATE,
FLOAT, or TIMESTAMP data type.

16-52 Oracle Database SQL Language Reference

CREATE TABLE

■
■

This clause is not supported for index-organized tables.
This clause is not supported for tables containing nested table columns, varray
columns, or XMLType columns.

■

You cannot create a domain index on an interval-partitioned table.

■

Interval partitioning is not supported at the subpartition level.

■

■

Serializable transactions do not work with interval partitioning. Trying to insert
data into a partition of an interval partitioned table that does not yet have a
segment causes an error.
In the VALUES clause:
–

You cannot specify MAXVALUE (an infinite upper bound), because doing so
would defeat the purpose of the automatic addition of partitions as needed.

–

You cannot specify NULL values for the partitioning key column.
See Also: Oracle Database VLDB and Partitioning Guide for more
information on interval partitioning

PARTITION partition
If you specify a partition name, then the name partition must conform to the rules
for naming schema objects and their part as described in "Database Object Naming
Rules" on page 3-111. If you omit partition, then the database generates a name as
described in "Notes on Partitioning in General" on page 16-51.
range_values_clause
Specify the noninclusive upper bound for the current partition. The value list is an
ordered list of literal values corresponding to the column list in the range_partitions
clause. You can substitute the keyword MAXVALUE for any literal in the value list.
MAXVALUE specifies a maximum value that will always sort higher than any other value,
including null.
Specifying a value other than MAXVALUE for the highest partition bound imposes an
implicit integrity constraint on the table.
Note: If table is partitioned on a DATE column, and if the date format
does not specify the first two digits of the year, then you must use the
TO_DATE function with the YYYY 4-character format mask for the year.
The RRRR format mask is not supported in this clause. The date format
is determined implicitly by NLS_TERRITORY or explicitly by NLS_DATE_
FORMAT. Refer to Oracle Database Globalization Support Guide for more
information on these initialization parameters.

Oracle Database Concepts for more information about
partition bounds and "Range Partitioning Example" on page 16-76

See Also:

table_partition_description
Use the table_partition_description to define the physical and storage
characteristics of the table.
The deferred_segment_creation clause, segment_attributes_clause and table_
compression clause have the same function as described for the physical_properties of
the table as a whole.

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-53

CREATE TABLE

The key_compression clause and OVERFLOW clause have the same function as described
for the index_org_table_clause.
The LOB_storage_clause lets you specify LOB storage
characteristics for one or more LOB items in this partition or in any range or list
subpartitions of this partition. If you do not specify the LOB_storage_clause for a LOB
item, then the database generates a name for each LOB data partition as described in
"Notes on Partitioning in General" on page 16-51.
LOB_storage_clause

varray_col_properties The varray_col_properties let you specify storage
characteristics for one or more varray items in this partition or in any range or list
subpartitions of this partition.
nested_table_col_properties The nested_table_col_properties let you specify
storage characteristics for one or more nested table storage table items in this partition
or in any range or list subpartitions of this partition. Storage characteristics specified in
this clause override any storage attributes specified at the table level.

partitioning_storage_clause
Use the partitioning_storage_clause to specify storage characteristics for hash
partitions and for range, list, and hash subpartitions.
Restrictions on partitioning_storage_clause

This clause is subject to the following

restrictions:
■

■

The OVERFLOW clause is relevant only for index-organized partitioned tables and is
valid only within the individual_hash_partitions clause. It is not valid for range
or hash partitions or for subpartitions of any type.
You can specify key_compression only for partitions of index-organized table, and
you can specify COMPRESS or NOCOMPRESS, but you cannot specify the prefix length
with integer.

Restrictions on Range Partitioning Range partitioning is subject to the restrictions

listed in "Restrictions on Partitioning in General" on page 16-52. The following
additional restrictions apply:
■
■

■

You cannot specify more than 16 partitioning key columns.
Partitioning key columns must be of type CHAR, NCHAR, VARCHAR2, NVARCHAR2,
VARCHAR, NUMBER, FLOAT, DATE, TIMESTAMP, TIMESTAMP WITH LOCAL TIMEZONE, or RAW.
You cannot specify NULL in the VALUES clause.

list_partitions
Use the list_partitions clause to partition the table on lists of literal values from
column. List partitioning is useful for controlling how individual rows map to specific
partitions.
The list_values_clause of each partition must have at least one
value. No value, including NULL, can appear in more than one partition. List partitions
are not ordered.
list_values_clause

If you specify the literal NULL for a partition value in the VALUES clause, then to access
data in that partition in subsequent queries, you must use an IS NULL condition in the
WHERE clause, rather than a comparison condition.

16-54 Oracle Database SQL Language Reference

CREATE TABLE

The DEFAULT keyword creates a partition into which the database will insert any row
that does not map to another partition. Therefore, you can specify DEFAULT for only
one partition, and you cannot specify any other values for that partition. Further, the
default partition must be the last partition you define. The use of DEFAULT is similar to
the use of MAXVALUE for range partitions.
The string comprising the list of values for each partition can be up to 4K bytes. The
total number of values for all partitions cannot exceed 64K-1.
The subclauses of the table_partition_description
have the same behavior as described for range partitions in table_partition_description
on page 16-55.

table_partition_description

Restrictions on List Partitioning List partitioning is subject to the restrictions listed
in "Restrictions on Partitioning in General" on page 16-52. The following additional
restrictions apply:
■
■

You can specify only one partitioning key column.
The partitioning key column must be of type CHAR, NCHAR, VARCHAR2, NVARCHAR2,
VARCHAR, NUMBER, FLOAT, DATE, TIMESTAMP, TIMESTAMP WITH LOCAL TIMEZONE, or RAW.

hash_partitions
Use the hash_partitions clause to specify that the table is to be partitioned using the
hash method. Oracle Database assigns rows to partitions using a hash function on
values found in columns designated as the partitioning key. You can specify individual
hash partitions, or you can specify how many hash partitions the database should
create.
column Specify an ordered list of columns used to determine into which partition a
row belongs (the partitioning key).
individual_hash_partitions

Use this clause to specify individual partitions by name.

The only clauses you can
specify in the partitioning_storage_clause are the TABLESPACE clause and table
compression.
Restriction on Specifying Individual Hash Partitions

If your enterprise has or will have databases using different
character sets, then use caution when partitioning on character
columns. The sort sequence of characters is not identical in all
character sets. Refer to Oracle Database Globalization Support Guide for
more information on character set support.
Note:

An alternative to defining individual partitions is to
specify the number of hash partitions. In this case, the database assigns partition
names of the form SYS_Pn. The STORE IN clause lets you specify one or more
tablespaces where the hash partition data is to be stored. The number of tablespaces
need not equal the number of partitions. If the number of partitions is greater than the
number of tablespaces, then the database cycles through the names of the tablespaces.

hash_partitions_by_quantity

For both methods of hash partitioning, for optimal load balancing you should specify
a number of partitions that is a power of 2. When you specify individual hash
partitions, you can specify both TABLESPACE and table compression in the
partitioning_storage_clause. When you specify hash partitions by quantity, you

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-55

CREATE TABLE

can specify only TABLESPACE. Hash partitions inherit all other attributes from
table-level defaults.
The table_compression clause has the same function as described for the table_
properties of the table as a whole.
The key_compression clause and OVERFLOW clause have the same function as described
for the index_org_table_clause.
Tablespace storage specified at the table level is overridden by tablespace storage
specified at the partition level, which in turn is overridden by tablespace storage
specified at the subpartition level.
In the individual_hash_partitions clause, the TABLESPACE clause of the
partitioning_storage_clause determines tablespace storage only for the individual
partition being created. In the hash_partitions_by_quantity clause, the STORE IN
clause determines placement of partitions as the table is being created and the default
storage location for subsequently added partitions.
See Also: Oracle Database VLDB and Partitioning Guide for more
information on hash partitioning

Hash partitioning is subject to the restrictions
listed in "Restrictions on Partitioning in General" on page 16-52. The following
additional restrictions apply:
Restrictions on Hash Partitioning

■
■

You cannot specify more than 16 partitioning key columns.
Partitioning key columns must be of type CHAR, NCHAR, VARCHAR2, NVARCHAR2,
VARCHAR, NUMBER, FLOAT, DATE, TIMESTAMP, TIMESTAMP WITH LOCAL TIMEZONE, or RAW.

composite_range_partitions
Use the composite_range_partitions clause to first partition table by range, and
then partition the partitions further into range, list, or hash subpartitions.
The INTERVAL clause has the same semantics for composite range partitioning that it
has for range partitioning. Refer to "INTERVAL Clause" on page 16-52 for more
information.
Specify subpartition_by_range, subpartition_by_list, or subpartition_by_hash to indicate the
type of subpartitioning you want for each composite range partition. Within these
clauses you can specify a subpartition template, which establishes default subpartition
characteristics for subpartitions created as part of this statement or subsequently
created subpartitions.
After establishing the type of subpartitioning you want for the table, and optionally a
subpartition template, you must define at least one range partition.
■

■

■

In the range_partition_desc, you must specify the range_values_clause, which has
the same requirements as for noncomposite range partitions.
Use the table_partition_description to define the physical and storage characteristics
of the each partition.
In the range_partition_desc, use range_subpartition_desc, list_
subpartition_desc, individual_hash_subparts, or hash_subparts_by_quantity
to specify characteristics for the individual subpartitions of the partition. The
values you specify in these clauses supersede for these subpartitions any values
you have specified in the subpartition_template.

16-56 Oracle Database SQL Language Reference

CREATE TABLE

■

The only characteristics you can specify for a list or hash subpartition or any LOB
subpartition are TABLESPACE and table_compression.

Restrictions on Composite Range Partitioning Regardless of the type of
subpartitioning, composite range partitioning is subject to the following restrictions:
■

■

■

The only physical attributes you can specify at the subpartition level are
TABLESPACE and table compression.
You cannot specify composite partitioning for an index-organized table. Therefore,
the OVERFLOW clause of the table_partition_description is not valid for
composite-partitioned tables.
You cannot specify composite partitioning for tables containing nested table
columns, varray columns, or XMLType columns.
"Composite-Partitioned Table Examples" on page 16-79 for
examples of composite range partitioning and Oracle Database VLDB
and Partitioning Guide for examples of composite list partitioning

See Also:

composite_list_partitions
Use the composite_list_partitions clause to first partition table by list, and then
partition the partitions further into range, list, or hash subpartitions.
Specify subpartition_by_range, subpartition_by_list, or subpartition_by_hash to indicate the
type of subpartitioning you want for each composite list partition. Within these clauses
you can specify a subpartition template, which establishes default subpartition
characteristics for subpartitions created as part of this statement and for subsequently
created subpartitions.
After establishing the type of subpartitioning you want for each composite partition,
and optionally defining a subpartition template, you must define at least one list
partition.
■

■

■

In the list_partition_desc, you must specify the list_values_clause, which has the
same requirements as for noncomposite list partitions.
Use the table_partition_description to define the physical and storage characteristics
of the each partition.
In the list_partition_desc, use range_subpartition_desc, list_subpartition_
desc, individual_hash_subparts, or hash_subparts_by_quantity to specify
characteristics for the individual subpartitions of the partition. The values you
specify in these clauses supersede the for these subpartitions any values you have
specified in the subpartition_template.

Restrictions on Composite List Partitioning Composite list partitioning is subject to

the same restrictions as described in "Restrictions on Composite Range Partitioning"
on page 16-57.

composite_hash_partitions
Use the composite_hash_partitions clause to first partition table using the hash
method, and then partition the partitions further into range, list, or hash subpartitions.
Specify subpartition_by_range, subpartition_by_list, or subpartition_by_hash to indicate the
type of subpartitioning you want for each composite hash partition. Within these
clauses you can specify a subpartition template, which establishes default subpartition
characteristics for subpartitions created as part of this statement or subsequently
created subpartitions.

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-57

CREATE TABLE

After establishing the type of subpartitioning you want for the table, and optionally a
subpartition template, you can define the hash partitions in one of the following ways:
■

■

■

Specify hash_partition_desc to define individual hash partitions. In the hash_
partition_desc, use the partitioning_storage_clause to define the storage
characteristics of the each partition. Use range_subpartition_desc, list_
subpartition_desc, or individual_hash_subparts to specify characteristics for
the individual subpartitions of each partition. The values you specify in these
clauses supersede for these subpartitions any values you have specified in the
subpartition_template.
Specify hash_partitions_by_quantity to specify the number of hash partitions. Each
partition will have subpartitions as described in the subpartition template. If you
do not specify a subpartition template, then each partition will have one
subpartition.
If you omit these clauses, then Oracle Database creates a table with one hash
partition. The partition will have subpartitions as described in the subpartition
template. If you do not specify a subpartition template, then the partition will have
one subpartition.

Restrictions on Composite Hash Partitioning Composite hash partitioning is subject
to the same restrictions as described in "Restrictions on Composite Range Partitioning"
on page 16-57.
subpartition_template The subpartition_template is an optional element of range,
list, and hash subpartitioning. The template lets you define default subpartitions for
each table partition. Oracle Database will create these default subpartition
characteristics in any partition for which you do not explicitly define subpartitions.
This clause is useful for creating symmetric partitions. You can override this clause by
explicitly defining subpartitions at the partition level, in range_subpartition_desc,
list_subpartition_desc, individual_hash_subparts, or hash_subparts_by_
quantity.

When defining subpartitions with a template, you can explicitly define range, list, or
hash subpartitions, or you can define a quantity of hash subpartitions.
■

■

To explicitly define subpartitions, use range_subpartition_desc, list_
subpartition_desc, or individual_hash_subparts. You must specify a name for
each subpartition. If you specify the LOB_partitioning_clause of the
partitioning_storage_clause, then you must specify LOB_segname.
To define a quantity of hash subpartitions, specify a positive integer for hash_
subpartition_quantity. The database creates that number of subpartitions in
each partition and assigns subpartition names of the form SYS_SUBPn.
When you specify tablespace storage for the subpartition
template, it does not override any tablespace storage you have
specified explicitly for the partitions of table. To specify tablespace
storage for subpartitions, do one of these things:

Note:

■

■

Omit tablespace storage at the partition level and specify
tablespace storage in the subpartition template.
Define individual subpartitions with specific tablespace
storage.

16-58 Oracle Database SQL Language Reference

CREATE TABLE

Restrictions on Subpartition Templates Subpartition templates are subject to the

following restrictions:
■

■

■

If you specify TABLESPACE for one LOB subpartition, then you must specify
TABLESPACE for all of the LOB subpartitions of that LOB column. You can specify
the same tablespace for more than one LOB subpartition.
If you specify separate LOB storage for list subpartitions using the partitioning_
storage_clause, either in the subpartition_template or when defining
individual subpartitions, then you must specify LOB_segname for both LOB and
varray columns.
For range-hash and list-hash composite-partitioned tables, you can override the
subpartition_template for an individual partition only by specifying the
individual_hash_subparts clause of range_partition_desc or list_partition_
desc. If you attempt to override the subpartition_template by specifying the
hash_subparts_by_quantity clause of range_partition_desc or list_
partition_desc, then an error occurs.

subpartition_by_range
Use the subpartition_by_range clause to indicate that the database should
subpartition by range each partition in table. The subpartitioning column list is
unrelated to the partitioning key but is subject to the same restrictions (see column on
page 16-52).
You can use the subpartition_template to specify default subpartition characteristic
values. See subpartition_template on page 16-58. The database uses these values for any
subpartition in this partition for which you do not explicitly specify the characteristic.
You can also define range subpartitions individually for each partition using the
range_subpartition_desc of range_partition_desc or list_partition_desc. If you
omit both subpartition_template and the range_subpartition_desc, then the
database creates a single MAXVALUE subpartition.
subpartition_by_list
Use the subpartition_by_list clause to indicate that the database should
subpartition each partition in table by literal values from column. You can specify only
one list subpartitioning key column.
You can use the subpartition_template to specify default subpartition characteristic
values. See subpartition_template on page 16-58. The database uses these values for any
subpartition in this partition for which you do not explicitly specify the characteristic.
You can also define list subpartitions individually for each partition using the list_
subpartition_desc of range_partition_desc or list_partition_desc. If you omit
both subpartition_template and the list_subpartition_desc, then the database
creates a single DEFAULT subpartition.
Restrictions on List Subpartitioning List subpartitioning is subject to the same
restrictions as described in Restrictions on Composite Range Partitioning on page 16-57.

subpartition_by_hash
Use the subpartition_by_hash clause to indicate that the database should
subpartition by hash each partition in table. The subpartitioning column list is
unrelated to the partitioning key but is subject to the same restrictions (see column on
page 16-55).

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-59

CREATE TABLE

You can define the subpartitions using the subpartition_template or the
SUBPARTITIONS integer clause. See subpartition_template on page 16-58. In either case,
for optimal load balancing you should specify a number of partitions that is a power
of 2.
If you specify SUBPARTITIONS integer, then you determine the default number of
subpartitions in each partition of table, and optionally one or more tablespaces in
which they are to be stored. The default value is 1. If you omit both this clause and
subpartition_template, then the database will create each partition with one hash
subpartition.
Notes on Composite Partitions The following notes apply to composite partitions:
■

■

■

For all subpartitions, you can use the range_subpartition_desc, list_
subpartition_desc, individual_hash_subparts, or hash_subparts_by_quantity
to specify individual subpartitions by name, and optionally some other
characteristics.
Alternatively, for list and hash subpartitions:
–

You can specify the number of subpartitions and optionally one or more
tablespaces where they are to be stored. In this case, Oracle Database assigns
subpartition names of the form SYS_SUBPn.

–

If you omit the subpartition description and if you have created a subpartition
template, then the database uses the template to create subpartitions. If you
have not created a subpartition template, then the database creates one
DEFAULT list subpartition or one hash subpartition.

For all types of subpartitions, if you omit the subpartitions description entirely,
then the database assigns subpartition names as follows:
–

If you have specified a subpartition template and you have specified partition
names, then the database generates subpartition names of the form
partition_name underscore (_) subpartition_name (for example, P1_SUB1).

–

If you have not specified a subpartition template or if you have specified a
subpartition template but did not specify partition names, then the database
generates subpartition names of the form SYS_SUBPn.

reference_partitioning
Use this clause to partition the table by reference. Partitioning by reference is a method
of equipartitioning the table being created (the child table) by a referential constraint
to an existing partitioned table (the parent table). When you partition a table by
reference, partition maintenance operations subsequently performed on the parent
table automatically cascade to the child table. Therefore, you cannot perform partition
maintenance operations on a reference-partitioned table directly.
constraint

The partitioning referential constraint must meet the following

conditions:
■

■

You must specify a referential integrity constraint defined on the table being
created, which must refer to a primary key or unique constraint on the parent
table. The constraint must be in ENABLE VALIDATE NOT DEFERRABLE state, which is
the default when you specify a referential integrity constraint during table
creation.
All foreign key columns referenced in constraint must be NOT NULL.

16-60 Oracle Database SQL Language Reference

CREATE TABLE

■

■

■

When you specify the constraint, you cannot specify the ON DELETE SET NULL clause
of the references_clause.
The parent table referenced in the constraint must be an existing partitioned table.
It can be partitioned by any method except interval partitioning.
The foreign and parent keys cannot contain any virtual columns that reference
PL/SQL functions or LOB columns.

Use this optional clause to specify partition names and to
define the physical and storage characteristics of the partition. The subclauses of the
table_partition_description have the same behavior as described for range
partitions in table_partition_description on page 16-55.
reference_partition_desc

Restrictions on Reference Partitioning Reference partitioning is subject to the

restrictions listed in "Restrictions on Partitioning in General" on page 16-52. The
following additional restrictions apply:
■

■

■

■

Restrictions for reference partitioning are derived from the partitioning strategy of
the parent table.
You cannot specify this clause for an index-organized table, an external table, or a
domain index storage table.
The parent table can be partitioned by reference, but constraint cannot be
self-referential. The table being created cannot be partitioned based on a reference
to itself.
If ROW MOVEMENT is enabled for the parent table, it must also be enabled for the child
table.
See Also: Oracle Database VLDB and Partitioning Guide for more
information on partitioning by reference and "Reference Partitioning
Example" on page 16-79

system_partitioning
Use this clause to create system partitions. System partitioning does not entail any
partitioning key columns, nor do system partitions have any range or list bounds or
hash algorithms. Rather, they provide a way to equipartition dependent tables such as
nested table or domain index storage tables with partitioned base tables.
■

■

■

If you specify only PARTITION BY SYSTEM, then the database creates one partition
with a system-generated name of the form SYS_Pn.
If you specify PARTITION BY SYSTEM PARTITIONS integer, then the database creates
as many partitions as you specify in integer, which can range from 1 to 1024K-1.
The description of the partition takes the same syntax as reference partitions, so
they share the reference_partition_desc. You can specify additional partition
attributes with the reference_partition_desc syntax. However, within the
table_partition_description, you cannot specify the OVERFLOW clause.

Restrictions on System Partitioning System partitioning is subject to the following
restrictions:
■

You cannot system partition an index-organized table or a table that is part of a
cluster.

■

Composite partitioning is not supported with system partitioning.

■

You cannot split a system partition.

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-61

CREATE TABLE

■
■

You cannot specify system partitioning in a CREATE TABLE ... AS SELECT statement.
To insert data into a system-partitioned table using an INSERT INTO ... AS subquery
statement, you must use partition-extended syntax to specify the partition into
which the values returned by the subquery will be inserted.
Refer to Oracle Database Data Cartridge Developer's Guide
for information on the uses for system partitioning and "References to
Partitioned Tables and Indexes" on page 3-119

See Also:

CACHE | NOCACHE | CACHE READS
Use these clauses to indicate how Oracle Database should store blocks in the buffer
cache. For LOB storage, you can specify CACHE, NOCACHE, or CACHE READS. For other
types of storage, you can specify only CACHE or NOCACHE.
If you omit these clauses, then:
■

In a CREATE TABLE statement, NOCACHE is the default.

■

In an ALTER TABLE statement, the existing value is not changed.

The behavior of CACHE and NOCACHE described in this section does not apply when
Oracle Database chooses to use direct reads or to perform table scans using parallel
query.
For data that is accessed frequently, this clause indicates that the blocks
retrieved for this table are placed at the most recently used end of the least recently
used (LRU) list in the buffer cache when a full table scan is performed. This attribute is
useful for small lookup tables.

CACHE

Oracle Database Concepts for more information on how the
database maintains the least recently used (LRU) list

See Also:

As a parameter in the LOB_storage_clause, CACHE specifies that the database places
LOB data values in the buffer cache for faster access. The database evaluates this
parameter in conjunction with the logging_clause. If you omit this clause, then the
default value for both BasicFiles and SecureFiles LOBs is NOCACHE LOGGING.
Restriction on CACHE You cannot specify CACHE for an index-organized table.
However, index-organized tables implicitly provide CACHE behavior.
NOCACHE For data that is not accessed frequently, this clause indicates that the

blocks retrieved for this table are placed at the least recently used end of the LRU list
in the buffer cache when a full table scan is performed. NOCACHE is the default for LOB
storage.
As a parameter in the LOB_storage_clause, NOCACHE specifies that the LOB values are
not brought into the buffer cache. NOCACHE is the default for LOB storage.
Restriction on NOCACHE

You cannot specify NOCACHE for an index-organized table.

CACHE READS CACHE READS applies only to LOB storage. It specifies that LOB values
are brought into the buffer cache only during read operations but not during write
operations.
logging_clause Use this clause to indicate whether the storage of data blocks should
be logged or not.

16-62 Oracle Database SQL Language Reference

CREATE TABLE

logging_clause on page 8-38 for a description of the
logging_clause when specified as part of LOB_parameters
See Also:

RESULT_CACHE Clause
Use this clause to determine whether the results of statements or query blocks that
name this table are considered for storage in the result cache. Two modes of result
caching are available:
■

■

DEFAULT: Result caching is not determined at the table level. The query is
considered for result caching if the RESULT_CACHE_MODE initialization parameter is
set to FORCE, or if that parameter is set to MANUAL and the RESULT_CACHE hint is
specified in the query. This is the default if you omit this clause.
FORCE: If all tables names in the query have this setting, then the query is always
considered for caching unless the NO_RESULT_CACHE hint is specified for the query.
If one or more tables named in the query are set to DEFAULT, then the effective table
annotation for that query is considered to be DEFAULT, with the semantics
described above.

You can query the RESULT_CACHE column of the DBA_, ALL_, and USER_TABLES data
dictionary views to learn the result cache mode of the table.
The RESULT_CACHE and NO_RESULT_CACHE SQL hints take precedence over these result
cache table annotations and the RESULT_CACHE_MODE initialization parameter. The
RESULT_CACHE_MODE setting of FORCE in turn takes precedence over this table
annotation clause.
The RESULT_CACHE_MODE setting of FORCE is not recommended,
as it can cause significant performance and latching overhead, as
database and clients will try to cache all queries.

Note:

See Also:
■

■

■

■

Oracle Call Interface Programmer's Guide and Oracle Database
Concepts for general information about result caching
Oracle Database Performance Tuning Guide for information about
using this clause
Oracle Database Reference for information about the RESULT_CACHE_
MODE initialization parameter and the *_TABLES data dictionary
views
"RESULT_CACHE Hint" on page 3-105 and "NO_RESULT_
CACHE Hint" on page 3-95 for information about the hints

parallel_clause
The parallel_clause lets you parallelize creation of the table and set the default
degree of parallelism for queries and the DML INSERT, UPDATE, DELETE, and MERGE after
table creation.

Note: The syntax of the parallel_clause supersedes syntax
appearing in earlier releases of Oracle. Superseded syntax is still
supported for backward compatibility but may result in slightly
different behavior from that documented.

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-63

CREATE TABLE

NOPARALLEL

Specify NOPARALLEL for serial execution. This is the default.

PARALLEL Specify PARALLEL if you want Oracle to select a degree of parallelism
equal to the number of CPUs available on all participating instances times the value of
the PARALLEL_THREADS_PER_CPU initialization parameter.
PARALLEL integer Specification of integer indicates the degree of parallelism,
which is the number of parallel threads used in the parallel operation. Each parallel
thread may use one or two parallel execution servers. Normally Oracle calculates the
optimum degree of parallelism, so it is not necessary for you to specify integer.
See Also: parallel_clause on page 8-41 for more information on this
clause

NOROWDEPENDENCIES | ROWDEPENDENCIES
This clause lets you specify whether table will use row-level dependency tracking.
With this feature, each row in the table has a system change number (SCN) that
represents a time greater than or equal to the commit time of the last transaction that
modified the row. You cannot change this setting after table is created.
ROWDEPENDENCIES Specify ROWDEPENDENCIES if you want to enable row-level
dependency tracking. This setting is useful primarily to allow for parallel propagation
in replication environments. It increases the size of each row by 6 bytes.
Restriction on the ROWDEPENDENCIES Clause Oracle does not support table
compression for tables that use row-level dependency tracking. If you specify both the
ROWDEPENDENCIES clause and the table_compression clause, then the table_
compression clause is ignored. To remove the ROWDEPENDENCIES attribute, you must
redefine the table using the DBMS_REDEFINITION package or recreate the table.

Specify NOROWDEPENDENCIES if you do not want table to
use the row-level dependency tracking feature. This is the default.

NOROWDEPENDENCIES

Oracle Database Advanced Replication for information about
the use of row-level dependency tracking in replication environments

See Also:

enable_disable_clause
The enable_disable_clause lets you specify whether Oracle Database should apply a
constraint. By default, constraints are created in ENABLE VALIDATE state.
Restrictions on Enabling and Disabling Constraints

Enabling and disabling

constraints are subject to the following restrictions:
■

■

■

To enable or disable any integrity constraint, you must have defined the constraint
in this or a previous statement.
You cannot enable a foreign key constraint unless the referenced unique or
primary key constraint is already enabled.
In the index_properties clause of the using_index_clause, the INDEXTYPE IS ...
clause is not valid in the definition of a constraint.
See Also: constraint on page 8-4 for more information on constraints
and "Creating a Table: ENABLE/DISABLE Examples" on page 16-72

16-64 Oracle Database SQL Language Reference

CREATE TABLE

Use this clause if you want the constraint to be applied to the data
in the table. This clause is described fully in "ENABLE Clause" on page 8-15 in the
documentation on constraints.

ENABLE Clause

DISABLE Clause Use this clause if you want to disable the integrity constraint. This
clause is described fully in "DISABLE Clause" on page 8-16 in the documentation on
constraints.

The UNIQUE clause lets you enable or disable the unique constraint defined
on the specified column or combination of columns.

UNIQUE

PRIMARY KEY The PRIMARY KEY clause lets you enable or disable the primary key
constraint defined on the table.
CONSTRAINT The CONSTRAINT clause lets you enable or disable the integrity
constraint named constraint_name.

This clause lets you either preserve or drop the index Oracle
Database has been using to enforce a unique or primary key constraint.

KEEP | DROP INDEX

Restriction on Preserving and Dropping Indexes You can specify this clause only
when disabling a unique or primary key constraint.

The using_index_clause lets you specify an index for Oracle
Database to use to enforce a unique or primary key constraint, or lets you instruct the
database to create the index used to enforce the constraint. This clause is discussed
fully in using_index_clause on page 8-17 in the documentation on constraints.

using_index_clause

See Also:
■

■

■

CREATE INDEX on page 14-60 for a description of index_
attributes, the global_partitioned_index and local_partitioned_index
clauses, NOSORT, and the logging_clause in relation to indexes
constraint on page 8-4 for information on the using_index_clause
and on PRIMARY KEY and UNIQUE constraints
"Explicit Index Control Example" on page 8-26 for an example of
using an index to enforce a constraint

CASCADE Specify CASCADE to disable any integrity constraints that depend on the
specified integrity constraint. To disable a primary or unique key that is part of a
referential integrity constraint, you must specify this clause.
Restriction on CASCADE

You can specify CASCADE only if you have specified

DISABLE.
row_movement_clause
The row_movement_clause lets you specify whether the database can move a table row.
It is possible for a row to move, for example, during table compression or an update
operation on partitioned data.

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-65

CREATE TABLE

Caution: If you need static rowids for data access, then do not enable
row movement. For a normal (heap-organized) table, moving a row
changes the rowid of the row. For a moved row in an index-organized
table, the logical rowid remains valid, although the physical guess
component of the logical rowid becomes inaccurate.

■
■

Specify ENABLE to allow the database to move a row, thus changing the rowid.
Specify DISABLE if you want to prevent the database from moving a row, thus
preventing a change of rowid.

If you omit this clause, then the database disables row movement.
Restriction on Row Movement You cannot specify this clause for a nonpartitioned

index-organized table.
flashback_archive_clause
You must have the FLASHBACK ARCHIVE object privilege on the specified flashback data
archive to specify this clause. Use this clause to enable or disable historical tracking for
the table.
■

Specify FLASHBACK ARCHIVE to enable tracking for the table. You can specify
flashback_archive to designate a particular flashback data archive for this table.
The flashback data archive you specify much already exist.
If you omit flashback_archive, then the database uses the default flashback data
archive designated for the system. If no default flashback data archive has been
designated for the system, then you must specify flashback_archive.

■

Specify NO FLASHBACK ARCHIVE to disable tracking for the table. This is the default.

Restrictions on flashback_archive_clause Flashback data archives are subject to the
following restrictions:
■

■

■

■

You cannot specify this clause for a nested table, clustered table, temporary table,
remote table, or external table.
You cannot specify this clause for a table compressed with Hybrid Columnar
Compression.
The table for which you are specifying this clause cannot contain any LONG or
nested table columns.
If you specify this clause and subsequently copy the table to a different
database—using the export and import utilities or the transportable tablespace
feature—then the copied table will not be enabled for tracking and the archived
data for the original table will not be available for the copied table.
See Also:
■

■

Oracle Database Advanced Application Developer's Guide for general
information on using flashback data archives
ALTER FLASHBACK ARCHIVE on page 10-74 for information on
changing the quota and retention attributes of the flashback data
archive, as well as adding or changing tablespace storage for the
flashback data archive

16-66 Oracle Database SQL Language Reference

CREATE TABLE

AS subquery
Specify a subquery to determine the contents of the table. The rows returned by the
subquery are inserted into the table upon its creation.
For object tables, subquery can contain either one expression corresponding to the
table type, or the number of top-level attributes of the table type. Refer to SELECT on
page 19-4 for more information.
If subquery returns the equivalent of part or all of an existing materialized view, then
the database may rewrite the query to use the materialized view in place of one or
more tables specified in subquery.
See Also: Oracle Database Data Warehousing Guide for more
information on materialized views and query rewrite

Oracle Database derives data types and lengths from the subquery. Oracle Database
follows the following rules for integrity constraints and other column and table
attributes:
■

■

■

■

■

Oracle Database automatically defines on columns in the new table any NOT NULL
constraints that have a state of NOT DEFERRABLE and VALIDATE, and were explicitly
created on the corresponding columns of the selected table if the subquery selects
the column rather than an expression containing the column. If any rows violate
the constraint, then the database does not create the table and returns an error.
NOT NULL constraints that were implicitly created by Oracle Database on columns
of the selected table (for example, for primary keys) are not carried over to the new
table.
In addition, primary keys, unique keys, foreign keys, check constraints,
partitioning criteria, indexes, and column default values are not carried over to the
new table.
If the selected table is partitioned, then you can choose whether the new table will
be partitioned the same way, partitioned differently, or not partitioned.
Partitioning is not carried over to the new table. Specify any desired partitioning
as part of the CREATE TABLE statement before the AS subquery clause.
A column that is encrypted using Transparent Data Encryption in the selected
table will not be encrypted in the new table unless you define the column in the
new table as encrypted at create time.
Oracle recommends that you encrypt sensitive columns before
populating them with data. This will avoid creating clear text copies
of sensitive data.

Note:

If all expressions in subquery are columns, rather than expressions, then you can omit
the columns from the table definition entirely. In this case, the names of the columns of
table are the same as the columns in subquery.
You can use subquery in combination with the TO_LOB function to convert the values in
a LONG column in another table to LOB values in a column of the table you are creating.

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-67

CREATE TABLE

See Also:
■

■

■

Oracle Database SecureFiles and Large Objects Developer's Guide for a
discussion of why and when to copy LONG data to a LOB
"Conversion Functions" on page 5-6 for a description of how to
use the TO_LOB function
SELECT on page 19-4 for more information on the order_by_
clause

parallel_clause If you specify the parallel_clause in this statement, then the
database will ignore any value you specify for the INITIAL storage parameter and will
instead use the value of the NEXT parameter.
See Also: storage_clause on page 8-46 for information on these
parameters
ORDER BY

The ORDER BY clause lets you order rows returned by the subquery.

When specified with CREATE TABLE, this clause does not necessarily order data across
the entire table. For example, it does not order across partitions. Specify this clause if
you intend to create an index on the same key as the ORDER BY key column. Oracle
Database will cluster data on the ORDER BY key so that it corresponds to the index key.
Restrictions on the Defining Query of a Table The table query is subject to the
following restrictions:
■

■

■

The number of columns in the table must equal the number of expressions in the
subquery.
The column definitions can specify only column names, default values, and
integrity constraints, not data types.
You cannot define a foreign key constraint in a CREATE TABLE statement that
contains AS subquery unless the table is reference partitioned and the constraint is
the table's partitioning referential constraint. In all other cases, you must create the
table without the constraint and then add it later with an ALTER TABLE statement.

object_table
The OF clause lets you explicitly create an object table of type object_type. The
columns of an object table correspond to the top-level attributes of type object_type.
Each row will contain an object instance, and each instance will be assigned a unique,
system-generated object identifier when a row is inserted. If you omit schema, then the
database creates the object table in your own schema.
Object tables, as well as XMLType tables, object views, and XMLType views, do not have
any column names specified for them. Therefore, Oracle defines a system-generated
pseudocolumn OBJECT_ID. You can use this column name in queries and to create
object views with the WITH OBJECT IDENTIFIER clause.
See Also:

"Object Column and Table Examples" on page 16-80

object_table_substitution
Use the object_table_substitution clause to specify whether row objects
corresponding to subtypes can be inserted into this object table.

16-68 Oracle Database SQL Language Reference

CREATE TABLE

NOT SUBSTITUTABLE AT ALL LEVELS indicates
that the object table being created is not substitutable. In addition, substitution is
disabled for all embedded object attributes and elements of embedded nested tables
and arrays. The default is SUBSTITUTABLE AT ALL LEVELS.

NOT SUBSTITUTABLE AT ALL LEVELS

See Also:
■

■

CREATE TYPE on page 17-3 for more information about creating
object types
"User-Defined Types" on page 3-29, "About User-Defined
Functions" on page 5-380, "About SQL Expressions" on page 6-1,
CREATE TYPE on page 17-3, and Oracle Database Object-Relational
Developer's Guide for more information about using REF types

object_properties
The properties of object tables are essentially the same as those of relational tables.
However, instead of specifying columns, you specify attributes of the object.
For attribute, specify the qualified column name of an item in an object.

oid_clause
The oid_clause lets you specify whether the object identifier of the object table should
be system generated or should be based on the primary key of the table. The default is
SYSTEM GENERATED.
Restrictions on the oid_clause This clause is subject to the following restrictions:
■

■

You cannot specify OBJECT IDENTIFIER IS PRIMARY KEY unless you have already
specified a PRIMARY KEY constraint for the table.
You cannot specify this clause for a nested table.
A primary key object identifier is locally unique but not
necessarily globally unique. If you require a globally unique identifier,
then you must ensure that the primary key is globally unique.

Note:

oid_index_clause
This clause is relevant only if you have specified the oid_clause as SYSTEM GENERATED.
It specifies an index, and optionally its storage characteristics, on the hidden object
identifier column.
For index, specify the name of the index on the hidden system-generated object
identifier column. If you omit index, then the database generates a name.

physical_properties and table_properties
The semantics of these clauses are documented in the corresponding sections under
relational tables. See physical_properties on page 16-32 and table_properties on
page 16-42.
XMLType_table
Use the XMLType_table syntax to create a table of data type XMLType. Most of the
clauses used to create an XMLType table have the same semantics that exist for object
tables. The clauses specific to XMLType tables are described in this section.

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-69

CREATE TABLE

Object tables, as well as XMLType tables, object views, and XMLType views, do not have
any column names specified for them. Therefore, Oracle defines a system-generated
pseudocolumn OBJECT_ID. You can use this column name in queries and to create
object views with the WITH OBJECT IDENTIFIER clause.

XMLSchema_spec
This clause lets you specify the URL of a registered XMLSchema, either in the
XMLSCHEMA clause or as part of the ELEMENT clause, and an XML element name.
You must specify an element, although the XMLSchema URL is optional. If you do
specify an XMLSchema URL, then you must already have registered the XMLSchema
using the DBMS_XMLSCHEMA package.
The optional ALLOW | DISALLOW clauses are valid only if you have specified BINARY XML
storage.
■

■

■

■

ALLOW NONSCHEMA indicates that non-schema-based documents can be stored in the
XMLType column.
DISALLOW NONSCHEMA indicates that non-schema-based documents cannot be stored
in the XMLType column. This is the default.
ALLOW ANYSCHEMA indicates that any schema-based document can be stored in the
XMLType column.
DISALLOW ANYSCHEMA indicates that any schema-based document cannot be stored
in the XMLType column. This is the default.
See Also:
■

■

■

Oracle Database PL/SQL Packages and Types Reference for
information on the DBMS_XMLSCHEMA package
Oracle XML DB Developer's Guide for information on creating and
working with XML data
"XMLType Table Examples" on page 16-75

Examples
16

Creating Tables: General Examples
This statement shows how the employees table owned by the sample human resources
(hr) schema was created. A hypothetical name is given to the table and constraints so
that you can duplicate this example in your test database:
CREATE TABLE employees_demo
( employee_id
NUMBER(6)
, first_name
VARCHAR2(20)
, last_name
VARCHAR2(25)
CONSTRAINT emp_last_name_nn_demo NOT NULL
, email
VARCHAR2(25)
CONSTRAINT emp_email_nn_demo
NOT NULL
, phone_number
VARCHAR2(20)
, hire_date
DATE DEFAULT SYSDATE
CONSTRAINT emp_hire_date_nn_demo NOT NULL
, job_id
VARCHAR2(10)
CONSTRAINT
emp_job_nn_demo NOT NULL
, salary
NUMBER(8,2)
CONSTRAINT
emp_salary_nn_demo NOT NULL
, commission_pct NUMBER(2,2)
, manager_id
NUMBER(6)
16-70 Oracle Database SQL Language Reference

CREATE TABLE

, department_id
, dn
, CONSTRAINT
, CONSTRAINT

NUMBER(4)
VARCHAR2(300)
emp_salary_min_demo
CHECK (salary > 0)
emp_email_uk_demo
UNIQUE (email)

) ;

This table contains twelve columns. The employee_id column is of data type NUMBER.
The hire_date column is of data type DATE and has a default value of SYSDATE. The
last_name column is of type VARCHAR2 and has a NOT NULL constraint, and so on.
To define the same employees_demo table in the
example tablespace with a small storage capacity, issue the following statement:
Creating a Table: Storage Example

CREATE TABLE employees_demo
( employee_id
NUMBER(6)
, first_name
VARCHAR2(20)
, last_name
VARCHAR2(25)
CONSTRAINT emp_last_name_nn_demo NOT NULL
, email
VARCHAR2(25)
CONSTRAINT emp_email_nn_demo
NOT NULL
, phone_number
VARCHAR2(20)
, hire_date
DATE DEFAULT SYSDATE
CONSTRAINT emp_hire_date_nn_demo NOT NULL
, job_id
VARCHAR2(10)
CONSTRAINT
emp_job_nn_demo NOT NULL
, salary
NUMBER(8,2)
CONSTRAINT
emp_salary_nn_demo NOT NULL
, commission_pct NUMBER(2,2)
, manager_id
NUMBER(6)
, department_id NUMBER(4)
, dn
VARCHAR2(300)
, CONSTRAINT
emp_salary_min_demo
CHECK (salary > 0)
, CONSTRAINT
emp_email_uk_demo
UNIQUE (email)
)
TABLESPACE example
STORAGE (INITIAL 8M);

Creating a Table: Temporary Table Example The following statement creates a
temporary table today_sales for use by sales representatives in the sample database.
Each sales representative session can store its own sales data for the day in the table.
The temporary data is deleted at the end of the session.
CREATE GLOBAL TEMPORARY TABLE today_sales
ON COMMIT PRESERVE ROWS
AS SELECT * FROM orders WHERE order_date = SYSDATE;

Creating a Table with Deferred Segment Creation: Example The following statement
creates a table with deferred segment creation. Oracle Database will not create a
segment for the data of this table until data is inserted into the table:
CREATE TABLE later (col1 NUMBER, col2 VARCHAR2(20))
SEGMENT CREATION DEFERRED;

The following statements create a type
hierarchy, which can be used to create a substitutable table. Type employee_t inherits
the name and ssn attributes from type person_t and in addition has department_id

Substitutable Table and Column Examples

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-71

CREATE TABLE

and salary attributes. Type part_time_emp_t inherits all of the attributes from
employee_t and, through employee_t, those of person_t and in addition has a num_hrs
attribute. Type part_time_emp_t is final by default, so no further subtypes can be
created under it.
CREATE TYPE person_t AS OBJECT (name VARCHAR2(100), ssn NUMBER)
NOT FINAL;
/
CREATE TYPE employee_t UNDER person_t
(department_id NUMBER, salary NUMBER) NOT FINAL;
/
CREATE TYPE part_time_emp_t UNDER employee_t (num_hrs NUMBER);
/

The following statement creates a substitutable table from the person_t type:
CREATE TABLE persons OF person_t;

The following statement creates a table with a substitutable column of type person_t:
CREATE TABLE books (title VARCHAR2(100), author person_t);

When you insert into persons or books, you can specify values for the attributes of
person_t or any of its subtypes. Examples of insert statements appear in "Inserting
into a Substitutable Tables and Columns: Examples" on page 18-67.
You can extract data from such tables using built-in functions and conditions. For
examples, see the functions TREAT on page 5-327 and SYS_TYPEID on page 5-289,
and the "IS OF type Condition" condition on page 7-25.
Creating a Table: Parallelism Examples The following statement creates a table using

an optimum number of parallel execution servers to scan employees and to populate
dept_80:
CREATE TABLE dept_80
PARALLEL
AS SELECT * FROM employees
WHERE department_id = 80;

Using parallelism speeds up the creation of the table, because the database uses
parallel execution servers to create the table. After the table is created, querying the
table is also faster, because the same degree of parallelism is used to access the table.
The following statement creates the same table serially. Subsequent DML and queries
on the table will also be serially executed.
CREATE TABLE dept_80
AS SELECT * FROM employees
WHERE department_id = 80;

Creating a Table: ENABLE/DISABLE Examples The following statement shows how
the sample table departments was created. The example defines a NOT NULL constraint,
and places it in ENABLE VALIDATE state. A hypothetical name is given to the table so
that you can duplicate this example in your test database:
CREATE TABLE departments_demo
( department_id
NUMBER(4)
, department_name VARCHAR2(30)
CONSTRAINT dept_name_nn
, manager_id
NUMBER(6)
16-72 Oracle Database SQL Language Reference

NOT NULL

CREATE TABLE

, location_id
, dn
) ;

NUMBER(4)
VARCHAR2(300)

The following statement creates the same departments_demo table but also defines a
disabled primary key constraint:
CREATE TABLE departments_demo
( department_id
NUMBER(4)
PRIMARY KEY DISABLE
, department_name VARCHAR2(30)
CONSTRAINT dept_name_nn NOT NULL
, manager_id
NUMBER(6)
, location_id
NUMBER(4)
, dn
VARCHAR2(300)
) ;

The following statement shows how the sample table
pm.print_media was created with a nested table column ad_textdocs_ntab:
Nested Table Example

CREATE TABLE print_media
( product_id
NUMBER(6)
, ad_id
NUMBER(6)
, ad_composite
BLOB
, ad_sourcetext
CLOB
, ad_finaltext
CLOB
, ad_fltextn
NCLOB
, ad_textdocs_ntab textdoc_tab
, ad_photo
BLOB
, ad_graphic
BFILE
, ad_header
adheader_typ
) NESTED TABLE ad_textdocs_ntab STORE AS textdocs_nestedtab;

Creating a Table: Multilevel Collection Example The following example shows how
an account manager might create a table of customers using two levels of nested
tables:
CREATE TYPE phone AS OBJECT (telephone NUMBER);
/
CREATE TYPE phone_list AS TABLE OF phone;
/
CREATE TYPE my_customers AS OBJECT (
cust_name VARCHAR2(25),
phones phone_list);
/
CREATE TYPE customer_list AS TABLE OF my_customers;
/
CREATE TABLE business_contacts (
company_name VARCHAR2(25),
company_reps customer_list)
NESTED TABLE company_reps STORE AS outer_ntab
(NESTED TABLE phones STORE AS inner_ntab);

The following variation of this example shows how to use the COLUMN_VALUE keyword
if the inner nested table has no column or attribute name:
CREATE TYPE phone AS TABLE OF NUMBER;
/
CREATE TYPE phone_list AS TABLE OF phone;
/
CREATE TABLE my_customers (
name VARCHAR2(25),

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-73

CREATE TABLE

phone_numbers phone_list)
NESTED TABLE phone_numbers STORE AS outer_ntab
(NESTED TABLE COLUMN_VALUE STORE AS inner_ntab);

Creating a Table: LOB Column Example The following statement is a variation of the
statement that created the pm.print_media table with some added LOB storage
characteristics:
CREATE TABLE print_media_new
( product_id
NUMBER(6)
, ad_id
NUMBER(6)
, ad_composite
BLOB
, ad_sourcetext
CLOB
, ad_finaltext
CLOB
, ad_fltextn
NCLOB
, ad_textdocs_ntab textdoc_tab
, ad_photo
BLOB
, ad_graphic
BFILE
, ad_header
adheader_typ
) NESTED TABLE ad_textdocs_ntab STORE AS textdocs_nestedtab_new
LOB (ad_sourcetext, ad_finaltext) STORE AS
(TABLESPACE example
STORAGE (INITIAL 6144)
CHUNK 4000
NOCACHE LOGGING);

In the example, the database rounds the value of CHUNK up to 4096 (the nearest
multiple of the block size of 2048).
The following statement is a variation of the sample
table hr.countries, which is index organized:

Index-Organized Table Example

CREATE TABLE countries_demo
( country_id
CHAR(2)
CONSTRAINT country_id_nn_demo NOT NULL
, country_name
VARCHAR2(40)
, currency_name
VARCHAR2(25)
, currency_symbol VARCHAR2(3)
, region
VARCHAR2(15)
, CONSTRAINT
country_c_id_pk_demo
PRIMARY KEY (country_id ) )
ORGANIZATION INDEX
INCLUDING
country_name
PCTTHRESHOLD 2
STORAGE
( INITIAL 4K )
OVERFLOW
STORAGE
( INITIAL 4K );

External Table Example The following statement creates an external table that
represents a subset of the sample table hr.departments. The opaque_format_spec is
shown in italics. Refer to Oracle Database Utilities for information on the ORACLE_LOADER
access driver and how to specify values for the opaque_format_spec.
CREATE TABLE dept_external (
deptno
NUMBER(6),
dname
VARCHAR2(20),
loc
VARCHAR2(25)
)

16-74 Oracle Database SQL Language Reference

CREATE TABLE

ORGANIZATION EXTERNAL
(TYPE oracle_loader
DEFAULT DIRECTORY admin
ACCESS PARAMETERS
(
RECORDS DELIMITED BY newline
BADFILE 'ulcase1.bad'
DISCARDFILE 'ulcase1.dis'
LOGFILE 'ulcase1.log'
SKIP 20
FIELDS TERMINATED BY "," OPTIONALLY ENCLOSED BY '"'
(
deptno
INTEGER EXTERNAL(6),
dname
CHAR(20),
loc
CHAR(25)
)
)
LOCATION ('ulcase1.ctl')
)
REJECT LIMIT UNLIMITED;

See Also: "Creating a Directory: Examples" on page 14-42 to see how
the admin directory was created

XMLType Examples
This section contains brief examples of creating an XMLType table or XMLType column.
For a more expanded version of these examples, refer to "Using XML in SQL
Statements" on page F-8.
XMLType Table Examples The following example creates a very simple XMLType table

with one implicit binary XML column:
CREATE TABLE xwarehouses OF XMLTYPE;

The following example creates an XMLSchema-based table. The XMLSchema must
already have been created (see "Using XML in SQL Statements" on page F-8 for more
information):
CREATE TABLE xwarehouses OF XMLTYPE
XMLSCHEMA "http://www.example.com/xwarehouses.xsd"
ELEMENT "Warehouse";

You can define constraints on an XMLSchema-based table, and you can also create
indexes on XMLSchema-based tables, which greatly enhance subsequent queries. You
can create object-relational views on XMLType tables, and you can create XMLType views
on object-relational tables.
See Also:
■

■

■

"Using XML in SQL Statements" on page F-8 for an example of
adding a constraint
"Creating an Index on an XMLType Table: Example" on page 14-81
for an example of creating an index
"Creating an XMLType View: Example" on page 17-25 for an
example of creating an XMLType view

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-75

CREATE TABLE

XMLType Column Examples The following example creates a table with an XMLType

column stored as a CLOB. This table does not require an XMLSchema, so the content
structure is not predetermined:
CREATE TABLE xwarehouses (
warehouse_id
NUMBER,
warehouse_spec
XMLTYPE)
XMLTYPE warehouse_spec STORE AS CLOB
(TABLESPACE example
STORAGE (INITIAL 6144)
CHUNK 4000
NOCACHE LOGGING);

The following example creates a similar table, but stores XMLType data in an object
relational XMLType column whose structure is determined by the specified schema:
CREATE TABLE xwarehouses (
warehouse_id
NUMBER,
warehouse_spec XMLTYPE)
XMLTYPE warehouse_spec STORE AS OBJECT RELATIONAL
XMLSCHEMA "http://www.example.com/xwarehouses.xsd"
ELEMENT "Warehouse";

The following example creates another similar table with an XMLType column stored as
a SecureFiles CLOB. This table does not require an XMLSchema, so the content structure
is not predetermined. SecureFiles LOBs require a tablespace with automatic
segment-space management, so the example uses the tablespace created in "Specifying
Segment Space Management for a Tablespace: Example" on page 16-97.
CREATE TABLE xwarehouses (
warehouse_id
NUMBER,
warehouse_spec XMLTYPE)
XMLTYPE
warehouse_spec STORE AS SECUREFILE CLOB
(TABLESPACE auto_seg_ts
STORAGE (INITIAL 6144)
CACHE);

Partitioning Examples
The sales table in the sample schema sh is partitioned
by range. The following example shows an abbreviated variation of the sales table.
Constraints and storage elements have been omitted from the example.
Range Partitioning Example

CREATE TABLE range_sales
( prod_id
NUMBER(6)
, cust_id
NUMBER
, time_id
DATE
, channel_id
CHAR(1)
, promo_id
NUMBER(6)
, quantity_sold NUMBER(3)
, amount_sold
NUMBER(10,2)
)
PARTITION BY RANGE (time_id)
(PARTITION SALES_Q1_1998 VALUES LESS
PARTITION SALES_Q2_1998 VALUES LESS
PARTITION SALES_Q3_1998 VALUES LESS
PARTITION SALES_Q4_1998 VALUES LESS
PARTITION SALES_Q1_1999 VALUES LESS
PARTITION SALES_Q2_1999 VALUES LESS
PARTITION SALES_Q3_1999 VALUES LESS

THAN
THAN
THAN
THAN
THAN
THAN
THAN

(TO_DATE('01-APR-1998','DD-MON-YYYY')),
(TO_DATE('01-JUL-1998','DD-MON-YYYY')),
(TO_DATE('01-OCT-1998','DD-MON-YYYY')),
(TO_DATE('01-JAN-1999','DD-MON-YYYY')),
(TO_DATE('01-APR-1999','DD-MON-YYYY')),
(TO_DATE('01-JUL-1999','DD-MON-YYYY')),
(TO_DATE('01-OCT-1999','DD-MON-YYYY')),

16-76 Oracle Database SQL Language Reference

CREATE TABLE

PARTITION
PARTITION
PARTITION
PARTITION
PARTITION

SALES_Q4_1999
SALES_Q1_2000
SALES_Q2_2000
SALES_Q3_2000
SALES_Q4_2000

VALUES
VALUES
VALUES
VALUES
VALUES

LESS
LESS
LESS
LESS
LESS

THAN
THAN
THAN
THAN
THAN

(TO_DATE('01-JAN-2000','DD-MON-YYYY')),
(TO_DATE('01-APR-2000','DD-MON-YYYY')),
(TO_DATE('01-JUL-2000','DD-MON-YYYY')),
(TO_DATE('01-OCT-2000','DD-MON-YYYY')),
(MAXVALUE))

;

For information about partitioned table maintenance operations, see Oracle Database
VLDB and Partitioning Guide.
The following example creates a variation of the
oe.customers table that is partitioned by interval on the credit_limit column. One
range partition is created to establish the transition point. All of the original data in the
table is within the bounds of the range partition. Then data is added that exceeds the
range partition, and the database creates a new interval partition.
Interval Partitioning Example

CREATE TABLE customers_demo (
customer_id number(6),
cust_first_name varchar2(20),
cust_last_name varchar2(20),
credit_limit number(9,2))
PARTITION BY RANGE (credit_limit)
INTERVAL (1000)
(PARTITION p1 VALUES LESS THAN (5001));
INSERT INTO customers_demo
(customer_id, cust_first_name, cust_last_name, credit_limit)
(select customer_id, cust_first_name, cust_last_name, credit_limit
from customers);

Query the USER_TAB_PARTITIONS data dictionary view before the database creates the
interval partition:
SELECT partition_name, high_value FROM user_tab_partitions
WHERE table_name = 'CUSTOMERS_DEMO';
PARTITION_NAME
HIGH_VALUE
------------------------------ --------------P1
5001

Insert data into the table that exceeds the high value of the range partition:
INSERT INTO customers_demo
VALUES (699, 'Fred', 'Flintstone', 5500);

Query the USER_TAB_PARTITIONS view again after the insert to learn the
system-generated name of the interval partition created to accommodate the inserted
data. (The system-generated name will vary for each session.)
SELECT partition_name, high_value FROM user_tab_partitions
WHERE table_name = 'CUSTOMERS_DEMO'
ORDER BY partition_name;
PARTITION_NAME
-----------------------------P1
SYS_P44

HIGH_VALUE
--------------5001
6001

The following statement shows how the sample table
oe.customers might have been created as a list-partitioned table. Some columns and
all constraints of the sample table have been omitted in this example.
List Partitioning Example

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-77

CREATE TABLE

CREATE TABLE list_customers
( customer_id
NUMBER(6)
, cust_first_name
VARCHAR2(20)
, cust_last_name
VARCHAR2(20)
, cust_address
CUST_ADDRESS_TYP
, nls_territory
VARCHAR2(30)
, cust_email
VARCHAR2(30))
PARTITION BY LIST (nls_territory) (
PARTITION asia VALUES ('CHINA', 'THAILAND'),
PARTITION europe VALUES ('GERMANY', 'ITALY', 'SWITZERLAND'),
PARTITION west VALUES ('AMERICA'),
PARTITION east VALUES ('INDIA'),
PARTITION rest VALUES (DEFAULT));

This statement creates a partitioned
table print_media_demo with two partitions p1 and p2, and a number of LOB columns.
The statement uses the sample table pm.print_media.

Partitioned Table with LOB Columns Example

CREATE TABLE print_media_demo
( product_id NUMBER(6)
, ad_id NUMBER(6)
, ad_composite BLOB
, ad_sourcetext CLOB
, ad_finaltext CLOB
, ad_fltextn NCLOB
, ad_textdocs_ntab textdoc_tab
, ad_photo BLOB
, ad_graphic BFILE
, ad_header adheader_typ
) NESTED TABLE ad_textdocs_ntab STORE AS textdocs_nestedtab_demo
LOB (ad_composite, ad_photo, ad_finaltext)
STORE AS(STORAGE (INITIAL 20M))
PARTITION BY RANGE (product_id)
(PARTITION p1 VALUES LESS THAN (3000) TABLESPACE tbs_01
LOB (ad_composite, ad_photo)
STORE AS (TABLESPACE tbs_02 STORAGE (INITIAL 10M))
NESTED TABLE ad_textdocs_ntab STORE AS nt_p1 (TABLESPACE example),
PARTITION P2 VALUES LESS THAN (MAXVALUE)
LOB (ad_composite, ad_finaltext)
STORE AS SECUREFILE (TABLESPACE auto_seg_ts)
NESTED TABLE ad_textdocs_ntab STORE AS nt_p2
)
TABLESPACE tbs_03;

Partition p1 will be in tablespace tbs_01. The LOB data partitions for ad_composite
and ad_photo will be in tablespace tbs_02. The LOB data partition for the remaining
LOB columns will be in tablespace tbs_01. The storage attribute INITIAL is specified
for LOB columns ad_composite and ad_photo. Other attributes will be inherited from
the default table-level specification. The default LOB storage attributes not specified at
the table level will be inherited from the tablespace tbs_02 for columns ad_composite
and ad_photo and from tablespace tbs_01 for the remaining LOB columns. LOB index
partitions will be in the same tablespaces as the corresponding LOB data partitions.
Other storage attributes will be based on values of the corresponding attributes of the
LOB data partitions and default attributes of the tablespace where the index partitions
reside. The nested table partition for ad_textdocs_ntab will be stored as nt_p1 in
tablespace example.
Partition p2 will be in the default tablespace tbs_03. The LOB data for ad_composite
and ad_finaltext will be in tablespace auto_seg_ts as SecureFiles LOBs. The LOB

16-78 Oracle Database SQL Language Reference

CREATE TABLE

data for the remaining LOB columns will be in tablespace tbs_03. The LOB index for
columns ad_composite and ad_finaltext will be in tablespace auto_seg_ts. The LOB
index for the remaining LOB columns will be in tablespace tbs_03. The nested table
partition for ad_textdocs_ntab will be stored as nt_p2 in the default tablespace tbs_
03.
The sample table oe.product_information is not
partitioned. However, you might want to partition such a large table by hash for
performance reasons, as shown in this example. The tablespace names are hypothetical
in this example.

Hash Partitioning Example

CREATE TABLE hash_products
( product_id
NUMBER(6)
PRIMARY KEY
, product_name
VARCHAR2(50)
, product_description VARCHAR2(2000)
, category_id
NUMBER(2)
, weight_class
NUMBER(1)
, warranty_period
INTERVAL YEAR TO MONTH
, supplier_id
NUMBER(6)
, product_status
VARCHAR2(20)
, list_price
NUMBER(8,2)
, min_price
NUMBER(8,2)
, catalog_url
VARCHAR2(50)
, CONSTRAINT
product_status_lov_demo
CHECK (product_status in ('orderable'
,'planned'
,'under development'
,'obsolete')
) )
PARTITION BY HASH (product_id)
PARTITIONS 4
STORE IN (tbs_01, tbs_02, tbs_03, tbs_04);

Reference Partitioning Example The next statement uses the hash_products
partitioned table created in the preceding example. It creates a variation of the
oe.order_items table that is partitioned by reference to the hash partitioning on the
product id of hash_products. The resulting child table will be created with five
partitions. For each row of the child table part_order_items, the database evaluates
the foreign key value (product_id) to determine the partition number of the parent
table hash_products to which the referenced key belongs. The part_order_items row
is placed in its corresponding partition.
CREATE TABLE part_order_items (
order_id
NUMBER(12) PRIMARY KEY,
line_item_id
NUMBER(3),
product_id
NUMBER(6) NOT NULL,
unit_price
NUMBER(8,2),
quantity
NUMBER(8),
CONSTRAINT product_id_fk
FOREIGN KEY (product_id) REFERENCES hash_products(product_id))
PARTITION BY REFERENCE (product_id_fk);

The table created in the "Range Partitioning
Example" on page 16-76 divides data by time of sale. If you plan to access recent data
according to distribution channel as well as time, then composite partitioning might be
more appropriate. The following example creates a copy of that range_sales table but
specifies range-hash composite partitioning. The partitions with the most recent data
are subpartitioned with both system-generated and user-defined subpartition names.
Constraints and storage attributes have been omitted from the example.
Composite-Partitioned Table Examples

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-79

CREATE TABLE

CREATE TABLE composite_sales
( prod_id
NUMBER(6)
, cust_id
NUMBER
, time_id
DATE
, channel_id
CHAR(1)
, promo_id
NUMBER(6)
, quantity_sold NUMBER(3)
, amount_sold
NUMBER(10,2)
)
PARTITION BY RANGE (time_id)
SUBPARTITION BY HASH (channel_id)
(PARTITION SALES_Q1_1998 VALUES LESS
PARTITION SALES_Q2_1998 VALUES LESS
PARTITION SALES_Q3_1998 VALUES LESS
PARTITION SALES_Q4_1998 VALUES LESS
PARTITION SALES_Q1_1999 VALUES LESS
PARTITION SALES_Q2_1999 VALUES LESS
PARTITION SALES_Q3_1999 VALUES LESS
PARTITION SALES_Q4_1999 VALUES LESS
PARTITION SALES_Q1_2000 VALUES LESS
PARTITION SALES_Q2_2000 VALUES LESS
SUBPARTITIONS 8,
PARTITION SALES_Q3_2000 VALUES LESS
(SUBPARTITION ch_c,
SUBPARTITION ch_i,
SUBPARTITION ch_p,
SUBPARTITION ch_s,
SUBPARTITION ch_t),
PARTITION SALES_Q4_2000 VALUES LESS
SUBPARTITIONS 4)
;

THAN
THAN
THAN
THAN
THAN
THAN
THAN
THAN
THAN
THAN

(TO_DATE('01-APR-1998','DD-MON-YYYY')),
(TO_DATE('01-JUL-1998','DD-MON-YYYY')),
(TO_DATE('01-OCT-1998','DD-MON-YYYY')),
(TO_DATE('01-JAN-1999','DD-MON-YYYY')),
(TO_DATE('01-APR-1999','DD-MON-YYYY')),
(TO_DATE('01-JUL-1999','DD-MON-YYYY')),
(TO_DATE('01-OCT-1999','DD-MON-YYYY')),
(TO_DATE('01-JAN-2000','DD-MON-YYYY')),
(TO_DATE('01-APR-2000','DD-MON-YYYY')),
(TO_DATE('01-JUL-2000','DD-MON-YYYY'))

THAN (TO_DATE('01-OCT-2000','DD-MON-YYYY'))

THAN (MAXVALUE)

The following examples creates a partitioned table of customers based on the sample
table oe.customers. In this example, the table is partitioned on the credit_limit
column and list subpartitioned on the nls_territory column. The subpartition
template determines the subpartitioning of any subsequently added partitions, unless
you override the template by defining individual subpartitions. This composite
partitioning makes it possible to query the table based on a credit limit range within a
specified region:
CREATE TABLE customers_part (
customer_id
NUMBER(6),
cust_first_name
VARCHAR2(20),
cust_last_name
VARCHAR2(20),
nls_territory
VARCHAR2(30),
credit_limit
NUMBER(9,2))
PARTITION BY RANGE (credit_limit)
SUBPARTITION BY LIST (nls_territory)
SUBPARTITION TEMPLATE
(SUBPARTITION east VALUES
('CHINA', 'JAPAN', 'INDIA', 'THAILAND'),
SUBPARTITION west VALUES
('AMERICA', 'GERMANY', 'ITALY', 'SWITZERLAND'),
SUBPARTITION other VALUES (DEFAULT))
(PARTITION p1 VALUES LESS THAN (1000),
PARTITION p2 VALUES LESS THAN (2500),
PARTITION p3 VALUES LESS THAN (MAXVALUE));

Object Column and Table Examples
Creating Object Tables: Examples
16-80 Oracle Database SQL Language Reference

Consider object type department_typ:

CREATE TABLE

CREATE TYPE department_typ AS OBJECT
( d_name
VARCHAR2(100),
d_address VARCHAR2(200) );
/

Object table departments_obj_t holds department objects of type department_typ:
CREATE TABLE departments_obj_t OF department_typ;

The following statement creates object table salesreps with a user-defined object type,
salesrep_typ:
CREATE OR REPLACE TYPE salesrep_typ AS OBJECT
( repId NUMBER,
repName VARCHAR2(64));
CREATE TABLE salesreps OF salesrep_typ;

Creating a Table with a User-Defined Object Identifier: Example This example
creates an object type and a corresponding object table whose object identifier is
primary key based:
CREATE TYPE employees_typ AS OBJECT
(e_no NUMBER, e_address CHAR(30));
/
CREATE TABLE employees_obj_t OF employees_typ (e_no PRIMARY KEY)
OBJECT IDENTIFIER IS PRIMARY KEY;

You can subsequently reference the employees_obj_t object table using either inline_
ref_constraint or out_of_line_ref_constraint syntax:
CREATE TABLE departments_t
(d_no
NUMBER,
mgr_ref REF employees_typ SCOPE IS employees_obj_t);
CREATE TABLE departments_t (
d_no NUMBER,
mgr_ref REF employees_typ
CONSTRAINT mgr_in_emp REFERENCES employees_obj_t);

Specifying Constraints on Type Columns: Example The following example shows

how to define constraints on attributes of an object type column:
CREATE TYPE address_t AS OBJECT
( hno
NUMBER,
street VARCHAR2(40),
city
VARCHAR2(20),
zip
VARCHAR2(5),
phone VARCHAR2(10) );
/
CREATE TYPE person AS OBJECT
( name
VARCHAR2(40),
dateofbirth DATE,
homeaddress address_t,
manager
REF person );
/
CREATE TABLE persons OF person
( homeaddress NOT NULL,

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-81

CREATE TABLE

UNIQUE (homeaddress.phone),
CHECK (homeaddress.zip IS NOT NULL),
CHECK (homeaddress.city <> 'San Francisco') );

16-82 Oracle Database SQL Language Reference

CREATE TABLESPACE

CREATE TABLESPACE
Purpose
16

Use the CREATE TABLESPACE statement to create a tablespace, which is an allocation of
space in the database that can contain schema objects.
■

■

■

A permanent tablespace contains persistent schema objects. Objects in permanent
tablespaces are stored in data files.
An undo tablespace is a type of permanent tablespace used by Oracle Database to
manage undo data if you are running your database in automatic undo
management mode. Oracle strongly recommends that you use automatic undo
management mode rather than using rollback segments for undo.
A temporary tablespace contains schema objects only for the duration of a session.
Objects in temporary tablespaces are stored in temp files.

When you create a tablespace, it is initially a read/write tablespace. You can
subsequently use the ALTER TABLESPACE statement to take the tablespace offline or
online, add data files or temp files to it, or make it a read-only tablespace.
You can also drop a tablespace from the database with the DROP TABLESPACE statement.
See Also:
■
■

Oracle Database Concepts for information on tablespaces
ALTER TABLESPACE on page 12-90 and DROP TABLESPACE on
page 18-9 for information on modifying and dropping tablespaces

Prerequisites
16

You must have the CREATE TABLESPACE system privilege. To create the SYSAUX
tablespace, you must have the SYSDBA system privilege.
Before you can create a tablespace, you must create a database to contain it, and the
database must be open.
See Also:

CREATE DATABASE on page 14-19

To use objects in a tablespace other than the SYSTEM tablespace:
■

■

If you are running the database in automatic undo management mode, then at
least one UNDO tablespace must be online.
If you are running the database in manual undo management mode, then at least
one rollback segment other than the SYSTEM rollback segment must be online.
Oracle strongly recommends that you run your database in
automatic undo management mode. For more information, refer to
Oracle Database Administrator's Guide.

Note:

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-83

CREATE TABLESPACE

Syntax
16

create_tablespace::=
BIGFILE
permanent_tablespace_clause

SMALLFILE
CREATE

temporary_tablespace_clause

;

undo_tablespace_clause

(permanent_tablespace_clause::= on page 16-84, temporary_tablespace_clause::= on
page 16-85, undo_tablespace_clause::= on page 16-86)
permanent_tablespace_clause::=
,
DATAFILE
TABLESPACE

file_specification

tablespace

MINIMUM

EXTENT

size_clause
K

BLOCKSIZE

integer

logging_clause
FORCE

LOGGING

ENCRYPTION

tablespace_encryption_spec
table_compression

storage_clause

DEFAULT
ONLINE
OFFLINE
extent_management_clause
segment_management_clause
flashback_mode_clause

If you specify the DEFAULT clause, then you must specify at
least one of the clauses table_compression or storage_clause.

Note:

(file_specification::= on page 8-29, size_clause::= on page 8-47, logging_clause::= on
page 16-85, tablespace_encryption_spec::= on page 16-85, table_compression::= on
page 16-11—part of CREATE TABLE, storage_clause::= on page 8-50, extent_management_
clause::= on page 16-85, segment_management_clause::= on page 16-85, flashback_mode_
clause::= on page 16-85)

16-84 Oracle Database SQL Language Reference

CREATE TABLESPACE

logging_clause::=
LOGGING
NOLOGGING
FILESYSTEM_LIKE_LOGGING

tablespace_encryption_spec::=
USING

’

encrypt_algorithm

’

extent_management_clause::=
AUTOALLOCATE
SIZE

size_clause

UNIFORM
EXTENT

MANAGEMENT

LOCAL

(size_clause::= on page 8-47)
segment_management_clause::=
AUTO
SEGMENT

SPACE

MANAGEMENT
MANUAL

flashback_mode_clause::=
ON
FLASHBACK
OFF

temporary_tablespace_clause::=
,
TEMPFILE
TEMPORARY

TABLESPACE

file_specification

tablespace

tablespace_group_clause

extent_management_clause

(file_specification::= on page 8-29, tablespace_group_clause on page 16-93, extent_
management_clause::=)
tablespace_group_clause::=
tablespace_group_name
TABLESPACE

GROUP
’

’

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-85

CREATE TABLESPACE

undo_tablespace_clause::=
,
DATAFILE
UNDO

TABLESPACE

file_specification

tablespace

extent_management_clause

tablespace_retention_clause

(file_specification::= on page 8-29, extent_management_clause::= on page 16-85, tablespace_
retention_clause::= on page 16-86)
tablespace_retention_clause::=
GUARANTEE
RETENTION
NOGUARANTEE

Semantics
16

BIGFILE | SMALLFILE
Use this clause to determine whether the tablespace is a bigfile or smallfile tablespace.
This clause overrides any default tablespace type setting for the database.
■

■

A bigfile tablespace contains only one data file or temp file, which can contain up
to approximately 4 billion (232) blocks. The minimum size of the single data file or
temp file is 256 kilobytes (KB) for a tablespace with 32K blocks or 8K blocks. The
maximum size of the single data file or temp file is 128 terabytes (TB) for a
tablespace with 32K blocks and 32TB for a tablespace with 8K blocks.
A smallfile tablespace is a traditional Oracle tablespace, which can contain 1022
data files or temp files, each of which can contain up to approximately 4 million
(222) blocks.

If you omit this clause, then Oracle Database uses the current default tablespace type
of permanent or temporary tablespace set for the database. If you specify BIGFILE for a
permanent tablespace, then the database by default creates a locally managed
tablespace with automatic segment-space management.
Restriction on Bigfile Tablespaces You can specify only one data file in the DATAFILE
clause or one temp file in the TEMPFILE clause.
See Also:
■

■

Oracle Database Administrator's Guide for more information on
using bigfile tablespaces
"Creating a Bigfile Tablespace: Example" on page 16-95

permanent_tablespace_clause
Use the following clauses to create a permanent tablespace. (Some of these clauses are
also used to create a temporary or undo tablespace.)

16-86 Oracle Database SQL Language Reference

CREATE TABLESPACE

tablespace
Specify the name of the tablespace to be created. The name must satisfy the
requirements listed in "Database Object Naming Rules" on page 3-111.
Note on the SYSAUX Tablespace SYSAUX is a required auxiliary system tablespace.

You must use the CREATE TABLESPACE statement to create the SYSAUX tablespace if you
are upgrading from a release prior to Oracle Database 11g. You must have the SYSDBA
system privilege to specify this clause, and you must have opened the database in
UPGRADE mode.
You must specify EXTENT MANAGEMENT LOCAL and SEGMENT SPACE MANAGEMENT AUTO for
the SYSAUX tablespace. The DATAFILE clause is optional only if you have enabled Oracle
Managed Files. See "DATAFILE | TEMPFILE Clause" on page 16-87 for the behavior of
the DATAFILE clause.
Take care to allocate sufficient space for the SYSAUX tablespace. For guidelines on
creating this tablespace, refer to Oracle Database Upgrade Guide.
Restrictions on the SYSAUX Tablespace

You cannot specify OFFLINE or TEMPORARY

for the SYSAUX tablespace.
DATAFILE | TEMPFILE Clause
Specify the data files to make up the permanent tablespace or the temp files to make
up the temporary tablespace. Use the datafile_tempfile_spec form of file_
specification to create regular data files and temp files in an operating system file
system or to create Oracle Automatic Storage Management (Oracle ASM) disk group
files.
You must specify the DATAFILE or TEMPFILE clause unless you have enabled Oracle
Managed Files by setting a value for the DB_CREATE_FILE_DEST initialization
parameter. For Oracle ASM disk group files, the parameter must be set to a multiple
file creation form of Oracle ASM filenames. If this parameter is set, then the database
creates a system-named 100 MB file in the default file destination specified in the
parameter. The file has AUTOEXTEND enabled and an unlimited maximum size.
Note:

Media recovery does not recognize temp files.

See Also:
■

■

Oracle Automatic Storage Management Administrator's Guide for
more information on using Oracle ASM
file_specification on page 8-29 for a full description, including the
AUTOEXTEND parameter and the multiple file creation form of
Oracle ASM filenames

Notes on Specifying Data Files and Temp Files
■

■

For operating systems that support raw devices, the REUSE keyword of datafile_
tempfile_spec has no meaning when specifying a raw device as a data file. Such a
CREATE TABLESPACE statement will succeed whether or not you specify REUSE.
You can create a tablespace within an Oracle ASM disk group by providing only
the disk group name in the datafile_tempfile_spec. In this case, Oracle ASM
creates a data file in the specified disk group with a system-generated filename.
The data file is auto-extensible with an unlimited maximum size and a default size
of 100 MB. You can use the autoextend_clause to override the default size.
SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-87

CREATE TABLESPACE

■

If you use one of the reference forms of the ASM_filename, which refers to an
existing file, then you must also specify REUSE.

On some operating systems, Oracle does not allocate space
for a temp file until the temp file blocks are actually accessed. This
delay in space allocation results in faster creation and resizing of
temp files, but it requires that sufficient disk space is available
when the temp files are later used. To avoid potential problems,
before you create or resize a temp file, ensure that the available disk
space exceeds the size of the new temp file or the increased size of a
resized temp file. The excess space should allow for anticipated
increases in disk space use by unrelated operations as well. Then
proceed with the creation or resizing operation.

Note:

See Also:
■

■

file_specification on page 8-29 for a full description, including the
AUTOEXTEND parameter
"Enabling Autoextend for a Tablespace: Example" on page 16-96
and "Creating Oracle Managed Files: Examples" on page 16-97

MINIMUM EXTENT Clause
This clause is valid only for a dictionary-managed tablespace. Specify the minimum
size of an extent in the tablespace. This clause lets you control free space fragmentation
in the tablespace by ensuring that the size of every used or free extent in a tablespace is
at least as large as, and is a multiple of, the value specified in the size_clause.
See Also: size_clause on page 8-47 for information on that clause and
Oracle Database VLDB and Partitioning Guide for more information
about using MINIMUM EXTENT to control fragmentation

BLOCKSIZE Clause
Use the BLOCKSIZE clause to specify a nonstandard block size for the tablespace. In
order to specify this clause, the DB_CACHE_SIZE and at least one DB_nK_CACHE_SIZE
parameter must be set, and the integer you specify in this clause must correspond with
the setting of one DB_nK_CACHE_SIZE parameter setting.
You cannot specify nonstandard block sizes for a
temporary tablespace or if you intend to assign this tablespace as the temporary
tablespace for any users.

Restriction on BLOCKSIZE

Oracle recommend that you do not store tablespaces with a 2K
block size on 4K sector size disks, because performance degradation
can result.

Note:

See Also: Oracle Database Reference for information on the DB_nK_
CACHE_SIZE parameter and Oracle Database Concepts for information on
multiple block sizes

16-88 Oracle Database SQL Language Reference

CREATE TABLESPACE

logging_clause
Specify the default logging attributes of all tables, indexes, materialized views,
materialized view logs, and partitions within the tablespace. LOGGING is the default.
This clause is not valid for a temporary or undo tablespace.
The tablespace-level logging attribute can be overridden by logging specifications at
the table, index, materialized view, materialized view log, and partition levels.
See Also:

logging_clause on page 8-38 for a full description of this

clause
FORCE LOGGING
Use this clause to put the tablespace into FORCE LOGGING mode. Oracle Database will
log all changes to all objects in the tablespace except changes to temporary segments,
overriding any NOLOGGING setting for individual objects. The database must be open
and in READ WRITE mode.
This setting does not exclude the NOLOGGING attribute. You can specify both FORCE
LOGGING and NOLOGGING. In this case, NOLOGGING is the default logging mode for objects
subsequently created in the tablespace, but the database ignores this default as long as
the tablespace or the database is in FORCE LOGGING mode. If you subsequently take the
tablespace out of FORCE LOGGING mode, then the NOLOGGING default is once again
enforced.

Note: FORCE LOGGING mode can have performance effects. Refer to
Oracle Database Administrator's Guide for information on when to
use this setting.
Restriction on Forced Logging You cannot specify FORCE LOGGING for an undo or
temporary tablespace.

ENCRYPTION Clause
Use this clause to specify the encryption properties of the tablespace. This clause does
not actually encrypt the tablespace. You must also specify the ENCRYPT keyword as part
of the DEFAULT storage_clause in this statement in order for the tablespace to be
encrypted. In addition, you must already have used ALTER SYSTEM SET ENCRYPTION
WALLET OPEN IDENTIFIED BY ... to load the TDE master key into database memory for
the duration of the instance, or establish a connection to the HSM to send the
encrypted table and tablespace keys to the HSM and receive them back decrypted. For
more information, see "SET ENCRYPTION WALLET Clause" on page 11-68.
You cannot decrypt a tablespace that has been created
encrypted. You must create an unencrypted tablespace and re-create
the database objects in the unencrypted tablespace.

Note:

tablespace_encryption_spec Specify USING 'encrypt_algorithm' to indicate the
name of the encryption algorithm to be used. Valid algorithms are AES256, AES192,
AES128 and 3DES168. If you omit this clause, then the database uses AES128.
See Also:

"Creating an Encrypted Tablespace: Example" on

page 16-96

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-89

CREATE TABLESPACE

DEFAULT Clause
The DEFAULT clause let you specify default parameters for the tablespace.
Use the table_compression clause to specify default
compression of data for all tables created in the tablespace. This clause is not valid for
a temporary tablespace. Refer to the table_compression clause of CREATE TABLE on
page 16-34 for the full semantics of this clause.

table_compression

storage_clause Use the storage_clause to specify storage parameters for all objects
created in the tablespace. This clause is not valid for a temporary tablespace or a
locally managed tablespace. For a dictionary-managed tablespace, you can specify the
following storage parameters with this clause: COMPRESS, INITIAL, NEXT, MINEXTENTS,
MAXEXTENTS, and PCTINCREASE. Refer to storage_clause on page 8-48 for more
information.
See Also:

"Creating Basic Tablespaces: Examples" on page 16-96

ONLINE | OFFLINE Clauses
Use these clauses to determine whether the tablespace is online or offline. This clause
is not valid for a temporary tablespace.
Specify ONLINE to make the tablespace available immediately after creation
to users who have been granted access to the tablespace. This is the default.

ONLINE

OFFLINE

Specify OFFLINE to make the tablespace unavailable immediately after

creation.
The data dictionary view DBA_TABLESPACES indicates whether each tablespace is online
or offline.
extent_management_clause
The extent_management_clause lets you specify how the extents of the tablespace will
be managed.
After you have specified extent management with this clause,
you can change extent management only by migrating the tablespace.

Note:

■

■

AUTOALLOCATE specifies that the tablespace is system managed. Users cannot
specify an extent size. You cannot specify AUTOALLOCATE for a temporary
tablespace.
UNIFORM specifies that the tablespace is managed with uniform extents of SIZE
bytes.The default SIZE is 1 megabyte. All extents of temporary tablespaces are of
uniform size, so this keyword is optional for a temporary tablespace. However,
you must specify UNIFORM in order to specify SIZE. You cannot specify UNIFORM for
an undo tablespace.

If you do not specify AUTOALLOCATE or UNIFORM, then the default is UNIFORM for
temporary tablespaces and AUTOALLOCATE for all other types of tablespaces.
If you do not specify the extent_management_clause, then Oracle Database interprets
the MINIMUM EXTENT clause and the DEFAULT storage_clause to determine extent
management.

16-90 Oracle Database SQL Language Reference

CREATE TABLESPACE

The DICTIONARY keyword is deprecated. It is still supported for
backward compatibility. However, Oracle recommends that you create
locally managed tablespaces. Locally managed tablespaces are much
more efficiently managed than dictionary-managed tablespaces. The
creation of new dictionary-managed tablespaces is scheduled for
desupport.

Note:

Oracle Database Concepts for a discussion of locally
managed tablespaces

See Also:

Restrictions on Extent Management

Extent management is subject to the following

restrictions:
■

■

A permanent locally managed tablespace can contain only permanent objects. If
you need a locally managed tablespace to store temporary objects, for example, if
you will assign it as a user's temporary tablespace, then use the temporary_
tablespace_clause.
If you specify this clause, then you cannot specify DEFAULT storage_clause,
MINIMUM EXTENT, or the temporary_tablespace_clause.
See Also: Oracle Database Administrator's Guide for information on
changing extent management by migrating tablespaces and "Creating
a Locally Managed Tablespace: Example" on page 16-96

segment_management_clause
The segment_management_clause is relevant only for permanent, locally managed
tablespaces. It lets you specify whether Oracle Database should track the used and free
space in the segments in the tablespace using free lists or bitmaps. This clause is not
valid for a temporary tablespace.
Specify AUTO if you want the database to manage the free space of segments in
the tablespace using a bitmap. If you specify AUTO, then the database ignores any
specification for PCTUSED, FREELIST, and FREELIST GROUPS in subsequent storage
specifications for objects in this tablespace. This setting is called automatic
segment-space management and is the default.
AUTO

MANUAL Specify MANUAL if you want the database to manage the free space of

segments in the tablespace using free lists. Oracle strongly recommends that you do
not use this setting and that you create tablespaces with automatic segment-space
management.
To determine the segment management of an existing tablespace, query the SEGMENT_
SPACE_MANAGEMENT column of the DBA_TABLESPACES or USER_TABLESPACES data
dictionary view.

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-91

CREATE TABLESPACE

Notes:
■

■

If you specify AUTO segment management, then:

If you set extent management to LOCAL UNIFORM, then you must
ensure that each extent contains at least 5 database blocks.
If you set extent management to LOCAL AUTOALLOCATE, and if the
database block size is 16K or greater, then Oracle manages
segment space by creating extents with a minimum size of 5
blocks rounded up to 64K.

Restrictions on Automatic Segment-Space Management This clause is subject to the
following restrictions:
■

You can specify this clause only for a permanent, locally managed tablespace.

■

You cannot specify this clause for the SYSTEM tablespace.
See Also:
■

■

■

Oracle Automatic Storage Management Administrator's Guide for
information on automatic segment-space management and when
to use it
Oracle Database Reference for information on the data dictionary
views
"Specifying Segment Space Management for a Tablespace:
Example" on page 16-97

flashback_mode_clause
Use this clause in conjunction with the ALTER DATABASE FLASHBACK clause to specify
whether the tablespace can participate in FLASHBACK DATABASE operations. This clause
is useful if you have the database in FLASHBACK mode but you do not want Oracle
Database to maintain Flashback log data for this tablespace.
This clause is not valid for temporary or undo tablespaces.
FLASHBACK ON Specify FLASHBACK ON to put the tablespace in FLASHBACK mode.
Oracle Database will save Flashback log data for this tablespace and the tablespace can
participate in a FLASHBACK DATABASE operation. If you omit the flashback_mode_
clause, then FLASHBACK ON is the default.
FLASHBACK OFF Specify FLASHBACK OFF to take the tablespace out of FLASHBACK
mode. Oracle Database will not save any Flashback log data for this tablespace. You
must take the data files in this tablespace offline or drop them prior to any subsequent
FLASHBACK DATABASE operation. Alternatively, you can take the entire tablespace
offline. In either case, the database does not drop existing Flashback logs.

The FLASHBACK mode of a tablespace is independent of the
FLASHBACK mode of an individual table.
Note:

16-92 Oracle Database SQL Language Reference

CREATE TABLESPACE

See Also:
■

■

■

Oracle Database Backup and Recovery User's Guide for information
on Oracle Flashback Database
ALTER DATABASE on page 10-8 and FLASHBACK DATABASE
on page 18-24 for information on setting the FLASHBACK mode of
the entire database and reverting the database to an earlier version
FLASHBACK TABLE on page 18-27 and flashback_query_clause on
page 19-17

temporary_tablespace_clause
Use this clause to create a locally managed temporary tablespace, which is an
allocation of space in the database that can contain transient data that persists only for
the duration of a session. This transient data cannot be recovered after process or
instance failure.
The transient data can be user-generated schema objects such as temporary tables or
system-generated data such as temp space used by hash joins and sort operations.
When a temporary tablespace, or a tablespace group of which this tablespace is a
member, is assigned to a particular user, then Oracle Database uses the tablespace for
sorting operations in transactions initiated by that user.
The TEMPFILE clause is described in "DATAFILE | TEMPFILE Clause" on page 16-87.
The extent_management_clause is described in extent_management_clause on
page 16-90.
See Also: Oracle Database Security Guide for information on assigning
temporary tablespaces to users

tablespace_group_clause
This clause is relevant only for temporary tablespaces. Use this clause to determine
whether tablespace is a member of a tablespace group. A tablespace group lets you
assign multiple temporary tablespaces to a single user and increases the addressability
of temporary tablespaces.
■

■

Specify a group name to indicate that tablespace is a member of this tablespace
group. The group name cannot be the same as tablespace or any other existing
tablespace. If the tablespace group already exists, then Oracle Database adds the
new tablespace to that group. If the tablespace group does not exist, then the
database creates the group and adds the new tablespace to that group.
Specify an empty string (' ') to indicate that tablespace is not a member of any
tablespace group.
See Also:
■

■

■

ALTER TABLESPACE on page 12-90 and "Adding a Temporary
Tablespace to a Tablespace Group: Example" on page 16-95 for
information on adding a tablespace to a tablespace group
CREATE USER on page 17-7 for information on assigning a
temporary tablespace to a user
Oracle Database Administrator's Guide for more information on
tablespace groups

The data stored in temporary tablespaces
persists only for the duration of a session. Therefore, only a subset of the CREATE

Restrictions on Temporary Tablespaces

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-93

CREATE TABLESPACE

TABLESPACE clauses are relevant for temporary tablespaces. The only clauses you can
specify for a temporary tablespace are the TEMPFILE clause, the tablespace_group_
clause, and the extent_management_clause.

undo_tablespace_clause
Specify UNDO to create an undo tablespace. When you run the database in automatic
undo management mode, Oracle Database manages undo space using the undo
tablespace instead of rollback segments. This clause is useful if you are now running in
automatic undo management mode but your database was not created in automatic
undo management mode.
Oracle Database always assigns an undo tablespace when you start up the database in
automatic undo management mode. If no undo tablespace has been assigned to this
instance, then the database uses the SYSTEM rollback segment. You can avoid this by
creating an undo tablespace, which the database will implicitly assign to the instance if
no other undo tablespace is currently assigned.
The DATAFILE clause is described in "DATAFILE | TEMPFILE Clause" on page 16-87.
extent_management_clause
It is unnecessary to specify the extent_management_clause when creating an undo
tablespace, because undo tablespaces must be locally managed tablespaces that use
AUTOALLOCATE extent management. If you do specify this clause, then you must specify
EXTENT MANAGEMENT LOCAL or EXTENT MANAGEMENT LOCAL AUTOALLOCATE, both of which
are the same as omitting this clause. Refer to extent_management_clause on page 16-90
for the full semantics of this clause.
tablespace_retention_clause
This clause is valid only for undo tablespaces.
■

■

RETENTION GUARANTEE specifies that Oracle Database should preserve unexpired
undo data in all undo segments of tablespace even if doing so forces the failure of
ongoing operations that need undo space in those segments. This setting is useful
if you need to issue an Oracle Flashback Query or an Oracle Flashback Transaction
Query to diagnose and correct a problem with the data.
RETENTION NOGUARANTEE returns the undo behavior to normal. Space occupied by
unexpired undo data in undo segments can be consumed if necessary by ongoing
transactions. This is the default.

Restrictions on Undo Tablespaces Undo tablespaces are subject to the following
restrictions:
■

■

You cannot create database objects in this tablespace. It is reserved for
system-managed undo data.
The only clauses you can specify for an undo tablespace are the DATAFILE clause
and the extent_management_clause to specify local AUTOALLOCATE extent
management. You cannot specify local UNIFORM extent management or dictionary
extent management using the extent_management_clause. All undo tablespaces
are created permanent, read/write, and in logging mode. Values for MINIMUM
EXTENT and DEFAULT STORAGE are system generated.

16-94 Oracle Database SQL Language Reference

CREATE TABLESPACE

See Also:
■

■

■

Oracle Database Administrator's Guide for information on automatic
undo management and undo tablespaces and Oracle Database
Reference for information on the UNDO_MANAGEMENT parameter
CREATE DATABASE on page 14-19 for information on creating an
undo tablespace during database creation, and ALTER
TABLESPACE on page 12-90 and DROP TABLESPACE on
page 18-9
"Creating an Undo Tablespace: Example" on page 16-95

Examples
16

These examples assume that your database is using 8K blocks.
Creating a Bigfile Tablespace: Example The following example creates a bigfile
tablespace bigtbs_01 with a data file bigtbs_f1.dbf of 20 MB:
CREATE BIGFILE TABLESPACE bigtbs_01
DATAFILE 'bigtbs_f1.dbf'
SIZE 20M AUTOEXTEND ON;

Creating an Undo Tablespace: Example The following example creates a 10 MB
undo tablespace undots1:
CREATE UNDO TABLESPACE undots1
DATAFILE 'undotbs_1a.dbf'
SIZE 10M AUTOEXTEND ON
RETENTION GUARANTEE;

This statement shows how the
temporary tablespace that serves as the default temporary tablespace for database
users in the sample database was created:

Creating a Temporary Tablespace: Example

CREATE TEMPORARY TABLESPACE temp_demo
TEMPFILE 'temp01.dbf' SIZE 5M AUTOEXTEND ON;

Assuming that the default database block size is 2K, and that each bit in the map
represents one extent, then each bit maps 2,500 blocks.
The following example sets the default location for data file creation and then creates a
tablespace with an Oracle-managed temp file in the default location. The temp file is
100 M and is autoextensible with unlimited maximum size. These are the default
values for Oracle Managed Files:
ALTER SYSTEM SET DB_CREATE_FILE_DEST = '$ORACLE_HOME/rdbms/dbs';
CREATE TEMPORARY TABLESPACE tbs_05;

Adding a Temporary Tablespace to a Tablespace Group: Example The following
statement creates the tbs_temp_02 temporary tablespace as a member of the tbs_grp_
01 tablespace group. If the tablespace group does not already exist, then Oracle
Database creates it during execution of this statement:
CREATE TEMPORARY TABLESPACE tbs_temp_02
TEMPFILE 'temp02.dbf' SIZE 5M AUTOEXTEND ON
TABLESPACE GROUP tbs_grp_01;

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-95

CREATE TABLESPACE

Creating Basic Tablespaces: Examples

This statement creates a tablespace named

tbs_01 with one data file:
CREATE TABLESPACE tbs_01
DATAFILE 'tbs_f2.dbf' SIZE 40M
ONLINE;

This statement creates tablespace tbs_03 with one data file and allocates every extent
as a multiple of 500K:
CREATE TABLESPACE tbs_03
DATAFILE 'tbs_f03.dbf' SIZE 20M
LOGGING;

Enabling Autoextend for a Tablespace: Example This statement creates a tablespace
named tbs_02 with one data file. When more space is required, 500 kilobyte extents
will be added up to a maximum size of 100 megabytes:
CREATE TABLESPACE tbs_02
DATAFILE 'diskb:tbs_f5.dbf' SIZE 500K REUSE
AUTOEXTEND ON NEXT 500K MAXSIZE 100M;

Creating a Locally Managed Tablespace: Example The following statement assumes
that the database block size is 2K.
CREATE TABLESPACE tbs_04 DATAFILE 'file_1.dbf' SIZE 10M
EXTENT MANAGEMENT LOCAL UNIFORM SIZE 128K;

This statement creates a locally managed tablespace in which every extent is 128K and
each bit in the bit map describes 64 blocks.
The following statement creates a locally managed tablespace with uniform extents
and shows an example of a table stored in that tablespace:
CREATE TABLESPACE lmt1 DATAFILE 'lmt_file2.dbf' SIZE 100m REUSE
EXTENT MANAGEMENT LOCAL UNIFORM SIZE 1M;
CREATE TABLE lmt_table1 (col1 NUMBER, col2 VARCHAR2(20))
TABLESPACE lmt1 STORAGE (INITIAL 2m);

The initial segment size of the table is 2M.
The following example creates a locally managed tablespace without uniform extents:
CREATE TABLESPACE lmt2 DATAFILE 'lmt_file3.dbf' SIZE 100m REUSE
EXTENT MANAGEMENT LOCAL;
CREATE TABLE lmt_table2 (col1 NUMBER, col2 VARCHAR2(20))
TABLESPACE lmt2 STORAGE (INITIAL 2m MAXSIZE 100m);

The initial segment size of the table is 2M. Oracle Database determines the size of each
extent and the total number of extents allocated to satisfy the initial segment size. The
segment's maximum size is limited to 100M.
In the following example, the first
statement enables encryption for the database by opening the wallet. The second
statement creates an encrypted tablespace.
Creating an Encrypted Tablespace: Example

ALTER SYSTEM SET ENCRYPTION WALLET OPEN IDENTIFIED BY "wallet_password";
CREATE TABLESPACE encrypt_ts
DATAFILE '$ORACLE_HOME/dbs/encrypt_df.dbf' SIZE 1M

16-96 Oracle Database SQL Language Reference

CREATE TABLESPACE

ENCRYPTION USING 'AES256'
DEFAULT STORAGE (ENCRYPT);

The following
example creates a tablespace with automatic segment-space management:

Specifying Segment Space Management for a Tablespace: Example
CREATE TABLESPACE auto_seg_ts DATAFILE 'file_2.dbf' SIZE 1M
EXTENT MANAGEMENT LOCAL
SEGMENT SPACE MANAGEMENT AUTO;

Creating Oracle Managed Files: Examples The following example sets the default
location for data file creation and creates a tablespace with a data file in the default
location. The data file is 100M and is autoextensible with an unlimited maximum size:
ALTER SYSTEM SET DB_CREATE_FILE_DEST = '$ORACLE_HOME/rdbms/dbs';
CREATE TABLESPACE omf_ts1;

The following example creates a tablespace with an Oracle-managed data file of 100M
that is not autoextensible:
CREATE TABLESPACE omf_ts2 DATAFILE AUTOEXTEND OFF;

SQL Statements: CREATE SYNONYM to CREATE TRIGGER 16-97

CREATE TRIGGER

CREATE TRIGGER
Purpose
16

Triggers are defined using PL/SQL. Therefore, this section provides some general
information but refers to Oracle Database PL/SQL Language Reference for details of
syntax and semantics.
Use the CREATE TRIGGER statement to create a database trigger, which is:
■
■

A stored PL/SQL block associated with a table, a schema, or the database or
An anonymous PL/SQL block or a call to a procedure implemented in PL/SQL or
Java

Oracle Database automatically executes a trigger when specified conditions occur.
See Also: ALTER TRIGGER on page 13-2 and DROP TRIGGER on
page 18-12

Prerequisites
16

To create a trigger in your own schema on a table in your own schema or on your own
schema (SCHEMA), you must have the CREATE TRIGGER system privilege.
To create a trigger in any schema on a table in any schema, or on another user's schema
(schema.SCHEMA), you must have the CREATE ANY TRIGGER system privilege.
In addition to the preceding privileges, to create a trigger on DATABASE, you must have
the ADMINISTER DATABASE TRIGGER system privilege.
If the trigger issues SQL statements or calls procedures or functions, then the owner of
the trigger must have the privileges necessary to perform these operations. These
privileges must be granted directly to the owner rather than acquired through roles.

Syntax
16

Triggers are defined using PL/SQL. Therefore, the syntax diagram in this book shows
only the SQL keywords. Refer to Oracle Database PL/SQL Language Reference for the
PL/SQL syntax, semantics, and examples.
create_trigger::=
OR

REPLACE

CREATE

TRIGGER

plsql_source

(plsql_source: See Oracle Database PL/SQL Language Reference.)

Semantics
16

OR REPLACE
Specify OR REPLACE to re-create the trigger if it already exists. Use this clause to change
the definition of an existing trigger without first dropping it.

plsql_source
See Oracle Database PL/SQL Language Reference for the semantics of plsql_source.

16-98 Oracle Database SQL Language Reference

17
71

SQL Statements: CREATE TYPE to
DROP ROLLBACK SEGMENT

This chapter contains the following SQL statements:
■

CREATE TYPE

■

CREATE TYPE BODY

■

CREATE USER

■

CREATE VIEW

■

DELETE

■

DISASSOCIATE STATISTICS

■

DROP CLUSTER

■

DROP CONTEXT

■

DROP DATABASE

■

DROP DATABASE LINK

■

DROP DIMENSION

■

DROP DIRECTORY

■

DROP DISKGROUP

■

DROP EDITION

■

DROP FLASHBACK ARCHIVE

■

DROP FUNCTION

■

DROP INDEX

■

DROP INDEXTYPE

■

DROP JAVA

■

DROP LIBRARY

■

DROP MATERIALIZED VIEW

■

DROP MATERIALIZED VIEW LOG

■

DROP OPERATOR

■

DROP OUTLINE

■

DROP PACKAGE

■

DROP PROCEDURE
SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-1

■

DROP PROFILE

■

DROP RESTORE POINT

■

DROP ROLE

■

DROP ROLLBACK SEGMENT

17-2 Oracle Database SQL Language Reference

CREATE TYPE

CREATE TYPE
Purpose
17

Object types are defined using PL/SQL. Therefore, this section provides some general
information but refers to Oracle Database PL/SQL Language Reference for details of
syntax and semantics.
Use the CREATE TYPE statement to create the specification of an object type, a SQLJ
object type, a named varying array (varray), a nested table type, or an incomplete
object type. You create object types with the CREATE TYPE and the CREATE TYPE BODY
statements. The CREATE TYPE statement specifies the name of the object type, its
attributes, methods, and other properties. The CREATE TYPE BODY statement contains the
code for the methods that implement the type.
Notes:
■

■

If you create an object type for which the type specification
declares only attributes but no methods, then you need not
specify a type body.
If you create a SQLJ object type, then you cannot specify a type
body. The implementation of the type is specified as a Java class.

An incomplete type is a type created by a forward type definition. It is called
"incomplete" because it has a name but no attributes or methods. It can be referenced
by other types, and so can be used to define types that refer to each other. However,
you must fully specify the type before you can use it to create a table or an object
column or a column of a nested table type.
See Also:
■

■

CREATE TYPE BODY on page 17-5 for information on creating
the member methods of a type
Oracle Database Object-Relational Developer's Guide for more
information about objects, incomplete types, varrays, and nested
tables

Prerequisites
17

To create a type in your own schema, you must have the CREATE TYPE system privilege.
To create a type in another user's schema, you must have the CREATE ANY TYPE system
privilege. You can acquire these privileges explicitly or be granted them through a role.
To create a subtype, you must have the UNDER ANY TYPE system privilege or the UNDER
object privilege on the supertype.
The owner of the type must be explicitly granted the EXECUTE object privilege in order
to access all other types referenced within the definition of the type, or the type owner
must be granted the EXECUTE ANY TYPE system privilege. The owner cannot obtain
these privileges through roles.
If the type owner intends to grant other users access to the type, then the owner must
be granted the EXECUTE object privilege on the referenced types with the GRANT OPTION
or the EXECUTE ANY TYPE system privilege with the ADMIN OPTION. Otherwise, the type
owner has insufficient privileges to grant access on the type to other users.

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-3

CREATE TYPE

Syntax
17

Types are defined using PL/SQL. Therefore, the syntax diagram in this book shows
only the SQL keywords. Refer to Oracle Database PL/SQL Language Reference for the
PL/SQL syntax, semantics, and examples.
create_type::=
OR

REPLACE

CREATE

TYPE

plsql_source

(plsql_source: See Oracle Database PL/SQL Language Reference.)

Semantics
17

OR REPLACE
Specify OR REPLACE to re-create the type if it already exists. Use this clause to change
the definition of an existing type without first dropping it.
Users previously granted privileges on the re-created object type can use and reference
the object type without being granted privileges again.
If any function-based indexes depend on the type, then Oracle Database marks the
indexes DISABLED.

plsql_source
See Oracle Database PL/SQL Language Reference for the syntax and semantics of the
plsql_source.

17-4 Oracle Database SQL Language Reference

CREATE TYPE BODY

CREATE TYPE BODY
Purpose
17

Type bodies are defined using PL/SQL. Therefore, this section provides some general
information but refers to Oracle Database PL/SQL Language Reference for details of
syntax and semantics.
Use the CREATE TYPE BODY to define or implement the member methods defined in the
object type specification. You create object types with the CREATE TYPE and the CREATE
TYPE BODY statements. The CREATE TYPE statement specifies the name of the object type,
its attributes, methods, and other properties. The CREATE TYPE BODY statement contains
the code for the methods that implement the type.
For each method specified in an object type specification for which you did not specify
the call_spec, you must specify a corresponding method body in the object type
body.
Note:

If you create a SQLJ object type, then specify it as a Java class.

See Also:
■

■

CREATE TYPE on page 17-3 for information on creating a type
specification
ALTER TYPE on page 13-4 for information on modifying a type
specification

Prerequisites
17

Every member declaration in the CREATE TYPE specification for object types must have
a corresponding construct in the CREATE TYPE or CREATE TYPE BODY statement.
To create or replace a type body in your own schema, you must have the CREATE TYPE
or the CREATE ANY TYPE system privilege. To create an object type in another user's
schema, you must have the CREATE ANY TYPE system privilege. To replace an object type
in another user's schema, you must have the DROP ANY TYPE system privilege.

Syntax
17

Type bodies are defined using PL/SQL. Therefore, the syntax diagram in this book
shows only the SQL keywords. Refer to Oracle Database PL/SQL Language Reference for
the PL/SQL syntax, semantics, and examples.
create_type_body::=
OR
CREATE

REPLACE
TYPE

BODY

plsql_source

(plsql_source: See Oracle Database PL/SQL Language Reference.)

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-5

CREATE TYPE BODY

Semantics
17

OR REPLACE
Specify OR REPLACE to re-create the type body if it already exists. Use this clause to
change the definition of an existing type body without first dropping it.
Users previously granted privileges on the re-created object type body can use and
reference the object type body without being granted privileges again.
You can use this clause to add new member subprogram definitions to specifications
added with the ALTER TYPE ... REPLACE statement.

plsql_source
See Oracle Database PL/SQL Language Reference for the syntax and semantics of the
plsql_source.

17-6 Oracle Database SQL Language Reference

CREATE USER

CREATE USER
Purpose
17

Use the CREATE USER statement to create and configure a database user, which is an
account through which you can log in to the database, and to establish the means by
which Oracle Database permits access by the user.
You can issue this statement in an Oracle Automatic Storage Management (Oracle
ASM) cluster to add a user and password combination to the password file that is local
to the Oracle ASM instance of the current node. Each node's Oracle ASM instance can
use this statement to update its own password file. The password file itself must have
been created by the ORAPWD utility.
You can enable a user to connect to the database through a proxy application or
application server. For syntax and discussion, refer to ALTER USER on page 13-6.

Prerequisites
17

You must have the CREATE USER system privilege. When you create a user with the
CREATE USER statement, the user's privilege domain is empty. To log on to Oracle
Database, a user must have the CREATE SESSION system privilege. Therefore, after
creating a user, you should grant the user at least the CREATE SESSION system privilege.
Refer to GRANT on page 18-33 for more information.
Only a user authenticated AS SYSASM can issue this command to modify the Oracle
ASM instance password file.

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-7

CREATE USER

Syntax
17

create_user::=
BY

password
certificate_DN
AS

’

’
kerberos_principal_name

CREATE

USER

user

IDENTIFIED

EXTERNALLY
directory_DN
AS

’

’

GLOBALLY

DEFAULT

TABLESPACE

tablespace
tablespace

TEMPORARY

TABLESPACE
tablespace_group_name
size_clause

QUOTA

ON

tablespace

UNLIMITED
PROFILE

profile

PASSWORD

EXPIRE
LOCK

ACCOUNT
UNLOCK
ENABLE

EDITIONS
;

(size_clause::= on page 8-47)

Semantics
17

user
Specify the name of the user to be created. This name can contain only characters from
your database character set and must follow the rules described in the section
"Database Object Naming Rules" on page 3-111. Oracle recommends that the user
name contain at least one single-byte character regardless of whether the database
character set also contains multibyte characters.

Oracle recommends that user names and passwords be
encoded in ASCII or EBCDIC characters only, depending on your
platform.

Note:

See Also:

"Creating a Database User: Example" on page 17-12

IDENTIFIED Clause
The IDENTIFIED clause lets you indicate how Oracle Database authenticates the user.

17-8 Oracle Database SQL Language Reference

CREATE USER

BY password
The BY password clause lets you creates a local user and indicates that the user must
specify password to log on to the database. Passwords are case sensitive. Any
subsequent CONNECT string used to connect this user to the database must specify the
password using the same case (upper, lower, or mixed) that is used in this CREATE USER
statement or a subsequent ALTER USER statement. Passwords can contain any
single-byte, multibyte, or special characters, or any combination of these, from your
database character set.
Oracle Database Security Guide for more information about
case-sensitive passwords, password complexity, and other password
guidelines

See Also:

Passwords must follow the rules described in the section "Database Object Naming
Rules" on page 3-111, unless you are using the Oracle Database password complexity
verification routine. That routine requires a more complex combination of characters
than the normal naming rules permit. You implement this routine with the
UTLPWDMG.SQL script, which is further described in Oracle Database Security Guide.

Oracle recommends that user names and passwords be
encoded in ASCII or EBCDIC characters only, depending on your
platform.

Note:

Oracle Database Security Guide to for a detailed discussion
of password management and protection

See Also:

EXTERNALLY Clause
Specify EXTERNALLY to create an external user. Such a user must be authenticated by an
external service, such as an operating system or a third-party service. In this case,
Oracle Database relies on authentication by the operating system or third-party service
to ensure that a specific external user has access to a specific database user.
AS 'certificate_DN' This clause is required for and used for SSL-authenticated
external users only. The certificate_DN is the distinguished name in the user's PKI
certificate in the user's wallet. The maximum length of certificate_DN is 1024
characters.
AS 'kerberos_principal_name' This clause is required for and used for
Kerberos-authenticated external users only. The maximum length of kerberos_
principal_name is 1024 characters.

Oracle strongly recommends that you do not use
IDENTIFIED EXTERNALLY with operating systems that have inherently
weak login security.
Caution:

Restrictions on Creating External Users The following restrictions apply to creating
external users:
■

The user SYS cannot be an external user.

■

Oracle ASM does not support the creation of external users.

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-9

CREATE USER

See Also:
■

■

Oracle Database Enterprise User Security Administrator's Guide for
more information on externally identified users
"Creating External Database Users: Examples" on page 17-13

GLOBALLY Clause
The GLOBALLY clause lets you create a global user. Such a user must be authorized by
the enterprise directory service (Oracle Internet Directory).
The directory_DN string can take one of two forms:
■

■

The X.509 name at the enterprise directory service that identifies this user. It
should be of the form CN=username,other_attributes, where other_attributes
is the rest of the user's distinguished name (DN) in the directory. This form creates
a private global schema.
A null string (' ') indicating that the enterprise directory service will map
authenticated global users to this database schema with the appropriate roles. This
form is the same as specifying the GLOBALLY keyword alone and creates a shared
global schema.

The maximum length of directory_DN is 1024 characters.
You can control the ability of an application server to connect as the specified user and
to activate that user's roles using the ALTER USER statement.
Restriction on Creating Global Users Oracle ASM does not support the creation of
global users.
See Also:
■

Oracle Database Security Guide for more information on global
users

■

ALTER USER on page 13-6

■

"Creating a Global Database User: Example" on page 17-13

DEFAULT TABLESPACE Clause
Specify the default tablespace for objects that are created in the user's schema. If you
omit this clause, then the user's objects are stored in the database default tablespace. If
no default tablespace has been specified for the database, then the user's objects are
stored in the SYSTEM tablespace.
Restriction on Default Tablespaces You cannot specify a locally managed temporary

tablespace, including an undo tablespace, or a dictionary-managed temporary
tablespace, as a user's default tablespace.
See Also:
■

■

CREATE TABLESPACE on page 16-83 for more information on
tablespaces in general and undo tablespaces in particular
Oracle Database Security Guide for more information on assigning
default tablespaces to users

17-10 Oracle Database SQL Language Reference

CREATE USER

TEMPORARY TABLESPACE Clause
Specify the tablespace or tablespace group for the user's temporary segments. If you
omit this clause, then the user's temporary segments are stored in the database default
temporary tablespace or, if none has been specified, in the SYSTEM tablespace.
■
■

Specify tablespace to indicate the user's temporary tablespace.
Specify tablespace_group_name to indicate that the user can save temporary
segments in any tablespace in the tablespace group specified by tablespace_
group_name.

Restrictions on Temporary Tablespace

This clause is subject to the following

restrictions:
■

■

The tablespace must be a temporary tablespace and must have a standard block
size.
The tablespace cannot be an undo tablespace or a tablespace with automatic
segment-space management.
See Also:
■

■

■

Oracle Database Administrator's Guide for information about
tablespace groups and Oracle Database Security Guide for
information on assigning temporary tablespaces to users
CREATE TABLESPACE on page 16-83 for more information on
undo tablespaces and segment management
"Assigning a Tablespace Group: Example" on page 13-12

QUOTA Clause
Use the QUOTA clause to specify the maximum amount of space the user can allocate in
the tablespace.
A CREATE USER statement can have multiple QUOTA clauses for multiple tablespaces.
UNLIMITED lets the user allocate space in the tablespace without bound.
The maximum amount of space that you can specify is 2 terabytes (TB). If you need
more space, then specify UNLIMITED.
Restriction on the QUOTA Clause You cannot specify this clause for a temporary
tablespace.
See Also: size_clause on page 8-47 for information on that clause and
Oracle Database Security Guide for more information on assigning
tablespace quotas

PROFILE Clause
Specify the profile you want to assign to the user. The profile limits the amount of
database resources the user can use. If you omit this clause, then Oracle Database
assigns the DEFAULT profile to the user.

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-11

CREATE USER

Oracle recommends that you use the Database Resource
Manager rather SQL profiles to establish database resource limits.
The Database Resource Manager offers a more flexible means of
managing and tracking resource use. For more information on the
Database Resource Manager, refer to Oracle Database Administrator's
Guide.

Note:

See Also: GRANT on page 18-33 and CREATE PROFILE on
page 15-50

PASSWORD EXPIRE Clause
Specify PASSWORD EXPIRE if you want the user's password to expire. This setting forces
the user or the DBA to change the password before the user can log in to the database.

ACCOUNT Clause
Specify ACCOUNT LOCK to lock the user's account and disable access. Specify ACCOUNT
UNLOCK to unlock the user's account and enable access to the account.
ENABLE EDITIONS
This clause is not reversible. Specify ENABLE EDITIONS to allow the user to create
multiple versions of editionable objects in this schema using editions. Editionable
objects in schemas that are not editions-enabled cannot be editioned.
Restriction on Enabling Editions You cannot enable editions for any schemas

supplied by Oracle except for the sample schemas in the seed database.

Examples
17

All of the following examples use the example tablespace, which exists in the seed
database and is accessible to the sample schemas.
Creating a Database User: Example If you create a new user with PASSWORD EXPIRE,

then the user's password must be changed before the user attempts to log in to the
database. You can create the user sidney by issuing the following statement:
CREATE USER sidney
IDENTIFIED BY out_standing1
DEFAULT TABLESPACE example
QUOTA 10M ON example
TEMPORARY TABLESPACE temp
QUOTA 5M ON system
PROFILE app_user
PASSWORD EXPIRE;

The user sidney has the following characteristics:
■

The password out_standing1

■

Default tablespace example, with a quota of 10 megabytes

■

Temporary tablespace temp

■

Access to the tablespace SYSTEM, with a quota of 5 megabytes

■

Limits on database resources defined by the profile app_user (which was created
in "Creating a Profile: Example" on page 15-54)

17-12 Oracle Database SQL Language Reference

CREATE USER

■

An expired password, which must be changed before sidney can log in to the
database

The following example creates an
external user, who must be identified by an external source before accessing the
database:

Creating External Database Users: Examples

CREATE USER app_user1
IDENTIFIED EXTERNALLY
DEFAULT TABLESPACE example
QUOTA 5M ON example
PROFILE app_user;

The user app_user1 has the following additional characteristics:
■

Default tablespace example

■

Default temporary tablespace example

■

■

5M of space on the tablespace example and unlimited quota on the temporary
tablespace of the database
Limits on database resources defined by the app_user profile

To create another user accessible only by an operating system account, prefix the user
name with the value of the initialization parameter OS_AUTHENT_PREFIX. For example,
if this value is "ops$", then you can create the externally identified user external_user
with the following statement:
CREATE USER ops$external_user
IDENTIFIED EXTERNALLY
DEFAULT TABLESPACE example
QUOTA 5M ON example
PROFILE app_user;

The following example creates a global
user. When you create a global user, you can specify the X.509 name that identifies this
user at the enterprise directory server:

Creating a Global Database User: Example

CREATE USER global_user
IDENTIFIED GLOBALLY AS 'CN=analyst, OU=division1, O=oracle, C=US'
DEFAULT TABLESPACE example
QUOTA 5M ON example;

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-13

CREATE VIEW

CREATE VIEW
Purpose
17

Use the CREATE VIEW statement to define a view, which is a logical table based on one
or more tables or views. A view contains no data itself. The tables upon which a view
is based are called base tables.
You can also create an object view or a relational view that supports LOBs, object
types, REF data types, nested table, or varray types on top of the existing view
mechanism. An object view is a view of a user-defined type, where each row contains
objects, each object with a unique object identifier.
You can also create XMLType views, which are similar to object views but display data
from XMLSchema-based tables of XMLType.
See Also:
■

■

■

Oracle Database Concepts, Oracle Database Advanced Application
Developer's Guide, and Oracle Database Administrator's Guide for
information on various types of views and their uses
Oracle XML DB Developer's Guide for information on XMLType
views
ALTER VIEW on page 13-14 and DROP VIEW on page 18-18 for
information on modifying a view and removing a view from the
database

Prerequisites
17

To create a view in your own schema, you must have the CREATE VIEW system
privilege. To create a view in another user's schema, you must have the CREATE ANY
VIEW system privilege.
To create a subview, you must have the UNDER ANY VIEW system privilege or the UNDER
object privilege on the superview.
The owner of the schema containing the view must have the privileges necessary to
either select, insert, update, or delete rows from all the tables or views on which the
view is based. The owner must be granted these privileges directly, rather than
through a role.
To use the basic constructor method of an object type when creating an object view,
one of the following must be true:
■

The object type must belong to the same schema as the view to be created.

■

You must have the EXECUTE ANY TYPE system privileges.

■

You must have the EXECUTE object privilege on that object type.
See Also: SELECT on page 19-4, INSERT on page 18-54, UPDATE on
page 19-73, and DELETE on page 17-26 for information on the
privileges required by the owner of a view on the base tables or views
of the view being created

17-14 Oracle Database SQL Language Reference

CREATE VIEW

Syntax
17

create_view::=
NO
OR

REPLACE

FORCE

EDITIONING

CREATE

VIEW

,
inline_constraint
alias
(

)
out_of_line_constraint

object_view_clause
schema

.

XMLType_view_clause
view

subquery_restriction_clause
AS

subquery

;

(inline_constraint::= on page 8-5 and out_of_line_constraint::= on page 8-5, object_view_
clause::= on page 17-15, XMLType_view_clause::= on page 17-15, subquery::= on
page 19-5—part of SELECT, subquery_restriction_clause::= on page 17-16)
object_view_clause::=
DEFAULT
IDENTIFIER
WITH
schema

OBJECT

,
ID

.

OF

(

type_name
schema

attribute

)

.

UNDER

superview

,
out_of_line_constraint
(

)
attribute

inline_constraint

(inline_constraint::= on page 8-5 and out_of_line_constraint::= on page 8-5)
XMLType_view_clause::=
DEFAULT

XMLSchema_spec
OF

XMLTYPE

IDENTIFIER
WITH

OBJECT

,
ID
(

expr

)

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-15

CREATE VIEW

XMLSchema_spec::=
XMLSCHEMA

XMLSchema_URL

element
ELEMENT
XMLSchema_URL

ALLOW

#

element

ALLOW
NONSCHEMA

ANYSCHEMA

DISALLOW

DISALLOW

subquery_restriction_clause::=
READ

CONSTRAINT

ONLY

constraint

WITH
CHECK

OPTION

Semantics
17

OR REPLACE
Specify OR REPLACE to re-create the view if it already exists. You can use this clause to
change the definition of an existing view without dropping, re-creating, and
regranting object privileges previously granted on it.
INSTEAD OF triggers defined on a conventional view are dropped when the view is
re-created. DML triggers defined on an editioning views are retained when an
editioning view is re-created. However, such triggers can be rendered permanently
invalid if the editioning view has changed so that it can no longer be compiled—for
example if an editioning view column referenced in the trigger definition has been
dropped.
If any materialized views are dependent on view, then those materialized views will
be marked UNUSABLE and will require a full refresh to restore them to a usable state.
Invalid materialized views cannot be used by query rewrite and cannot be refreshed
until they are recompiled.
You cannot replace a conventional view with an editioning view or an editioning view
with a conventional view. See Oracle Database Advanced Application Developer's Guide
for more information on editioning views.
See Also:
■

■

■

ALTER MATERIALIZED VIEW on page 11-3 for information on
refreshing invalid materialized views
Oracle Database Concepts for information on materialized views in
general
CREATE TRIGGER on page 16-98 for more information about the
INSTEAD OF clause

FORCE
Specify FORCE if you want to create the view regardless of whether the base tables of
the view or the referenced object types exist or the owner of the schema containing the
view has privileges on them. These conditions must be true before any SELECT, INSERT,
UPDATE, or DELETE statements can be issued against the view.

17-16 Oracle Database SQL Language Reference

CREATE VIEW

If the view definition contains any constraints, CREATE VIEW ... FORCE fails if the base
table does not exist or the referenced object type does not exist. CREATE VIEW ... FORCE
also fails if the view definition names a constraint that does not exist.

NO FORCE
Specify NOFORCE if you want to create the view only if the base tables exist and the
owner of the schema containing the view has privileges on them. This is the default.

EDITIONING
Use this clause to create an editioning view. An editioning view is a single-table view
that selects all rows from the base table and displays a subset of the base table
columns. You can use an editioning view to isolate an application from DDL changes
to the base table during administrative operations such as upgrades. You can obtain
information about the relationship of existing editioning view to their base tables by
querying the USER_, ALL_, and DBA_EDITIONING_VIEW data dictionary views.
The owner of an editioning view must be editions-enabled. Refer to ENABLE
EDITIONS on page 13-10 for more information.
Notes on Editioning Views

Editioning views differ from conventional views in

several important ways:
■

■

■

Editioning views are intended only to select and provide aliases for a subset of
columns in a table. Therefore, the syntax for creating an editioning view is more
limited than the syntax for creating a conventional view. Any violation of the
restrictions that follow causes the creation of the view to fail, even if you specify
FORCE.
You can create DML triggers on editioning views. In this case, the database
considers the editioning view to be the base object of the trigger. Such triggers fire
when a DML operation target the editioning view itself. They do not fire if the
DML operation targets the base table.
You cannot create INSTEAD OF triggers on editioning views.

Restrictions on Editioning Views Editioning views are subject to the following
restrictions:
■

Within any edition, whether published or not, you can create only one editioning
view for any single table.

■

You cannot specify the object_view_clause or the XMLType_view_clause.

■

You cannot define a constraint WITH CHECK OPTION on an editioning view.

■

■

■

In the select list of the defining subquery, you can specify only simple references to
the columns of the base table, and you can specify each column of the base table
only once in the select list. The asterisk wildcard symbol * and t_alias.* are
supported to designate all columns of a base table.
The FROM clause of the defining subquery of the view can reference only a single
existing database table. Joins are not permitted. The base table must be in the same
schema as the view being created. You cannot use a synonym to identify the table,
but you can specify a table alias.
The following clauses of the defining subquery are not valid for editioning views:
subquery_factoring_clause, DISTINCT or UNIQUE, where_clause, hierarchical_
query_clause, group_by_clause, HAVING condition, model_clause, or the set
operators (UNION, INTERSECT, or MINUS)

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-17

CREATE VIEW

See Also:
■

■

Oracle Database Advanced Application Developer's Guide for detailed
information about editioning views
CREATE EDITION on page 14-51 for information about editions,
including an example of an editioning view

schema
Specify the schema to contain the view. If you omit schema, then Oracle Database
creates the view in your own schema.

view
Specify the name of the view or the object view. The name must satisfy the
requirements listed in "Database Object Naming Rules" on page 3-111.
If a view has INSTEAD OF triggers, then any views created on it
must have INSTEAD OF triggers, even if the views are inherently updatable.

Restriction on Views

See Also:

"Creating a View: Example" on page 17-22

alias
Specify names for the expressions selected by the defining query of the view. The
number of aliases must match the number of expressions selected by the view. Aliases
must follow the rules for naming Oracle Database schema objects. Aliases must be
unique within the view.
If you omit the aliases, then the database derives them from the columns or column
aliases in the query. For this reason, you must use aliases if the query contains
expressions rather than only column names. Also, you must specify aliases if the view
definition includes constraints.
Restriction on View Aliases

You cannot specify an alias when creating an object

view.
"Syntax for Schema Objects and Parts in SQL Statements"
on page 3-115

See Also:

inline_constraint and out_of_line_constraint
You can specify constraints on views and object views. You define the constraint at the
view level using the out_of_line_constraint clause. You define the constraint as part
of column or attribute specification using the inline_constraint clause after the
appropriate alias.
Oracle Database does not enforce view constraints. For a full discussion of view
constraints, including restrictions, refer to "View Constraints" on page 8-19.
See Also: "Creating a View with Constraints: Example" on
page 17-23

object_view_clause
The object_view_clause lets you define a view on an object type.
See Also:

"Creating an Object View: Example" on page 17-24

17-18 Oracle Database SQL Language Reference

CREATE VIEW

OF type_name Clause
Use this clause to explicitly create an object view of type type_name. The columns of
an object view correspond to the top-level attributes of type type_name. Each row will
contain an object instance and each instance will be associated with an object identifier
as specified in the WITH OBJECT IDENTIFIER clause. If you omit schema, then the
database creates the object view in your own schema.
Object tables, as well as XMLType tables, object views, and XMLType views, do not have
any column names specified for them. Therefore, Oracle Database defines a
system-generated pseudocolumn OBJECT_ID. You can use this column name in queries
and to create object views with the WITH OBJECT IDENTIFIER clause.
WITH OBJECT IDENTIFIER Clause
Use the WITH OBJECT IDENTIFIER clause to specify a top-level (root) object view. This
clause lets you specify the attributes of the object type that will be used as a key to
identify each row in the object view. In most cases these attributes correspond to the
primary key columns of the base table. You must ensure that the attribute list is unique
and identifies exactly one row in the view. The WITH OBJECT IDENTIFIER and WITH
OBJECT ID clauses can be used interchangeably and are provided for semantic clarity.
Restrictions on Object Views
■

■

Object views are subject to the following restrictions:

If you try to dereference or pin a primary key REF that resolves to more than one
instance in the object view, then the database returns an error.
You cannot specify this clause if you are creating a subview, because subviews
inherit object identifiers from superviews.
Note: The database8i, Release 8.0 syntax WITH OBJECT OID is replaced
with this syntax for clarity. The keywords WITH OBJECT OID are
supported for backward compatibility, but Oracle recommends that
you use the new syntax WITH OBJECT IDENTIFIER.

If the object view is defined on an object table or an object view, then you can omit this
clause or specify DEFAULT.
DEFAULT Specify DEFAULT if you want the database to use the intrinsic object

identifier of the underlying object table or object view to uniquely identify each row.
For attribute, specify an attribute of the object type from which the
database should create the object identifier for the object view.

attribute

UNDER Clause
Use the UNDER clause to specify a subview based on an object superview.
Restrictions on Subviews

Subviews are subject to the following restrictions:

■

You must create a subview in the same schema as the superview.

■

The object type type_name must be the immediate subtype of superview.

■

You can create only one subview of a particular type under the same superview.

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-19

CREATE VIEW

See Also:
■

CREATE TYPE on page 17-3 for information about creating objects

■

Oracle Database Reference for information on data dictionary views

AS subquery
Specify a subquery that identifies columns and rows of the table(s) that the view is
based on. The select list of the subquery can contain up to 1000 expressions.
If you create views that refer to remote tables and views, then the database links you
specify must have been created using the CONNECT TO clause of the CREATE DATABASE
LINK statement, and you must qualify them with a schema name in the view subquery.
If you create a view with the flashback_query_clause in the defining query, then the
database does not interpret the AS OF expression at create time but rather each time a
user subsequently queries the view.
"Creating a Join View: Example" on page 17-23 and Oracle
Database Advanced Application Developer's Guide for more information
on Oracle Flashback Query

See Also:

Restrictions on the Defining Query of a View The view query is subject to the
following restrictions:
■
■

■

■

■

The subquery cannot select the CURRVAL or NEXTVAL pseudocolumns.
If the subquery selects the ROWID, ROWNUM, or LEVEL pseudocolumns, then those
columns must have aliases in the view subquery.
If the subquery uses an asterisk (*) to select all columns of a table, and you later
add new columns to the table, then the view will not contain those columns until
you re-create the view by issuing a CREATE OR REPLACE VIEW statement.
For object views, the number of elements in the subquery select list must be the
same as the number of top-level attributes for the object type. The data type of
each of the selecting elements must be the same as the corresponding top-level
attribute.
You cannot specify the SAMPLE clause.

The preceding restrictions apply to materialized views as well.
Notes on Updatable Views

The following notes apply to updatable views:

An updatable view is one you can use to insert, update, or delete base table rows. You
can create a view to be inherently updatable, or you can create an INSTEAD OF trigger
on any view to make it updatable.
To learn whether and in what ways the columns of an inherently updatable view can
be modified, query the USER_UPDATABLE_COLUMNS data dictionary view. The
information displayed by this view is meaningful only for inherently updatable views.
For a view to be inherently updatable, the following conditions must be met:
■

■

Each column in the view must map to a column of a single table. For example, if a
view column maps to the output of a TABLE clause (an unnested collection), then
the view is not inherently updatable.
The view must not contain any of the following constructs:
A set operator
A DISTINCT operator

17-20 Oracle Database SQL Language Reference

CREATE VIEW

■

■

An aggregate or analytic function
A GROUP BY, ORDER BY, MODEL, CONNECT BY, or START WITH clause
A collection expression in a SELECT list
A subquery in a SELECT list
A subquery designated WITH READ ONLY
Joins, with some exceptions, as documented in Oracle Database Administrator's
Guide
In addition, if an inherently updatable view contains pseudocolumns or
expressions, then you cannot update base table rows with an UPDATE statement
that refers to any of these pseudocolumns or expressions.
If you want a join view to be updatable, then all of the following conditions must
be true:
–

The DML statement must affect only one table underlying the join.

–

For an INSERT statement, the view must not be created WITH CHECK OPTION, and
all columns into which values are inserted must come from a key-preserved
table. A key-preserved table is one for which every primary key or unique key
value in the base table is also unique in the join view.

–

For an UPDATE statement, the view must not be created WITH CHECK OPTION, and
all columns updated must be extracted from a key-preserved table.

–

For a DELETE statement, if the join results in more than one key-preserved
table, then Oracle Database deletes from the first table named in the FROM
clause, whether or not the view was created WITH CHECK OPTION.
See Also:
■

■

Oracle Database Administrator's Guide for more information on
updatable views
"Creating an Updatable View: Example" on page 17-23, "Creating
a Join View: Example" on page 17-23 for an example of updatable
join views and key-preserved tables, and Oracle Database PL/SQL
Language Reference for an example of an INSTEAD OF trigger on a
view that is not inherently updatable

XMLType_view_clause
Use this clause to create an XMLType view, which displays data from an
XMLSchema-based table of type XMLType. The XMLSchema_spec indicates the
XMLSchema to be used to map the XML data to its object-relational equivalents. The
XMLSchema must already have been created before you can create an XMLType view.
The WITH OBJECT IDENTIFIER and WITH OBJECT ID clauses can be used interchangeably
and are provided for semantic clarity.
Object tables, as well as XMLType tables, object views, and XMLType views, do not have
any column names specified for them. Therefore, Oracle Database defines a
system-generated pseudocolumn OBJECT_ID. You can use this column name in queries
and to create object views with the WITH OBJECT IDENTIFIER clause.
See Also:
■

■

Oracle XML DB Developer's Guide for information on XMLType
views and XMLSchemas
"Creating an XMLType View: Example" on page 17-25 and
"Creating a View on an XMLType Table: Example" on page 17-25

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-21

CREATE VIEW

subquery_restriction_clause
Use the subquery_restriction_clause to restrict the defining query of the view in
one of the following ways:
WITH READ ONLY

Specify WITH READ ONLY to indicate that the table or view cannot

be updated.
WITH CHECK OPTION Specify WITH CHECK OPTION to indicate that Oracle Database
prohibits any changes to the table or view that would produce rows that are not
included in the subquery. When used in the subquery of a DML statement, you can
specify this clause in a subquery in the FROM clause but not in subquery in the WHERE
clause.
CONSTRAINT constraint Specify the name of the READ ONLY or CHECK OPTION
constraint. If you omit this identifier, then Oracle automatically assigns the constraint a
name of the form SYS_Cn, where n is an integer that makes the constraint name unique
within the database.

For tables, WITH CHECK OPTION guarantees that inserts and
updates result in tables that the defining table subquery can select. For
views, WITH CHECK OPTION cannot make this guarantee if:

Note:

■

■

There is a subquery within the defining query of this view or any
view on which this view is based or
INSERT, UPDATE, or DELETE operations are performed using
INSTEAD OF triggers.

Restriction on the subquery_restriction_clause

You cannot specify this clause if you

have specify an ORDER BY clause.
See Also:

"Creating a Read-Only View: Example" on page 17-24

Examples
17

The following statement creates a view of the sample table
employees named emp_view. The view shows the employees in department 20 and
their annual salary:
Creating a View: Example

CREATE VIEW emp_view AS
SELECT last_name, salary*12 annual_salary
FROM employees
WHERE department_id = 20;

The view declaration need not define a name for the column based on the expression
salary*12, because the subquery uses a column alias (annual_salary) for this
expression.
Creating an Editioning View: Example The following statement creates an editioning
view of the orders table:
CREATE EDITIONING VIEW ed_orders_view (o_id, o_date, o_status)
AS SELECT order_id, order_date, order_status FROM orders
WITH READ ONLY;

17-22 Oracle Database SQL Language Reference

CREATE VIEW

You can use this view to isolate an application from DDL changes to the orders table
during an administrative operation such as an upgrade. You can create a DML trigger
on this view, so that the trigger fires when a DML operation targets the view itself, but
does not fire if the DML operation targets the orders table.
Creating a View with Constraints: Example The following statement creates a
restricted view of the sample table hr.employees and defines a unique constraint on
the email view column and a primary key constraint for the view on the emp_id view
column:
CREATE VIEW emp_sal (emp_id, last_name,
email UNIQUE RELY DISABLE NOVALIDATE,
CONSTRAINT id_pk PRIMARY KEY (emp_id) RELY DISABLE NOVALIDATE)
AS SELECT employee_id, last_name, email FROM employees;

The following statement creates an updatable
view named clerk of all clerks in the employees table. Only the employees' IDs, last
names, department numbers, and jobs are visible in this view, and these columns can
be updated only in rows where the employee is a kind of clerk:

Creating an Updatable View: Example

CREATE VIEW clerk AS
SELECT employee_id, last_name, department_id, job_id
FROM employees
WHERE job_id = 'PU_CLERK'
or job_id = 'SH_CLERK'
or job_id = 'ST_CLERK';

This view lets you change the job_id of a purchasing clerk to purchasing manager
(PU_MAN):
UPDATE clerk SET job_id = 'PU_MAN' WHERE employee_id = 118;

The next example creates the same view WITH CHECK OPTION. You cannot subsequently
insert a new row into clerk if the new employee is not a clerk. You can update an
employee's job_id from one type of clerk to another type of clerk, but the update in
the preceding statement would fail, because the view cannot access employees with
non-clerk job_id.
CREATE VIEW clerk AS
SELECT employee_id, last_name, department_id, job_id
FROM employees
WHERE job_id = 'PU_CLERK'
or job_id = 'SH_CLERK'
or job_id = 'ST_CLERK'
WITH CHECK OPTION;

Creating a Join View: Example A join view is one whose view subquery contains a
join. If at least one column in the join has a unique index, then it may be possible to
modify one base table in a join view. You can query USER_UPDATABLE_COLUMNS to see
whether the columns in a join view are updatable. For example:
CREATE VIEW locations_view AS
SELECT d.department_id, d.department_name, l.location_id, l.city
FROM departments d, locations l
WHERE d.location_id = l.location_id;
SELECT column_name, updatable
FROM user_updatable_columns
WHERE table_name = 'LOCATIONS_VIEW'
ORDER BY column_name, updatable;

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-23

CREATE VIEW

COLUMN_NAME
-----------------------------DEPARTMENT_ID
DEPARTMENT_NAME
LOCATION_ID
CITY

UPD
--YES
YES
NO
NO

In the preceding example, the primary key index on the location_id column of the
locations table is not unique in the locations_view view. Therefore, locations is not
a key-preserved table and columns from that base table are not updatable.
INSERT INTO locations_view VALUES
(999, 'Entertainment', 87, 'Roma');
INSERT INTO locations_view VALUES
*
ERROR at line 1:
ORA-01776: cannot modify more than one base table through a join view

You can insert, update, or delete a row from the departments base table, because all
the columns in the view mapping to the departments table are marked as updatable
and because the primary key of departments is retained in the view.
INSERT INTO locations_view (department_id, department_name)
VALUES (999, 'Entertainment');
1 row created.

For you to insert into the table using the view, the view must
contain all NOT NULL columns of all tables in the join, unless you have
specified DEFAULT values for the NOT NULL columns.
Note:

See Also: Oracle Database Administrator's Guide for more information
on updating join views

The following statement creates a read-only
view named customer_ro of the oe.customers table. Only the customers' last names,
language, and credit limit are visible in this view:
Creating a Read-Only View: Example

CREATE VIEW customer_ro (name, language, credit)
AS SELECT cust_last_name, nls_language, credit_limit
FROM customers
WITH READ ONLY;

Creating an Object View: Example The following example shows the creation of the
type inventory_typ in the oc schema, and the oc_inventories view that is based on
that type:
CREATE TYPE inventory_typ
OID '82A4AF6A4CD4656DE034080020E0EE3D'
AS OBJECT
( product_id
NUMBER(6)
, warehouse
warehouse_typ
, quantity_on_hand
NUMBER(8)
) ;
/
CREATE OR REPLACE VIEW oc_inventories OF inventory_typ
WITH OBJECT OID (product_id)
AS SELECT i.product_id,
17-24 Oracle Database SQL Language Reference

CREATE VIEW

warehouse_typ(w.warehouse_id, w.warehouse_name, w.location_id),
i.quantity_on_hand
FROM inventories i, warehouses w
WHERE i.warehouse_id=w.warehouse_id;

Creating a View on an XMLType Table: Example The following example builds a
regular view on the XMLType table xwarehouses, which was created in "Examples" on
page 16-70:
CREATE VIEW warehouse_view AS
SELECT VALUE(p) AS warehouse_xml
FROM xwarehouses p;

You select from such a view as follows:
SELECT e.warehouse_xml.getclobval()
FROM warehouse_view e
WHERE EXISTSNODE(warehouse_xml, '//Docks') =1;

In some cases you may have an
object-relational table upon which you would like to build an XMLType view. The
following example creates an object-relational table resembling the XMLType column
warehouse_spec in the sample table oe.warehouses, and then creates an XMLType view
of that table:
Creating an XMLType View: Example

CREATE TABLE warehouse_table
(
WarehouseID
NUMBER,
Area
NUMBER,
Docks
NUMBER,
DockType
VARCHAR2(100),
WaterAccess
VARCHAR2(10),
RailAccess
VARCHAR2(10),
Parking
VARCHAR2(20),
VClearance
NUMBER
);
INSERT INTO warehouse_table
VALUES(5, 103000,3,'Side Load','false','true','Lot',15);
CREATE VIEW warehouse_view OF XMLTYPE
XMLSCHEMA "http://www.example.com/xwarehouses.xsd"
ELEMENT "Warehouse"
WITH OBJECT ID
(extract(OBJECT_VALUE, '/Warehouse/Area/text()').getnumberval())
AS SELECT XMLELEMENT("Warehouse",
XMLFOREST(WarehouseID as "Building",
area as "Area",
docks as "Docks",
docktype as "DockType",
wateraccess as "WaterAccess",
railaccess as "RailAccess",
parking as "Parking",
VClearance as "VClearance"))
FROM warehouse_table;

You query this view as follows:
SELECT VALUE(e) FROM warehouse_view e;

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-25

DELETE

DELETE
Purpose
17

Use the DELETE statement to remove rows from:
■

An unpartitioned or partitioned table

■

The unpartitioned or partitioned base table of a view

■

The unpartitioned or partitioned container table of a writable materialized view

■

The unpartitioned or partitioned master table of an updatable materialized view

Prerequisites
17

For you to delete rows from a table, the table must be in your own schema or you must
have the DELETE object privilege on the table.
For you to delete rows from an updatable materialized view, the materialized view
must be in your own schema or you must have the DELETE object privilege on the
materialized view.
For you to delete rows from the base table of a view, the owner of the schema
containing the view must have the DELETE object privilege on the base table. Also, if
the view is in a schema other than your own, then you must have the DELETE object
privilege on the view.
The DELETE ANY TABLE system privilege also allows you to delete rows from any table
or table partition or from the base table of any view.
You must also have the SELECT object privilege on the object from which you want to
delete if:
■
■

The object is on a remote database or
The SQL92_SECURITY initialization parameter is set to TRUE and the DELETE
operation references table columns, such as the columns in a where_clause

You cannot delete rows from a table if a function-based index on the table has become
invalid. You must first validate the function-based index.

Syntax
17

delete::=
hint

FROM

t_alias

dml_table_expression_clause

DELETE
ONLY

where_clause

returning_clause

(

dml_table_expression_clause

)

error_logging_clause
;

(DML_table_expression_clause::= on page 17-27, where_clause::= on page 17-27, returning_
clause::= on page 17-27, error_logging_clause::= on page 17-28)

17-26 Oracle Database SQL Language Reference

DELETE

DML_table_expression_clause::=
partition_extension_clause
@
schema

dblink

table

.

@

view

dblink

materialized view
subquery_restriction_clause
(

subquery

)

table_collection_expression

(partition_extension_clause::= on page 17-27, subquery::= on page 19-5, subquery_
restriction_clause::= on page 17-27, table_collection_expression::= on page 17-27)
partition_extension_clause::=
(

partition

)

PARTITION

,
FOR

(

(

partition_key_value

subpartition

)

)

SUBPARTITION

,
FOR

(

subpartition_key_value

)

subquery_restriction_clause::=

READ

CONSTRAINT

ONLY

constraint

WITH
CHECK

OPTION

table_collection_expression::=
(
TABLE

(

collection_expression

+

)

)

where_clause::=
WHERE

condition

returning_clause::=
,

,

RETURN
expr

INTO

data_item

RETURNING

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-27

DELETE

error_logging_clause::=
schema

.

INTO
LOG

table

(

simple_expression

)

ERRORS

integer
REJECT

LIMIT
UNLIMITED

Semantics
17

hint
Specify a comment that passes instructions to the optimizer on choosing an execution
plan for the statement.
See Also: "Hints" on page 3-74 for the syntax and description of
hints

from_clause
Use the FROM clause to specify the database objects from which you are deleting rows.
The ONLY syntax is relevant only for views. Use the ONLY clause if the view in the FROM
clause belongs to a view hierarchy and you do not want to delete rows from any of its
subviews.

DML_table_expression_clause
Use this clause to specify the objects from which data is being deleted.
schema
Specify the schema containing the table or view. If you omit schema, then Oracle
Database assumes the table or view is in your own schema.
table | view | materialized view | subquery
Specify the name of a table, view, materialized view, or the column or columns
resulting from a subquery, from which the rows are to be deleted.
When you delete rows from an updatable view, Oracle Database deletes rows from the
base table.
You cannot delete rows from a read-only materialized view. If you delete rows from a
writable materialized view, then the database removes the rows from the underlying
container table. However, the deletions are overwritten at the next refresh operation. If
you delete rows from an updatable materialized view that is part of a materialized
view group, then the database also removes the corresponding rows from the master
table.
If table or the base table of view or the master table of materialized_view contains
one or more domain index columns, then this statement executes the appropriate
indextype delete routine.
See Also: Oracle Database Data Cartridge Developer's Guide for more
information on these routines

17-28 Oracle Database SQL Language Reference

DELETE

Issuing a DELETE statement against a table fires any DELETE triggers defined on the
table.
All table or index space released by the deleted rows is retained by the table and index.
partition_extension_clause
Specify the name or partition key value of the partition or subpartition targeted for
deletes within the object.
You need not specify the partition name when deleting values from a partitioned
object. However, in some cases, specifying the partition name is more efficient than a
complicated where_clause.
See Also: "References to Partitioned Tables and Indexes" on
page 3-119 and "Deleting Rows from a Partition: Example" on
page 17-32

dblink
Specify the complete or partial name of a database link to a remote database where the
object is located. You can delete rows from a remote object only if you are using Oracle
Database distributed functionality.
See Also: "References to Objects in Remote Databases" on page 3-117
for information on referring to database links and "Deleting Rows
from a Remote Database: Example" on page 17-32

If you omit dblink, then the database assumes that the object is located on the local
database.
subquery_restriction_clause
The subquery_restriction_clause lets you restrict the subquery in one of the
following ways:
WITH READ ONLY

Specify WITH READ ONLY to indicate that the table or view cannot

be updated.
WITH CHECK OPTION Specify WITH CHECK OPTION to indicate that Oracle Database
prohibits any changes to the table or view that would produce rows that are not
included in the subquery. When used in the subquery of a DML statement, you can
specify this clause in a subquery in the FROM clause but not in subquery in the WHERE
clause.
CONSTRAINT constraint Specify the name of the CHECK OPTION constraint. If you
omit this identifier, then Oracle automatically assigns the constraint a name of the
form SYS_Cn, where n is an integer that makes the constraint name unique within the
database.
See Also: "Using the WITH CHECK OPTION Clause: Example" on
page 19-48

table_collection_expression
The table_collection_expression lets you inform Oracle that the value of
collection_expression should be treated as a table for purposes of query and DML
operations. The collection_expression can be a subquery, a column, a function, or a
collection constructor. Regardless of its form, it must return a collection value—that is,

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-29

DELETE

a value whose type is nested table or varray. This process of extracting the elements of
a collection is called collection unnesting.
The optional plus (+) is relevant if you are joining the TABLE collection expression with
the parent table. The + creates an outer join of the two, so that the query returns rows
from the outer table even if the collection expression is null.
In earlier releases of Oracle, when collection_expression
was a subquery, table_collection_expression was expressed as
THE subquery. That usage is now deprecated.
Note:

You can use a table_collection_expression in a correlated subquery to delete rows
with values that also exist in another table.
See Also:

"Table Collections: Examples" on page 19-54

collection_expression Specify a subquery that selects a nested table column from the
object from which you are deleting.
Restrictions on the dml_table_expression_clause Clause This clause is subject to
the following restrictions:
■

■

■

■

You cannot execute this statement if table or the base or master table of view or
materialized_view contains any domain indexes marked IN_PROGRESS or FAILED.
You cannot insert into a partition if any affected index partitions are marked
UNUSABLE.
You cannot specify the ORDER BY clause in the subquery of the DML_table_
expression_clause.
You cannot delete from a view except through INSTEAD OF triggers if the defining
query of the view contains one of the following constructs:

A set operator
A DISTINCT operator
An aggregate or analytic function
A GROUP BY, ORDER BY, MODEL, CONNECT BY, or START WITH clause
A collection expression in a SELECT list
A subquery in a SELECT list
A subquery designated WITH READ ONLY
Joins, with some exceptions, as documented in Oracle Database Administrator's
Guide
If you specify an index, index partition, or index subpartition that has been marked
UNUSABLE, then the DELETE statement will fail unless the SKIP_UNUSABLE_INDEXES
initialization parameter has been set to true.
See Also:

ALTER SESSION on page 11-45

where_clause
Use the where_clause to delete only rows that satisfy the condition. The condition can
reference the object from which you are deleting and can contain a subquery. You can
delete rows from a remote object only if you are using Oracle Database distributed
functionality. Refer to Chapter 7, "Conditions" for the syntax of condition.
If this clause contains a subquery that refers to remote objects, then the DELETE
operation can run in parallel as long as the reference does not loop back to an object on

17-30 Oracle Database SQL Language Reference

DELETE

the local database. However, if the subquery in the DML_table_expression_clause
refers to any remote objects, then the DELETE operation will run serially without
notification. Refer to the parallel_clause on page 16-63 in the CREATE TABLE
documentation for additional information.
If you omit dblink, then the database assumes that the table or view is located on the
local database.
If you omit the where_clause, then the database deletes all rows of the object.
t_alias Provide a correlation name for the table, view, materialized view, subquery,
or collection value to be referenced elsewhere in the statement. This alias is required if
the DML_table_expression_clause references any object type attributes or object type
methods. Table aliases are generally used in DELETE statements with correlated queries.

returning_clause
This clause lets you return values from deleted columns, and thereby eliminate the
need to issue a SELECT statement following the DELETE statement.
The returning clause retrieves the rows affected by a DML statement. You can specify
this clause for tables and materialized views and for views with a single base table.
When operating on a single row, a DML statement with a returning_clause can
retrieve column expressions using the affected row, rowid, and REFs to the affected
row and store them in host variables or PL/SQL variables.
When operating on multiple rows, a DML statement with the returning_clause stores
values from expressions, rowids, and REFs involving the affected rows in bind arrays.
expr

Each item in the expr list must be a valid expression syntax.

The INTO clause indicates that the values of the changed rows are to be stored
in the variable(s) specified in data_item list.

INTO

Each data_item is a host variable or PL/SQL variable that stores the
retrieved expr value.

data_item

For each expression in the RETURNING list, you must specify a corresponding
type-compatible PL/SQL variable or host variable in the INTO list.
Restrictions The following restrictions apply to the RETURNING clause:
■

■

The expr is restricted as follows:
–

For UPDATE and DELETE statements each expr must be a simple expression or a
single-set aggregate function expression. You cannot combine simple
expressions and single-set aggregate function expressions in the same
returning_clause. For INSERT statements, each expr must be a simple
expression. Aggregate functions are not supported in an INSERT statement
RETURNING clause.

–

Single-set aggregate function expressions cannot include the DISTINCT
keyword.

If the expr list contains a primary key column or other NOT NULL column, then the
update statement fails if the table has a BEFORE UPDATE trigger defined on it.

■

You cannot specify the returning_clause for a multitable insert.

■

You cannot use this clause with parallel DML or with remote objects.

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-31

DELETE

■
■

You cannot retrieve LONG types with this clause.
You cannot specify this clause for a view on which an INSTEAD OF trigger has been
defined.
See Also:
■

■

Oracle Database PL/SQL Language Reference for information on
using the BULK COLLECT clause to return multiple values to
collection variables
"Using the RETURNING Clause: Example" on page 17-32

error_logging_clause
The error_logging_clause has the same behavior in DELETE statement as it does in an
INSERT statement. Refer to the INSERT statement error_logging_clause on page 18-65 for
more information.
See Also: "Inserting Into a Table with Error Logging: Example" on
page 18-66

Examples
17

The following statement deletes all rows from the sample
table oe.product_descriptions where the value of the language_id column is AR:

Deleting Rows: Examples

DELETE FROM product_descriptions
WHERE language_id = 'AR';

The following statement deletes from the sample table hr.employees purchasing
clerks whose commission rate is less than 10%:
DELETE FROM employees
WHERE job_id = 'SA_REP'
AND commission_pct < .2;

The following statement has the same effect as the preceding example, but uses a
subquery:
DELETE FROM (SELECT * FROM employees)
WHERE job_id = 'SA_REP'
AND commission_pct < .2;

Deleting Rows from a Remote Database: Example The following statement deletes
specified rows from the locations table owned by the user hr on a database accessible
by the database link remote:
DELETE FROM hr.locations@remote
WHERE location_id > 3000;

Deleting Nested Table Rows: Example For an example that deletes nested table
rows, refer to "Table Collections: Examples" on page 19-54.

The following example removes rows from
partition sales_q1_1998 of the sh.sales table:

Deleting Rows from a Partition: Example

DELETE FROM sales PARTITION (sales_q1_1998)
WHERE amount_sold > 1000;

Using the RETURNING Clause: Example

17-32 Oracle Database SQL Language Reference

DELETE

The following example returns column salary from the deleted rows and stores the
result in bind variable :bnd1. The bind variable must already have been declared.
DELETE FROM employees
WHERE job_id = 'SA_REP'
AND hire_date + TO_YMINTERVAL('01-00') < SYSDATE
RETURNING salary INTO :bnd1;

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-33

DISASSOCIATE STATISTICS

DISASSOCIATE STATISTICS
Purpose
17

Use the DISASSOCIATE STATISTICS statement to disassociate default statistics or a
statistics type from columns, standalone functions, packages, types, domain indexes,
or indextypes.
See Also: ASSOCIATE STATISTICS on page 13-25 for more
information on statistics type associations

Prerequisites
17

To issue this statement, you must have the appropriate privileges to alter the
underlying table, function, package, type, domain index, or indextype.

Syntax
17

disassociate_statistics::=
DISASSOCIATE

STATISTICS

FROM

,
schema

.

COLUMNS

table

.

column

,
schema

.

FUNCTIONS

function
,
schema

.

PACKAGES

package
FORCE

,
schema

;
.

TYPES

type
,
schema

.

INDEXES

index
,
schema

INDEXTYPES

.
indextype

17-34 Oracle Database SQL Language Reference

DISASSOCIATE STATISTICS

Semantics
17

FROM COLUMNS | FUNCTIONS | PACKAGES | TYPES | INDEXES |
INDEXTYPES
Specify one or more columns, standalone functions, packages, types, domain indexes,
or indextypes from which you are disassociating statistics.
If you do not specify schema, then Oracle Database assumes the object is in your own
schema.
If you have collected user-defined statistics on the object, then the statement fails
unless you specify FORCE.

FORCE
Specify FORCE to remove the association regardless of whether any statistics exist for
the object using the statistics type. If statistics do exist, then the statistics are deleted
before the association is deleted.
When you drop an object with which a statistics type has been
associated, Oracle Database automatically disassociates the statistics
type with the FORCE option and drops all statistics that have been
collected with the statistics type.

Note:

Examples
17

This statement disassociates statistics from the
emp_mgmt package. See Oracle Database PL/SQL Language Reference for the example that
creates this package in the hr schema.
Disassociating Statistics: Example

DISASSOCIATE STATISTICS FROM PACKAGES hr.emp_mgmt;

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-35

DROP CLUSTER

DROP CLUSTER
Purpose
17

Use the DROP CLUSTER clause to remove a cluster from the database.
Caution: When you drop a cluster, any tables in the recycle bin that
were once part of that cluster are purged from the recycle bin and can
no longer be recovered with a FLASHBACK TABLE operation.

You cannot uncluster an individual table. Instead you must perform these steps:
1.

Create a new table with the same structure and contents as the old one, but with
no CLUSTER clause.

2.

Drop the old table.

3.

Use the RENAME statement to give the new table the name of the old one.

4.

Grant privileges on the new unclustered table. Grants on the old clustered table do
not apply.
CREATE TABLE on page 16-6, DROP TABLE on
page 18-5, RENAME on page 18-85, GRANT on page 18-33 for
information on these steps

See Also:

Prerequisites
17

The cluster must be in your own schema or you must have the DROP ANY CLUSTER
system privilege.

Syntax
17

drop_cluster::=
CASCADE
schema
DROP

CLUSTER

.

INCLUDING

CONSTRAINTS

TABLES

cluster

Semantics
17

schema
Specify the schema containing the cluster. If you omit schema, then the database
assumes the cluster is in your own schema.

cluster
Specify the name of the cluster to be dropped. Dropping a cluster also drops the
cluster index and returns all cluster space, including data blocks for the index, to the
appropriate tablespace(s).

INCLUDING TABLES
Specify INCLUDING TABLES to drop all tables that belong to the cluster.

17-36 Oracle Database SQL Language Reference

;

DROP CLUSTER

CASCADE CONSTRAINTS
Specify CASCADE CONSTRAINTS to drop all referential integrity constraints from tables
outside the cluster that refer to primary and unique keys in tables of the cluster. If you
omit this clause and such referential integrity constraints exist, then the database
returns an error and does not drop the cluster.

Examples
17

Dropping a Cluster: Examples The following examples drop the clusters created in
the "Examples" section of CREATE CLUSTER on page 14-7.

The following statements drops the language cluster:
DROP CLUSTER language;

The following statement drops the personnel cluster as well as tables dept_10 and
dept_20 and any referential integrity constraints that refer to primary or unique keys
in those tables:
DROP CLUSTER personnel
INCLUDING TABLES
CASCADE CONSTRAINTS;

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-37

DROP CONTEXT

DROP CONTEXT
Purpose
17

Use the DROP CONTEXT statement to remove a context namespace from the database.
Removing a context namespace does not invalidate any context under that namespace
that has been set for a user session. However, the context will be invalid when the user
next attempts to set that context.
See Also: CREATE CONTEXT on page 14-9 and Oracle Database
Security Guide for more information on contexts

Prerequisites
17

You must have the DROP ANY CONTEXT system privilege.

Syntax
17

drop_context::=
DROP

CONTEXT

namespace

;

Semantics
17

namespace
Specify the name of the context namespace to drop. You cannot drop the built-in
namespace USERENV.
SYS_CONTEXT on page 5-279 for information on the
USERENV namespace
See Also:

Examples
17

Dropping an Application Context: Example The following statement drops the
context created in CREATE CONTEXT on page 14-9:
DROP CONTEXT hr_context;

17-38 Oracle Database SQL Language Reference

DROP DATABASE

DROP DATABASE
Purpose
17

Caution:

You cannot roll back a DROP DATABASE statement.

Use the DROP DATABASE statement to drop the database. This statement is useful when
you want to drop a test database or drop an old database after successful migration to
a new host.
See Also: Oracle Database Backup and Recovery User's Guide for more
information on dropping the database

Prerequisites
17

You must have the SYSDBA system privilege to issue this statement. The database must
be mounted in exclusive and restricted mode, and it must be closed.

Syntax
17

drop_database::=
DROP

DATABASE

;

Semantics
17

When you issue this statement, Oracle Database drops the database and deletes all
control files and data files listed in the control file. If the database used a server
parameter file (spfile), then the spfile is also deleted.
Archived logs and backups are not removed, but you can use Recovery Manager
(RMAN) to remove them. If the database is on raw disks, then this statement does not
delete the actual raw disk special files.

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-39

DROP DATABASE LINK

DROP DATABASE LINK
Purpose
17

Use the DROP DATABASE LINK statement to remove a database link from the database.
See Also: CREATE DATABASE LINK on page 14-31 for information
on creating database links

Prerequisites
17

A private database link must be in your own schema. To drop a PUBLIC database link,
you must have the DROP PUBLIC DATABASE LINK system privilege.

Syntax
17

drop_database_link::=
PUBLIC
DROP

DATABASE

LINK

dblink

;

Semantics
17

PUBLIC
You must specify PUBLIC to drop a PUBLIC database link.

dblink
Specify the name of the database link to be dropped.
Restriction on Dropping Database Links You cannot drop a database link in another
user's schema, and you cannot qualify dblink with the name of a schema, because
periods are permitted in names of database links. Therefore, Oracle Database
interprets the entire name, such as ralph.linktosales, as the name of a database link
in your schema rather than as a database link named linktosales in the schema
ralph.

Examples
17

Dropping a Database Link: Example The following statement drops the public
database link named remote, which was created in "Defining a Public Database Link:
Example" on page 14-34:
DROP PUBLIC DATABASE LINK remote;

17-40 Oracle Database SQL Language Reference

DROP DIMENSION

DROP DIMENSION
Purpose
17

Use the DROP DIMENSION statement to remove the named dimension.
This statement does not invalidate materialized views that use relationships specified
in dimensions. However, requests that have been rewritten by query rewrite may be
invalidated, and subsequent operations on such views may execute more slowly.
See Also:
■

■

CREATE DIMENSION on page 14-36 and ALTER DIMENSION
on page 10-48 for information on creating and modifying a
dimension
Oracle Database Concepts for general information about dimensions

Prerequisites
17

The dimension must be in your own schema or you must have the DROP ANY DIMENSION
system privilege to use this statement.

Syntax
17

drop_dimension::=
schema
DROP

DIMENSION

.
dimension

;

Semantics
17

schema
Specify the name of the schema in which the dimension is located. If you omit schema,
then Oracle Database assumes the dimension is in your own schema.

dimension
Specify the name of the dimension you want to drop. The dimension must already
exist.

Examples
17

Dropping a Dimension: Example This example drops the sh.customers_dim
dimension:
DROP DIMENSION customers_dim;

See Also: "Creating a Dimension: Examples" on page 14-39 and
"Modifying a Dimension: Examples" on page 10-50 for examples of
creating and modifying this dimension

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-41

DROP DIRECTORY

DROP DIRECTORY
Purpose
17

Use the DROP DIRECTORY statement to remove a directory object from the database.
See Also: CREATE DIRECTORY on page 14-41 for information on
creating a directory

Prerequisites
17

To drop a directory, you must have the DROP ANY DIRECTORY system privilege.
Caution: Do not drop a directory when files in the associated file
system are being accessed by PL/SQL or OCI programs.

Syntax
17

drop_directory::=
DROP

DIRECTORY

directory_name

;

Semantics
17

directory_name
Specify the name of the directory database object to be dropped.
Oracle Database removes the directory object but does not delete the associated
operating system directory on the server file system.

Examples
17

Dropping a Directory: Example The following statement drops the directory object

bfile_dir:
DROP DIRECTORY bfile_dir;

See Also:

"Creating a Directory: Examples" on page 14-42

17-42 Oracle Database SQL Language Reference

DROP DISKGROUP

DROP DISKGROUP
This SQL statement is valid only if you are using Oracle ASM
and you have started an Oracle ASM instance. You must issue this
statement from within the Oracle ASM instance, not from a normal
database instance. For information on starting an Oracle ASM
instance, refer to Oracle Automatic Storage Management Administrator's
Guide.

Note:

Purpose
17

The DROP DISKGROUP statement lets you drop an Oracle ASM disk group along with all
the files in the disk group. Oracle ASM first ensures that no files in the disk group are
open. It then drops the disk group and all its member disks and clears the disk header.
See Also:
■

■

CREATE DISKGROUP on page 14-43 and ALTER DISKGROUP
on page 10-51 for information on creating and modifying disk
groups
Oracle Automatic Storage Management Administrator's Guide for
information on Oracle ASM and using disks groups to simplify
database administration

Prerequisites
17

You must have the SYSASM system privilege and you must have an Oracle ASM
instance started, from which you issue this statement. The disk group to be dropped
must be mounted.

Syntax
17

drop_diskgroup::=
FORCE
INCLUDING
CONTENTS
EXCLUDING
DROP

DISKGROUP

diskgroup_name

;

Semantics
17

diskgroup_name
Specify the name of the disk group you want to drop.

INCLUDING CONTENTS
Specify INCLUDING CONTENTS to confirm that Oracle ASM should drop all the files in
the disk group. You must specify this clause if the disk group contains any files.

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-43

DROP DISKGROUP

EXCLUDING CONTENTS
Specify EXCLUDING CONTENTS to ensure that Oracle ASM drops the disk group only
when the disk group is empty. This is the default. If the disk group is not empty, then
an error will be returned.

FORCE
This clause clears the headers on the disk belonging to a disk group that cannot be
mounted by the Oracle ASM instance. The disk group cannot be mounted by any
instance of the database.
The Oracle ASM instance first determines whether the disk group is being used by any
other Oracle ASM instance using the same storage subsystem. If it is being used, and if
the disk group is in the same cluster, or on the same node, then the statement fails. If
the disk group is in a different cluster, then the system further checks to determine
whether the disk group is mounted by any instance in the other cluster. If it is
mounted elsewhere, then the statement fails. However, this latter check is not as
definitive as the checks for disk groups in the same cluster. Therefore, use this clause
with caution.

Examples
17

Dropping a Diskgroup: Example The following statement drops the Oracle ASM disk
group dgroup_01, which was created in "Creating a Diskgroup: Example" on
page 14-50, and all of the files in the disk group:
DROP DISKGROUP dgroup_01 INCLUDING CONTENTS;

17-44 Oracle Database SQL Language Reference

DROP EDITION

DROP EDITION
Purpose
17

Use the DROP EDITION statement to drop an edition, along with all actual editionable
objects it contains. An actual editionable object is an editionable object that has been
created or modified in an edition.
See Also: CREATE EDITION on page 14-51 for a listing of
editionable object types

Prerequisites
17

You must have the DROP ANY EDITION system privilege, granted either directly or
through a role.

Syntax
17

drop_edition::=
CASCADE
DROP

EDITION

edition

;

Semantics
17

When successful, this statement drops the specified edition, including versions of any
objects associated with that edition. Versions of the same objects in other editions are
not dropped. Objects that are not editionable, or that are editionable but have not been
actualized in the current edition, are not dropped.
You must specify CASCADE if the specified edition contains any actual editionable
objects.
This statement is subject to the following conditions and restrictions:
■
■

■
■

The specified edition cannot have both a parent edition and a child edition.
The specified edition cannot contain any actual editionable objects that are
inherited by a child edition, even if you specify CASCADE.
DROP EDITION will fail if you attempt to drop the default edition.
DROP EDITION will fail if you attempt to drop the root edition and the recycle bin
contains at least one object that used to be in that edition before it was dropped.
Under these circumstances, even DROP EDITION CASCADE will fail. In this case, you
can purge all objects from the recycle bin with the PURGE DBA_RECYCLEBIN
statement and then drop the edition. Refer to PURGE on page 18-83 for more
information.
DROP EDITION will also fail if you attempt to drop the leaf edition and the recycle
bin contains at least one object that used to be in that edition before it was
dropped. However, under these circumstances, DROP EDITION CASCADE will
succeed.
The only type of editioned object that might be in the recycle bin is a trigger.

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-45

DROP EDITION

Examples
17

For examples that use this statement, refer to CREATE EDITION on page 14-51.

17-46 Oracle Database SQL Language Reference

DROP FLASHBACK ARCHIVE

DROP FLASHBACK ARCHIVE
Purpose
17

Use the DROP FLASHBACK ARCHIVE clause to remove a flashback data archive from the
system. This statement removes the flashback data archive and all the historical data in
it, but does not drop the tablespaces that were used by the flashback data archive.

Prerequisites
17

You must have the FLASHBACK ARCHIVE ADMINISTER system privilege to drop a
flashback data archive.

Syntax
17

drop_flashback_archive::=
DROP

FLASHBACK

ARCHIVE

flashback_archive

;

Semantics
17

flashback_archive
Specify the name of the flashback data archive you want to drop.
See Also: CREATE FLASHBACK ARCHIVE on page 14-55 for
information on creating flashback data archives and for some simple
examples of using flashback data archives

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-47

DROP FUNCTION

DROP FUNCTION
Purpose
17

Functions are defined using PL/SQL. Refer to Oracle Database PL/SQL Language
Reference for complete information on creating, altering, and dropping functions.
Use the DROP FUNCTION statement to remove a standalone stored function from the
database.
Do not use this statement to remove a function that is part of a
package. Instead, either drop the entire package using the DROP
PACKAGE statement or redefine the package without the function using
the CREATE PACKAGE statement with the OR REPLACE clause.
Note:

Prerequisites
17

The function must be in your own schema or you must have the DROP ANY PROCEDURE
system privilege.

Syntax
17

drop_function::=
schema
DROP

.

FUNCTION

function_name

;

Semantics
17

schema
Specify the schema containing the function. If you omit schema, then Oracle Database
assumes the function is in your own schema.

function_name
Specify the name of the function to be dropped.
Oracle Database invalidates any local objects that depend on, or call, the dropped
function. If you subsequently reference one of these objects, then the database tries to
recompile the object and returns an error if you have not re-created the dropped
function.
If any statistics types are associated with the function, then the database disassociates
the statistics types with the FORCE option and drops any user-defined statistics
collected with the statistics type.
See Also:
■

■

Oracle Database Concepts for more information on how Oracle
Database maintains dependencies among schema objects,
including remote objects
ASSOCIATE STATISTICS on page 13-25 and DISASSOCIATE
STATISTICS on page 17-34 for more information on statistics type
associations

17-48 Oracle Database SQL Language Reference

DROP FUNCTION

Examples
17

The following statement drops the function
SecondMax in the sample schema oe and invalidates all objects that depend upon
SecondMax:
Dropping a Function: Example

DROP FUNCTION oe.SecondMax;

Oracle Database PL/SQL Language Reference for the example
that creates the SecondMax function

See Also:

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-49

DROP INDEX

DROP INDEX
Purpose
17

Use the DROP INDEX statement to remove an index or domain index from the database.
When you drop a global partitioned index, a range-partitioned index, or a
hash-partitioned index, all the index partitions are also dropped. If you drop a
composite-partitioned index, then all the index partitions and subpartitions are also
dropped.
In addition, when you drop a domain index:
■
■

Oracle Database invokes the appropriate routine.
If any statistics are associated with the domain index, then Oracle Database
disassociates the statistics types with the FORCE clause and removes the
user-defined statistics collected with the statistics type.
See Also:

Oracle Database Data Cartridge Developer's Guide for information on
the routines

■

CREATE INDEX on page 14-60 and ALTER INDEX on page 10-78
for information on creating and modifying an index

■

The domain_index_clause of CREATE INDEX on page 14-60 for
more information on domain indexes

■

ASSOCIATE STATISTICS on page 13-25 and DISASSOCIATE
STATISTICS on page 17-34 for more information on statistics type
associations

■

Prerequisites
17

The index must be in your own schema or you must have the DROP ANY INDEX system
privilege.

Syntax
17

drop_index::=
schema
DROP

INDEX

.

FORCE
index

;

Semantics
17

schema
Specify the schema containing the index. If you omit schema, then Oracle Database
assumes the index is in your own schema.

index
Specify the name of the index to be dropped. When the index is dropped, all data
blocks allocated to the index are returned to the tablespace that contained the index.

17-50 Oracle Database SQL Language Reference

DROP INDEX

You cannot drop a domain index if the index or any
of its index partitions is marked IN_PROGRESS.

Restriction on Dropping Indexes

FORCE
FORCE applies only to domain indexes. This clause drops the domain index even if the
indextype routine invocation returns an error or the index is marked IN PROGRESS.
Without FORCE, you cannot drop a domain index if its indextype routine invocation
returns an error or the index is marked IN PROGRESS.

Examples
17

This statement drops an index named ord_customer_
ix_demo, which was created in "Compressing an Index: Example" on page 14-80:

Dropping an Index: Example

DROP INDEX ord_customer_ix_demo;

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-51

DROP INDEXTYPE

DROP INDEXTYPE
Purpose
17

Use the DROP INDEXTYPE statement to drop an indextype as well as any association with
a statistics type.
See Also: CREATE INDEXTYPE on page 14-87 for more information
on indextypes

Prerequisites
17

The indextype must be in your own schema or you must have the DROP ANY INDEXTYPE
system privilege.

Syntax
17

drop_indextype::=
schema
DROP

INDEXTYPE

.

FORCE
indextype

;

Semantics
17

schema
Specify the schema containing the indextype. If you omit schema, then Oracle Database
assumes the indextype is in your own schema.

indextype
Specify the name of the indextype to be dropped.
If any statistics types have been associated with indextype, then the database
disassociates the statistics type from the indextype and drops any statistics that have
been collected using the statistics type.
See Also: ASSOCIATE STATISTICS on page 13-25 and
DISASSOCIATE STATISTICS on page 17-34 for more information on
statistics associations

FORCE
Specify FORCE to drop the indextype even if the indextype is currently being referenced
by one or more domain indexes. Oracle Database marks those domain indexes
INVALID. Without FORCE, you cannot drop an indextype if any domain indexes
reference the indextype.

Examples
17

Dropping an Indextype: Example The following statement drops the indextype

position_indextype, created in "Using Extensible Indexing" on page F-1, and marks
INVALID any domain indexes defined on this indextype:
DROP INDEXTYPE position_indextype FORCE;

17-52 Oracle Database SQL Language Reference

DROP JAVA

DROP JAVA
Purpose
17

Use the DROP JAVA statement to drop a Java source, class, or resource schema object.
See Also:
■

■

CREATE JAVA on page 14-91 for information on creating Java
objects
Oracle Database Java Developer's Guide for more information on
resolving Java sources, classes, and resources

Prerequisites
17

The Java source, class, or resource must be in your own schema or you must have the
DROP ANY PROCEDURE system privilege. You also must have the EXECUTE object privilege
on Java classes to use this command.

Syntax
17

drop_java::=
SOURCE
DR0P

JAVA

schema

CLASS

.
object_name

;

RESOURCE

Semantics
17

JAVA SOURCE
Specify SOURCE to drop a Java source schema object and all Java class schema objects
derived from it.

JAVA CLASS
Specify CLASS to drop a Java class schema object.

JAVA RESOURCE
Specify RESOURCE to drop a Java resource schema object.

object_name
Specify the name of an existing Java class, source, or resource schema object. Enclose
the object_name in double quotation marks to preserve lower- or mixed-case names.

Examples
17

The following statement drops the Java
class Agent, created in "Creating a Java Class Object: Example" on page 14-95:

Dropping a Java Class Object: Example
DROP JAVA CLASS "Agent";

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-53

DROP LIBRARY

DROP LIBRARY
Purpose
17

Use the DROP LIBRARY statement to remove an external procedure library from the
database.
See Also: CREATE LIBRARY on page 15-2 for information on
creating a library

Prerequisites
17

You must have the DROP ANY LIBRARY system privilege.

Syntax
17

drop_library::=
DROP

LIBRARY

library_name

;

Semantics
17

library_name
Specify the name of the external procedure library being dropped.

Examples
17

Dropping a Library: Example
DROP LIBRARY ext_lib;

17-54 Oracle Database SQL Language Reference

The following statement drops the ext_lib library:

DROP MATERIALIZED VIEW

DROP MATERIALIZED VIEW
Purpose
17

Use the DROP MATERIALIZED VIEW statement to remove an existing materialized view
from the database.
When you drop a materialized view, Oracle Database does not place it in the recycle
bin. Therefore, you cannot subsequently either purge or undrop the materialized view.
The keyword SNAPSHOT is supported in place of MATERIALIZED
VIEW for backward compatibility.
Note:

See Also:
■

■

■

■

CREATE MATERIALIZED VIEW on page 15-4 for more
information on the various types of materialized views
ALTER MATERIALIZED VIEW on page 11-3 for information on
modifying a materialized view
Oracle Database Advanced Replication for information on
materialized views in a replication environment
Oracle Database Data Warehousing Guide for information on
materialized views in a data warehousing environment

Prerequisites
17

The materialized view must be in your own schema or you must have the DROP ANY
MATERIALIZED VIEW system privilege. You must also have the privileges to drop the
internal table, views, and index that the database uses to maintain the materialized
view data.
See Also: DROP TABLE on page 18-5, DROP VIEW on page 18-18,
and DROP INDEX on page 17-50 for information on privileges
required to drop objects that the database uses to maintain the
materialized view

Syntax
17

drop_materialized_view::=
schema
DROP

MATERIALIZED

VIEW

.

PRESERVE

TABLE

materialized_view

;

Semantics
17

schema
Specify the schema containing the materialized view. If you omit schema, then Oracle
Database assumes the materialized view is in your own schema.

materialized_view
Specify the name of the existing materialized view to be dropped.
SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-55

DROP MATERIALIZED VIEW

■

■

■

■

If you drop a simple materialized view that is the least recently refreshed
materialized view of a master table, then the database automatically purges from
the master table materialized view log only the rows needed to refresh the
dropped materialized view.
If you drop a materialized view that was created on a prebuilt table, then the
database drops the materialized view, and the prebuilt table reverts to its identity
as a table.
When you drop a master table, the database does not automatically drop
materialized views based on the table. However, the database returns an error
when it tries to refresh a materialized view based on a master table that has been
dropped.
If you drop a materialized view, then any compiled requests that were rewritten to
use the materialized view will be invalidated and recompiled automatically. If the
materialized view was prebuilt on a table, then the table is not dropped, but it can
no longer be maintained by the materialized view refresh mechanism.

PRESERVE TABLE Clause
This clause lets you retain the materialized view container table and its contents after
the materialized view object is dropped. The resulting table has the same name as the
dropped materialized view.
Oracle Database removes all metadata associated with the materialized view.
However, indexes created on the container table automatically during creation of the
materialized view are preserved, with one exception: the index created during the
creation of a rowid materialized view is dropped. Also, if the materialized view has
any nested table columns, then the storage tables for those columns are preserved,
along with their metadata.
This clause is not valid for
materialized views that have been imported from releases earlier than Oracle9i, when
these objects were called "snapshots".

Restriction on the PRESERVE TABLE Clause

Examples
17

Dropping a Materialized View: Examples The following statement drops the
materialized view emp_data in the sample schema hr:
DROP MATERIALIZED VIEW emp_data;

The following statement drops the sales_by_month_by_state materialized view and
the underlying table of the materialized view, unless the underlying table was
registered in the CREATE MATERIALIZED VIEW statement with the ON PREBUILT TABLE
clause:
DROP MATERIALIZED VIEW sales_by_month_by_state;

17-56 Oracle Database SQL Language Reference

DROP MATERIALIZED VIEW LOG

DROP MATERIALIZED VIEW LOG
Purpose
17

Use the DROP MATERIALIZED VIEW LOG statement to remove a materialized view log
from the database.
The keyword SNAPSHOT is supported in place of MATERIALIZED
VIEW for backward compatibility.
Note:

See Also:
■

■

■

■

CREATE MATERIALIZED VIEW on page 15-4 and ALTER
MATERIALIZED VIEW on page 11-3 for more information on
materialized views
CREATE MATERIALIZED VIEW LOG on page 15-27 for
information on materialized view logs
Oracle Database Advanced Replication for information on
materialized views in a replication environment
Oracle Database Data Warehousing Guide for information on
materialized views in a data warehousing environment

Prerequisites
17

To drop a materialized view log, you must have the privileges needed to drop a table.
See Also:

DROP TABLE on page 18-5

Syntax
17

drop_materialized_view_log::=
schema
DROP

MATERIALIZED

VIEW

LOG

ON

.
table

;

Semantics
17

schema
Specify the schema containing the materialized view log and its master table. If you
omit schema, then Oracle Database assumes the materialized view log and master table
are in your own schema.

table
Specify the name of the master table associated with the materialized view log to be
dropped.
After you drop a materialized view log, some materialized views based on the
materialized view log master table can no longer be fast refreshed. These materialized
views include rowid materialized views, primary key materialized views, and
subquery materialized views.

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-57

DROP MATERIALIZED VIEW LOG

See Also: Oracle Database Data Warehousing Guide for a description of
these types of materialized views

Examples
17

The following statement drops the
materialized view log on the oe.customers master table:

Dropping a Materialized View Log: Example
DROP MATERIALIZED VIEW LOG ON customers;

17-58 Oracle Database SQL Language Reference

DROP OPERATOR

DROP OPERATOR
Purpose
17

Use the DROP OPERATOR statement to drop a user-defined operator.
See Also:
■

■

■

CREATE OPERATOR on page 15-35 and ALTER OPERATOR on
page 11-25 for information on creating and modifying operators
"User-Defined Operators" on page 4-9 and Oracle Database Data
Cartridge Developer's Guide for more information on operators in
general
ALTER INDEXTYPE on page 10-97 for information on dropping
an operator of a user-defined indextype

Prerequisites
17

The operator must be in your schema or you must have the DROP ANY OPERATOR system
privilege.

Syntax
17

drop_operator::=
schema
DROP

.

FORCE

OPERATOR

operator

;

Semantics
17

schema
Specify the schema containing the operator. If you omit schema, then Oracle Database
assumes the operator is in your own schema.

operator
Specify the name of the operator to be dropped.

FORCE
Specify FORCE to drop the operator even if it is currently being referenced by one or
more schema objects, such as indextypes, packages, functions, procedures, and so on.
The database marks any such dependent objects INVALID. Without FORCE, you cannot
drop an operator if any schema objects reference it.

Examples
17

Dropping a User-Defined Operator: Example The following statement drops the

operator eq_op:
DROP OPERATOR eq_op;

Because the FORCE clause is not specified, this operation will fail if any of the bindings
of this operator are referenced by an indextype.
SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-59

DROP OUTLINE

DROP OUTLINE
Purpose
17

Oracle strongly recommends that you use SQL plan
management for new applications. SQL plan management creates SQL
plan baselines, which offer superior SQL performance stability
compared with stored outlines.

Note:

You can migrate existing stored outlines to SQL plan baselines by
using the MIGRATE_STORED_OUTLINE function of the DBMS_SPM package
or Enterprise Manager DB Control. When the migration is complete,
the stored outlines are marked as migrated and can be removed. You
can drop all migrated stored outlines on your system by using the
DROP_MIGRATED_STORED_OUTLINE function of the DBMS_SPM package.
See Also: Oracle Database Performance Tuning Guide for more
information about SQL plan management and Oracle Database PL/SQL
Packages and Types Reference for information about the DBMS_SPM
package
Use the DROP OUTLINE statement to drop a stored outline.
See Also:

CREATE OUTLINE on page 15-38 for information on creating an
outline

■

Oracle Database Performance Tuning Guide for more information on
outlines in general

■

Prerequisites
17

To drop an outline, you must have the DROP ANY OUTLINE system privilege.

Syntax
17

drop_outline::=
DROP

OUTLINE

outline

;

Semantics
17

outline
Specify the name of the outline to be dropped.
After the outline is dropped, if the SQL statement for which the stored outline was
created is compiled, then the optimizer generates a new execution plan without the
influence of the outline.

17-60 Oracle Database SQL Language Reference

DROP OUTLINE

Examples
17

Dropping an Outline: Example The following statement drops the stored outline
called salaries.
DROP OUTLINE salaries;

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-61

DROP PACKAGE

DROP PACKAGE
Purpose
17

Packages are defined using PL/SQL. Refer to Oracle Database PL/SQL Language
Reference for complete information on creating, altering, and dropping packages.
Use the DROP PACKAGE statement to remove a stored package from the database. This
statement drops the body and specification of a package.
Do not use this statement to remove a single object from a
package. Instead, re-create the package without the object using the
CREATE PACKAGE and CREATE PACKAGE BODY statements with the OR
REPLACE clause.
Note:

Prerequisites
17

The package must be in your own schema or you must have the DROP ANY PROCEDURE
system privilege.

Syntax
17

drop_package::=
BODY
DROP

schema

.

PACKAGE

package

;

Semantics
17

BODY
Specify BODY to drop only the body of the package. If you omit this clause, then Oracle
Database drops both the body and specification of the package.
When you drop only the body of a package but not its specification, the database does
not invalidate dependent objects. However, you cannot call one of the procedures or
stored functions declared in the package specification until you re-create the package
body.

schema
Specify the schema containing the package. If you omit schema, then the database
assumes the package is in your own schema.

package
Specify the name of the package to be dropped.
Oracle Database invalidates any local objects that depend on the package specification.
If you subsequently reference one of these objects, then the database tries to recompile
the object and returns an error if you have not re-created the dropped package.
If any statistics types are associated with the package, then the database disassociates
the statistics types with the FORCE clause and drops any user-defined statistics
collected with the statistics types.

17-62 Oracle Database SQL Language Reference

DROP PACKAGE

See Also: ASSOCIATE STATISTICS on page 13-25 and
DISASSOCIATE STATISTICS on page 17-34

Examples
17

Dropping a Package: Example The following statement drops the specification and
body of the emp_mgmt package, invalidating all objects that depend on the
specification. See Oracle Database PL/SQL Language Reference for the example that
creates this package.
DROP PACKAGE emp_mgmt;

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-63

DROP PROCEDURE

DROP PROCEDURE
Purpose
17

Procedures are defined using PL/SQL. Refer to Oracle Database PL/SQL Language
Reference for complete information on creating, altering, and dropping procedures.
Use the DROP PROCEDURE statement to remove a standalone stored procedure from the
database. Do not use this statement to remove a procedure that is part of a package.
Instead, either drop the entire package using the DROP PACKAGE statement, or redefine
the package without the procedure using the CREATE PACKAGE statement with the OR
REPLACE clause.

Prerequisites
17

The procedure must be in your own schema or you must have the DROP ANY PROCEDURE
system privilege.

Syntax
17

drop_procedure::=
schema
DROP

PR0CEDURE

.
procedure

;

Semantics
17

schema
Specify the schema containing the procedure. If you omit schema, then Oracle
Database assumes the procedure is in your own schema.

procedure
Specify the name of the procedure to be dropped.
When you drop a procedure, Oracle Database invalidates any local objects that depend
upon the dropped procedure. If you subsequently reference one of these objects, then
the database tries to recompile the object and returns an error message if you have not
re-created the dropped procedure.

Examples
17

The following statement drops the procedure
remove_emp owned by the user hr and invalidates all objects that depend upon
remove_emp:
Dropping a Procedure: Example

DROP PROCEDURE hr.remove_emp;

17-64 Oracle Database SQL Language Reference

DROP PROFILE

DROP PROFILE
Purpose
17

Use the DROP PROFILE statement to remove a profile from the database. You can drop
any profile except the DEFAULT profile.
See Also: CREATE PROFILE on page 15-50 and ALTER PROFILE on
page 11-32 on creating and modifying a profile

Prerequisites
17

You must have the DROP PROFILE system privilege.

Syntax
17

drop_profile::=
CASCADE
DROP

PROFILE

profile

;

Semantics
17

profile
Specify the name of the profile to be dropped.

CASCADE
Specify CASCADE to deassign the profile from any users to whom it is assigned. Oracle
Database automatically assigns the DEFAULT profile to such users. You must specify
this clause to drop a profile that is currently assigned to users.

Examples
17

The following statement drops the profile app_user,
which was created in "Creating a Profile: Example" on page 15-54. Oracle Database
drops the profile app_user and assigns the DEFAULT profile to any users currently
assigned the app_user profile:

Dropping a Profile: Example

DROP PROFILE app_user CASCADE;

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-65

DROP RESTORE POINT

DROP RESTORE POINT
Purpose
17

Use the DROP RESTORE POINT statement to remove a normal restore point or a
guaranteed restore point from the database.
■

■

You need not drop normal restore points. The database automatically drops the
oldest restore points when necessary, as described in the semantics for restore_point
on page 15-57. However, you can drop a normal restore point if you want to reuse
the name.
Guaranteed restore points are not dropped automatically. Therefore, if you want to
remove a guaranteed restore point from the database, then you must do so
explicitly using this statement.
CREATE RESTORE POINT on page 15-56, FLASHBACK
DATABASE on page 18-24, and FLASHBACK TABLE on page 18-27
for information on creating and using restore points
See Also:

Prerequisites
17

To drop a normal restore point, you must have either the SELECT ANY DICTIONARY or the
FLASHBACK ANY TABLE system privilege. To drop a guaranteed restore point, you must
have the SYSDBA system privilege.

Syntax
17

DROP

RESTORE

POINT

restore_point

;

Semantics
17

restore_point

Specify the name of the restore point you want to drop.

Examples
17

Dropping a Restore Point: Example The following example drops the good_data
restore point, which was created in "Creating and Using a Restore Point: Example" on
page 15-57:
DROP RESTORE POINT good_data;

17-66 Oracle Database SQL Language Reference

DROP ROLE

DROP ROLE
Purpose
17

Use the DROP ROLE statement to remove a role from the database. When you drop a
role, Oracle Database revokes it from all users and roles to whom it has been granted
and removes it from the database. User sessions in which the role is already enabled
are not affected. However, no new user session can enable the role after it is dropped.
See Also:
■

■

CREATE ROLE on page 15-59 and ALTER ROLE on page 11-38 for
information on creating roles and changing the authorization
needed to enable a role
SET ROLE on page 19-61 for information on disabling roles for the
current session

Prerequisites
17

You must have been granted the role with the ADMIN OPTION or you must have the DROP
ANY ROLE system privilege.

Syntax
17

drop_role::=
DROP

ROLE

role

;

Semantics
17

role
Specify the name of the role to be dropped.

Examples
17

To drop the role dw_manager, which was created in
"Creating a Role: Example" on page 15-60, issue the following statement:

Dropping a Role: Example
DROP ROLE dw_manager;

SQL Statements: CREATE TYPE to DROP ROLLBACK SEGMENT 17-67

DROP ROLLBACK SEGMENT

DROP ROLLBACK SEGMENT
Purpose
17

Use the DROP ROLLBACK SEGMENT to remove a rollback segment from the database. When
you drop a rollback segment, all space allocated to the rollback segment returns to the
tablespace.
Note: If your database is running in automatic undo mode, then this
is the only valid operation on rollback segments. In that mode, you
cannot create or alter a rollback segment.

Prerequisites
17

You must have the DROP ROLLBACK SEGMENT system privilege, and the rollback segment
must be offline.

Syntax
17

drop_rollback_segment::=
DROP

ROLLBACK

SEGMENT

rollback_segment

;

Semantics
17

rollback_segment
Specify the name the rollback segment to be dropped.
Restrictions on Dropping Rollback Segments

This statement is subject to the

following restrictions:
■

■

You can drop a rollback segment only if it is offline. To determine whether a
rollback segment is offline, query the data dictionary view DBA_ROLLBACK_SEGS.
Offline rollback segments have the value AVAILABLE in the STATUS column. You can
take a rollback segment offline with the OFFLINE clause of the ALTER ROLLBACK
SEGMENT statement.
You cannot drop the SYSTEM rollback segment.

Examples
17

Dropping a Rollback Segment: Example The following syntax drops the rollback
segment created in "Creating a Rollback Segment: Example" on page 15-64:
DROP ROLLBACK SEGMENT rbs_one;

17-68 Oracle Database SQL Language Reference

18
SQL Statements: DROP SEQUENCE to
ROLLBACK
18

This chapter contains the following SQL statements:
■

DROP SEQUENCE

■

DROP SYNONYM

■

DROP TABLE

■

DROP TABLESPACE

■

DROP TRIGGER

■

DROP TYPE

■

DROP TYPE BODY

■

DROP USER

■

DROP VIEW

■

EXPLAIN PLAN

■

FLASHBACK DATABASE

■

FLASHBACK TABLE

■

GRANT

■

INSERT

■

LOCK TABLE

■

MERGE

■

NOAUDIT

■

PURGE

■

RENAME

■

REVOKE

■

ROLLBACK

SQL Statements: DROP SEQUENCE to ROLLBACK 18-1

DROP SEQUENCE

DROP SEQUENCE
Purpose
18

Use the DROP SEQUENCE statement to remove a sequence from the database.
You can also use this statement to restart a sequence by dropping and then re-creating
it. For example, if you have a sequence with a current value of 150 and you would like
to restart the sequence with a value of 27, then you can drop the sequence and then
re-create it with the same name and a START WITH value of 27.
CREATE SEQUENCE on page 15-67 and ALTER
SEQUENCE on page 11-43 for more information on creating and
modifying a sequence

See Also:

Prerequisites
18

The sequence must be in your own schema or you must have the DROP ANY SEQUENCE
system privilege.

Syntax
18

drop_sequence::=
schema
DROP

.

SEQUENCE

sequence_name

;

Semantics
18

schema
Specify the schema containing the sequence. If you omit schema, then Oracle Database
assumes the sequence is in your own schema.

sequence_name
Specify the name of the sequence to be dropped.

Examples
18

The following statement drops the sequence
customers_seq owned by the user oe, which was created in "Creating a Sequence:
Example" on page 15-70. To issue this statement, you must either be connected as user
oe or have the DROP ANY SEQUENCE system privilege:
Dropping a Sequence: Example

DROP SEQUENCE oe.customers_seq;

18-2 Oracle Database SQL Language Reference

DROP SYNONYM

DROP SYNONYM
Purpose
18

Use the DROP SYNONYM statement to remove a synonym from the database or to change
the definition of a synonym by dropping and re-creating it.
See Also: CREATE SYNONYM on page 16-2 for more information
on synonyms

Prerequisites
18

To drop a private synonym, either the synonym must be in your own schema or you
must have the DROP ANY SYNONYM system privilege.
To drop a PUBLIC synonym, you must have the DROP PUBLIC SYNONYM system privilege.

Syntax
18

drop_synonym::=
PUBLIC
DROP

schema
SYNONYM

.

FORCE
synonym

;

Semantics
18

PUBLIC
You must specify PUBLIC to drop a public synonym. You cannot specify schema if you
have specified PUBLIC.

schema
Specify the schema containing the synonym. If you omit schema, then Oracle Database
assumes the synonym is in your own schema.

synonym
Specify the name of the synonym to be dropped.
If you drop a synonym for the master table of a materialized view, and if the defining
query of the materialized view specified the synonym rather than the actual table
name, then Oracle Database marks the materialized view unusable.
If an object type synonym has any dependent tables or user-defined types, then you
cannot drop the synonym unless you also specify FORCE.

FORCE
Specify FORCE to drop the synonym even if it has dependent tables or user-defined
types.

SQL Statements: DROP SEQUENCE to ROLLBACK 18-3

DROP SYNONYM

Oracle does not recommend that you specify FORCE to drop
object type synonyms with dependencies. This operation can result in
invalidation of other user-defined types or marking UNUSED the table
columns that depend on the synonym. For information about type
dependencies, see Oracle Database Object-Relational Developer's Guide.

Caution:

Examples
18

Dropping a Synonym: Example To drop the public synonym named customers,
which was created in "Oracle Database Resolution of Synonyms: Example" on
page 16-4, issue the following statement:
DROP PUBLIC SYNONYM customers;

18-4 Oracle Database SQL Language Reference

DROP TABLE

DROP TABLE
Purpose
18

Use the DROP TABLE statement to move a table or object table to the recycle bin or to
remove the table and all its data from the database entirely.
Caution: Unless you specify the PURGE clause, the DROP TABLE
statement does not result in space being released back to the
tablespace for use by other objects, and the space continues to count
toward the user's space quota.

For an external table, this statement removes only the table metadata in the database.
It has no affect on the actual data, which resides outside of the database.
When you drop a table that is part of a cluster, the table is moved to the recycle bin.
However, if you subsequently drop the cluster, then the table is purged from the
recycle bin and can no longer be recovered with a FLASHBACK TABLE operation.
Dropping a table invalidates dependent objects and removes object privileges on the
table. If you want to re-create the table, then you must regrant object privileges on the
table, re-create the indexes, integrity constraints, and triggers for the table, and
respecify its storage parameters. Truncating has none of these effects. Therefore,
removing rows with the TRUNCATE statement can be more efficient than dropping and
re-creating a table.
See Also:

CREATE TABLE on page 16-6 and ALTER TABLE on page 12-2 for
information on creating and modifying tables

■

TRUNCATE TABLE on page 19-69 and DELETE on page 17-26 for
information on removing data from a table

■

FLASHBACK TABLE on page 18-27 for information on retrieving
a dropped table from the recycle bin

■

Prerequisites
18

The table must be in your own schema or you must have the DROP ANY TABLE system
privilege.
You can perform DDL operations (such as ALTER TABLE, DROP TABLE, CREATE INDEX) on
a temporary table only when no session is bound to it. A session becomes bound to a
temporary table by performing an INSERT operation on the table. A session becomes
unbound to the temporary table by issuing a TRUNCATE statement or at session
termination, or, for a transaction-specific temporary table, by issuing a COMMIT or
ROLLBACK statement.

Syntax
18

drop_table::=
schema
DROP

TABLE

.

CASCADE
table

CONSTRAINTS

PURGE
;

SQL Statements: DROP SEQUENCE to ROLLBACK 18-5

DROP TABLE

Semantics
18

schema
Specify the schema containing the table. If you omit schema, then Oracle Database
assumes the table is in your own schema.

table
Specify the name of the table to be dropped. Oracle Database automatically performs
the following operations:
■
■

■
■

■

All rows from the table are dropped.
All table indexes and domain indexes are dropped, as well as any triggers defined
on the table, regardless of who created them or whose schema contains them. If
table is partitioned, then any corresponding local index partitions are also
dropped.
All the storage tables of nested tables and LOBs of table are dropped.
When you drop a range-, hash-, or list-partitioned table, then the database drops
all the table partitions. If you drop a composite-partitioned table, then all the
partitions and subpartitions are also dropped.
When you drop a partitioned table with the PURGE keyword, the statement
executes as a series of subtransactions, each of which drops a subset of partitions
or subpartitions and their metadata. This division of the drop operation into
subtransactions optimizes the processing of internal system resource consumption
(for example, the library cache), especially for the dropping of very large
partitioned tables. As soon as the first subtransaction commits, the table is marked
UNUSABLE. If any of the subtransactions fails, then the only operation allowed on
the table is another DROP TABLE ... PURGE statement. Such a statement will resume
work from where the previous DROP TABLE statement failed, assuming that you
have corrected any errors that the previous operation encountered.
You can list the tables marked UNUSABLE by such a drop operation by querying the
status column of the *_TABLES, *_PART_TABLES, *_ALL_TABLES, or *_OBJECT_
TABLES data dictionary views, as appropriate.
See Also: Oracle Database VLDB and Partitioning Guide for more
information on dropping partitioned tables.

■

■

■

For an index-organized table, any mapping tables defined on the index-organized
table are dropped.
For a domain index, the appropriate drop routines are invoked. Refer to Oracle
Database Data Cartridge Developer's Guide for more information on these routines.
If any statistics types are associated with the table, then the database disassociates
the statistics types with the FORCE clause and removes any user-defined statistics
collected with the statistics type.
See Also: ASSOCIATE STATISTICS on page 13-25 and
DISASSOCIATE STATISTICS on page 17-34 for more information on
statistics type associations

■

If the table is not part of a cluster, then the database returns all data blocks
allocated to the table and its indexes to the tablespaces containing the table and its
indexes.

18-6 Oracle Database SQL Language Reference

DROP TABLE

To drop a cluster and all its the tables, use the DROP CLUSTER statement with the
INCLUDING TABLES clause to avoid dropping each table individually. See DROP
CLUSTER on page 17-36.
■

If the table is a base table for a view, a container or master table of a materialized
view, or if it is referenced in a stored procedure, function, or package, then the
database invalidates these dependent objects but does not drop them. You cannot
use these objects unless you re-create the table or drop and re-create the objects so
that they no longer depend on the table.
If you choose to re-create the table, then it must contain all the columns selected by
the subqueries originally used to define the materialized views and all the
columns referenced in the stored procedures, functions, or packages. Any users
previously granted object privileges on the views, stored procedures, functions, or
packages need not be regranted these privileges.
If the table is a master table for a materialized view, then the materialized view can
still be queried, but it cannot be refreshed unless the table is re-created so that it
contains all the columns selected by the defining query of the materialized view.
If the table has a materialized view log, then the database drops this log and any
other direct-path INSERT refresh information associated with the table.

Restrictions on Dropping Tables
■

■

■

You cannot directly drop the storage table of a nested table. Instead, you must
drop the nested table column using the ALTER TABLE ... DROP COLUMN clause.
You cannot drop the parent table of a reference-partitioned table. You must first
drop all reference-partitioned child tables.
You cannot drop a table that uses a flashback data archive for historical tracking.
You must first disable the table's use of the flashback data archive.

CASCADE CONSTRAINTS
Specify CASCADE CONSTRAINTS to drop all referential integrity constraints that refer to
primary and unique keys in the dropped table. If you omit this clause, and such
referential integrity constraints exist, then the database returns an error and does not
drop the table.

PURGE
Specify PURGE if you want to drop the table and release the space associated with it in a
single step. If you specify PURGE, then the database does not place the table and its
dependent objects into the recycle bin.
You cannot roll back a DROP TABLE statement with the PURGE
clause, nor can you recover the table if you have dropped it with the
PURGE clause.
Caution:

Using this clause is equivalent to first dropping the table and then purging it from the
recycle bin. This clause lets you save one step in the process. It also provides enhanced
security if you want to prevent sensitive material from appearing in the recycle bin.
See Also: Oracle Database Administrator's Guide for information on
the recycle bin and naming conventions for objects in the recycle bin

SQL Statements: DROP SEQUENCE to ROLLBACK 18-7

DROP TABLE

Examples
18

The following statement drops the oe.list_customers
table created in "List Partitioning Example" on page 16-77.

Dropping a Table: Example

DROP TABLE list_customers PURGE;

18-8 Oracle Database SQL Language Reference

DROP TABLESPACE

DROP TABLESPACE
Purpose
18

Use the DROP TABLESPACE statement to remove a tablespace from the database.
When you drop a tablespace, Oracle Database does not place it in the recycle bin.
Therefore, you cannot subsequently either purge or undrop the tablespace.
CREATE TABLESPACE on page 16-83 and ALTER
TABLESPACE on page 12-90 for information on creating and
modifying a tablespace

See Also:

Prerequisites
18

You must have the DROP TABLESPACE system privilege. You cannot drop a tablespace if
it contains any rollback segments holding active transactions.

Syntax
18

drop_tablespace::=
DROP

TABLESPACE

tablespace

AND
DATAFILES
KEEP
INCLUDING

CASCADE

CONSTRAINTS

CONTENTS
;

Semantics
18

tablespace
Specify the name of the tablespace to be dropped.
You can drop a tablespace regardless of whether it is online or offline. Oracle
recommends that you take the tablespace offline before dropping it to ensure that no
SQL statements in currently running transactions access any of the objects in the
tablespace.
You cannot drop the SYSTEM tablespace. You can drop the SYSAUX tablespace only if
you have the SYSDBA system privilege and you have started the database in UPGRADE
mode.
You may want to alert any users who have been assigned the tablespace as either a
default or temporary tablespace. After the tablespace has been dropped, these users
cannot allocate space for objects or sort areas in the tablespace. You can reassign users
new default and temporary tablespaces with the ALTER USER statement.
Any objects that were previously dropped from the tablespace and moved to the
recycle bin are purged from the recycle bin. Oracle Database removes from the data
dictionary all metadata about the tablespace and all data files and temp files in the
tablespace. The database also automatically drops from the operating system any
Oracle-managed data files and temp files in the tablespace. Other data files and temp

SQL Statements: DROP SEQUENCE to ROLLBACK 18-9

DROP TABLESPACE

files are not removed from the operating system unless you specify INCLUDING
CONTENTS AND DATAFILES.
You cannot use this statement to drop a tablespace group. However, if tablespace is
the only tablespace in a tablespace group, then Oracle Database removes the
tablespace group from the data dictionary as well.
Restrictions on Dropping Tablespaces Dropping tablespaces is subject to the
following restrictions:
■

■

■

■

■

■

You cannot drop a tablespace that contains a domain index or any objects created
by a domain index.
You cannot drop an undo tablespace if it is being used by any instance or if it
contains any undo data needed to roll back uncommitted transactions.
You cannot drop a tablespace that has been designated as the default tablespace
for the database. You must first reassign another tablespace as the default
tablespace and then drop the old default tablespace.
You cannot drop a temporary tablespace if it is part of the database default
temporary tablespace group. You must first remove the tablespace from the
database default temporary tablespace group and then drop it.
You cannot drop a temporary tablespace if it contains segments that are in use by
existing sessions. In this case, no error is raised. The database waits until there are
no segments in use by existing sessions and then drops the tablespace.
You cannot drop a tablespace, even with the INCLUDING CONTENTS and CASCADE
CONSTRAINTS clauses, if doing so would disable a primary key or unique constraint
in another tablespace. For example, if the tablespace being dropped contains a
primary key index, but the primary key column itself is in a different tablespace,
then you cannot drop the tablespace until you have manually disabled the
primary key constraint in the other tablespace.
See Also: Oracle Database Data Cartridge Developer's Guide and Oracle
Database Concepts for more information on domain indexes

INCLUDING CONTENTS
Specify INCLUDING CONTENTS to drop all the contents of the tablespace. You must
specify this clause to drop a tablespace that contains any database objects. If you omit
this clause, and the tablespace is not empty, then the database returns an error and
does not drop the tablespace.
DROP TABLESPACE fails, even if you specify INCLUDING CONTENTS, if the tablespace
contains some, but not all, of the partitions or subpartitions of a single table. If all the
partitions or subpartitions of a partitioned table reside in tablespace, then DROP
TABLESPACE ... INCLUDING CONTENTS drops tablespace, as well as any associated index
segments, LOB data and index segments, and nested table data and index segments of
table in other tablespace(s).
For a partitioned index-organized table, if all the primary key index segments are in
this tablespace, then this clause will also drop any overflow segments that exist in
other tablespaces, as well as any associated mapping table in other tablespaces. If some
of the primary key index segments are not in this tablespace, then the statement will
fail. In that case, before you can drop the tablespace, you must use ALTER TABLE ... MOVE
PARTITION to move those primary key index segments into this tablespace, drop the
partitions whose overflow data segments are not in this tablespace, and drop the
partitioned index-organized table.

18-10 Oracle Database SQL Language Reference

DROP TABLESPACE

If the tablespace contains a master table of a materialized view, then the database
invalidates the materialized view.
If the tablespace contains a materialized view log, then the database drops the log and
any other direct-path INSERT refresh information associated with the table.

AND DATAFILES
When you specify INCLUDING CONTENTS, the AND DATAFILES clause lets you instruct the
database to delete the associated operating system files as well. Oracle Database writes
a message to the alert log for each operating system file deleted. This clause is not
needed for Oracle Managed Files, because they are removed from the system even if
you do not specify AND DATAFILES.

KEEP DATAFILES
When you specify INCLUDING CONTENTS, the KEEP DATAFILES clause lets you instruct the
database to leave untouched the associated operating system files, including Oracle
Managed Files. You must specify this clause if you are using Oracle Managed Files and
you do not want the associated operating system files removed by the INCLUDING
CONTENTS clause.

CASCADE CONSTRAINTS
Specify CASCADE CONSTRAINTS to drop all referential integrity constraints from tables
outside tablespace that refer to primary and unique keys of tables inside tablespace.
If you omit this clause and such referential integrity constraints exist, then Oracle
Database returns an error and does not drop the tablespace.

Examples
18

The following statement drops the tbs_01
tablespace and drops all referential integrity constraints that refer to primary and
unique keys inside tbs_01:

Dropping a Tablespace: Example

DROP TABLESPACE tbs_01
INCLUDING CONTENTS
CASCADE CONSTRAINTS;

Deleting Operating System Files: Example The following example drops the tbs_02
tablespace and deletes all associated operating system data files:
DROP TABLESPACE tbs_02
INCLUDING CONTENTS AND DATAFILES;

SQL Statements: DROP SEQUENCE to ROLLBACK 18-11

DROP TRIGGER

DROP TRIGGER
Purpose
18

Triggers are defined using PL/SQL. Refer to Oracle Database PL/SQL Language Reference
for complete information on creating, altering, and dropping triggers.
Use the DROP TRIGGER statement to remove a database trigger from the database.
See Also: CREATE TRIGGER on page 16-98 and ALTER TRIGGER
on page 13-2

Prerequisites
18

The trigger must be in your own schema or you must have the DROP ANY TRIGGER
system privilege. To drop a trigger on DATABASE in another user's schema, you must
also have the ADMINISTER DATABASE TRIGGER system privilege.

Syntax
18

drop_trigger::=
schema
DROP

TRIGGER

.
trigger

;

Semantics
18

schema
Specify the schema containing the trigger. If you omit schema, then Oracle Database
assumes the trigger is in your own schema.

trigger
Specify the name of the trigger to be dropped. Oracle Database removes it from the
database and does not fire it again.

Examples
18

Dropping a Trigger: Example The following statement drops the salary_check
trigger in the schema hr:
DROP TRIGGER hr.salary_check;

18-12 Oracle Database SQL Language Reference

DROP TYPE

DROP TYPE
Purpose
18

Object types are defined using PL/SQL. Refer to Oracle Database PL/SQL Language
Reference for complete information on creating, altering, and dropping object types.
Use the DROP TYPE statement to drop the specification and body of an object type, a
varray, or a nested table type.

Prerequisites
18

The object type, varray, or nested table type must be in your own schema or you must
have the DROP ANY TYPE system privilege.

Syntax
18

drop_type::=
FORCE
schema
DROP

.

TYPE

VALIDATE
type_name

;

Semantics
18

schema
Specify the schema containing the type. If you omit schema, then Oracle Database
assumes the type is in your own schema.

type_name
Specify the name of the object, varray, or nested table type to be dropped. You can
drop only types with no type or table dependencies.
If type_name is a supertype, then this statement will fail unless you also specify FORCE.
If you specify FORCE, then the database invalidates all subtypes depending on this
supertype.
If type_name is a statistics type, then this statement will fail unless you also specify
FORCE. If you specify FORCE, then the database first disassociates all objects that are
associated with type_name and then drops type_name.
See Also: ASSOCIATE STATISTICS on page 13-25 and
DISASSOCIATE STATISTICS on page 17-34 for more information on
statistics types

If type_name is an object type that has been associated with a statistics type, then the
database first attempts to disassociate type_name from the statistics type and then
drops type_name. However, if statistics have been collected using the statistics type,
then the database will be unable to disassociate type_name from the statistics type, and
this statement will fail.
If type_name is an implementation type for an indextype, then the indextype will be
marked INVALID.

SQL Statements: DROP SEQUENCE to ROLLBACK 18-13

DROP TYPE

If type_name has a public synonym defined on it, then the database will also drop the
synonym.
Unless you specify FORCE, you can drop only object types, nested tables, or varray
types that are standalone schema objects with no dependencies. This is the default
behavior.
See Also:

CREATE INDEXTYPE on page 14-87

FORCE
Specify FORCE to drop the type even if it has dependent database objects. Oracle
Database marks UNUSED all columns dependent on the type to be dropped, and those
columns become inaccessible.
Oracle does not recommend that you specify FORCE to drop
object types with dependencies. This operation is not recoverable and
could cause the data in the dependent tables or columns to become
inaccessible.

Caution:

VALIDATE
If you specify VALIDATE when dropping a type, then Oracle Database checks for stored
instances of this type within substitutable columns of any of its supertypes. If no such
instances are found, then the database completes the drop operation.
This clause is meaningful only for subtypes. Oracle recommends the use of this option
to safely drop subtypes that do not have any explicit type or table dependencies.

Examples
18

Dropping an Object Type: Example

The following statement removes object type person_t. See Oracle Database PL/SQL
Language Reference for the example that creates this object type. Any columns that are
dependent on person_t are marked UNUSED and become inaccessible.
DROP TYPE person_t FORCE;

18-14 Oracle Database SQL Language Reference

DROP TYPE BODY

DROP TYPE BODY
Purpose
18

Object types are defined using PL/SQL. Refer to Oracle Database PL/SQL Language
Reference for complete information on creating, altering, and dropping object types.
Use the DROP TYPE BODY statement to drop the body of an object type, varray, or nested
table type. When you drop a type body, the object type specification still exists, and
you can re-create the type body. Prior to re-creating the body, you can still use the
object type, although you cannot call the member functions.

Prerequisites
18

The object type body must be in your own schema or you must have the DROP ANY TYPE
system privilege.

Syntax
18

drop_type_body::=
schema
DROP

TYPE

.

BODY

type_name

;

Semantics
18

schema
Specify the schema containing the object type. If you omit schema, then Oracle
Database assumes the object type is in your own schema.

type_name
Specify the name of the object type body to be dropped.
Restriction on Dropping Type Bodies

You can drop a type body only if it has no type

or table dependencies.

Examples
18

Dropping an Object Type Body: Example The following statement removes object
type body data_typ1. See Oracle Database PL/SQL Language Reference for the example
that creates this object type.
DROP TYPE BODY data_typ1;

SQL Statements: DROP SEQUENCE to ROLLBACK 18-15

DROP USER

DROP USER
Purpose
18

Use the DROP USER statement to remove a database user and optionally remove the
user's objects.
In an Oracle Automatic Storage Management (Oracle ASM) cluster, a user
authenticated AS SYSASM can use this clause to remove a user from the password file
that is local to the Oracle ASM instance of the current node.
When you drop a user, Oracle Database also purges all of that user's schema objects
from the recycle bin.
Do not attempt to drop the users SYS or SYSTEM. Doing so
will corrupt your database.

Caution:

See Also: CREATE USER on page 17-7 and ALTER USER on
page 13-6 for information on creating and modifying a user

Prerequisites
18

You must have the DROP USER system privilege. In an Oracle ASM cluster, you must be
authenticated AS SYSASM.

Syntax
18

drop_user::=
CASCADE
DROP

USER

user

;

Semantics
18

user
Specify the user to be dropped. Oracle Database does not drop users whose schemas
contain objects unless you specify CASCADE or unless you first explicitly drop the user's
objects.
Restriction on Dropping Users You cannot drop a user whose schema contains a
table that uses a flashback data archive for historical tracking. You must first disable
the table's use of the flashback data archive.

CASCADE
Specify CASCADE to drop all objects in the user's schema before dropping the user. You
must specify this clause to drop a user whose schema contains any objects.
■

If the user's schema contains tables, then Oracle Database drops the tables and
automatically drops any referential integrity constraints on tables in other schemas
that refer to primary and unique keys on these tables.

18-16 Oracle Database SQL Language Reference

DROP USER

■

If this clause results in tables being dropped, then the database also drops all
domain indexes created on columns of those tables and invokes appropriate drop
routines.
See Also: Oracle Database Data Cartridge Developer's Guide for more
information on these routines

■

■

Oracle Database invalidates, but does not drop, the following objects in other
schemas:
–

Views or synonyms for objects in the dropped user's schema

–

Stored procedures, functions, or packages that query objects in the dropped
user's schema

Oracle Database does not drop materialized views in other schemas that are based
on tables in the dropped user's schema. However, because the base tables no
longer exist, the materialized views in the other schemas can no longer be
refreshed.

■

Oracle Database drops all triggers in the user's schema.

■

Oracle Database does not drop roles created by the user.
Oracle Database also drops with FORCE all types owned by
the user. See the FORCE keyword of DROP TYPE on page 18-14.

Caution:

Restriction on Dropping Users You cannot drop a user whose schema contains a
table that uses a flashback data archive. You must first disable use of the flashback
data archive.

Examples
18

Dropping a Database User: Example If user Sidney's schema contains no objects,
then you can drop sidney by issuing the statement:
DROP USER sidney;

If Sidney's schema contains objects, then you must use the CASCADE clause to drop
sidney and the objects:
DROP USER sidney CASCADE;

SQL Statements: DROP SEQUENCE to ROLLBACK 18-17

DROP VIEW

DROP VIEW
Purpose
18

Use the DROP VIEW statement to remove a view or an object view from the database.
You can change the definition of a view by dropping and re-creating it.
See Also: CREATE VIEW on page 17-14 and ALTER VIEW on
page 13-14 for information on creating and modifying a view

Prerequisites
18

The view must be in your own schema or you must have the DROP ANY VIEW system
privilege.

Syntax
18

drop_view::=
schema
DROP

.

VIEW

CASCADE

CONSTRAINTS

view

;

Semantics
18

schema
Specify the schema containing the view. If you omit schema, then Oracle Database
assumes the view is in your own schema.

view
Specify the name of the view to be dropped.
Oracle Database does not drop views, materialized views, and synonyms that are
dependent on the view but marks them INVALID. You can drop them or redefine views
and synonyms, or you can define other views in such a way that the invalid views and
synonyms become valid again.
If any subviews have been defined on view, then the database invalidates the
subviews as well. To determine whether the view has any subviews, query the
SUPERVIEW_NAME column of the USER_, ALL_, or DBA_VIEWS data dictionary views.
See Also:
■

■

CREATE TABLE on page 16-6 and CREATE SYNONYM on
page 16-2
ALTER MATERIALIZED VIEW on page 11-3 for information on
revalidating invalid materialized views

CASCADE CONSTRAINTS
Specify CASCADE CONSTRAINTS to drop all referential integrity constraints that refer to
primary and unique keys in the view to be dropped. If you omit this clause, and such
constraints exist, then the DROP statement fails.

18-18 Oracle Database SQL Language Reference

DROP VIEW

Examples
18

Dropping a View: Example The following statement drops the emp_view view, which
was created in "Creating a View: Example" on page 17-22:
DROP VIEW emp_view;

SQL Statements: DROP SEQUENCE to ROLLBACK 18-19

EXPLAIN PLAN

EXPLAIN PLAN
Purpose
18

Use the EXPLAIN PLAN statement to determine the execution plan Oracle Database
follows to execute a specified SQL statement. This statement inserts a row describing
each step of the execution plan into a specified table. You can also issue the EXPLAIN
PLAN statement as part of the SQL trace facility.
This statement also determines the cost of executing the statement. If any domain
indexes are defined on the table, then user-defined CPU and I/O costs will also be
inserted.
The definition of a sample output table PLAN_TABLE is available in a SQL script on your
distribution media. Your output table must have the same column names and data
types as this table. The common name of this script is UTLXPLAN.SQL. The exact name
and location depend on your operating system.
Oracle Database provides information on cached cursors through several dynamic
performance views:
■

For information on the work areas used by SQL cursors, query V$SQL_WORKAREA.

■

For information on the execution plan for a cached cursor, query V$SQL_PLAN.

■

■

■

For execution statistics at each step or operation of an execution plan of cached
cursors (for example, number of produced rows, number of blocks read), query
V$SQL_PLAN_STATISTICS.
For a selective precomputed join of the preceding three views, query V$SQL_PLAN_
STATISTICS_ALL.
Execution statistics at each step or operation of an execution plan of cached
cursors are displayed in V$SQL_PLAN_MONITOR if the statement execution is
monitored. You can force monitoring using the MONITOR hint.
See Also:
■

■

Oracle Database Performance Tuning Guide for information on the
output of EXPLAIN PLAN, how to use the SQL trace facility, and
how to generate and interpret execution plans
Oracle Database Reference for information on dynamic performance
views

Prerequisites
18

To issue an EXPLAIN PLAN statement, you must have the privileges necessary to insert
rows into an existing output table that you specify to hold the execution plan.
You must also have the privileges necessary to execute the SQL statement for which
you are determining the execution plan. If the SQL statement accesses a view, then you
must have privileges to access any tables and views on which the view is based. If the
view is based on another view that is based on a table, then you must have privileges
to access both the other view and its underlying table.
To examine the execution plan produced by an EXPLAIN PLAN statement, you must
have the privileges necessary to query the output table.
The EXPLAIN PLAN statement is a data manipulation language (DML) statement, rather
than a data definition language (DDL) statement. Therefore, Oracle Database does not

18-20 Oracle Database SQL Language Reference

EXPLAIN PLAN

implicitly commit the changes made by an EXPLAIN PLAN statement. If you want to
keep the rows generated by an EXPLAIN PLAN statement in the output table, then you
must commit the transaction containing the statement.
See Also: INSERT on page 18-54 and SELECT on page 19-4 for
information on the privileges you need to populate and query the
plan table

Syntax
18

explain_plan::=
SET
EXPLAIN

STATEMENT_ID

=

string

PLAN

schema
INTO

.

@

dblink

table
FOR

statement

;

Semantics
18

SET STATEMENT_ID Clause
Specify a value for the STATEMENT_ID column for the rows of the execution plan in the
output table. You can then use this value to identify these rows among others in the
output table. Be sure to specify a STATEMENT_ID value if your output table contains
rows from many execution plans. If you omit this clause, then the STATEMENT_ID value
defaults to null.

INTO table Clause
Specify the name of the output table, and optionally its schema and database. This
table must exist before you use the EXPLAIN PLAN statement.
If you omit schema, then the database assumes the table is in your own schema.
The dblink can be a complete or partial name of a database link to a remote Oracle
Database where the output table is located. You can specify a remote output table only
if you are using Oracle Database distributed functionality. If you omit dblink, then the
database assumes the table is on your local database. See "References to Objects in
Remote Databases" on page 3-117 for information on referring to database links.
If you omit INTO altogether, then the database assumes an output table named PLAN_
TABLE in your own schema on your local database.

FOR statement Clause
Specify a SELECT, INSERT, UPDATE, DELETE, MERGE, CREATE TABLE, CREATE INDEX, or ALTER
INDEX ... REBUILD statement for which the execution plan is generated.
Notes on EXPLAIN PLAN The following notes apply to EXPLAIN PLAN:
■

If statement includes the parallel_clause, then the resulting execution plan will
indicate parallel execution. However, EXPLAIN PLAN actually inserts the statement
into the plan table, so that the parallel DML statement you submit is no longer the
first DML statement in the transaction. This violates the Oracle Database
restriction of one parallel DML statement in a single transaction, and the statement

SQL Statements: DROP SEQUENCE to ROLLBACK 18-21

EXPLAIN PLAN

will be executed serially. To maintain parallel execution of the statements, you
must commit or roll back the EXPLAIN PLAN statement, and then submit the parallel
DML statement.
■

To determine the execution plan for an operation on a temporary table, EXPLAIN
PLAN must be run from the same session, because the data in temporary tables is
session specific.

Examples
18

The following statement determines the execution plan
and cost for an UPDATE statement and inserts rows describing the execution plan into
the specified plan_table table with the STATEMENT_ID value of 'Raise in Tokyo':
EXPLAIN PLAN Examples

EXPLAIN PLAN
SET STATEMENT_ID = 'Raise in Tokyo'
INTO plan_table
FOR UPDATE employees
SET salary = salary * 1.10
WHERE department_id =
(SELECT department_id FROM departments
WHERE location_id = 1700);

The following SELECT statement queries the plan_table table and returns the
execution plan and the cost:
SELECT id, LPAD(' ',2*(LEVEL-1))||operation operation, options,
object_name, object_alias, position
FROM plan_table
START WITH id = 0 AND statement_id = 'Raise in Tokyo'
CONNECT BY PRIOR id = parent_id AND statement_id = 'Raise in Tokyo'
ORDER BY id;

The query returns this execution plan:
ID OPERATION
--- -------------------0 UPDATE STATEMENT
1
UPDATE
2
INDEX
3
TABLE ACCESS
4
INDEX

OPTIONS
OBJECT_NAME
OBJECT_ALIAS
POSITION
-------------------- -------------------- -------------------- -------4
EMPLOYEES
1
RANGE SCAN
EMP_DEPARTMENT_IX
EMPLOYEES@UPD$1
1
BY INDEX ROWID
DEPARTMENTS
DEPARTMENTS@SEL$1
1
RANGE SCAN
DEPT_LOCATION_IX
DEPARTMENTS@SEL$1
1

The value in the POSITION column of the first row shows that the statement has a cost
of 4.
EXPLAIN PLAN: Partitioned Example The sample table sh.sales is partitioned on
the time_id column. Partition sales_q3_2000 contains time values less than Oct. 1,
2000, and there is a local index sales_time_bix on the time_id column.

Consider the query:
EXPLAIN PLAN FOR
SELECT * FROM sales
WHERE time_id BETWEEN :h AND '01-OCT-2000';

where :h represents an already declared bind variable. EXPLAIN PLAN executes this
query with PLAN_TABLE as the output table. The basic execution plan, including
partitioning information, is obtained with the following query:
SELECT operation, options, partition_start, partition_stop,
18-22 Oracle Database SQL Language Reference

EXPLAIN PLAN

partition_id
FROM plan_table;

SQL Statements: DROP SEQUENCE to ROLLBACK 18-23

FLASHBACK DATABASE

FLASHBACK DATABASE
Purpose
18

Use the FLASHBACK DATABASE statement to return the database to a past time or system
change number (SCN). This statement provides a fast alternative to performing
incomplete database recovery.
Following a FLASHBACK DATABASE operation, in order to have write access to the flashed
back database, you must reopen it with an ALTER DATABASE OPEN RESETLOGS statement.
See Also: Oracle Database Backup and Recovery User's Guide for more
information on FLASHBACK DATABASE

Prerequisites
18

You must have the SYSDBA system privilege. A fast recovery area must have been
prepared for the database. The database must have been put in FLASHBACK mode with
an ALTER DATABASE FLASHBACK ON statement unless you are flashing the database back
to a guaranteed restore point. The database must be mounted but not open. In
addition:
■
■

■

The database must run in ARCHIVELOG mode.
The database must be mounted, but not open, with a current control file. The
control file cannot be a backup or re-created. When the database control file is
restored from backup or re-created, all existing flashback log information is
discarded.
The database must contain no online tablespaces for which flashback functionality
was disabled with the SQL statement ALTER TABLESPACE ... FLASHBACK OFF.
See Also:
■

■

Oracle Database Backup and Recovery User's Guide and the ALTER
DATABASE ... flashback_mode_clause on page 10-41 for information on
putting the database in FLASHBACK mode
CREATE RESTORE POINT on page 15-56 for information on
restore points and guaranteed restore points

Syntax
18

flashback_database::=
SCN
expr
TO
STANDBY
FLASHBACK

TIMESTAMP
RESTORE

database
DATABASE

POINT

restore_point
;

SCN
expr
TO

BEFORE

TIMESTAMP
RESETLOGS

18-24 Oracle Database SQL Language Reference

FLASHBACK DATABASE

Semantics
18

When you issue a FLASHBACK DATABASE statement, Oracle Database first verifies that all
required archived and online redo logs are available. If they are available, then it
reverts all currently online data files in the database to the SCN or time specified in
this statement.
■

■

■

The amount of Flashback data retained in the database is controlled by the DB_
FLASHBACK_RETENTION_TARGET initialization parameter and the size of the fast
recovery area. You can determine how far back you can flash back the database by
querying the V$FLASHBACK_DATABASE_LOG view.
If insufficient data remains in the database to perform the flashback, then you can
use standard recovery procedures to recover the database to a past point in time.
If insufficient data remains for a set of data files, then the database returns an
error. In this case, you can take those data files offline and reissue the statement to
revert the remainder of the database. You can then attempt to recover the offline
data files using standard recovery procedures.
See Also: Oracle Database Backup and Recovery User's Guide for more
information on recovering data files

STANDBY
Specify STANDBY to revert the standby database to an earlier SCN or time. If the
database is not a standby database, then the database returns an error. If you omit this
clause, then database can be either a primary or a standby database.
See Also: Oracle Data Guard Concepts and Administration for
information on how you can use FLASHBACK DATABASE on a standby
database to achieve different delays

TO SCN Clause
Specify a system change number (SCN):
■
■

TO SCN reverts the database back to its state at the specified SCN.
TO BEFORE SCN reverts the database back to its state at the system change number
just preceding the specified SCN.

You can determine the current SCN by querying the CURRENT_SCN column of the
V$DATABASE view. This in turn lets you save the SCN to a spool file, for example, before
running a high-risk batch job.

TO TIMESTAMP Clause
Specify a valid datetime expression.
■
■

TO TIMESTAMP reverts the database back to its state at the specified timestamp.
TO BEFORE TIMESTAMP reverts the database back to its state one second before the
specified timestamp.

You can represent the timestamp as an offset from a determinate value, such as
SYSDATE, or as an absolute system timestamp.

TO RESTORE POINT Clause
Specify this clause to flash back the database to the specified restore point. If you have
not enabled flashback database, then this is the only clause you can specify in this
FLASHBACK DATABASE statement. If the database is not in FLASHBACK mode, as described

SQL Statements: DROP SEQUENCE to ROLLBACK 18-25

FLASHBACK DATABASE

in the "Prerequisites" section above, then this is the only clause you can specify for this
statement.

RESETLOGS
Specify TO BEFORE RESETLOGS to flash the database back to just before the last resetlogs
operation (ALTER DATABASE OPEN RESETLOGS).
See Also: Oracle Database Backup and Recovery User's Guide for more
information about this clause

Examples
18

Assuming that you have prepared a fast recovery area for the database and enabled
media recovery, enable database FLASHBACK mode and open the database with the
following statements:
STARTUP MOUNT
ALTER DATABASE FLASHBACK ON;
ALTER DATABASE OPEN;

With your database open for at least a day, you can flash back the database one day
with the following statements:
SHUTDOWN DATABASE
STARTUP MOUNT
FLASHBACK DATABASE TO TIMESTAMP SYSDATE-1;

18-26 Oracle Database SQL Language Reference

FLASHBACK TABLE

FLASHBACK TABLE
Purpose
18

Use the FLASHBACK TABLE statement to restore an earlier state of a table in the event of
human or application error. The time in the past to which the table can be flashed back
is dependent on the amount of undo data in the system. Also, Oracle Database cannot
restore a table to an earlier state across any DDL operations that change the structure
of the table.
Oracle strongly recommends that you run your database in
automatic undo mode by leaving the UNDO_MANAGEMENT initialization
parameter set to AUTO, which is the default. In addition, set the UNDO_
RETENTION initialization parameter to an interval large enough to
include the oldest data you anticipate needing. For more information
refer to the documentation on the UNDO_MANAGEMENT and UNDO_
RETENTION initialization parameters.

Note:

You cannot roll back a FLASHBACK TABLE statement. However, you can issue another
FLASHBACK TABLE statement and specify a time just prior to the current time. Therefore,
it is advisable to record the current SCN before issuing a FLASHBACK TABLE clause.
See Also:
■

■

■

FLASHBACK DATABASE on page 18-24 for information on
reverting the entire database to an earlier version
the flashback_query_clause of SELECT on page 19-17 for information
on retrieving past data from a table
Oracle Database Backup and Recovery User's Guide for additional
information on using the FLASHBACK TABLE statement

Prerequisites
18

To flash back a table to an earlier SCN or timestamp, you must have either the
FLASHBACK object privilege on the table or the FLASHBACK ANY TABLE system privilege. In
addition, you must have the SELECT, INSERT, DELETE, and ALTER object privileges on
the table.
Row movement must be enabled for all tables in the Flashback list unless you are
flashing back the table TO BEFORE DROP. That operation is called a flashback drop
operation, and it uses dropped data in the recycle bin rather than undo data. Refer to
row_movement_clause on page 10-41 for information on enabling row movement.
To flash back a table to a restore point, you must have the SELECT ANY DICTIONARY or
FLASHBACK ANY TABLE system privilege or the SELECT_CATALOG_ROLE role.
To flash back a table to before a DROP TABLE operation, you need only the privileges
necessary to drop the table.

SQL Statements: DROP SEQUENCE to ROLLBACK 18-27

FLASHBACK TABLE

Syntax
18

flashback_table::=
,
schema
FLASHBACK

.

TABLE

table

ENABLE

SCN

TRIGGERS

expr

DISABLE

TIMESTAMP
RESTORE

POINT

restore_point

TO

;
RENAME
BEFORE

TO

table

DROP

Semantics
18

During an Oracle Flashback Table operation, Oracle Database acquires exclusive DML
locks on all the tables specified in the Flashback list. These locks prevent any
operations on the tables while they are reverting to their earlier state.
The Flashback Table operation is executed in a single transaction, regardless of the
number of tables specified in the Flashback list. Either all of the tables revert to the
earlier state or none of them do. If the Flashback Table operation fails on any table,
then the entire statement fails.
At the completion of the Flashback Table operation, the data in table is consistent
with table at the earlier time. However, FLASHBACK TABLE TO SCN or TIMESTAMP does
not preserve rowids, and FLASHBACK TABLE TO BEFORE DROP does not recover referential
constraints.
Oracle Database does not revert statistics associated with table to their earlier form.
Indexes on table that exist currently are reverted and reflect the state of the table at
the Flashback point. If the index exists now but did not yet exist at the Flashback point,
then the database updates the index to reflect the state of the table at the Flashback
point. However, indexes that were dropped during the interval between the Flashback
point and the current time are not restored.

schema
Specify the schema containing the table. If you omit schema, then the database assumes
the table is in your own schema.

table
Specify the name of one or more tables containing data you want to revert to an earlier
version.
Restrictions on Flashing Back Tables This statement is subject to the following
restrictions:
■

Flashback Table operations are not valid for the following type objects: tables that
are part of a cluster, materialized views, Advanced Queuing (AQ) tables, static
data dictionary tables, system tables, remote tables, object tables, nested tables, or
individual table partitions or subpartitions.

18-28 Oracle Database SQL Language Reference

FLASHBACK TABLE

■

The following DDL operations change the structure of a table, so that you cannot
subsequently use the TO SCN or TO TIMESTAMP clause to flash the table back to a
time preceding the operation: upgrading, moving, or truncating a table; adding a
constraint to a table, adding a table to a cluster; modifying or dropping a column;
changing a column encryption key; adding, dropping, merging, splitting,
coalescing, or truncating a partition or subpartition (with the exception of adding
a range partition).

TO SCN Clause
Specify the system change number (SCN) corresponding to the point in time to which
you want to return the table. The expr must evaluate to a number representing a valid
SCN.

TO TIMESTAMP Clause
Specify a timestamp value corresponding to the point in time to which you want to
return the table. The expr must evaluate to a valid timestamp in the past. The table
will be flashed back to a time within approximately 3 seconds of the specified
timestamp.

TO RESTORE POINT Clause
Specify a restore point to which you want to flash back the table. The restore point
must already have been created.
CREATE RESTORE POINT on page 15-56 for information
on creating restore points

See Also:

ENABLE | DISABLE TRIGGERS
By default, Oracle Database disables all enabled triggers defined on table during the
Flashback Table operation and then reenables them after the Flashback Table operation
is complete. Specify ENABLE TRIGGERS if you want to override this default behavior and
keep the triggers enabled during the Flashback process.
This clause affects only those database triggers defined on table that are already
enabled. To enable currently disabled triggers selectively, use the ALTER TABLE ...
enable_disable_clause before you issue the FLASHBACK TABLE statement with the
ENABLE TRIGGERS clause.

TO BEFORE DROP Clause
Use this clause to retrieve from the recycle bin a table that has been dropped, along
with all possible dependent objects. The table must have resided in a locally managed
tablespace other than the SYSTEM tablespace.
See Also:
■

■

Oracle Database Administrator's Guide for information on the
recycle bin and naming conventions for objects in the recycle bin
PURGE on page 18-83 for information on removing objects
permanently from the recycle bin

You can specify either the original user-specified name of the table or the
system-generated name Oracle Database assigned to the object when it was dropped.
■

System-generated recycle bin object names are unique. Therefore, if you specify
the system-generated name, then the database retrieves that specified object.

SQL Statements: DROP SEQUENCE to ROLLBACK 18-29

FLASHBACK TABLE

To see the contents of your recycle bin, query the USER_RECYCLEBIN data dictionary
view. You can use the RECYCLEBIN synonym instead. The following two statements
return the same rows:
SELECT * FROM RECYCLEBIN;
SELECT * FROM USER_RECYCLEBIN;
■

If you specify the user-specified name, and if the recycle bin contains more than
one object of that name, then the database retrieves the object that was moved to
the recycle bin most recently. If you want to retrieve an older version of the table,
then do one of these things:
–

Specify the system-generated recycle bin name of the table you want to
retrieve.

–

Issue additional FLASHBACK TABLE ... TO BEFORE DROP statements until you
retrieve the table you want.

Oracle Database attempts to preserve the original table name. If a new table of the
same name has been created in the same schema since the original table was dropped,
then the database returns an error unless you also specify the RENAME TO clause.
RENAME TO Clause Use this clause to specify a new name for the table being
retrieved from the recycle bin.
Notes on Flashing Back Dropped Tables The following notes apply to flashing back
dropped tables:
■

■

Oracle Database retrieves all indexes defined on the table retrieved from the
recycle bin except for bitmap join indexes and domain indexes. (Bitmap join
indexes and domain indexes are not put in the recycle bin during a DROP TABLE
operation, so cannot be retrieved.)
The database also retrieves all triggers and constraints defined on the table except
for referential integrity constraints that reference other tables.
The retrieved indexes, triggers, and constraints have recycle bin names. Therefore
it is advisable to query the USER_RECYCLEBIN view before issuing a FLASHBACK
TABLE ... TO BEFORE DROP statement so that you can rename the retrieved triggers
and constraints to more usable names.

■

■

■

When you drop a table, all materialized view logs defined on the table are also
dropped but are not placed in the recycle bin. Therefore, the materialized view
logs cannot be flashed back along with the table.
When you drop a table, any indexes on the table are dropped and put into the
recycle bin along with the table. If subsequent space pressures arise, then the
database reclaims space from the recycle bin by first purging indexes. In this case,
when you flash back the table, you may not get back all of the indexes that were
defined on the table.
You cannot flash back a table if it has been purged, either by a user or by Oracle
Database as a result of some space reclamation operation.

Examples
18

Restoring a Table to an Earlier State: Examples The examples below create a new
table, employees_test, with row movement enabled, update values within the new
table, and issue the FLASHBACK TABLE statement.

18-30 Oracle Database SQL Language Reference

FLASHBACK TABLE

Create table employees_test, with row movement enabled, from table employees of
the sample hr schema:
CREATE TABLE employees_test
AS SELECT * FROM employees;

As a benchmark, list those salaries less than 2500:
SELECT salary
FROM employees_test
WHERE salary < 2500;
SALARY
---------2400
2200
2100
2400
2200

To allow time for the SCN to propagate to the mapping table
used by the FLASHBACK TABLE statement, wait a minimum of 5 minutes
prior to issuing the following statement. This wait would not be
necessary if a previously existing table were being used in this
example.
Note:

Enable row movement for the table:
ALTER TABLE employees_test
ENABLE ROW MOVEMENT;

Issue a 10% salary increase to those employees earning less than 2500:
UPDATE employees_test
SET salary = salary * 1.1
WHERE salary < 2500;
5 rows updated.
COMMIT;

As a second benchmark, list those salaries that remain less than 2500 following the
10% increase:
SELECT salary
FROM employees_test
WHERE salary < 2500;
SALARY
---------2420
2310
2420

Restore the table employees_test to its state prior to the current system time. The
unrealistic duration of 1 minute is used so that you can test this series of examples
quickly. Under normal circumstances a much greater interval would have elapsed.
FLASHBACK TABLE employees_test
TO TIMESTAMP (SYSTIMESTAMP - INTERVAL '1' minute);

SQL Statements: DROP SEQUENCE to ROLLBACK 18-31

FLASHBACK TABLE

List those salaries less than 2500. After the FLASHBACK TABLE statement issued above,
this list should match the list in the first benchmark.
SELECT salary
FROM employees_test
WHERE salary < 2500;
SALARY
---------2400
2200
2100
2400
2200

Retrieving a Dropped Table: Example If you accidentally drop the pm.print_media
table and want to retrieve it, then issue the following statement:
FLASHBACK TABLE print_media TO BEFORE DROP;

If another print_media table has been created in the pm schema, then use the RENAME TO
clause to rename the retrieved table:
FLASHBACK TABLE print_media TO BEFORE DROP RENAME TO print_media_old;

If you know that the employees table has been dropped multiple times, and you want
to retrieve the oldest version, then query the USER_RECYLEBIN table to determine the
system-generated name, and then use that name in the FLASHBACK TABLE statement.
(System-generated names in your database will differ from those shown here.)
SELECT object_name, droptime FROM user_recyclebin
WHERE original_name = 'PRINT_MEDIA';
OBJECT_NAME
-----------------------------RB$$45703$TABLE$0
RB$$45704$TABLE$0
RB$$45705$TABLE$0

18-32 Oracle Database SQL Language Reference

DROPTIME
------------------2003-06-03:15:26:39
2003-06-12:12:27:27
2003-07-08:09:28:01

GRANT

GRANT
Purpose
18

Use the GRANT statement to grant:
■

■

■

Roles to users and roles. The granted roles can be either user-defined (local or
external) or predefined. For a list of predefined roles, refer to Oracle Database
Security Guide.
System privileges to users and roles. Table 18–1 lists the system privileges
(organized by the database object operated upon).
Object privileges for a particular object to users, roles, and PUBLIC. Table 18–2 lists
object privileges (organized by the database object operated upon).
Note: Global roles (created with IDENTIFIED GLOBALLY) are granted
through enterprise roles and cannot be granted using the GRANT
statement.

Notes on Authorizing Database Users You can authorize database users through

means other than the database and the GRANT statement.
■

■

Many Oracle Database privileges are granted through supplied PL/SQL and Java
packages. For information on those privileges, refer to the documentation for the
appropriate package.
Some operating systems have facilities that let you grant roles to Oracle Database
users with the initialization parameter OS_ROLES. If you choose to grant roles to
users through operating system facilities, then you cannot also grant roles to users
with the GRANT statement, although you can use the GRANT statement to grant
system privileges to users and system privileges and roles to other roles.

Note on Oracle Automatic Storage Management A user authenticated AS SYSASM can

use this statement to grant the system privileges SYSASM, SYSOPER, and SYSDBA to a user
in the Oracle ASM password file of the current node.
Note on Editionable Objects A GRANT operation to grant object privileges on an
editionable object actualizes the object in the current edition. See Oracle Database
Advanced Application Developer's Guide for more information about editions and
editionable objects.
See Also:
■

■

■

CREATE USER on page 17-7 and CREATE ROLE on page 15-59
for definitions of local, global, and external privileges
Oracle Database Security Guide for information about other
authorization methods and for information about privileges
REVOKE on page 18-87 for information on revoking grants

Prerequisites
18

To grant a system privilege, one of the following conditions must be met:
■

You must have been granted the GRANT ANY PRIVILEGE system privilege. In this
case, if you grant the system privilege to a role, then a user to whom the role has

SQL Statements: DROP SEQUENCE to ROLLBACK 18-33

GRANT

been granted does not have the privilege unless the role is enabled in user's
session.
■

You must have been granted the system privilege with the ADMIN OPTION. In this
case, if you grant the system privilege to a role, then a user to whom the role has
been granted has the privilege regardless whether the role is enabled in the user's
session.

To grant a role, you must either have been granted the role with the ADMIN OPTION or
have been granted the GRANT ANY ROLE system privilege, or you must have created the
role.
To grant an object privilege, you must own the object, or the owner of the object must
have granted you the object privileges with the GRANT OPTION, or you must have been
granted the GRANT ANY OBJECT PRIVILEGE system privilege. If you have the GRANT ANY
OBJECT PRIVILEGE, then you can grant the object privilege only if the object owner
could have granted the same object privilege. In this case, the GRANTOR column of the
DBA_TAB_PRIVS view displays the object owner rather than the user who issued the
GRANT statement.

Syntax
18

grant::=
grant_system_privileges
GRANT

;
grant_object_privileges

(grant_system_privileges::= on page 18-34, grant_object_privileges::= on page 18-34)
grant_system_privileges::=
,
system_privilege

WITH

role
ALL

TO

ADMIN

OPTION

grantee_clause

PRIVILEGES

(grantee_clause::= on page 18-35)
grant_object_privileges::=
,
,
object_privilege

(

column

)
on_object_clause

PRIVILEGES
ALL

WITH
TO

HIERARCHY

OPTION

WITH

GRANT

OPTION

grantee_clause

(on_object_clause::= on page 18-35, grantee_clause::= on page 18-35)

18-34 Oracle Database SQL Language Reference

GRANT

on_object_clause::=
schema

.
object

DIRECTORY
EDITION

directory_name
edition_name

ON

schema
MINING

.

MODEL

mining_model_name
schema

SOURCE

.

JAVA

object
RESOURCE

grantee_clause::=
,
IDENTIFIED

BY

password

user
role
PUBLIC

Semantics
18

grant_system_privileges
Use these clauses to grant system privileges.
system_privilege
Specify the system privilege you want to grant. Table 18–1 lists the system privileges,
organized by the database object operated upon.
■

■

If you grant a privilege to a user, then the database adds the privilege to the user's
privilege domain. The user can immediately exercise the privilege. Oracle
recommends that you only grant the ANY privileges to trusted users.
If you grant a privilege to a role, then the database adds the privilege to the
privilege domain of the role. Users who have been granted and have enabled the
role can immediately exercise the privilege. Other users who have been granted
the role can enable the role and exercise the privilege.
Granting a System Privilege to a User: Example on
page 18-52 and "Granting System Privileges to a Role: Example" on
page 18-52

See Also:

■

If you grant a privilege to PUBLIC, then the database adds the privilege to the
privilege domains of each user. All users can immediately perform operations
authorized by the privilege. Oracle recommends against granting system
privileges to PUBLIC.

Oracle Database provides the ALL PRIVILEGES shortcut for granting all the system
privileges listed in Table 18–1 on page 18-40, except the SELECT ANY DICTIONARY, ALTER
DATABASE LINK, and ALTER PUBLIC DATABASE LINK privileges.

SQL Statements: DROP SEQUENCE to ROLLBACK 18-35

GRANT

role
Specify the role you want to grant. You can grant an Oracle Database predefined role
or a user-defined role.
■

If you grant a role to a user, then the database makes the role available to the user.
The user can immediately enable the role and exercise the privileges in the
privilege domain of the role.
In the case of a secure application role, you need not grant such a role directly to
the user. You can let the associated PL/SQL package do this, assuming the user
passes appropriate security policies. For more information, see the CREATE ROLE
semantics for USING package on page 15-60 and Oracle Database Security Guide

■

■

If you grant a role to another role, then the database adds the privilege domain of
the granted role to the privilege domain of the grantee role. Users who have been
granted the grantee role can enable it and exercise the privileges in the granted
role's privilege domain.
If you grant a role to PUBLIC, then the database makes the role available to all
users. All users can immediately enable the role and exercise the privileges in the
privilege domain of the role.
See Also:
■

■
■

Oracle Database Security Guide for information on the Oracle
predefined roles
"Granting a Role to a Role: Example" on page 18-52
CREATE ROLE on page 15-59 for information on creating a
user-defined role

IDENTIFIED BY Clause
This clause is valid only when granting system privileges, not when granting object
privileges. Use the IDENTIFIED BY clause to specify a new password for an existing
user or to create a nonexistent user. This clause is not valid if the grantee is a role or
PUBLIC. If the user specified in the grantee_clause does not exist, then the database
creates the user with the password and with the privileges and roles specified in this
clause.
See Also: CREATE USER on page 17-7 for restrictions on usernames
and passwords

WITH ADMIN OPTION
Specify WITH ADMIN OPTION to enable the grantee to:
■

Grant the privilege or role to another user or role, unless the role is a GLOBAL role

■

Revoke the privilege or role from another user or role

■

Alter the privilege or role to change the authorization needed to access it

■

Drop the privilege or role

If you grant a system privilege or role to a user without specifying WITH ADMIN OPTION,
and then subsequently grant the privilege or role to the user WITH ADMIN OPTION, then
the user has the ADMIN OPTION on the privilege or role.
To revoke the ADMIN OPTION on a system privilege or role from a user, you must revoke
the privilege or role from the user altogether and then grant the privilege or role to the
user without the ADMIN OPTION.
18-36 Oracle Database SQL Language Reference

GRANT

See Also: "Granting a Role with the Admin Option: Example" on
page 18-52

grantee_clause
TO grantee_clause identifies users or roles to which the system privilege, role, or
object privilege is granted.
Restriction on Grantees A user, role, or PUBLIC cannot appear more than once in the

TO grantee_clause.
PUBLIC Specify PUBLIC to grant the privileges to all users. Oracle recommends
against granting system privileges to PUBLIC.
Restrictions on Granting System Privileges and Roles Privileges and roles are
subject to the following restrictions:
■

A privilege or role cannot appear more than once in the list of privileges and roles
to be granted.

■

You cannot grant a role to itself.

■

You cannot grant a role IDENTIFIED GLOBALLY to anything.

■

You cannot grant a role IDENTIFIED EXTERNALLY to a global user or global role.

■

■

You cannot grant roles circularly. For example, if you grant the role banker to the
role teller, then you cannot subsequently grant teller to banker.
You cannot grant an IDENTIFIED BY role, IDENTIFIED USING role, or IDENTIFIED
EXTERNALLY role to another role.

grant_object_privileges
Use these clauses to grant object privileges.
object_privilege
Specify the object privilege you want to grant. Table 18–2 lists the object privileges,
organized by the type of object on which they can be granted. When you grant an
object privilege on a editionable object, either to a user or to a role, the object is
actualized in the edition in which the grant is made. Refer to CREATE EDITION on
page 14-51 for information on editionable object types and editions.
To grant SELECT on a view to another user, either you must
own all of the objects underlying the view or you must have been
granted the SELECT object privilege WITH GRANT OPTION on all of those
underlying objects. This is true even if the grantee already has SELECT
privileges on those underlying objects.

Note:

Restriction on Object Privileges A privilege cannot appear more than once in the list
of privileges to be granted.

ALL [PRIVILEGES]
Specify ALL to grant all the privileges for the object that you have been granted with
the GRANT OPTION. The user who owns the schema containing an object automatically
has all privileges on the object with the GRANT OPTION. The keyword PRIVILEGES is
provided for semantic clarity and is optional.

SQL Statements: DROP SEQUENCE to ROLLBACK 18-37

GRANT

column
Specify the table or view column on which privileges are to be granted. You can
specify columns only when granting the INSERT, REFERENCES, or UPDATE privilege. If
you do not list columns, then the grantee has the specified privilege on all columns in
the table or view.
For information on existing column object grants, query the USER_, ALL_, or DBA_COL_
PRIVS data dictionary view.
Oracle Database Reference for information on the data
dictionary views and "Granting Multiple Object Privileges on
Individual Columns: Example" on page 18-53

See Also:

on_object_clause
The on_object_clause identifies the object on which the privileges are granted.
Directory schema objects, editions, data mining models, and Java source and resource
schema objects are identified separately because they reside in separate namespaces.
If you can make this grant only because you have the GRANT ANY OBJECT PRIVILEGE
system privilege—that is, you are not the owner of object, nor do you have object_
privilege on object WITH GRANT OPTION—then the effect of this grant is that you are
acting on behalf of the object owner. The *_TAB_PRIVS data dictionary views will
reflect that this grant was made by the owner of object.
See Also:
■
■

"Granting Object Privileges to a Role: Example" on page 18-52
"Revoke Operations that Use GRANT ANY OBJECT PRIVILEGE:
Example" on page 18-94 for more information on using the GRANT
ANY OBJECT PRIVILEGE system privilege for revoke operations

WITH GRANT OPTION
Specify WITH GRANT OPTION to enable the grantee to grant the object privileges to other
users and roles.
If you grant an object privilege to a user without specifying WITH GRANT OPTION, and
then subsequently grant the privilege to the user WITH GRANT OPTION, then the user has
the GRANT OPTION on the privilege.
To revoke the GRANT OPTION on an object privilege from a user, you must revoke the
privilege from the user altogether and then grant the privilege to the user without the
GRANT OPTION.
You can specify WITH GRANT OPTION
only when granting to a user or to PUBLIC, not when granting to a role.

Restriction on Granting WITH GRANT OPTION

WITH HIERARCHY OPTION
Specify WITH HIERARCHY OPTION to grant the specified object privilege on all subobjects
of object, such as subviews created under a view, including subobjects created
subsequent to this statement.
This clause is meaningful only in combination with the SELECT object privilege.
Specify the schema object on which the privileges are to be granted. If you do
not qualify object with schema, then the database assumes the object is in your own
schema. The object can be one of the following types:

object

18-38 Oracle Database SQL Language Reference

GRANT

■

Table, view, or materialized view

■

Sequence

■

Procedure, function, or package

■

User-defined type

■

Synonym for any of the preceding items

■

Directory, library, operator, or indextype

■

Java source, class, or resource

You cannot grant privileges directly to a single partition of a partitioned table.
See Also: "Granting Object Privileges on a Table to a User: Example"
on page 18-53, "Granting Object Privileges on a View: Example" on
page 18-53, and "Granting Object Privileges to a Sequence in Another
Schema: Example" on page 18-53
DIRECTORY directory_name Specify a directory schema object on which privileges
are to be granted. You cannot qualify directory_name with a schema name.
See Also: CREATE DIRECTORY on page 14-41 and "Granting an
Object Privilege on a Directory: Example" on page 18-52

The JAVA clause lets you specify a Java source or
resource schema object on which privileges are to be granted.

JAVA SOURCE | RESOURCE

See Also:

CREATE JAVA on page 14-91

Listings of System and Object Privileges
18

When you grant a privilege on ANY object, such as CREATE ANY
CLUSTER, the result is determined by the value of the O7_DICTIONARY_
ACCESSIBILITY initialization parameter. By default, this parameter is
set to FALSE, so that ANY privileges give the grantee access to that type
of object in all schemas except the SYS schema. If you set O7_
DICTIONARY_ACCESSIBILITY to TRUE, then the ANY privileges also give
the grantee access, in the SYS schema, to all objects except Oracle
Scheduler objects. For security reasons, Oracle recommends that you
use this setting only with great caution.
Note:

SQL Statements: DROP SEQUENCE to ROLLBACK 18-39

GRANT

Table 18–1

System Privileges (Organized by the Database Object Operated Upon)

System Privilege Name

Operations Authorized

Advisor Framework Privileges: All of the
advisor framework privileges are part of the
DBA role.

—

ADVISOR

Access the advisor framework through PL/SQL packages such as
DBMS_ADVISOR and DBMS_SQLTUNE.
Refer to Oracle Database PL/SQL Packages and Types Reference for
information on these packages.

ADMINISTER SQL TUNING SET

Create, drop, select (read), load (write), and delete a SQL tuning
set owned by the grantee through the DBMS_SQLTUNE package.

ADMINISTER ANY SQL TUNING SET

Create, drop, select (read), load (write), and delete a SQL tuning
set owned by any user through the DBMS_SQLTUNE package.

CREATE ANY SQL PROFILE

Accept a SQL Profile recommended by the SQL Tuning Advisor,
which is accessed through Enterprise Manager or by the DBMS_
SQLTUNE package.
Note: This privilege has been deprecated in favor of ADMINISTER
SQL MANAGEMENT OBJECT.

ALTER ANY SQL PROFILE

Alter the attributes of an existing SQL Profile.
Note: This privilege has been deprecated in favor of ADMINISTER
SQL MANAGEMENT OBJECT.

DROP ANY SQL PROFILE

Drop an existing SQL Profile.
Note: This privilege has been deprecated in favor of ADMINISTER
SQL MANAGEMENT OBJECT.

ADMINISTER SQL MANAGEMENT OBJECT

Create, alter, and drop a SQL Profile owned by any user through
the DBMS_SQLTUNE package.

CLUSTERS:

—

CREATE CLUSTER

Create clusters in the grantee's schema.

CREATE ANY CLUSTER

Create a cluster in any schema. Behaves similarly to CREATE ANY
TABLE.

ALTER ANY CLUSTER

Alter clusters in any schema.

DROP ANY CLUSTER

Drop clusters in any schema.

CONTEXTS:

—

CREATE ANY CONTEXT

Create any context namespace.

DROP ANY CONTEXT

Drop any context namespace.

DATA REDACTION:

—

EXEMPT REDACTION POLICY

Bypass any existing Oracle Data Redaction policies and view
actual data from tables or views on which Data Redaction policies
are defined.
Note: This privilege is available starting with Oracle Database 11g
Release 2 (11.2.0.4).

DATABASE:

—

ALTER DATABASE

Alter the database.

ALTER SYSTEM

Issue ALTER SYSTEM statements.

AUDIT SYSTEM

Issue AUDIT statements.

18-40 Oracle Database SQL Language Reference

GRANT

Table 18–1 (Cont.) System Privileges (Organized by the Database Object Operated Upon)
System Privilege Name

Operations Authorized

DATABASE LINKS:

—

CREATE DATABASE LINK

Create private database links in the grantee's schema.

CREATE PUBLIC DATABASE LINK

Create public database links.

ALTER DATABASE LINK

Modify a fixed-user database link when the password of the
connection or authentication user changes.

ALTER PUBLIC DATABASE LINK

Modify a public fixed-user database link when the password of
the connection or authentication user changes.

DROP PUBLIC DATABASE LINK

Drop public database links.

DEBUGGING:

—

DEBUG CONNECT SESSION

Connect the current session to a debugger.

DEBUG ANY PROCEDURE

Debug all PL/SQL and Java code in any database object. Display
information on all SQL statements executed by the application.
Note: Granting this privilege is equivalent to granting the DEBUG
object privilege on all applicable objects in the database.

DICTIONARIES:

—

ANALYZE ANY DICTIONARY

Analyze any data dictionary object.

DIMENSIONS:

—

CREATE DIMENSION

Create dimensions in the grantee's schema.

CREATE ANY DIMENSION

Create dimensions in any schema.

ALTER ANY DIMENSION

Alter dimensions in any schema.

DROP ANY DIMENSION

Drop dimensions in any schema.

DIRECTORIES:

—

CREATE ANY DIRECTORY

Create directory database objects.

DROP ANY DIRECTORY

Drop directory database objects.

EDITIONS:

—

CREATE ANY EDITION

Create editions.

DROP ANY EDITION

Drop editions.

FLASHBACK DATA ARCHIVES:

—

FLASHBACK ARCHIVE ADMINISTER

Create, alter, or drop any flashback data archive.

INDEXES:

—

CREATE ANY INDEX

Create in any schema a domain index or an index on any table in
any schema.

ALTER ANY INDEX

Alter indexes in any schema.

DROP ANY INDEX

Drop indexes in any schema.

INDEXTYPES:

—

CREATE INDEXTYPE

Create an indextype in the grantee's schema.

CREATE ANY INDEXTYPE

Create an indextype in any schema and create a comment on an
indextype in any schema.

ALTER ANY INDEXTYPE

Modify indextypes in any schema.

DROP ANY INDEXTYPE

Drop an indextype in any schema.

SQL Statements: DROP SEQUENCE to ROLLBACK 18-41

GRANT

Table 18–1 (Cont.) System Privileges (Organized by the Database Object Operated Upon)
System Privilege Name

Operations Authorized

EXECUTE ANY INDEXTYPE

Reference an indextype in any schema.

JOB SCHEDULER OBJECTS:

The following privileges are needed to execute procedures in the
DBMS_SCHEDULER package. This privileges do not apply to
lightweight jobs, which are not database objects. Refer to Oracle
Database Administrator's Guide for more information about
lightweight jobs.

CREATE JOB

Create jobs, schedules, or programs in the grantee's schema.

CREATE ANY JOB

Create, alter, or drop jobs, chains, schedules, programs, or
credentials in any schema except SYS.
Caution: This extremely powerful privilege allows the grantee to
execute code as any other user. It should be granted with caution.

CREATE EXTERNAL JOB

Create in the grantee's schema an executable scheduler job that
runs on the operating system.

EXECUTE ANY PROGRAM

Use any program in a job in the grantee's schema.

EXECUTE ANY CLASS

Specify any job class in a job in the grantee's schema.

MANAGE SCHEDULER

Create, alter, or drop any job class, window, or window group.

LIBRARIES:

Caution: CREATE LIBARARY, CREATE ANY LIBRARY, ALTER ANY
LIBRARY, and EXECUTE ANY LIBRARY are extremely powerful
privileges that should be granted only to trusted users. Refer to
Oracle Database Security Guide before granting these privileges.

CREATE LIBRARY

Create external procedure or function libraries in the grantee's
schema.

CREATE ANY LIBRARY

Create external procedure or function libraries in any schema.

ALTER ANY LIBRARY

Alter external procedure or function libraries in any schema.

DROP ANY LIBRARY

Drop external procedure or function libraries in any schema.

EXECUTE ANY LIBRARY

Use external procedure or function libraries in any schema.

MATERIALIZED VIEWS:

—

CREATE MATERIALIZED VIEW

Create a materialized view in the grantee's schema.

CREATE ANY MATERIALIZED VIEW

Create materialized views in any schema.

ALTER ANY MATERIALIZED VIEW

Alter materialized views in any schema.

DROP ANY MATERIALIZED VIEW

Drop materialized views in any schema.

QUERY REWRITE

This privilege has been deprecated. No privileges are needed for a
user to enable rewrite for a materialized view that references
tables or views in the user's own schema.

GLOBAL QUERY REWRITE

Enable rewrite using a materialized view when that materialized
view references tables or views in any schema.

ON COMMIT REFRESH

Create a refresh-on-commit materialized view on any table in the
database.
Alter a refresh-on-demand materialized on any table in the
database to refresh-on-commit.

FLASHBACK ANY TABLE

Issue a SQL Flashback Query on any table, view, or materialized
view in any schema. This privilege is not needed to execute the
DBMS_FLASHBACK procedures.

MINING MODELS:

—

18-42 Oracle Database SQL Language Reference

GRANT

Table 18–1 (Cont.) System Privileges (Organized by the Database Object Operated Upon)
System Privilege Name

Operations Authorized

CREATE MINING MODEL

Create mining models in the grantee's schema using the DBMS_
DATA_MINING.CREATE_MODEL procedure.

CREATE ANY MINING MODEL

Create mining models in any schema using the DBMS_DATA_
MINING.CREATE_MODEL procedure.

ALTER ANY MINING MODEL

Change the mining model name or the associated cost matrix of
any model in any schema by using the applicable DBMS_DATA_
MINING procedures.

DROP ANY MINING MODEL

Drop any mining model in any schema by using the DBMS_DATA_
MINING.DROP_MODEL procedure.

SELECT ANY MINING MODEL

Score or view any model in any schema. Scoring is done either
with the PREDICTION family of SQL functions or with the DBMS_
DATA_MINING.APPLY procedure. Viewing the model is done with
the DBMS_DATA_MINING.GET_MODEL_DETAILS_* procedures.

COMMENT ANY MINING MODEL

Create a comment on any model in any schema using the SQL
COMMENT statement.

OLAP CUBES:

The following privileges are valid when you are using Oracle
Database with the OLAP option.

CREATE CUBE

Create an OLAP cube in the grantee's schema.

CREATE ANY CUBE

Create an OLAP cube in any schema.

ALTER ANY CUBE

Alter an OLAP cube in any schema.

DROP ANY CUBE

Drop any OLAP cube in any schema.

SELECT ANY CUBE

Query or view any OLAP cube in any schema.

UPDATE ANY CUBE

Update any cube in any schema.

OLAP CUBE MEASURE FOLDERS:

The following privileges are valid when you are using Oracle
Database with the OLAP option.

CREATE MEASURE FOLDER

Create an OLAP measure folder in the grantee's schema.

CREATE ANY MEASURE FOLDER

Create an OLAP measure folder in any schema.

DELETE ANY MEASURE FOLDER

Delete from any OLAP measure folder in any schema.

DROP ANY MEASURE FOLDER

Drop any measure folder in any schema.

INSERT ANY MEASURE FOLDER

Insert a measure into any measure folder in any schema.

OLAP CUBE DIMENSIONS:

The following privileges are valid when you are using Oracle
Database with the OLAP option.

CREATE CUBE DIMENSION

Create an OLAP cube dimension in the grantee's schema.

CREATE ANY CUBE DIMENSION

Create an OLAP cube dimension in any schema.

ALTER ANY CUBE DIMENSION

Alter an OLAP cube dimension in any schema.

DELETE ANY CUBE DIMENSION

Delete from an OLAP cube dimension in any schema.

DROP ANY CUBE DIMENSION

Drop an OLAP cube dimension in any schema.

INSERT ANY CUBE DIMENSION

Insert into an OLAP cube dimension in any schema.

SELECT ANY CUBE DIMENSION

View or query an OLAP cube dimension in any schema.

UPDATE ANY CUBE DIMENSION

Update an OLAP cube dimension in any schema.

OLAP CUBE BUILD PROCESSES:

—

CREATE CUBE BUILD PROCESS

Create an OLAP cube build process in the grantee's schema.

SQL Statements: DROP SEQUENCE to ROLLBACK 18-43

GRANT

Table 18–1 (Cont.) System Privileges (Organized by the Database Object Operated Upon)
System Privilege Name

Operations Authorized

CREATE ANY CUBE BUILD PROCESS

Create an OLAP cube build process in any schema.

DROP ANY CUBE BUILD PROCESS

Drop an OLAP cube build process in any schema.

UPDATE ANY CUBE BUILD PROCESS

Update an OLAP cube build process in any schema.

OPERATORS:

—

CREATE OPERATOR

Create an operator and its bindings in the grantee's schema.

CREATE ANY OPERATOR

Create an operator and its bindings in any schema and create a
comment on an operator in any schema.

ALTER ANY OPERATOR

Modify an operator in any schema.

DROP ANY OPERATOR

Drop an operator in any schema.

EXECUTE ANY OPERATOR

Reference an operator in any schema.

OUTLINES:

—

CREATE ANY OUTLINE

Create public outlines that can be used in any schema that uses
outlines.

ALTER ANY OUTLINE

Modify outlines.

DROP ANY OUTLINE

Drop outlines.

PLAN MANAGEMENT:

—

ADMINISTER SQL MANAGEMENT OBJECT

Perform controlled manipulation of plan history and SQL plan
baselines maintained for various SQL statements.

PROCEDURES:

—

CREATE PROCEDURE

Create stored procedures, functions, and packages in the grantee's
schema.

CREATE ANY PROCEDURE

Create stored procedures, functions, and packages in any schema.

ALTER ANY PROCEDURE

Alter stored procedures, functions, or packages in any schema.

DROP ANY PROCEDURE

Drop stored procedures, functions, or packages in any schema.

EXECUTE ANY PROCEDURE

Execute procedures or functions, either standalone or packaged.
Reference public package variables in any schema.

PROFILES:

—

CREATE PROFILE

Create profiles.

ALTER PROFILE

Alter profiles.

DROP PROFILE

Drop profiles.

ROLES:

—

CREATE ROLE

Create roles.

ALTER ANY ROLE

Alter any role in the database.

DROP ANY ROLE

Drop roles.

GRANT ANY ROLE

Grant any role in the database.

ROLLBACK SEGMENTS:

—

CREATE ROLLBACK SEGMENT

Create rollback segments.

ALTER ROLLBACK SEGMENT

Alter rollback segments.

DROP ROLLBACK SEGMENT

Drop rollback segments.

18-44 Oracle Database SQL Language Reference

GRANT

Table 18–1 (Cont.) System Privileges (Organized by the Database Object Operated Upon)
System Privilege Name

Operations Authorized

SEQUENCES:

—

CREATE SEQUENCE

Create sequences in the grantee's schema.

CREATE ANY SEQUENCE

Create sequences in any schema.

ALTER ANY SEQUENCE

Alter any sequence in the database.

DROP ANY SEQUENCE

Drop sequences in any schema.

SELECT ANY SEQUENCE

Reference sequences in any schema.

SESSIONS:

—

CREATE SESSION

Connect to the database.

ALTER RESOURCE COST

Set costs for session resources.

ALTER SESSION

Enable and disable the SQL trace facility.

RESTRICTED SESSION

Logon after the instance is started using the SQL*Plus STARTUP
RESTRICT statement.

SNAPSHOTS:

See MATERIALIZED VIEWS

SYNONYMS:

Caution: CREATE PUBLIC SYNONYM and DROP PUBLIC SYNONYM are
extremely powerful privileges that should be granted only to
trusted users. Refer to Oracle Database Security Guide before
granting these privileges.

CREATE SYNONYM

Create synonyms in the grantee's schema.

CREATE ANY SYNONYM

Create private synonyms in any schema.

CREATE PUBLIC SYNONYM

Create public synonyms.

DROP ANY SYNONYM

Drop private synonyms in any schema.

DROP PUBLIC SYNONYM

Drop public synonyms.

TABLES:

Note: For external tables, the only valid privileges are CREATE ANY
TABLE, ALTER ANY TABLE, DROP ANY TABLE, and SELECT ANY TABLE.

CREATE TABLE

Create a table in the grantee's schema.

CREATE ANY TABLE

Create a table in any schema. The owner of the schema containing
the table must have space quota on the tablespace to contain the
table.

ALTER ANY TABLE

Alter any table or view in any schema.

BACKUP ANY TABLE

Use the Export utility to incrementally export objects from the
schema of other users.

DELETE ANY TABLE

Delete rows from tables, table partitions, or views in any schema.

DROP ANY TABLE

Drop or truncate tables or table partitions in any schema.

INSERT ANY TABLE

Insert rows into tables and views in any schema.

LOCK ANY TABLE

Lock tables and views in any schema.

SELECT ANY TABLE

Query tables, views, or materialized views in any schema.

FLASHBACK ANY TABLE

Issue a SQL Flashback Query on any table, view, or materialized
view in any schema. This privilege is not needed to execute the
DBMS_FLASHBACK procedures.

UPDATE ANY TABLE

Update rows in tables and views in any schema.

SQL Statements: DROP SEQUENCE to ROLLBACK 18-45

GRANT

Table 18–1 (Cont.) System Privileges (Organized by the Database Object Operated Upon)
System Privilege Name

Operations Authorized

TABLESPACES:

—

CREATE TABLESPACE

Create tablespaces.

ALTER TABLESPACE

Alter tablespaces.

DROP TABLESPACE

Drop tablespaces.

MANAGE TABLESPACE

Take tablespaces offline and online and begin and end tablespace
backups.

UNLIMITED TABLESPACE

Use an unlimited amount of any tablespace. This privilege
overrides any specific quotas assigned. If you revoke this
privilege from a user, then the user's schema objects remain but
further tablespace allocation is denied unless authorized by
specific tablespace quotas. You cannot grant this system privilege
to roles.

TRIGGERS:

—

CREATE TRIGGER

Create a database trigger in the grantee's schema.

CREATE ANY TRIGGER

Create database triggers in any schema.

ALTER ANY TRIGGER

Enable, disable, or compile database triggers in any schema.

DROP ANY TRIGGER

Drop database triggers in any schema.

ADMINISTER DATABASE TRIGGER

Create a trigger on DATABASE. You must also have the CREATE
TRIGGER or CREATE ANY TRIGGER system privilege.

TYPES:

—

CREATE TYPE

Create object types and object type bodies in the grantee's schema.

CREATE ANY TYPE

Create object types and object type bodies in any schema.

ALTER ANY TYPE

Alter object types in any schema.

DROP ANY TYPE

Drop object types and object type bodies in any schema.

EXECUTE ANY TYPE

Use and reference object types and collection types in any schema,
and invoke methods of an object type in any schema if you make
the grant to a specific user. If you grant EXECUTE ANY TYPE to a role,
then users holding the enabled role will not be able to invoke
methods of an object type in any schema.

UNDER ANY TYPE

Create subtypes under any nonfinal object types.

USERS:

—

CREATE USER

Create users. This privilege also allows the creator to:

ALTER USER

DROP USER

■

Assign quotas on any tablespace.

■

Set default and temporary tablespaces.

■

Assign a profile as part of a CREATE USER statement.

Alter any user. This privilege authorizes the grantee to:
■

Change another user's password or authentication method.

■

Assign quotas on any tablespace.

■

Set default and temporary tablespaces.

■

Assign a profile and default roles.

Drop users

18-46 Oracle Database SQL Language Reference

GRANT

Table 18–1 (Cont.) System Privileges (Organized by the Database Object Operated Upon)
System Privilege Name

Operations Authorized

VIEWS:

—

CREATE VIEW

Create views in the grantee's schema.

CREATE ANY VIEW

Create views in any schema.

DROP ANY VIEW

Drop views in any schema.

UNDER ANY VIEW

Create subviews under any object views.

FLASHBACK ANY TABLE

Issue a SQL Flashback Query on any table, view, or materialized
view in any schema. This privilege is not needed to execute the
DBMS_FLASHBACK procedures.

MERGE ANY VIEW

If a user has been granted the MERGE ANY VIEW privilege, then for
any query issued by that user, the optimizer can use view merging
to improve query performance without performing the checks
that would otherwise be performed to ensure that view merging
does not violate any security intentions of the view creator. See
also Oracle Database Reference for information on the OPTIMIZER_
SECURE_VIEW_MERGING parameter and Oracle Database Performance
Tuning Guide for information on view merging.

MISCELLANEOUS:

—

ANALYZE ANY

Analyze any table, cluster, or index in any schema.

AUDIT ANY

Audit any object in any schema using AUDIT schema_objects
statements.

BECOME USER

Allows users of the Data Pump Import utility (impdp) and the
original Import utility (imp) to assume the identity of another
user in order to perform operations that cannot be directly
performed by a third party (for example, loading objects such as
object privilege grants).
Allows Streams administrators to create or alter capture users and
apply users in a Streams environment. By default this privilege is
part of the DBA role. Database Vault removes this privileges from
the DBA role. Therefore, this privilege is needed by Streams only
in an environment where Database Vault is installed.

CHANGE NOTIFICATION

Create a registration on queries and receive database change
notifications in response to DML or DDL changes to the objects
associated with the registered queries. Refer to Oracle Database
Advanced Application Developer's Guide for more information on
database change notification.

COMMENT ANY TABLE

Comment on any table, view, or column in any schema.

EXEMPT ACCESS POLICY

Bypass fine-grained access control.
Caution: This is a very powerful system privilege, as it lets the
grantee bypass application-driven security policies. Database
administrators should use caution when granting this privilege.

FORCE ANY TRANSACTION

Force the commit or rollback of any in-doubt distributed
transaction in the local database.
Induce the failure of a distributed transaction.

FORCE TRANSACTION

Force the commit or rollback of the grantee's in-doubt distributed
transactions in the local database.

SQL Statements: DROP SEQUENCE to ROLLBACK 18-47

GRANT

Table 18–1 (Cont.) System Privileges (Organized by the Database Object Operated Upon)
System Privilege Name

Operations Authorized

GRANT ANY OBJECT PRIVILEGE

Grant any object privilege that the object owner is permitted to
grant.
Revoke any object privilege that was granted by the object owner
or by some other user with the GRANT ANY OBJECT PRIVILEGE
privilege.

GRANT ANY PRIVILEGE

Grant any system privilege.

RESUMABLE

Enable resumable space allocation.

SELECT ANY DICTIONARY

Query any data dictionary object in the SYS schema. This privilege
lets you selectively override the default FALSE setting of the O7_
DICTIONARY_ACCESSIBILITY initialization parameter.

SELECT ANY TRANSACTION

Query the contents of the FLASHBACK_TRANSACTION_QUERY view.
Caution: This is a very powerful system privilege, as it lets the
grantee view all data in the database, including past data. This
privilege should be granted only to users who need to use the
Oracle Flashback Transaction Query feature.
Perform STARTUP and SHUTDOWN operations.

SYSDBA

ALTER DATABASE: open, mount, back up, or change character set.
CREATE DATABASE.
ARCHIVELOG and RECOVERY.
CREATE SPFILE.
Includes the RESTRICTED SESSION privilege.
Perform STARTUP and SHUTDOWN operations.

SYSOPER

ALTER DATABASE: open, mount, or back up.
ARCHIVELOG and RECOVERY.
CREATE SPFILE.
Includes the RESTRICTED SESSION privilege.

Table 18–2

Object Privileges (Organized by the Database Object Operated Upon)

Object Privilege Name

Operations Authorized

DIRECTORY PRIVILEGES

The following directory privileges provide secured access to the files stored in
the operating system directory to which the directory object serves as a
pointer. The directory object contains the full path name of the operating
system directory where the files reside. Because the files are actually stored
outside the database, Oracle Database server processes also need to have
appropriate file permissions on the file system server. Granting object
privileges on the directory database object to individual database users, rather
than on the operating system, allows the database to enforce security during
file operations.

READ

Read files in the directory.

WRITE

Write files in the directory. This privilege is useful only in connection with
external tables. It allows the grantee to determine whether the external table
agent can write a log file or a bad file to the directory.
Restriction: This privilege does not allow the grantee to write to a BFILE.

18-48 Oracle Database SQL Language Reference

GRANT

Table 18–2 (Cont.) Object Privileges (Organized by the Database Object Operated Upon)
Object Privilege Name

Operations Authorized

EXECUTE

Execute a preprocessor program that resides in the directory. A preprocessor
program converts data to a supported format when loading data records from
an external table with the ORACLE_LOADER access driver. Refer to Oracle
Database Utilities for more information. This privilege does not implicitly allow
READ access on the external table data.

EDITION PRIVILEGE

The following edition privilege authorizes the use of an edition.

USE

Use an edition.

INDEXTYPE PRIVILEGE

The following indextype privilege authorizes operations on indextypes.

EXECUTE

Reference an indextype.

FLASHBACK DATA ARCHIVE
PRIVILEGE

The following flashback data archive privilege authorizes operations on
flashback data archives.

FLASHBACK ARCHIVE

Enable or disable historical tracking for a table.

LIBRARY PRIVILEGE

The following library privilege authorizes operations on a library.

EXECUTE

Use and reference the specified object and invoke its methods.
Caution: This extremely powerful privilege should be granted only to trusted
users. Refer to Oracle Database Security Guide before granting this privilege.

MATERIALIZED VIEW
PRIVILEGES

The following materialized view privileges authorize operations on a
materialized view. The DELETE, INSERT, and UPDATE privileges can be granted
only to updatable materialized views.

ON COMMIT REFRESH

Create a refresh-on-commit materialized view on the specified table.

QUERY REWRITE

Create a materialized view for query rewrite using the specified table.

SELECT

Query the materialized view with the SELECT statement.

MINING MODEL PRIVILEGES

The following mining model privileges authorize operations on a mining
model. These privileges are not required for models within the users own
schema.

ALTER

Change the mining model name or the associated cost matrix using the
applicable DBMS_DATA_MINING procedures.

SELECT

Score or view the mining model. Scoring is done with the PREDICTION family
of SQL functions or with the DBMS_DATA_MINING.APPLY procedure. Viewing
the model is done with the DBMS_DATA_MINING.GET_MODEL_DETAILS_*
procedures.

OBJECT TYPE PRIVILEGES

The following object type privileges authorize operations on a database
object type.

DEBUG

Access, through a debugger, all public and nonpublic variables, methods, and
types defined on the object type.
Place a breakpoint or stop at a line or instruction boundary within the type
body.

EXECUTE

Use and reference the specified object and invoke its methods.
Access, through a debugger, public variables, types, and methods defined on
the object type.

UNDER

Create a subtype under this type. You can grant this object privilege only if
you have the UNDER ANY TYPE privilege WITH GRANT OPTION on the immediate
supertype of this type.

OLAP PRIVILEGES

The following object privileges are valid if you are using Oracle Database with
the OLAP option.

SQL Statements: DROP SEQUENCE to ROLLBACK 18-49

GRANT

Table 18–2 (Cont.) Object Privileges (Organized by the Database Object Operated Upon)
Object Privilege Name

Operations Authorized

INSERT

Insert members into the OLAP cube dimension or measures into the measures
folder.

ALTER

Change the definition of the OLAP cube dimension or cube.

DELETE

Delete members from the OLAP cube dimension or measures from the
measures folder.

SELECT

View or query the OLAP cube or cube dimension.

UPDATE

Update measure values of the OLAP cube or attribute values of the cube
dimension.

OPERATOR PRIVILEGE

The following operator privilege authorizes operations on user-defined
operators.

EXECUTE

Reference an operator.

PROCEDURE, FUNCTION,
PACKAGE PRIVILEGES

The following procedure, function, and package privileges authorize
operations on procedures, functions, and packages. These privileges also
apply to Java sources, classes, and resources, which Oracle Database treats as
though they were procedures for purposes of granting object privileges.

DEBUG

Access, through a debugger, all public and nonpublic variables, methods, and
types defined on the object.
Place a breakpoint or stop at a line or instruction boundary within the
procedure, function, or package. This privilege grants access to the
declarations in the method or package specification and body.

EXECUTE

Execute the procedure or function directly, or access any program object
declared in the specification of a package, or compile the object implicitly
during a call to a currently invalid or uncompiled function or procedure. This
privilege does not allow the grantee to explicitly compile using ALTER
PROCEDURE or ALTER FUNCTION. For explicit compilation you need the
appropriate ALTER system privilege.
Access, through a debugger, public variables, types, and methods defined on
the procedure, function, or package. This privilege grants access to the
declarations in the method or package specification only.
Job scheduler objects are created using the DBMS_SCHEDULER package. After
these objects are created, you can grant the EXECUTE object privilege on job
scheduler classes and programs. You can also grant ALTER privilege on job
scheduler jobs, programs, and schedules.
Note: Users do not need this privilege to execute a procedure, function, or
package indirectly.

SCHEDULER PRIVILEGES

Job scheduler objects are created using the DBMS_SCHEDULER package. After
these objects are created, you can grant the following privileges.

EXECUTE

Operations on job classes, programs, chains, and credentials.

ALTER

Modifications to jobs, programs, chains, credentials, and schedules.

SEQUENCE PRIVILEGES

The following sequence privileges authorize operations on a sequence.

ALTER

Change the sequence definition with the ALTER SEQUENCE statement.

SELECT

Examine and increment values of the sequence with the CURRVAL and NEXTVAL
pseudocolumns.

SYNONYM PRIVILEGES

Synonym privileges are the same as the privileges for the target object.
Granting a privilege on a synonym is equivalent to granting the privilege on
the base object. Similarly, granting a privilege on a base object is equivalent to
granting the privilege on all synonyms for the object. If you grant to a user a
privilege on a synonym, then the user can use either the synonym name or the
base object name in the SQL statement that exercises the privilege.

18-50 Oracle Database SQL Language Reference

GRANT

Table 18–2 (Cont.) Object Privileges (Organized by the Database Object Operated Upon)
Object Privilege Name

Operations Authorized

TABLE PRIVILEGES

The following table privileges authorize operations on a table. Any one of
following object privileges allows the grantee to lock the table in any lock
mode with the LOCK TABLE statement.
Note: For external tables, the only valid object privileges are ALTER and
SELECT.

ALTER
DEBUG

DELETE

Change the table definition with the ALTER TABLE statement.
Access, through a debugger:
■

PL/SQL code in the body of any triggers defined on the table

■

Information on SQL statements that reference the table directly

Remove rows from the table with the DELETE statement.
Note: You must grant the SELECT privilege on the table along with the DELETE
privilege if the table is on a remote database.

INDEX

Create an index on the table with the CREATE INDEX statement.

INSERT

Add new rows to the table with the INSERT statement.
Note: You must grant the SELECT privilege on the table along with the INSERT
privilege if the table is on a remote database.

REFERENCES

Create a constraint that refers to the table. You cannot grant this privilege to a
role.

SELECT

Query the table with the SELECT statement.

UPDATE

Change data in the table with the UPDATE statement.
Note: You must grant the SELECT privilege on the table along with the UPDATE
privilege if the table is on a remote database.

VIEW PRIVILEGES

The following view privileges authorize operations on a view. Any one of the
following object privileges allows the grantee to lock the view in any lock
mode with the LOCK TABLE statement.
To grant a privilege on a view, you must have that privilege with the GRANT
OPTION on all of the base tables of the view.

DEBUG

Access, through a debugger:
■

PL/SQL code in the body of any triggers defined on the view

■

Information on SQL statements that reference the view directly

DELETE

Remove rows from the view with the DELETE statement.

INSERT

Add new rows to the view with the INSERT statement.

MERGE VIEW

This object privilege has the same behavior as the system privilege MERGE ANY
VIEW on page 18-47, except that the privilege is limited to the views specified
in the ON clause. For any query issued by the grantee on the specified views,
the optimizer can use view merging to improve query performance without
performing the checks that would otherwise be performed to ensure that view
merging does not violate any security intentions of the view creator.

REFERENCES

Define foreign key constraints on the view.

SELECT

Query the view with the SELECT statement.
See Also: object_privilege on page 18-37 for additional information on
granting this object privilege on a view

UNDER

Create a subview under this view. You can grant this object privilege only if
you have the UNDER ANY VIEW privilege WITH GRANT OPTION on the immediate
superview of this view.

UPDATE

Change data in the view with the UPDATE statement.

SQL Statements: DROP SEQUENCE to ROLLBACK 18-51

GRANT

Examples
18

To grant the CREATE SESSION system
privilege to the sample user hr, allowing hr to log on to Oracle Database, issue the
following statement:

Granting a System Privilege to a User: Example

GRANT CREATE SESSION
TO hr;

Granting System Privileges to a Role: Example The following statement grants
appropriate system privileges to a data warehouse manager role, which was created in
the "Creating a Role: Example" on page 15-60:
GRANT
CREATE ANY MATERIALIZED VIEW
, ALTER ANY MATERIALIZED VIEW
, DROP ANY MATERIALIZED VIEW
, QUERY REWRITE
, GLOBAL QUERY REWRITE
TO dw_manager
WITH ADMIN OPTION;

The dw_manager privilege domain now contains the system privileges related to
materialized views.
To grant the dw_manager role with
the ADMIN OPTION to the sample user sh, issue the following statement:

Granting a Role with the Admin Option: Example
GRANT dw_manager
TO sh
WITH ADMIN OPTION;

User sh can now perform the following operations with the dw_manager role:
■

Enable the role and exercise any privileges in the privilege domain of the role,
including the CREATE MATERIALIZED VIEW system privilege

■

Grant and revoke the role to and from other users

■

Drop the role

Granting Object Privileges to a Role: Example The following example grants the

SELECT object privileges to a data warehouse user role, which was created in the
"Creating a Role: Example" on page 15-60:
GRANT SELECT ON sh.sales TO warehouse_user;

The following statement grants the warehouse_
user role to the dw_manager role. Both roles were created in the "Creating a Role:
Example" on page 15-60:
Granting a Role to a Role: Example

GRANT warehouse_user TO dw_manager;

The dw_manager role now contains all of the privileges in the domain of the
warehouse_user role.
To grant READ on directory
bfile_dir to user hr, with the GRANT OPTION, issue the following statement:
Granting an Object Privilege on a Directory: Example
GRANT READ ON DIRECTORY bfile_dir TO hr
WITH GRANT OPTION;

18-52 Oracle Database SQL Language Reference

GRANT

Granting Object Privileges on a Table to a User: Example To grant all privileges on
the table oe.bonuses, which was created in "Merging into a Table: Example" on
page 18-77, to the user hr with the GRANT OPTION, issue the following statement:
GRANT ALL ON bonuses TO hr
WITH GRANT OPTION;

The user hr can subsequently perform the following operations:
■

Exercise any privilege on the bonuses table

■

Grant any privilege on the bonuses table to another user or role

To grant SELECT and UPDATE
privileges on the view emp_view, which was created in "Creating a View: Example" on
page 17-22, to all users, issue the following statement:

Granting Object Privileges on a View: Example

GRANT SELECT, UPDATE
ON emp_view TO PUBLIC;

All users can subsequently query and update the view of employee details.
To grant
SELECT privilege on the customers_seq sequence in the schema oe to the user hr, issue
the following statement:
Granting Object Privileges to a Sequence in Another Schema: Example

GRANT SELECT
ON oe.customers_seq TO hr;

The user hr can subsequently generate the next value of the sequence with the
following statement:
SELECT oe.customers_seq.NEXTVAL
FROM DUAL;

To grant to
user oe the REFERENCES privilege on the employee_id column and the UPDATE privilege
on the employee_id, salary, and commission_pct columns of the employees table in
the schema hr, issue the following statement:
Granting Multiple Object Privileges on Individual Columns: Example

GRANT REFERENCES (employee_id),
UPDATE (employee_id, salary, commission_pct)
ON hr.employees
TO oe;

The user oe can subsequently update values of the employee_id, salary, and
commission_pct columns. User oe can also define referential integrity constraints that
refer to the employee_id column. However, because the GRANT statement lists only
these columns, oe cannot perform operations on any of the other columns of the
employees table.
For example, oe can create a table with a constraint:
CREATE TABLE dependent
(dependno
NUMBER,
dependname VARCHAR2(10),
employee
NUMBER
CONSTRAINT in_emp REFERENCES hr.employees(employee_id) );

The constraint in_emp ensures that all dependents in the dependent table correspond
to an employee in the employees table in the schema hr.

SQL Statements: DROP SEQUENCE to ROLLBACK 18-53

INSERT

INSERT
Purpose
18

Use the INSERT statement to add rows to a table, the base table of a view, a partition of
a partitioned table or a subpartition of a composite-partitioned table, or an object table
or the base table of an object view.

Prerequisites
18

For you to insert rows into a table, the table must be in your own schema or you must
have the INSERT object privilege on the table.
For you to insert rows into the base table of a view, the owner of the schema
containing the view must have the INSERT object privilege on the base table. Also, if
the view is in a schema other than your own, then you must have the INSERT object
privilege on the view.
If you have the INSERT ANY TABLE system privilege, then you can also insert rows into
any table or the base table of any view.
You must also have the SELECT object privilege on the table into which you want to
insert rows if the table is on a remote database.
Conventional and Direct-Path INSERT
You can use the INSERT statement to insert data into a table, partition, or view in two
ways: conventional INSERT and direct-path INSERT. When you issue a conventional
INSERT statement, Oracle Database reuses free space in the table into which you are
inserting and maintains referential integrity constraints. With direct-path INSERT, the
database appends the inserted data after existing data in the table. Data is written
directly into data files, bypassing the buffer cache. Free space in the existing data is not
reused. This alternative enhances performance during insert operations and is similar
to the functionality of the Oracle direct-path loader utility, SQL*Loader. When you
insert into a table that has been created in parallel mode, direct-path INSERT is the
default.
The manner in which the database generates redo and undo data depends in part on
whether you are using conventional or direct-path INSERT:
■

■

Conventional INSERT always generates maximal redo and undo for changes to
both data and metadata, regardless of the logging setting of the table and the
archivelog and force logging settings of the database.
Direct-path INSERT generates both redo and undo for metadata changes, because
these are needed for operation recovery. For data changes, undo and redo are
generated as follows:
■
■

■

■

Direct-path INSERT always bypasses undo generation for data changes.
If the database is not in ARCHIVELOG or FORCE LOGGING mode, then no redo is
generated for data changes, regardless of the logging setting of the table.
If the database is in ARCHIVELOG mode (but not in FORCE LOGGING mode), then
direct-path INSERT generates data redo for LOGGING tables but not for
NOLOGGING tables.
If the database is in ARCHIVELOG and FORCE LOGGING mode, then direct-path
SQL generate data redo for both LOGGING and NOLOGGING tables.

18-54 Oracle Database SQL Language Reference

INSERT

Direct-path INSERT is subject to a number of restrictions. If any of these restrictions is
violated, then Oracle Database executes conventional INSERT serially without
returning any message, unless otherwise noted:
■

■

■

You can have multiple direct-path INSERT statements in a single transaction, with
or without other DML statements. However, after one DML statement alters a
particular table, partition, or index, no other DML statement in the transaction can
access that table, partition, or index.
Queries that access the same table, partition, or index are allowed before the
direct-path INSERT statement, but not after it.
If any serial or parallel statement attempts to access a table that has already been
modified by a direct-path INSERT in the same transaction, then the database
returns an error and rejects the statement.

■

The target table cannot be of a cluster.

■

The target table cannot contain object type columns.

■

■

■

■
■

Direct-path INSERT is not supported for an index-organized table (IOT) if it is not
partitioned, if it has a mapping table, or if it is reference by a materialized view.
Direct-path INSERT into a single partition of an index-organized table (IOT), or into
a partitioned IOT with only one partition, will be done serially, even if the IOT was
created in parallel mode or you specify the APPEND or APPEND_VALUES hint.
However, direct-path INSERT operations into a partitioned IOT will honor parallel
mode as long as the partition-extended name is not used and the IOT has more
than one partition.
The target table cannot have any triggers or referential integrity constraints
defined on it.
The target table cannot be replicated.
A transaction containing a direct-path INSERT statement cannot be or become
distributed.

You cannot query or modify direct-path inserted data immediately after the insert is
complete. If you attempt to do so, an ORA-12838 error is generated. You must first issue
a COMMIT statement before attempting to read or modify the newly-inserted data.
See Also:
■

■
■

Oracle Database Administrator's Guide for a more complete
description of direct-path INSERT
Oracle Database Utilities for information on SQL*Loader
Oracle Database Performance Tuning Guide for information on how
to tune parallel direct-path INSERT

Syntax
18

insert::=
hint

single_table_insert

INSERT

;
multi_table_insert

(single_table_insert::= on page 18-56, multi_table_insert::= on page 18-56)

SQL Statements: DROP SEQUENCE to ROLLBACK 18-55

INSERT

single_table_insert::=
returning_clause
error_logging_clause

values_clause
insert_into_clause
subquery

(insert_into_clause::= on page 18-56, values_clause::= on page 18-56, returning_clause::=
on page 18-56, subquery::= on page 19-5, error_logging_clause::= on page 18-57)
insert_into_clause::=
,
t_alias
INTO

(

column

)

dml_table_expression_clause

(DML_table_expression_clause::= on page 18-57)
values_clause::=
,
expr
VALUES

(

)
DEFAULT

returning_clause::=
,

,

RETURN
expr

INTO

data_item

RETURNING

multi_table_insert::=
values_clause
ALL

error_logging_clause

insert_into_clause
subquery

conditional_insert_clause

(insert_into_clause::= on page 18-56, values_clause::= on page 18-56, conditional_insert_
clause::= on page 18-56, subquery::= on page 19-5, error_logging_clause::= on page 18-57)
conditional_insert_clause::=
ALL
FIRST

values_clause
WHEN

condition

THEN

values_clause
ELSE

insert_into_clause

error_logging_clause

insert_into_clause

(insert_into_clause::= on page 18-56, values_clause::= on page 18-56)

18-56 Oracle Database SQL Language Reference

error_logging_clause

INSERT

DML_table_expression_clause::=
partition_extension_clause
@
schema

dblink

table

.

@

view

dblink

materialized view
subquery_restriction_clause
(

subquery

)

table_collection_expression

(partition_extension_clause::= on page 18-72, subquery::= on page 19-5—part of SELECT,
subquery_restriction_clause::= on page 18-57, table_collection_expression::= on page 18-57)
partition_extension_clause::=
(

partition

)

PARTITION

,
FOR

(

(

partition_key_value

subpartition

)

)

SUBPARTITION

,
FOR

(

subpartition_key_value

)

subquery_restriction_clause::=

READ

CONSTRAINT

ONLY

constraint

WITH
CHECK

OPTION

table_collection_expression::=
(
TABLE

(

collection_expression

+

)

)

error_logging_clause::=
schema
INTO
LOG

.
table

(

simple_expression

)

ERRORS

integer
REJECT

LIMIT
UNLIMITED

SQL Statements: DROP SEQUENCE to ROLLBACK 18-57

INSERT

Semantics
18

hint
Specify a comment that passes instructions to the optimizer on choosing an execution
plan for the statement.
For a multitable insert, if you specify the PARALLEL hint for any target table, then the
entire multitable insert statement is parallelized even if the target tables have not been
created or altered with PARALLEL specified. If you do not specify the PARALLEL hint,
then the insert operation will not be parallelized unless all target tables were created or
altered with PARALLEL specified.
See Also:
■

"Hints" on page 3-74 for the syntax and description of hints

■

"Restrictions on Multitable Inserts" on page 18-63

single_table_insert
In a single-table insert, you insert values into one row of a table, view, or materialized
view by specifying values explicitly or by retrieving the values through a subquery.
You can use the flashback_query_clause in subquery to insert past data into table.
Refer to the flashback_query_clause of SELECT on page 19-17 for more information on
this clause.
Restriction on Single-table Inserts If you retrieve values through a subquery, then
the select list of the subquery must have the same number of columns as the column
list of the INSERT statement. If you omit the column list, then the subquery must
provide values for every column in the table.
See Also:

"Inserting Values into Tables: Examples" on page 18-66

insert_into_clause
Use the INSERT INTO clause to specify the target object or objects into which the
database is to insert data.
DML_table_expression_clause
Use the INTO DML_table_expression_clause to specify the objects into which data is
being inserted.
schema Specify the schema containing the table, view, or materialized view. If you
omit schema, then the database assumes the object is in your own schema.

Specify the name of the table or object
table, view or object view, materialized view, or the column or columns returned by a
subquery, into which rows are to be inserted. If you specify a view or object view, then
the database inserts rows into the base table of the view.

table | view | materialized_view | subquery

You cannot insert rows into a read-only materialized view. If you insert rows into a
writable materialized view, then the database inserts the rows into the underlying
container table. However, the insertions are overwritten at the next refresh operation.
If you insert rows into an updatable materialized view that is part of a materialized
view group, then the database also inserts the corresponding rows into the master
table.

18-58 Oracle Database SQL Language Reference

INSERT

If any value to be inserted is a REF to an object table, and if the object table has a
primary key object identifier, then the column into which you insert the REF must be a
REF column with a referential integrity or SCOPE constraint to the object table.
If table, or the base table of view, contains one or more domain index columns, then
this statement executes the appropriate indextype insert routine.
Issuing an INSERT statement against a table fires any INSERT triggers defined on the
table.
See Also: Oracle Database Data Cartridge Developer's Guide for more
information on these routines
Restrictions on the DML_table_expression_clause

This clause is subject to the

following restrictions:
■

■

■

■

■

■

You cannot execute this statement if table or the base table of view contains any
domain indexes marked IN_PROGRESS or FAILED.
You cannot insert into a partition if any affected index partitions are marked
UNUSABLE.
With regard to the ORDER BY clause of the subquery in the DML_table_expression_
clause, ordering is guaranteed only for the rows being inserted, and only within
each extent of the table. Ordering of new rows with respect to existing rows is not
guaranteed.
If a view was created using the WITH CHECK OPTION, then you can insert into the
view only rows that satisfy the defining query of the view.
If a view was created using a single base table, then you can insert rows into the
view and then retrieve those values using the returning_clause.
You cannot insert rows into a view except with INSTEAD OF triggers if the defining
query of the view contains one of the following constructs:
A set operator
A DISTINCT operator
An aggregate or analytic function
A GROUP BY, ORDER BY, MODEL, CONNECT BY, or START WITH clause
A collection expression in a SELECT list
A subquery in a SELECT list
A subquery designated WITH READ ONLY
Joins, with some exceptions, as documented in Oracle Database Administrator's
Guide

■

If you specify an index, index partition, or index subpartition that has been
marked UNUSABLE, then the INSERT statement will fail unless the SKIP_UNUSABLE_
INDEXES session parameter has been set to TRUE. Refer to ALTER SESSION on
page 11-45 for information on the SKIP_UNUSABLE_INDEXES session parameter.

Specify the name or partition key value of the partition
or subpartition within table, or the base table of view, targeted for inserts.

partition_extension_clause

If a row to be inserted does not map into a specified partition or subpartition, then the
database returns an error.
Restriction on Target Partitions and Subpartitions This clause is not valid for object

tables or object views.

SQL Statements: DROP SEQUENCE to ROLLBACK 18-59

INSERT

See Also: "References to Partitioned Tables and Indexes" on
page 3-119

Specify a complete or partial name of a database link to a remote database
where the table or view is located. You can insert rows into a remote table or view only
if you are using Oracle Database distributed functionality.

dblink

If you omit dblink, then Oracle Database assumes that the table or view is on the local
database.
See Also:
■

■

"Syntax for Schema Objects and Parts in SQL Statements" on
page 3-115 and "References to Objects in Remote Databases" on
page 3-117 for information on referring to database links
"Inserting into a Remote Database: Example" on page 18-67

subquery_restriction_clause Use the subquery_restriction_clause to restrict the
subquery in one of the following ways:
WITH READ ONLY

Specify WITH READ ONLY to indicate that the table or view cannot

be updated.
WITH CHECK OPTION Specify WITH CHECK OPTION to indicate that Oracle Database
prohibits any changes to the table or view that would produce rows that are not
included in the subquery. When used in the subquery of a DML statement, you can
specify this clause in a subquery in the FROM clause but not in subquery in the WHERE
clause.
CONSTRAINT constraint Specify the name of the CHECK OPTION constraint. If you
omit this identifier, then Oracle automatically assigns the constraint a name of the
form SYS_Cn, where n is an integer that makes the constraint name unique within the
database.
See Also: "Using the WITH CHECK OPTION Clause: Example" on
page 19-48

The table_collection_expression lets you inform
Oracle that the value of collection_expression should be treated as a table for
purposes of query and DML operations. The collection_expression can be a
subquery, a column, a function, or a collection constructor. Regardless of its form, it
must return a collection value—that is, a value whose type is nested table or varray.
This process of extracting the elements of a collection is called collection unnesting.
table_collection_expression

The optional plus (+) is relevant if you are joining the TABLE collection expression with
the parent table. The + creates an outer join of the two, so that the query returns rows
from the outer table even if the collection expression is null.
In earlier releases of Oracle, when collection_expression
was a subquery, table_collection_expression was expressed as THE
subquery. That usage is now deprecated.
Note:

See Also:

"Table Collections: Examples" on page 19-54

18-60 Oracle Database SQL Language Reference

INSERT

t_alias
Specify a correlation name, which is an alias for the table, view, materialized view, or
subquery to be referenced elsewhere in the statement.
Restriction on Table Aliases You cannot specify t_alias during a multitable insert.

column
Specify a column of the table, view, or materialized view. In the inserted row, each
column in this list is assigned a value from the values_clause or the subquery.
If you omit one or more of the table's columns from this list, then the column value of
that column for the inserted row is the column default value as specified when the
table was created or last altered. If any omitted column has a NOT NULL constraint and
no default value, then the database returns an error indicating that the constraint has
been violated and rolls back the INSERT statement. Refer to CREATE TABLE on
page 16-6 for more information on default column values.
If you omit the column list altogether, then the values_clause or query must specify
values for all columns in the table.

values_clause
For a single-table insert operation, specify a row of values to be inserted into the table
or view. You must specify a value in the values_clause for each column in the column
list. If you omit the column list, then the values_clause must provide values for every
column in the table.
For a multitable insert operation, each expression in the values_clause must refer to
columns returned by the select list of the subquery. If you omit the values_clause,
then the select list of the subquery determines the values to be inserted, so it must
have the same number of columns as the column list of the corresponding insert_
into_clause. If you do not specify a column list in the insert_into_clause, then the
computed row must provide values for all columns in the target table.
For both types of insert operations, if you specify a column list in the insert_into_
clause, then the database assigns to each column in the list a corresponding value
from the values clause or the subquery. You can specify DEFAULT for any value in the
values_clause. If you have specified a default value for the corresponding column of
the table or view, then that value is inserted. If no default value for the corresponding
column has been specified, then the database inserts null. Refer to "About SQL
Expressions" on page 6-1 and SELECT on page 19-4 for syntax of valid expressions.
Restrictions on Inserted Values
■

The value are subject to the following restrictions:

You cannot insert a BFILE value until you have initialized the BFILE locator to null
or to a directory name and filename.
See Also:
■

■

■

BFILENAME on page 5-29 for information on initializing BFILE
values and for an example of inserting into a BFILE
Oracle Database SecureFiles and Large Objects Developer's Guide for
information on initializing BFILE locators

When inserting into a list-partitioned table, you cannot insert a value into the
partitioning key column that does not already exist in the partition_key_value
list of one of the partitions.

SQL Statements: DROP SEQUENCE to ROLLBACK 18-61

INSERT

■
■

You cannot specify DEFAULT when inserting into a view.
If you insert string literals into a RAW column, then during subsequent queries
Oracle Database will perform a full table scan rather than using any index that
might exist on the RAW column.
See Also:
■

■

"Using XML in SQL Statements" on page F-8 for information on
inserting values into an XMLType table
"Inserting into a Substitutable Tables and Columns: Examples" on
page 18-67, "Inserting Using the TO_LOB Function: Example" on
page 18-67, "Inserting Sequence Values: Example" on page 18-67,
and "Inserting Using Bind Variables: Example" on page 18-67

returning_clause
The returning clause retrieves the rows affected by a DML statement. You can specify
this clause for tables and materialized views and for views with a single base table.
When operating on a single row, a DML statement with a returning_clause can
retrieve column expressions using the affected row, rowid, and REFs to the affected
row and store them in host variables or PL/SQL variables.
When operating on multiple rows, a DML statement with the returning_clause stores
values from expressions, rowids, and REFs involving the affected rows in bind arrays.
expr

Each item in the expr list must be a valid expression syntax.

INTO The INTO clause indicates that the values of the changed rows are to be stored
in the variable(s) specified in data_item list.

Each data_item is a host variable or PL/SQL variable that stores the
retrieved expr value.

data_item

For each expression in the RETURNING list, you must specify a corresponding
type-compatible PL/SQL variable or host variable in the INTO list.
Restrictions The following restrictions apply to the RETURNING clause:
■

■

The expr is restricted as follows:
–

For UPDATE and DELETE statements each expr must be a simple expression or a
single-set aggregate function expression. You cannot combine simple
expressions and single-set aggregate function expressions in the same
returning_clause. For INSERT statements, each expr must be a simple
expression. Aggregate functions are not supported in an INSERT statement
RETURNING clause.

–

Single-set aggregate function expressions cannot include the DISTINCT
keyword.

If the expr list contains a primary key column or other NOT NULL column, then the
update statement fails if the table has a BEFORE UPDATE trigger defined on it.

■

You cannot specify the returning_clause for a multitable insert.

■

You cannot use this clause with parallel DML or with remote objects.

■

You cannot retrieve LONG types with this clause.

18-62 Oracle Database SQL Language Reference

INSERT

■

You cannot specify this clause for a view on which an INSTEAD OF trigger has been
defined.
Oracle Database PL/SQL Language Reference for information
on using the BULK COLLECT clause to return multiple values to
collection variables

See Also:

multi_table_insert
In a multitable insert, you insert computed rows derived from the rows returned from
the evaluation of a subquery into one or more tables.
Table aliases are not defined by the select list of the subquery. Therefore, they are not
visible in the clauses dependent on the select list. For example, this can happen when
trying to refer to an object column in an expression. To use an expression with a table
alias, you must put the expression into the select list with a column alias, and then
refer to the column alias in the VALUES clause or WHEN condition of the multitable insert.
ALL into_clause
Specify ALL followed by multiple insert_into_clauses to perform an unconditional
multitable insert. Oracle Database executes each insert_into_clause once for each
row returned by the subquery.
conditional_insert_clause
Specify the conditional_insert_clause to perform a conditional multitable insert.
Oracle Database filters each insert_into_clause through the corresponding WHEN
condition, which determines whether that insert_into_clause is executed. Each
expression in the WHEN condition must refer to columns returned by the select list of the
subquery. A single multitable insert statement can contain up to 127 WHEN clauses.
ALL If you specify ALL, the default value, then the database evaluates each WHEN
clause regardless of the results of the evaluation of any other WHEN clause. For each
WHEN clause whose condition evaluates to true, the database executes the
corresponding INTO clause list.

If you specify FIRST, then the database evaluates each WHEN clause in the order
in which it appears in the statement. For the first WHEN clause that evaluates to true, the
database executes the corresponding INTO clause and skips subsequent WHEN clauses
for the given row.
FIRST

ELSE clause
■

■

For a given row, if no WHEN clause evaluates to true, then:

If you have specified an ELSE clause, then the database executes the INTO clause list
associated with the ELSE clause.
If you did not specify an else clause, then the database takes no action for that row.
See Also:

"Multitable Inserts: Examples" on page 18-68

Restrictions on Multitable Inserts Multitable inserts are subject to the following

restrictions:
■

■

You can perform multitable inserts only on tables, not on views or materialized
views.
You cannot perform a multitable insert into a remote table.

SQL Statements: DROP SEQUENCE to ROLLBACK 18-63

INSERT

■

■

■
■

You cannot specify a TABLE collection expression when performing a multitable
insert.
Multitable inserts are not parallelized if any target table is index organized or if
any target table has a bitmap index defined on it.
Plan stability is not supported for multitable insert statements.
You cannot specify a sequence in any part of a multitable insert statement. A
multitable insert is considered a single SQL statement. Therefore, the first
reference to NEXTVAL generates the next number, and all subsequent references in
the statement return the same number.

subquery
Specify a subquery that returns rows that are inserted into the table. The subquery can
refer to any table, view, or materialized view, including the target tables of the INSERT
statement. If the subquery selects no rows, then the database inserts no rows into the
table.
You can use subquery in combination with the TO_LOB function to convert the values in
a LONG column to LOB values in another column in the same or another table.
■

■

To migrate LONG values to LOB values in another column in a view, you must
perform the migration on the base table and then add the LOB column to the view.
To migrate LONG values on a remote table to LOB values in a local table, you must
perform the migration on the remote table using the TO_LOB function, and then
perform an INSERT ... subquery operation to copy the LOB values from the remote
table into the local table.

Notes on Inserting with a Subquery The following notes apply when inserting with
a subquery:
■

If subquery returns the partial or total equivalent of a materialized view, then the
database may use the materialized view for query rewrite in place of one or more
tables specified in subquery.
See Also: Oracle Database Data Warehousing Guide for more
information on materialized views and query rewrite

■

If subquery refers to remote objects, then the INSERT operation can run in parallel
as long as the reference does not loop back to an object on the local database.
However, if the subquery in the DML_table_expression_clause refers to any
remote objects, then the INSERT operation will run serially without notification. See
parallel_clause on page 16-63 for more information.
See Also:
■
■

■

■

"Inserting Values with a Subquery: Example" on page 18-66
BFILENAME on page 5-29 for an example of inserting into a
BFILE
Oracle Database SecureFiles and Large Objects Developer's Guide for
information on initializing BFILEs
"About SQL Expressions" on page 6-1 and SELECT on page 19-4
for syntax of valid expressions

18-64 Oracle Database SQL Language Reference

INSERT

error_logging_clause
The error_logging_clause lets you capture DML errors and the log column values of
the affected rows and save them in an error logging table.
INTO table Specify the name of the error logging table. If you omit this clause, then
the database assigns the default name generated by the DBMS_ERRLOG package. The
default error log table name is ERR$_ followed by the first 25 characters of the name of
the table upon which the DML operation is being executed.
simple_expression Specify the value to be used as a statement tag, so that you can
identify the errors from this statement in the error logging table. The expression can be
either a text literal, a number literal, or a general SQL expression such as a bind
variable. You can also use a function expression if you convert it to a text literal — for
example, TO_CHAR(SYSDATE).
REJECT LIMIT This clause lets you specify an integer as an upper limit for the
number of errors to be logged before the statement terminates and rolls back any
changes made by the statement. The default rejection limit is zero. For parallel DML
operations, the reject limit is applied to each parallel server.

Restrictions on DML Error Logging
■
The following conditions cause the statement to fail and roll back without
invoking the error logging capability:

■

–

Violated deferred constraints.

–

Any direct-path INSERT or MERGE operation that raises a unique constraint or
index violation.

–

Any update operation UPDATE or MERGE that raises a unique constraint or index
violation.

You cannot track errors in the error logging table for LONG, LOB, or object type
columns. However, the table that is the target of the DML operation can contain
these types of columns.
–

If you create or modify the corresponding error logging table so that it
contains a column of an unsupported type, and if the name of that column
corresponds to an unsupported column in the target DML table, then the DML
statement fails at parse time.

–

If the error logging table does not contain any unsupported column types,
then all DML errors are logged until the reject limit of errors is reached. For
rows on which errors occur, column values with corresponding columns in the
error logging table are logged along with the control information.
See Also:
■

■

Oracle Database PL/SQL Packages and Types Reference for
information on using the create_error_log procedure of the
DBMS_ERRLOG package and Oracle Database Administrator's Guide for
general information on DML error logging.
"Inserting Into a Table with Error Logging: Example" on
page 18-66

SQL Statements: DROP SEQUENCE to ROLLBACK 18-65

INSERT

Examples
18

Inserting Values into Tables: Examples The following statement inserts a row into
the sample table departments:
INSERT INTO departments
VALUES (280, 'Recreation', 121, 1700);

If the departments table had been created with a default value of 121 for the manager_
id column, then you could issue the same statement as follows:
INSERT INTO departments
VALUES (280, 'Recreation', DEFAULT, 1700);

The following statement inserts a row with six columns into the employees table. One
of these columns is assigned NULL and another is assigned a number in scientific
notation:
INSERT INTO employees (employee_id, last_name, email,
hire_date, job_id, salary, commission_pct)
VALUES (207, 'Gregory', 'pgregory@example.com',
sysdate, 'PU_CLERK', 1.2E3, NULL);

The following statement has the same effect as the preceding example, but uses a
subquery in the DML_table_expression_clause:
INSERT INTO
(SELECT employee_id, last_name, email, hire_date, job_id,
salary, commission_pct FROM employees)
VALUES (207, 'Gregory', 'pgregory@example.com',
sysdate, 'PU_CLERK', 1.2E3, NULL);

Inserting Values with a Subquery: Example The following statement copies

employees whose commission exceeds 25% of their salary into the bonuses table,
which was created in "Merging into a Table: Example" on page 18-77:
INSERT INTO bonuses
SELECT employee_id, salary*1.1
FROM employees
WHERE commission_pct > 0.25;

The following statements create
a raises table in the sample schema hr, create an error logging table using the DBMS_
ERRLOG package, and populate the raises table with data from the employees table.
One of the inserts violates the check constraint on raises, and that row can be seen in
errlog. If more than ten errors had occurred, then the statement would have aborted,
rolling back any insertions made:
Inserting Into a Table with Error Logging: Example

CREATE TABLE raises (emp_id NUMBER, sal NUMBER
CONSTRAINT check_sal CHECK(sal > 8000));
EXECUTE DBMS_ERRLOG.CREATE_ERROR_LOG('raises', 'errlog');
INSERT INTO raises
SELECT employee_id, salary*1.1 FROM employees
WHERE commission_pct > .2
LOG ERRORS INTO errlog ('my_bad') REJECT LIMIT 10;
SELECT ORA_ERR_MESG$, ORA_ERR_TAG$, emp_id, sal FROM errlog;
ORA_ERR_MESG$

18-66 Oracle Database SQL Language Reference

ORA_ERR_TAG$

EMP_ID SAL

INSERT

--------------------------- -------------------- ------ ------ORA-02290: check constraint my_bad
161
7700
(HR.SYS_C004266) violated

Inserting into a Remote Database: Example The following statement inserts a row
into the employees table owned by the user hr on the database accessible by the
database link remote:
INSERT INTO employees@remote
VALUES (8002, 'Juan', 'Fernandez', 'juanf@example.com', NULL,
TO_DATE('04-OCT-1992', 'DD-MON-YYYY'), 'SH_CLERK', 3000,
NULL, 121, 20);

Inserting Sequence Values: Example The following statement inserts a new row
containing the next value of the departments_seq sequence into the departments
table:
INSERT INTO departments
VALUES (departments_seq.nextval, 'Entertainment', 162, 1400);

The following example returns the values
of the inserted rows into output bind variables :bnd1 and :bnd2. The bind variables
must first be declared.

Inserting Using Bind Variables: Example

INSERT INTO employees
(employee_id, last_name, email, hire_date, job_id, salary)
VALUES
(employees_seq.nextval, 'Doe', 'john.doe@example.com',
SYSDATE, 'SH_CLERK', 2400)
RETURNING salary*12, job_id INTO :bnd1, :bnd2;

The following
example inserts into the persons table, which is created in "Substitutable Table and
Column Examples" on page 16-71. The first statement uses the root type person_t. The
second insert uses the employee_t subtype of person_t, and the third insert uses the
part_time_emp_t subtype of employee_t:
Inserting into a Substitutable Tables and Columns: Examples

INSERT INTO persons VALUES (person_t('Bob', 1234));
INSERT INTO persons VALUES (employee_t('Joe', 32456, 12, 100000));
INSERT INTO persons VALUES (
part_time_emp_t('Tim', 5678, 13, 1000, 20));

The following example inserts into the books table, which was created in
"Substitutable Table and Column Examples" on page 16-71. Notice that specification of
the attribute values is identical to that for the substitutable table example:
INSERT INTO books VALUES (
'An Autobiography', person_t('Bob', 1234));
INSERT INTO books VALUES (
'Business Rules', employee_t('Joe', 3456, 12, 10000));
INSERT INTO books VALUES (
'Mixing School and Work',
part_time_emp_t('Tim', 5678, 13, 1000, 20));

You can extract data from substitutable tables and columns using built-in functions
and conditions. For examples, see the functions TREAT on page 5-327 and SYS_
TYPEID on page 5-289, and "IS OF type Condition" on page 7-25.
The following example copies LONG
data to a LOB column in the following long_tab table:

Inserting Using the TO_LOB Function: Example

SQL Statements: DROP SEQUENCE to ROLLBACK 18-67

INSERT

CREATE TABLE long_tab (pic_id NUMBER, long_pics LONG RAW);

First you must create a table with a LOB.
CREATE TABLE lob_tab (pic_id NUMBER, lob_pics BLOB);

Next, use an INSERT ... SELECT statement to copy the data in all rows for the LONG
column into the newly created LOB column:
INSERT INTO lob_tab
SELECT pic_id, TO_LOB(long_pics) FROM long_tab;

When you are confident that the migration has been successful, you can drop the
long_pics table. Alternatively, if the table contains other columns, then you can
simply drop the LONG column from the table as follows:
ALTER TABLE long_tab DROP COLUMN long_pics;

Multitable Inserts: Examples The following example uses the multitable insert
syntax to insert into the sample table sh.sales some data from an input table with a
different structure.

A number of NOT NULL constraints on the sales table have been
disabled for purposes of this example, because the example ignores a
number of table columns for the sake of brevity.

Note:

The input table looks like this:
SELECT * FROM sales_input_table;
PRODUCT_ID CUSTOMER_ID WEEKLY_ST SALES_SUN SALES_MON SALES_TUE SALES_WED SALES_THU SALES_FRI SALES_SAT
---------- ----------- --------- ---------- ---------- ---------- -------------------- ---------- ---------111
222 01-OCT-00
100
200
300
400
500
600
700
222
333 08-OCT-00
200
300
400
500
600
700
800
333
444 15-OCT-00
300
400
500
600
700
800
900

The multitable insert statement looks like this:
INSERT ALL
INTO sales (prod_id, cust_id, time_id, amount)
VALUES (product_id, customer_id, weekly_start_date, sales_sun)
INTO sales (prod_id, cust_id, time_id, amount)
VALUES (product_id, customer_id, weekly_start_date+1, sales_mon)
INTO sales (prod_id, cust_id, time_id, amount)
VALUES (product_id, customer_id, weekly_start_date+2, sales_tue)
INTO sales (prod_id, cust_id, time_id, amount)
VALUES (product_id, customer_id, weekly_start_date+3, sales_wed)
INTO sales (prod_id, cust_id, time_id, amount)
VALUES (product_id, customer_id, weekly_start_date+4, sales_thu)
INTO sales (prod_id, cust_id, time_id, amount)
VALUES (product_id, customer_id, weekly_start_date+5, sales_fri)
INTO sales (prod_id, cust_id, time_id, amount)
VALUES (product_id, customer_id, weekly_start_date+6, sales_sat)
SELECT product_id, customer_id, weekly_start_date, sales_sun,
sales_mon, sales_tue, sales_wed, sales_thu, sales_fri, sales_sat
FROM sales_input_table;

Assuming these are the only rows in the sales table, the contents now look like this:
SELECT * FROM sales

18-68 Oracle Database SQL Language Reference

INSERT

ORDER BY prod_id, cust_id, time_id;
PROD_ID
CUST_ID TIME_ID
C
PROMO_ID QUANTITY_SOLD
AMOUNT
COST
---------- ---------- --------- - ---------- ------------- ---------- ---------111
222 01-OCT-00
100
111
222 02-OCT-00
200
111
222 03-OCT-00
300
111
222 04-OCT-00
400
111
222 05-OCT-00
500
111
222 06-OCT-00
600
111
222 07-OCT-00
700
222
333 08-OCT-00
200
222
333 09-OCT-00
300
222
333 10-OCT-00
400
222
333 11-OCT-00
500
222
333 12-OCT-00
600
222
333 13-OCT-00
700
222
333 14-OCT-00
800
333
444 15-OCT-00
300
333
444 16-OCT-00
400
333
444 17-OCT-00
500
333
444 18-OCT-00
600
333
444 19-OCT-00
700
333
444 20-OCT-00
800
333
444 21-OCT-00
900

The next examples insert into multiple tables. Suppose you want to provide to sales
representatives some information on orders of various sizes. The following example
creates tables for small, medium, large, and special orders and populates those tables
with data from the sample table oe.orders:
CREATE TABLE small_orders
(order_id
NUMBER(12)
NOT NULL,
customer_id
NUMBER(6)
NOT NULL,
order_total
NUMBER(8,2),
sales_rep_id
NUMBER(6)
);
CREATE TABLE medium_orders AS SELECT * FROM small_orders;
CREATE TABLE large_orders AS SELECT * FROM small_orders;
CREATE TABLE special_orders
(order_id
NUMBER(12)
customer_id
NUMBER(6)
order_total
NUMBER(8,2),
sales_rep_id
NUMBER(6),
credit_limit
NUMBER(9,2),
cust_email
VARCHAR2(30)
);

NOT NULL,
NOT NULL,

The first multitable insert populates only the tables for small, medium, and large
orders:
INSERT ALL
WHEN order_total <= 100000 THEN
INTO small_orders
WHEN order_total > 100000 AND order_total <= 200000 THEN
INTO medium_orders
WHEN order_total > 200000 THEN
INTO large_orders
SQL Statements: DROP SEQUENCE to ROLLBACK 18-69

INSERT

SELECT order_id, order_total, sales_rep_id, customer_id
FROM orders;

You can accomplish the same thing using the ELSE clause in place of the insert into the
large_orders table:
INSERT ALL
WHEN order_total <= 100000 THEN
INTO small_orders
WHEN order_total > 100000 AND order_total <= 200000 THEN
INTO medium_orders
ELSE
INTO large_orders
SELECT order_id, order_total, sales_rep_id, customer_id
FROM orders;

The next example inserts into the small, medium, and large tables, as in the preceding
example, and also puts orders greater than 290,000 into the special_orders table. This
table also shows how to use column aliases to simplify the statement:
INSERT ALL
WHEN ottl <= 100000 THEN
INTO small_orders
VALUES(oid, ottl, sid, cid)
WHEN ottl > 100000 and ottl <= 200000 THEN
INTO medium_orders
VALUES(oid, ottl, sid, cid)
WHEN ottl > 200000 THEN
into large_orders
VALUES(oid, ottl, sid, cid)
WHEN ottl > 290000 THEN
INTO special_orders
SELECT o.order_id oid, o.customer_id cid, o.order_total ottl,
o.sales_rep_id sid, c.credit_limit cl, c.cust_email cem
FROM orders o, customers c
WHERE o.customer_id = c.customer_id;

Finally, the next example uses the FIRST clause to put orders greater than 290,000 into
the special_orders table and exclude those orders from the large_orders table:
INSERT FIRST
WHEN ottl <= 100000 THEN
INTO small_orders
VALUES(oid, ottl, sid, cid)
WHEN ottl > 100000 and ottl <= 200000 THEN
INTO medium_orders
VALUES(oid, ottl, sid, cid)
WHEN ottl > 290000 THEN
INTO special_orders
WHEN ottl > 200000 THEN
INTO large_orders
VALUES(oid, ottl, sid, cid)
SELECT o.order_id oid, o.customer_id cid, o.order_total ottl,
o.sales_rep_id sid, c.credit_limit cl, c.cust_email cem
FROM orders o, customers c
WHERE o.customer_id = c.customer_id;

18-70 Oracle Database SQL Language Reference

LOCK TABLE

LOCK TABLE
Purpose
18

Use the LOCK TABLE statement to lock one or more tables, table partitions, or table
subpartitions in a specified mode. This lock manually overrides automatic locking and
permits or denies access to a table or view by other users for the duration of your
operation.
Some forms of locks can be placed on the same table at the same time. Other locks
allow only one lock for a table.
A locked table remains locked until you either commit your transaction or roll it back,
either entirely or to a savepoint before you locked the table.
A lock never prevents other users from querying the table. A query never places a lock
on a table. Readers never block writers and writers never block readers.
See Also:
■

Oracle Database Concepts for a complete description of the
interaction of lock modes

■

COMMIT on page 13-49

■

ROLLBACK on page 18-96

■

SAVEPOINT on page 19-2

Prerequisites
18

The table or view must be in your own schema or you must have the LOCK ANY TABLE
system privilege, or you must have any object privilege on the table or view.

Syntax
18

lock_table::=
,
partition_extension_clause
schema
LOCK

.

@

table

dblink

TABLE
view

NOWAIT
WAIT
IN

lockmode

MODE

integer
;

SQL Statements: DROP SEQUENCE to ROLLBACK 18-71

LOCK TABLE

partition_extension_clause::=
(

partition

)

PARTITION

,
FOR
(

(

partition_key_value

subpartition

)

)

SUBPARTITION

,
FOR

(

subpartition_key_value

)

Semantics
18

schema
Specify the schema containing the table or view. If you omit schema, then Oracle
Database assumes the table or view is in your own schema.

table / view
Specify the name of the table or view to be locked.
If you specify view, then Oracle Database locks the base tables of the view.
If you specify the partition_extension_clause, then Oracle Database first acquires
an implicit lock on the table. The table lock is the same as the lock you specify for the
partition or subpartition, with two exceptions:
■

■

If you specify a SHARE lock for the subpartition, then the database acquires an
implicit ROW SHARE lock on the table.
If you specify an EXCLUSIVE lock for the subpartition, then the database acquires
an implicit ROW EXCLUSIVE lock on the table.

If you specify PARTITION and table is composite-partitioned, then the database
acquires locks on all the subpartitions of the partition.
Restriction on Locking Tables If view is part of a hierarchy, then it must be the root

of the hierarchy.

dblink
Specify a database link to a remote Oracle Database where the table or view is located.
You can lock tables and views on a remote database only if you are using Oracle
distributed functionality. All tables locked by a LOCK TABLE statement must be on the
same database.
If you omit dblink, then Oracle Database assumes the table or view is on the local
database.
See Also: "References to Objects in Remote Databases" on page 3-117
for information on specifying database links

lockmode Clause
Specify one of the following modes:
ROW SHARE ROW SHARE permits concurrent access to the locked table but prohibits
users from locking the entire table for exclusive access. ROW SHARE is synonymous with

18-72 Oracle Database SQL Language Reference

LOCK TABLE

SHARE UPDATE, which is included for compatibility with earlier versions of Oracle
Database.
ROW EXCLUSIVE is the same as ROW SHARE, but it also prohibits
locking in SHARE mode. ROW EXCLUSIVE locks are automatically obtained when
updating, inserting, or deleting.

ROW EXCLUSIVE

SHARE UPDATE
SHARE

See ROW SHARE on page 18-72.

SHARE permits concurrent queries but prohibits updates to the locked table.

SHARE ROW EXCLUSIVE SHARE ROW EXCLUSIVE is used to look at a whole table and
to allow others to look at rows in the table but to prohibit others from locking the table
in SHARE mode or from updating rows.
EXCLUSIVE

EXCLUSIVE permits queries on the locked table but prohibits any other

activity on it.

NOWAIT
Specify NOWAIT if you want the database to return control to you immediately if the
specified table, partition, or table subpartition is already locked by another user. In this
case, the database returns a message indicating that the table, partition, or subpartition
is already locked by another user.

WAIT
Use the WAIT clause to indicate that the LOCK TABLE statement should wait up to the
specified number of seconds to acquire a DML lock. There is no limit on the value of
integer.
If you specify neither NOWAIT nor WAIT, then the database waits indefinitely until the
table is available, locks it, and returns control to you. When the database is executing
DDL statements concurrently with DML statements, a timeout or deadlock can
sometimes result. The database detects such timeouts and deadlocks and returns an
error.
See Also: Oracle Database Administrator's Guide for more information
about locking tables

Examples
18

The following statement locks the employees table in
exclusive mode but does not wait if another user already has locked the table:

Locking a Table: Example
LOCK TABLE employees
IN EXCLUSIVE MODE
NOWAIT;

The following statement locks the remote employees table that is accessible through
the database link remote:
LOCK TABLE employees@remote
IN SHARE MODE;

SQL Statements: DROP SEQUENCE to ROLLBACK 18-73

MERGE

MERGE
Purpose
18

Use the MERGE statement to select rows from one or more sources for update or
insertion into a table or view. You can specify conditions to determine whether to
update or insert into the target table or view.
This statement is a convenient way to combine multiple operations. It lets you avoid
multiple INSERT, UPDATE, and DELETE DML statements.
MERGE is a deterministic statement. You cannot update the same row of the target table
multiple times in the same MERGE statement.
In previous releases of Oracle Database, when you created an
Oracle Virtual Private Database policy on an application that included
the MERGE INTO statement, the MERGE INTO statement would be
prevented with an ORA-28132: Merge into syntax does not
support security policies error, due to the presence of the Virtual
Private Database policy. Beginning with Oracle Database 11g Release 2
(11.2.0.2), you can create policies on applications that include MERGE
INTO operations. To do so, in the DBMS_RLS.ADD_POLICY statement_
types parameter, include the INSERT, UPDATE, and DELETE statements,
or just omit the statement_types parameter altogether. Refer to Oracle
Database Security Guide for more information on enforcing policies on
specific SQL statement types.
Note:

Prerequisites
18

You must have the INSERT and UPDATE object privileges on the target table and the
SELECT object privilege on the source table. To specify the DELETE clause of the merge_
update_clause, you must also have the DELETE object privilege on the target table.

Syntax
18

merge::=
hint
MERGE

schema

.

t_alias

table

INTO
view

schema

.

table
view

USING

t_alias
ON

(

condition

)

subquery

merge_update_clause

merge_insert_clause

error_logging_clause
;

18-74 Oracle Database SQL Language Reference

MERGE

You must specify at least one of the clauses merge_update_
clause or merge_insert_clause.

Note:

(merge_update_clause::= on page 18-75, merge_insert_clause::= on page 18-75, error_
logging_clause::= on page 18-75
merge_update_clause::=
,
expr
WHEN

MATCHED

THEN

UPDATE

SET

column

=
DEFAULT

where_clause

DELETE

where_clause

merge_insert_clause::=
,
(
WHEN

NOT

MATCHED

THEN

column

)

INSERT

,
where_clause

expr
VALUES

(

)
DEFAULT

where_clause::=
WHERE

condition

error_logging_clause::=
schema
INTO
LOG

.
table

(

simple_expression

)

ERRORS

integer
REJECT

LIMIT
UNLIMITED

Semantics
18

INTO Clause
Use the INTO clause to specify the target table or view you are updating or inserting
into. In order to merge data into a view, the view must be updatable. Refer to "Notes
on Updatable Views" on page 17-20 for more information.

SQL Statements: DROP SEQUENCE to ROLLBACK 18-75

MERGE

Restriction on Target Views

You cannot specify a target view on which an INSTEAD

OF trigger has been defined.

USING Clause
Use the USING clause to specify the source of the data to be updated or inserted. The
source can be a table, view, or the result of a subquery.

ON Clause
Use the ON clause to specify the condition upon which the MERGE operation either
updates or inserts. For each row in the target table for which the search condition is
true, Oracle Database updates the row with corresponding data from the source table.
If the condition is not true for any rows, then the database inserts into the target table
based on the corresponding source table row.
Restrictions on the ON Clause In previous releases of Oracle Database, when you
created an Oracle Virtual Private Database policy on an application that included the
MERGE INTO statement, the MERGE INTO statement would be prevented with an
ORA-28132: Merge into syntax does not support security policies error, due to
the presence of the Virtual Private Database policy. Beginning with Oracle Database
11g Release 2 (11.2.0.2), you can create policies on applications that include MERGE INTO
operations. To do so, in the DBMS_RLS.ADD_POLICY statement_types parameter, include
the INSERT, UPDATE, and DELETE statements, or just omit the statement_types
parameter altogether. Refer to Oracle Database Security Guide for more information on
enforcing policies on specific SQL statement types.

merge_update_clause
The merge_update_clause specifies the new column values of the target table. Oracle
performs this update if the condition of the ON clause is true. If the update clause is
executed, then all update triggers defined on the target table are activated.
Specify the where_clause if you want the database to execute the update operation
only if the specified condition is true. The condition can refer to either the data source
or the target table. If the condition is not true, then the database skips the update
operation when merging the row into the table.
Specify the DELETE where_clause to clean up data in a table while populating or
updating it. The only rows affected by this clause are those rows in the destination
table that are updated by the merge operation. The DELETE WHERE condition evaluates
the updated value, not the original value that was evaluated by the UPDATE SET ...
WHERE condition. If a row of the destination table meets the DELETE condition but is not
included in the join defined by the ON clause, then it is not deleted. Any delete triggers
defined on the target table will be activated for each row deletion.
You can specify this clause by itself or with the merge_insert_clause. If you specify
both, then they can be in either order.
Restrictions on the merge_update_clause

This clause is subject to the following

restrictions:
■

You cannot update a column that is referenced in the ON condition clause.

■

You cannot specify DEFAULT when updating a view.

merge_insert_clause
The merge_insert_clause specifies values to insert into the column of the target table
if the condition of the ON clause is false. If the insert clause is executed, then all insert

18-76 Oracle Database SQL Language Reference

MERGE

triggers defined on the target table are activated. If you omit the column list after the
INSERT keyword, then the number of columns in the target table must match the
number of values in the VALUES clause.
To insert all of the source rows into the table, you can use a constant filter predicate in
the ON clause condition. An example of a constant filter predicate is ON (0=1). Oracle
Database recognizes such a predicate and makes an unconditional insert of all source
rows into the table. This approach is different from omitting the merge_update_clause.
In that case, the database still must perform a join. With constant filter predicate, no
join is performed.
Specify the where_clause if you want Oracle Database to execute the insert operation
only if the specified condition is true. The condition can refer only to the data source
table. Oracle Database skips the insert operation for all rows for which the condition is
not true.
You can specify this clause by itself or with the merge_update_clause. If you specify
both, then they can be in either order.
Restriction on the merge_insert_clause

You cannot specify DEFAULT when inserting

into a view.

error_logging_clause
The error_logging_clause has the same behavior in a MERGE statement as in an INSERT
statement. Refer to the INSERT statement error_logging_clause on page 18-65 for more
information.
See Also:

"Inserting Into a Table with Error Logging: Example" on

page 18-66

Examples
18

Merging into a Table: Example The following example uses the bonuses table in the

sample schema oe with a default bonus of 100. It then inserts into the bonuses table all
employees who made sales, based on the sales_rep_id column of the oe.orders
table. Finally, the human resources manager decides that employees with a salary of
$8000 or less should receive a bonus. Those who have not made sales get a bonus of
1% of their salary. Those who already made sales get an increase in their bonus equal
to 1% of their salary. The MERGE statement implements these changes in one step:
CREATE TABLE bonuses (employee_id NUMBER, bonus NUMBER DEFAULT 100);
INSERT INTO bonuses(employee_id)
(SELECT e.employee_id FROM employees e, orders o
WHERE e.employee_id = o.sales_rep_id
GROUP BY e.employee_id);
SELECT * FROM bonuses ORDER BY employee_id;
EMPLOYEE_ID
BONUS
----------- ---------153
100
154
100
155
100
156
100
158
100
159
100
160
100

SQL Statements: DROP SEQUENCE to ROLLBACK 18-77

MERGE

161
163

100
100

MERGE INTO bonuses D
USING (SELECT employee_id, salary, department_id FROM employees
WHERE department_id = 80) S
ON (D.employee_id = S.employee_id)
WHEN MATCHED THEN UPDATE SET D.bonus = D.bonus + S.salary*.01
DELETE WHERE (S.salary > 8000)
WHEN NOT MATCHED THEN INSERT (D.employee_id, D.bonus)
VALUES (S.employee_id, S.salary*.01)
WHERE (S.salary <= 8000);
SELECT * FROM bonuses ORDER BY employee_id;
EMPLOYEE_ID
BONUS
----------- ---------153
180
154
175
155
170
159
180
160
175
161
170
164
72
165
68
166
64
167
62
171
74
172
73
173
61
179
62

18-78 Oracle Database SQL Language Reference

NOAUDIT

NOAUDIT
Purpose
18

Use the NOAUDIT statement to stop auditing operations previously enabled by the
AUDIT statement.
The NOAUDIT statement must have the same syntax as the previous AUDIT statement.
Further, it reverses the effects only of that particular statement. For example, suppose
one AUDIT statement A enables auditing for a specific user. A second statement B
enables auditing for all users. A NOAUDIT statement C to disable auditing for all users
reverses statement B. However, statement C leaves statement A in effect and continues
to audit the user that statement A specified.
See Also:

AUDIT on page 13-29 for more information on auditing

Prerequisites
18

To stop auditing of SQL statements, you must have the AUDIT SYSTEM system privilege.
To stop auditing of schema objects, you must be the owner of the object on which you
stop auditing or you must have the AUDIT ANY system privilege. In addition, if the
object you chose for auditing is a directory, then even if you created it, you must have
the AUDIT ANY system privilege.

Syntax
18

noaudit::=
auditing_by_clause
audit_operation_clause
NOAUDIT

NOT
WHENEVER

SUCCESSFUL

audit_schema_object_clause

;

NETWORK

(audit_operation_clause::= on page 18-79, auditing_by_clause::= on page 18-80, audit_
schema_object_clause::= on page 18-80)
audit_operation_clause::=
,
sql_statement_shortcut
ALL
ALL

STATEMENTS
,

system_privilege
ALL

PRIVILEGES

SQL Statements: DROP SEQUENCE to ROLLBACK 18-79

NOAUDIT

auditing_by_clause::=
,
BY

user

audit_schema_object_clause::=
,
sql_operation
auditing_on_clause
ALL

auditing_on_clause::=
schema

.
object

DIRECTORY

directory_name

ON

schema
MINING

.

MODEL

model

DEFAULT

Semantics
18

audit_operation_clause
Use the audit_operation_clause to stop auditing of a particular SQL statement.
statement_option For sql_statement_shortcut, specify the shortcut for the SQL

statements for which auditing is to be stopped. Refer to Table 13–1 on page 13-34 and
Table 13–2 on page 13-37 for a list of the SQL statement shortcuts and the SQL
statements they audit.
Specify ALL to stop auditing of all statement options currently being audited
because of an earlier AUDIT ALL ... statement. You cannot use this clause to reverse an
earlier AUDIT ALL STATEMENTS ... statement.
ALL

ALL STATEMENTS Specify ALL STATEMENTS to reverse an earlier AUDIT ALL

STATEMENTS ... statement. You cannot use this clause to reverse an earlier AUDIT ALL
... statement.
For system_privilege, specify the system privilege for which
auditing is to be stopped. Refer to Table 18–1 on page 18-40 for a list of the system
privileges and the statements they authorize.

system_privilege

ALL PRIVILEGES Specify ALL PRIVILEGES to stop auditing of all system privileges
currently being audited.

auditing_by_clause
Use the auditing_by_clause to stop auditing only for SQL statements issued by the
specified users in their subsequent sessions. If you omit this clause, then Oracle
Database stops auditing for all users' statements, except for the situation described for
WHENEVER SUCCESSFUL.

18-80 Oracle Database SQL Language Reference

NOAUDIT

audit_schema_object_clause
Use the audit_schema_object_clause to stop auditing of a particular database object.
sql_operation For sql_operation, specify the type of operation for which auditing is
to be stopped on the object specified in the ON clause. Refer to Table 13–3 on page 13-38
for a list of these options.
ALL Specify ALL as a shortcut equivalent to specifying all SQL operations applicable
for the type of object.
auditing_on_clause The auditing_on_clause lets you specify the particular schema
object for which auditing is to be stopped.
■

■

■

For object, specify the object name of a table, view, sequence, stored procedure,
function, or package, materialized view, or library. If you do not qualify object
with schema, then Oracle Database assumes the object is in your own schema.
Refer to AUDIT on page 13-29 for information on auditing specific schema objects.
The DIRECTORY clause lets you specify the name of the directory on which auditing
is to be stopped.
Specify DEFAULT to remove the specified object options as default object options for
subsequently created objects.

NETWORK

Use this clause to discontinue auditing of database link usage and logins.

WHENEVER [NOT] SUCCESSFUL Specify WHENEVER SUCCESSFUL to stop auditing
only for SQL statements and operations on schema objects that complete successfully.

Specify WHENEVER NOT SUCCESSFUL to stop auditing only for statements and operations
that result in Oracle Database errors.
If you omit this clause, then the database stops auditing for all statements or
operations, regardless of success or failure.

Examples
18

Stop Auditing of SQL Statements Related to Roles: Example If you have chosen
auditing for every SQL statement that creates or drops a role, then you can stop
auditing of such statements by issuing the following statement:
NOAUDIT ROLE;

Stop Auditing of Updates or Queries on Objects Owned by a Particular User:
Example If you have chosen auditing for any statement that queries or updates any

table issued by the users hr and oe, then you can stop auditing for queries by hr by
issuing the following statement:
NOAUDIT SELECT TABLE BY hr;

The preceding statement stops auditing only queries by hr, so the database continues
to audit queries and updates by oe as well as updates by hr.
Stop Auditing of Statements Authorized by a Particular Object Privilege: Example

To stop auditing on all statements that are authorized by DELETE ANY TABLE system
privilege, issue the following statement:
NOAUDIT DELETE ANY TABLE;

SQL Statements: DROP SEQUENCE to ROLLBACK 18-81

NOAUDIT

If you have chosen
auditing for every SQL statement that queries the employees table in the schema hr,
then you can stop auditing for such queries by issuing the following statement:

Stop Auditing of Queries on a Particular Object: Example

NOAUDIT SELECT
ON hr.employees;

You can stop
auditing for queries that complete successfully by issuing the following statement:
Stop Auditing of Queries that Complete Successfully: Example
NOAUDIT SELECT
ON hr.employees
WHENEVER SUCCESSFUL;

This statement stops auditing only for successful queries. Oracle Database continues to
audit queries resulting in Oracle Database errors.

18-82 Oracle Database SQL Language Reference

PURGE

PURGE
Purpose
18

Use the PURGE statement to remove a table or index from your recycle bin and release
all of the space associated with the object, or to remove the entire recycle bin, or to
remove part of all of a dropped tablespace from the recycle bin.
You cannot roll back a PURGE statement, nor can you
recover an object after it is purged.

Caution:

To see the contents of your recycle bin, query the USER_RECYCLEBIN data dictionary
view. You can use the RECYCLEBIN synonym instead. The following two statements
return the same rows:
SELECT * FROM RECYCLEBIN;
SELECT * FROM USER_RECYCLEBIN;

See Also:
■

■

■

Oracle Database Administrator's Guide for information on the
recycle bin and naming conventions for objects in the recycle bin
FLASHBACK TABLE on page 18-27 for information on retrieving
dropped tables from the recycle bin
Oracle Database Reference for information on using the RECYCLEBIN
initialization parameter to control whether dropped tables go into
the recycle bin

Prerequisites
18

The database object must reside in your own schema or you must have the DROP ANY ...
system privilege for the type of object to be purged, or you must have the SYSDBA
system privilege.

Syntax
18

purge::=
TABLE

table

INDEX

index

RECYCLEBIN
PURGE

;
DBA_RECYCLEBIN
USER
TABLESPACE

user

tablespace

SQL Statements: DROP SEQUENCE to ROLLBACK 18-83

PURGE

Semantics
18

TABLE or INDEX
Specify the name of the table or index in the recycle bin that you want to purge. You
can specify either the original user-specified name or the system-generated name
Oracle Database assigned to the object when it was dropped.
■

■

If you specify the user-specified name, and if the recycle bin contains more than
one object of that name, then the database purges the object that has been in the
recycle bin the longest.
System-generated recycle bin object names are unique. Therefore, if you specify
the system-generated name, then the database purges that specified object.

When the database purges a table, all table partitions, LOBs and LOB partitions,
indexes, and other dependent objects of that table are also purged.

RECYCLEBIN
Use this clause to purge the current user's recycle bin. Oracle Database will remove all
objects from the user's recycle bin and release all space associated with objects in the
recycle bin.

DBA_RECYCLEBIN
This clause is valid only if you have SYSDBA system privilege. It lets you remove all
objects from the system-wide recycle bin, and is equivalent to purging the recycle bin
of every user. This operation is useful, for example, before backward migration.

TABLESPACE tablespace
Use this clause to purge all the objects residing in the specified tablespace from the
recycle bin.
Use this clause to reclaim space in a tablespace for a specified user. This
operation is useful when a particular user is running low on disk quota for the
specified tablespace.

USER user

Examples
18

Remove a File From Your Recycle Bin: Example The following statement removes

the table test from the recycle bin. If more than one version of test resides in the
recycle bin, then Oracle Database removes the version that has been there the longest:
PURGE TABLE test;

To determine system-generated name of the table you want removed from your
recycle bin, issue a SELECT statement on your recycle bin. Using that object name, you
can remove the table by issuing a statement similar to the following statement. (The
system-generated name will differ from the one shown in the example.)
PURGE TABLE RB$$33750$TABLE$0;

Remove the Contents of Your Recycle Bin: Example To remove the entire contents
of your recycle bin, issue the following statement:
PURGE RECYCLEBIN;

18-84 Oracle Database SQL Language Reference

RENAME

RENAME
Purpose
18

Caution:

You cannot roll back a RENAME statement.

Use the RENAME statement to rename a table, view, sequence, or private synonym.
■

■

Oracle Database automatically transfers integrity constraints, indexes, and grants
on the old object to the new object.
Oracle Database invalidates all objects that depend on the renamed object, such as
views, synonyms, and stored procedures and functions that refer to a renamed
table.
See Also: CREATE SYNONYM on page 16-2 and DROP SYNONYM
on page 18-3

Prerequisites
18

The object must be in your own schema.

Syntax
18

rename::=
RENAME

old_name

TO

new_name

;

Semantics
18

old_name
Specify the name of an existing table, view, sequence, or private synonym.

new_name
Specify the new name to be given to the existing object. The new name must not
already be used by another schema object in the same namespace and must follow the
rules for naming schema objects.
Restrictions on Renaming Objects Renaming objects is subject to the following
restrictions:
■

■

You cannot rename a public synonym. Instead, drop the public synonym and then
re-create the public synonym with the new name.
You cannot rename a type synonym that has any dependent tables or dependent
valid user-defined object types.
See Also:

"Database Object Naming Rules" on page 3-111

SQL Statements: DROP SEQUENCE to ROLLBACK 18-85

RENAME

Example
18

Renaming a Database Object: Example The following example uses a copy of the
sample table hr.departments. To change the name of table departments_new to emp_
departments, issue the following statement:
RENAME departments_new TO emp_departments;

You cannot use this statement directly to rename columns. However, you can rename a
column using the ALTER TABLE ... rename_column_clause.
See Also:

rename_column_clause on page 12-52

Another way to rename a column is to use the RENAME statement together with the
CREATE TABLE statement with AS subquery. This method is useful if you are changing
the structure of a table rather than only renaming a column. The following statements
re-create the sample table hr.job_history, renaming a column from department_id to
dept_id:
CREATE TABLE temporary
(employee_id, start_date, end_date, job_id, dept_id)
AS SELECT
employee_id, start_date, end_date, job_id, department_id
FROM job_history;
DROP TABLE job_history;
RENAME temporary TO job_history;

Any integrity constraints defined on table job_history will be lost in the preceding
example. You will have to redefine them on the new job_history table using an ALTER
TABLE statement.

18-86 Oracle Database SQL Language Reference

REVOKE

REVOKE
Purpose
18

Use the REVOKE statement to:
■

Revoke system privileges from users and roles

■

Revoke roles from users and roles

■

Revoke object privileges for a particular object from users and roles

Note on Oracle Automatic Storage Management A user authenticated AS SYSASM can

use this statement to revoke the system privileges SYSASM, SYSOPER, and SYSDBA from a
user in the Oracle ASM password file of the current node.
Note on Editionable Objects A REVOKE operation to revoke object privileges on an
editionable object actualizes the object in the current edition. See Oracle Database
Advanced Application Developer's Guide for more information about editions and
editionable objects.
See Also:
■

■

GRANT on page 18-33 for information on granting system
privileges and roles
Table 18–2 on page 18-48 for a listing of the object privileges for
each type of object

Prerequisites
18

To revoke a system privilege, you must have been granted the privilege with the
ADMIN OPTION. You can revoke any privilege if you have the GRANT ANY PRIVILEGE
system privilege.
To revoke a role, you must have been granted the role with the ADMIN OPTION. You can
revoke any role if you have the GRANT ANY ROLE system privilege.
To revoke an object privilege, one of the following conditions must be met:
■
■

You must previously have granted the object privilege to the user or role.
You must have the GRANT ANY OBJECT PRIVILEGE system privilege. In this case, you
can revoke any object privilege that was granted by the object owner or on behalf
of the owner by a user with the GRANT ANY OBJECT PRIVILEGE. However, you
cannot revoke an object privilege that was granted by way of a WITH GRANT OPTION
grant.
See Also: "Revoke Operations that Use GRANT ANY OBJECT
PRIVILEGE: Example" on page 18-94

The REVOKE statement can revoke only privileges and roles that were previously
granted directly with a GRANT statement. You cannot use this statement to revoke:
■

Privileges or roles not granted to the revokee

■

Roles or object privileges granted through the operating system

■

Privileges or roles granted to the revokee through roles

SQL Statements: DROP SEQUENCE to ROLLBACK 18-87

REVOKE

Syntax
18

revoke::=
revoke_system_privilege
REVOKE

;
revoke_object_privileges

(revoke_system_privileges::= on page 18-88, revoke_object_privileges::= on page 18-88)
revoke_system_privileges::=
,
system_privilege
role

FROM

ALL

grantee_clause

PRIVILEGES

(grantee_clause::= on page 18-88)
revoke_object_privileges::=
,
object_privilege
on_object_clause

PRIVILEGES
ALL

CASCADE

CONSTRAINTS

FORCE
FROM

grantee_clause

(on_object_clause::= on page 18-89, grantee_clause::= on page 18-88)
grantee_clause::=
,
IDENTIFIED

BY

password

user
role
PUBLIC

18-88 Oracle Database SQL Language Reference

REVOKE

on_object_clause::=
schema

.
object

DIRECTORY
EDITION

directory_name
edition_name

ON

schema
MINING

.

MODEL

mining_model_name

SOURCE

schema

.

JAVA

object
RESOURCE

Semantics
18

revoke_system_privileges
Use these clauses to revoke system privileges.
system_privilege
Specify the system privilege to be revoked. Refer to Table 18–1 on page 18-40 for a list
of the system privileges.
If you revoke a system privilege from a user, then the database removes the privilege
from the user's privilege domain. Effective immediately, the user cannot exercise the
privilege.
If you revoke a system privilege from a role, then the database removes the privilege
from the privilege domain of the role. Effective immediately, users with the role
enabled cannot exercise the privilege. Also, other users who have been granted the
role and subsequently enable the role cannot exercise the privilege.
See Also: "Revoking a System Privilege from a User: Example" on
page 18-92 and "Revoking a System Privilege from a Role: Example"
on page 18-93

If you revoke a system privilege from PUBLIC, then the database removes the privilege
from the privilege domain of each user who has been granted the privilege through
PUBLIC. Effective immediately, such users can no longer exercise the privilege.
However, the privilege is not revoked from users who have been granted the privilege
directly or through roles.
Oracle Database provides a shortcut for specifying all system privileges at once:
Specify ALL PRIVILEGES to revoke all the system privileges listed in Table 18–1 on
page 18-40.
Restriction on Revoking System Privileges A system privilege cannot appear more

than once in the list of privileges to be revoked.
role
Specify the role to be revoked.
If you revoke a role from a user, then the database makes the role unavailable to the
user. If the role is currently enabled for the user, then the user can continue to exercise
the privileges in the role's privilege domain as long as it remains enabled. However,
the user cannot subsequently enable the role.
SQL Statements: DROP SEQUENCE to ROLLBACK 18-89

REVOKE

If you revoke a role from another role, then the database removes the privilege domain
of the revoked role from the privilege domain of the revokee role. Users who have
been granted and have enabled the revokee role can continue to exercise the privileges
in the privilege domain of the revoked role as long as the revokee role remains
enabled. However, other users who have been granted the revokee role and
subsequently enable it cannot exercise the privileges in the privilege domain of the
revoked role.
See Also: "Revoking a Role from a User: Example" on page 18-92
and "Revoking a Role from a Role: Example" on page 18-93

If you revoke a role from PUBLIC, then the database makes the role unavailable to all
users who have been granted the role through PUBLIC. Any user who has enabled the
role can continue to exercise the privileges in its privilege domain as long as it remains
enabled. However, users cannot subsequently enable the role. The role is not revoked
from users who have been granted the role directly or through other roles.
A system role cannot appear more than once
in the list of roles to be revoked. For information on the predefined roles, refer to
Oracle Database Security Guide.

Restriction on Revoking System Roles

grantee_clause
FROM grantee_clause identifies users or roles from which the system privilege, role, or
object privilege is to be revoked.
PUBLIC

Specify PUBLIC to revoke the privileges or roles from all users.

revoke_object_privileges
Use these clauses to revoke object privileges.
object_privilege
Specify the object privilege to be revoked. You can substitute any of the object
privileges described on Table 18–2 on page 18-48.
Each privilege authorizes some operation. By revoking a
privilege, you prevent the revokee from performing that operation.
However, multiple users may grant the same privilege to the same
user, role, or PUBLIC. To remove the privilege from the grantee's
privilege domain, all grantors must revoke the privilege. If even one
grantor does not revoke the privilege, then the grantee can still
exercise the privilege by virtue of that grant.

Note:

If you revoke an object privilege from a user, then the database removes the privilege
from the user's privilege domain. Effective immediately, the user cannot exercise the
privilege.
■

■

If that user has granted that privilege to other users or roles, then the database also
revokes the privilege from those other users or roles.
If that user's schema contains a procedure, function, or package that contains SQL
statements that exercise the privilege, then the procedure, function, or package can
no longer be executed.

18-90 Oracle Database SQL Language Reference

REVOKE

■

■

If that user's schema contains a view on that object, then the database invalidates
the view.
If you revoke the REFERENCES object privilege from a user who has exercised the
privilege to define referential integrity constraints, then you must specify the
CASCADE CONSTRAINTS clause.

If you revoke an object privilege from a role, then the database removes the privilege
from the privilege domain of the role. Effective immediately, users with the role
enabled cannot exercise the privilege. Other users who have been granted the role
cannot exercise the privilege after enabling the role.
If you revoke an object privilege from PUBLIC, then the database removes the privilege
from the privilege domain of each user who has been granted the privilege through
PUBLIC. Effective immediately, all such users are restricted from exercising the
privilege. However, the privilege is not revoked from users who have been granted the
privilege directly or through roles.
ALL [PRIVILEGES]
Specify ALL to revoke all object privileges that you have granted to the revokee. (The
keyword PRIVILEGES is provided for semantic clarity and is optional.)
If no privileges have been granted on the object, then the database takes no action and
does not return an error.
Restriction on Revoking Object Privileges A privilege cannot appear more than once

in the list of privileges to be revoked. A user, a role, or PUBLIC cannot appear more
than once in the FROM clause.
See Also: "Revoking an Object Privilege from a User: Example" on
page 18-93, "Revoking Object Privileges from PUBLIC: Example" on
page 18-93, and "Revoking All Object Privileges from a User:
Example" on page 18-93

CASCADE CONSTRAINTS
This clause is relevant only if you revoke the REFERENCES privilege or ALL
[PRIVILEGES]. It drops any referential integrity constraints that the revokee has defined
using the REFERENCES privilege, which might have been granted either explicitly or
implicitly through a grant of ALL [PRIVILEGES].
See Also: "Revoking an Object Privilege with CASCADE
CONSTRAINTS: Example" on page 18-94

FORCE
Specify FORCE to revoke the EXECUTE object privilege on user-defined type objects with
table or type dependencies. You must use FORCE to revoke the EXECUTE object privilege
on user-defined type objects with table dependencies.
If you specify FORCE, then all privileges are revoked, all dependent objects are marked
INVALID, data in dependent tables becomes inaccessible, and all dependent
function-based indexes are marked UNUSABLE. Regranting the necessary type privilege
will revalidate the table.
Oracle Database Concepts for detailed information about
type dependencies and user-defined object privileges

See Also:

SQL Statements: DROP SEQUENCE to ROLLBACK 18-91

REVOKE

on_object_clause
The on_object_clause identifies the objects on which privileges are to be revoked.
object Specify the object on which the object privileges are to be revoked. This object
can be:
■
■

■

A table, view, sequence, procedure, stored function, package, or materialized view
A synonym for a table, view, sequence, procedure, stored function, package,
materialized view, or user-defined type
A library, indextype, or user-defined operator

If you do not qualify object with schema, then the database assumes the object is in
your own schema.
See Also: "Revoking an Object Privilege on a Sequence from a User:
Example" on page 18-93

If you revoke the SELECT object privilege on the containing table or materialized view
of a materialized view, whether the privilege was granted with or without the GRANT
OPTION, then the database invalidates the materialized view.
If you revoke the SELECT object privilege on any of the master tables of a materialized
view, whether the privilege was granted with or without the GRANT OPTION, then the
database invalidates both the materialized view and its containing table or
materialized view.
DIRECTORY directory_name Specify the directory object on which privileges are to
be revoked. You cannot qualify directory_name with schema. The object must be a
directory.
See Also: CREATE DIRECTORY on page 14-41 and "Revoking an
Object Privilege on a Directory from a User: Example" on page 18-94
EDITION Specify the name of the edition on which the USE object privilege is to be
revoked. You cannot qualify edition_name with schema.

The JAVA clause lets you specify a Java source or
resource schema object on which privileges are to be revoked.

JAVA SOURCE | RESOURCE

Examples
18

Revoking a System Privilege from a User: Example The following statement revokes
the DROP ANY TABLE system privilege from the users hr and oe:
REVOKE DROP ANY TABLE
FROM hr, oe;

The users hr and oe can no longer drop tables in schemas other than their own.
Revoking a Role from a User: Example

The following statement revokes the role dw_

manager from the user sh:
REVOKE dw_manager
FROM sh;

The user sh can no longer enable the dw_manager role.

18-92 Oracle Database SQL Language Reference

REVOKE

Revoking a System Privilege from a Role: Example The following statement revokes

the CREATE TABLESPACE system privilege from the dw_manager role:
REVOKE CREATE TABLESPACE
FROM dw_manager;

Enabling the dw_manager role no longer allows users to create tablespaces.
Revoking a Role from a Role: Example

To revoke the role dw_user from the role dw_

manager, issue the following statement:
REVOKE dw_user
FROM dw_manager;

The dw_user role privileges are no longer granted to dw_manager.
You can grant DELETE, INSERT,
SELECT, and UPDATE privileges on the table orders to the user hr with the following
statement:
Revoking an Object Privilege from a User: Example

GRANT ALL
ON orders TO hr;

To revoke the DELETE privilege on orders from hr, issue the following statement:
REVOKE DELETE
ON orders FROM hr;

To revoke the remaining
privileges on orders that you granted to hr, issue the following statement:

Revoking All Object Privileges from a User: Example
REVOKE ALL
ON orders FROM hr;

You can grant SELECT and
UPDATE privileges on the view emp_details_view to all users by granting the privileges
to the role PUBLIC:
Revoking Object Privileges from PUBLIC: Example

GRANT SELECT, UPDATE
ON emp_details_view TO public;

The following statement revokes UPDATE privilege on emp_details_view from all users:
REVOKE UPDATE
ON emp_details_view FROM public;

Users can no longer update the emp_details_view view, although users can still query
it. However, if you have also granted the UPDATE privilege on emp_details_view to any
users, either directly or through roles, then these users retain the privilege.
Revoking an Object Privilege on a Sequence from a User: Example You can grant
the user oe the SELECT privilege on the departments_seq sequence in the schema hr
with the following statement:
GRANT SELECT
ON hr.departments_seq TO oe;

To revoke the SELECT privilege on departments_seq from oe, issue the following
statement:
REVOKE SELECT
ON hr.departments_seq FROM oe;

SQL Statements: DROP SEQUENCE to ROLLBACK 18-93

REVOKE

However, if the user hr has also granted SELECT privilege on departments to sh, then
sh can still use departments by virtue of hr's grant.
Revoking an Object Privilege with CASCADE CONSTRAINTS: Example You can
grant to oe the privileges REFERENCES and UPDATE on the employees table in the schema
hr with the following statement:
GRANT REFERENCES, UPDATE
ON hr.employees TO oe;

The user oe can exercise the REFERENCES privilege to define a constraint in his or her
own dependent table that refers to the employees table in the schema hr:
CREATE TABLE dependent
(dependno
NUMBER,
dependname VARCHAR2(10),
employee
NUMBER
CONSTRAINT in_emp REFERENCES hr.employees(employee_id) );

You can revoke the REFERENCES privilege on hr.employees from oe by issuing the
following statement that contains the CASCADE CONSTRAINTS clause:
REVOKE REFERENCES
ON hr.employees
FROM oe
CASCADE CONSTRAINTS;

Revoking oe's REFERENCES privilege on hr.employees causes Oracle Database to drop
the in_emp constraint, because oe required the privilege to define the constraint.
However, if oe has also been granted the REFERENCES privilege on hr.employees by a
user other than you, then the database does not drop the constraint. oe still has the
privilege necessary for the constraint by virtue of the other user's grant.
Revoking an Object Privilege on a Directory from a User: Example You can revoke

the READ object privilege on directory bfile_dir from hr by issuing the following
statement:
REVOKE READ ON DIRECTORY bfile_dir FROM hr;

Suppose
that the database administrator has granted GRANT ANY OBJECT PRIVILEGE to user sh.
Now suppose that user hr grants the update privilege on the employees table to oe:

Revoke Operations that Use GRANT ANY OBJECT PRIVILEGE: Example

CONNECT hr
GRANT UPDATE ON employees TO oe WITH GRANT OPTION;

This grant gives user oe the right to pass the object privilege along to another user:
CONNECT oe
GRANT UPDATE ON hr.employees TO pm;

User sh, who has the GRANT ANY OBJECT PRIVILEGE, can now act on behalf of user hr
and revoke the update privilege from user oe, because oe was granted the privilege by
hr:
CONNECT sh
REVOKE UPDATE ON hr.employees FROM oe;

18-94 Oracle Database SQL Language Reference

REVOKE

User sh cannot revoke the update privilege from user pm explicitly, because pm received
the grant neither from the object owner (hr), nor from sh, nor from another user with
GRANT ANY OBJECT PRIVILEGE, but from user oe. However, the preceding statement
cascades, removing all privileges that depend on the one revoked. Therefore the object
privilege is implicitly revoked from pm as well.

SQL Statements: DROP SEQUENCE to ROLLBACK 18-95

ROLLBACK

ROLLBACK
Purpose
18

Use the ROLLBACK statement to undo work done in the current transaction or to
manually undo the work done by an in-doubt distributed transaction.
Oracle recommends that you explicitly end transactions in
application programs using either a COMMIT or ROLLBACK statement. If
you do not explicitly commit the transaction and the program
terminates abnormally, then Oracle Database rolls back the last
uncommitted transaction.

Note:

See Also:
■
■

■

■

Oracle Database Concepts for information on transactions
Oracle Database Heterogeneous Connectivity User's Guide for
information on distributed transactions
SET TRANSACTION on page 19-64 for information on setting
characteristics of the current transaction
COMMIT on page 13-49 and SAVEPOINT on page 19-2

Prerequisites
18

To roll back your current transaction, no privileges are necessary.
To manually roll back an in-doubt distributed transaction that you originally
committed, you must have the FORCE TRANSACTION system privilege. To manually roll
back an in-doubt distributed transaction originally committed by another user, you
must have the FORCE ANY TRANSACTION system privilege.

Syntax
18

rollback::=
SAVEPOINT
TO
WORK

FORCE

savepoint
string

ROLLBACK

;

Semantics
18

WORK
The keyword WORK is optional and is provided for SQL standard compatibility.

TO SAVEPOINT Clause
Specify the savepoint to which you want to roll back the current transaction. If you
omit this clause, then the ROLLBACK statement rolls back the entire transaction.
Using ROLLBACK without the TO SAVEPOINT clause performs the following operations:

18-96 Oracle Database SQL Language Reference

ROLLBACK

■

Ends the transaction

■

Undoes all changes in the current transaction

■

Erases all savepoints in the transaction

■

Releases any transaction locks
See Also:

SAVEPOINT on page 19-2

Using ROLLBACK with the TO SAVEPOINT clause performs the following operations:
■

■

■

Rolls back just the portion of the transaction after the savepoint. It does not end
the transaction.
Erases all savepoints created after that savepoint. The named savepoint is retained,
so you can roll back to the same savepoint multiple times. Prior savepoints are
also retained.
Releases all table and row locks acquired since the savepoint. Other transactions
that have requested access to rows locked after the savepoint must continue to
wait until the transaction is committed or rolled back. Other transactions that have
not already requested the rows can request and access the rows immediately.

Restriction on In-doubt Transactions You cannot manually roll back an in-doubt

transaction to a savepoint.

FORCE Clause
Specify FORCE to manually roll back an in-doubt distributed transaction. The
transaction is identified by the string containing its local or global transaction ID. To
find the IDs of such transactions, query the data dictionary view DBA_2PC_PENDING.
A ROLLBACK statement with a FORCE clause rolls back only the specified transaction.
Such a statement does not affect your current transaction.
See Also: Oracle Database Administrator's Guide for more information
on distributed transactions and rolling back in-doubt transactions

Examples
18

Rolling Back Transactions: Examples

The following statement rolls back your entire

current transaction:
ROLLBACK;

The following statement rolls back your current transaction to savepoint banda_sal:
ROLLBACK TO SAVEPOINT banda_sal;

See "Creating Savepoints: Example" on page 19-2 for a full version of the preceding
example.
The following statement manually rolls back an in-doubt distributed transaction:
ROLLBACK WORK
FORCE '25.32.87';

SQL Statements: DROP SEQUENCE to ROLLBACK 18-97

ROLLBACK

18-98 Oracle Database SQL Language Reference

19
SQL Statements: SAVEPOINT to UPDATE
19

This chapter contains the following SQL statements:
■

SAVEPOINT

■

SELECT

■

SET CONSTRAINT[S]

■

SET ROLE

■

SET TRANSACTION

■

TRUNCATE CLUSTER

■

TRUNCATE TABLE

■

UPDATE

SQL Statements: SAVEPOINT to UPDATE 19-1

SAVEPOINT

SAVEPOINT
Purpose
19

Use the SAVEPOINT statement to create a name for a system change number (SCN), to
which you can later roll back.
See Also:
■
■

■

Oracle Database Concepts for information on savepoints.
ROLLBACK on page 18-96 for information on rolling back
transactions
SET TRANSACTION on page 19-64 for information on setting
characteristics of the current transaction

Prerequisites
19

None.

Syntax
19

savepoint::=
SAVEPOINT

savepoint

;

Semantics
19

savepoint
Specify the name of the savepoint to be created.
Savepoint names must be distinct within a given transaction. If you create a second
savepoint with the same identifier as an earlier savepoint, then the earlier savepoint is
erased. After a savepoint has been created, you can either continue processing, commit
your work, roll back the entire transaction, or roll back to the savepoint.

Example
19

Creating Savepoints: Example To update the salary for Banda and Greene in the
sample table hr.employees, check that the total department salary does not exceed
314,000, then reenter the salary for Greene:
UPDATE employees
SET salary = 7000
WHERE last_name = 'Banda';
SAVEPOINT banda_sal;
UPDATE employees
SET salary = 12000
WHERE last_name = 'Greene';
SAVEPOINT greene_sal;
SELECT SUM(salary) FROM employees;
ROLLBACK TO SAVEPOINT banda_sal;

19-2 Oracle Database SQL Language Reference

SAVEPOINT

UPDATE employees
SET salary = 11000
WHERE last_name = 'Greene';
COMMIT;

SQL Statements: SAVEPOINT to UPDATE 19-3

SELECT

SELECT
Purpose
19

Use a SELECT statement or subquery to retrieve data from one or more tables, object
tables, views, object views, or materialized views.
If part or all of the result of a SELECT statement is equivalent to an existing
materialized view, then Oracle Database may use the materialized view in place of one
or more tables specified in the SELECT statement. This substitution is called query
rewrite. It takes place only if cost optimization is enabled and the QUERY_REWRITE_
ENABLED parameter is set to TRUE. To determine whether query rewrite has occurred,
use the EXPLAIN PLAN statement.
See Also:
■

■

■

Chapter 9, "SQL Queries and Subqueries" for general information
on queries and subqueries
Oracle Database Data Warehousing Guide for more information on
materialized views and query rewrite
EXPLAIN PLAN on page 18-20

Prerequisites
19

For you to select data from a table or materialized view, the table or materialized view
must be in your own schema or you must have the SELECT privilege on the table or
materialized view.
For you to select rows from the base tables of a view:
■
■

You must have the SELECT privilege on the view, and
Whoever owns the schema containing the view must have the SELECT privilege on
the base tables.

The SELECT ANY TABLE system privilege also allows you to select data from any table or
any materialized view or the base table of any view.
To issue an Oracle Flashback Query using the flashback_query_clause, you must
have the SELECT privilege on the objects in the select list. In addition, either you must
have FLASHBACK object privilege on the objects in the select list, or you must have
FLASHBACK ANY TABLE system privilege.

Syntax
19

select::=
for_update_clause
subquery

;

(for_update_clause::= on page 19-13)

19-4 Oracle Database SQL Language Reference

SELECT

subquery::=
query_block
ALL
order_by_clause

UNION
subquery

INTERSECT

subquery

MINUS
(

subquery

)

(query_block::= on page 19-5, order_by_clause::= on page 19-13)
query_block::=
DISTINCT
UNIQUE
subquery_factoring_clause

hint

ALL

SELECT

select_list

,
table_reference
FROM

where_clause

hierarchical_query_clause

group_by_clause

join_clause
(

join_clause

)

model_clause

(subquery_factoring_clause::= on page 19-5, select_list::= on page 19-6, table_reference::=
on page 19-6, join_clause::= on page 19-8, where_clause::= on page 19-9, hierarchical_
query_clause::= on page 19-10, group_by_clause::= on page 19-10, model_clause::= on
page 19-11)
subquery_factoring_clause::=
,
,
(
WITH

query_name

c_alias

)

search_clause
AS

(

subquery

cycle_clause

)

SQL Statements: SAVEPOINT to UPDATE 19-5

SELECT

search_clause::=
,

DEPTH
SEARCH

FIRST

BY

ASC

NULLS

FIRST

DESC

NULLS

LAST

c_alias

BREADTH

SET

ordering_column

cycle_clause::=
,
CYCLE

c_alias

SET

cycle_mark_c_alias

TO

cycle_value

DEFAULT

no_cycle_value

select_list::=
t_alias

.
*
,

query_name
schema

.

table

.*

view
materialized view
AS
c_alias
expr

table_reference::=
ONLY

(

query_table_expression

)
flashback_query_clause

pivot_clause

t_alias

unpivot_clause
query_table_expression

(query_table_expression::= on page 19-7, flashback_query_clause::= on page 19-6, pivot_
clause::= on page 19-7, unpivot_clause::= on page 19-8)
flashback_query_clause::=
SCN
VERSIONS

expr

BETWEEN
TIMESTAMP

MINVALUE

SCN
AS

OF

expr
AND

expr
TIMESTAMP

19-6 Oracle Database SQL Language Reference

MAXVALUE

SELECT

query_table_expression::=
query_name
partition_extension_clause
@
schema

dblink

table

.

sample_clause
@

view

dblink

materialized view
subquery_restriction_clause
(

subquery

)

table_collection_expression

(subquery_restriction_clause::= on page 19-8, table_collection_expression::= on page 19-8)
pivot_clause::=
XML
PIVOT

,
AS
alias
(

aggregate_function

(

expr

)

pivot_for_clause

pivot_in_clause

)

pivot_for_clause::=
column
FOR

,
(

column

)

pivot_in_clause::=
,
AS

expr

alias
,

(
IN

(

subquery

expr

)
)

,
ANY

SQL Statements: SAVEPOINT to UPDATE 19-7

SELECT

unpivot_clause::=
INCLUDE
NULLS

column

EXCLUDE
UNPIVOT

(

,
(

pivot_for_clause

column

)

unpivot_in_clause::=
,
literal
AS

,

column
IN

(

(

literal

)

,
(

)

column

)

sample_clause::=
BLOCK

SEED

SAMPLE

(

sample_percent

)

partition_extension_clause::=
(

partition

)

PARTITION

,
FOR
(

(

partition_key_value

subpartition

)

)

SUBPARTITION

,
FOR

(

subpartition_key_value

subquery_restriction_clause::=
READ

CONSTRAINT

ONLY

constraint

WITH
CHECK

OPTION

table_collection_expression::=
(
TABLE

(

collection_expression

+

)

join_clause::=
inner_cross_join_clause
table_reference
outer_join_clause

19-8 Oracle Database SQL Language Reference

)

)

(

seed_value

)

unpivot_in_clause

)

SELECT

inner_cross_join_clause::=
ON

INNER
JOIN

condition

table_reference

,
USING

(

column

)

CROSS
JOIN

INNER

table_reference

NATURAL

(table_reference::= on page 19-6)
outer_join_clause::=
query_partition_clause

NATURAL
outer_join_type

ON

JOIN

condition
,

query_partition_clause

USING

(

column

)

table_reference

(query_partition_clause::= on page 19-9, outer_join_type::= on page 19-9, table_reference::=
on page 19-6)
query_partition_clause::=
,
expr
PARTITION

BY

,
(

expr

)

outer_join_type::=
FULL

OUTER

LEFT
RIGHT

where_clause::=
WHERE

condition

SQL Statements: SAVEPOINT to UPDATE 19-9

SELECT

hierarchical_query_clause::=
NOCYCLE
CONNECT

BY

START

WITH

condition

condition
NOCYCLE

START

WITH

condition

CONNECT

BY

condition

(condition can be any condition as described in Chapter 7, "Conditions")
group_by_clause::=
,
expr
GROUP

BY

HAVING

condition

rollup_cube_clause
grouping_sets_clause

(rollup_cube_clause::= on page 19-10, grouping_sets_clause::= on page 19-10)
rollup_cube_clause::=
ROLLUP
(

grouping_expression_list

)

CUBE

(grouping_expression_list::= on page 19-10)
grouping_sets_clause::=
,
rollup_cube_clause
GROUPING

SETS

(

)
grouping_expression_list

(rollup_cube_clause::= on page 19-10, grouping_expression_list::= on page 19-10)
grouping_expression_list::=
,
expression_list

expression_list::=
,
expr
,
expr
(

)

19-10 Oracle Database SQL Language Reference

SELECT

model_clause::=
cell_reference_options

return_rows_clause

reference_model

MODEL

main_model

(cell_reference_options::= on page 19-11, return_rows_clause::= on page 19-11, reference_
model::= on page 19-11, main_model::= on page 19-11)
cell_reference_options::=
IGNORE

DIMENSION
NAV

UNIQUE

KEEP

SINGLE

REFERENCE

return_rows_clause::=
UPDATED
RETURN

ROWS
ALL

reference_model::=
cell_reference_options
REFERENCE

reference_model_name

ON

(

subquery

)

model_column_clauses

(model_column_clauses::= on page 19-11, cell_reference_options::= on page 19-11)
main_model::=
MAIN

main_model_name

cell_reference_options
model_column_clauses

model_rules_clause

(model_column_clauses::= on page 19-11, cell_reference_options::= on page 19-11, model_
rules_clause::= on page 19-12)
model_column_clauses::=
,
c_alias
PARTITION

BY

(

expr

)

,

,
c_alias

DIMENSION

BY

(

expr

c_alias
)

MEASURES

(

expr

)

model_column::=
AS
c_alias
expr

SQL Statements: SAVEPOINT to UPDATE

19-11

SELECT

model_rules_clause::=
UPDATE
AUTOMATIC

ALL

ORDER

UPSERT

SEQUENTIAL

model_iterate_clause

RULES

,
UPDATE
ALL
UPSERT

order_by_clause

(

cell_assignment

=

expr

)

(model_iterate_clause::= on page 19-12, cell_assignment::= on page 19-12, order_by_
clause::= on page 19-13)
model_iterate_clause::=
UNTIL
ITERATE

(

number

(

condition

)

)

cell_assignment::=
,
condition
expr
measure_column

[

single_column_for_loop

]

multi_column_for_loop

(single_column_for_loop::= on page 19-12, multi_column_for_loop::= on page 19-12)
single_column_for_loop::=
,
literal
IN

(

)
subquery

FOR

dimension_column
LIKE

pattern

INCREMENT
FROM

literal

TO

literal

literal
DECREMENT

multi_column_for_loop::=
,
,
,
(
FOR

(

dimension_column

)

IN

(

)
)

subquery

19-12 Oracle Database SQL Language Reference

literal

SELECT

order_by_clause::=
,

expr

SIBLINGS
ORDER

BY

ASC

NULLS

FIRST

DESC

NULLS

LAST

position
c_alias

for_update_clause::=
,
schema

.

NOWAIT

table
.

WAIT

view
OF
FOR

column

SKIP

integer
LOCKED

UPDATE

Semantics
19

subquery_factoring_clause
The WITH query_name clause lets you assign a name to a subquery block. You can then
reference the subquery block multiple places in the query by specifying query_name.
Oracle Database optimizes the query by treating the query name as either an inline
view or as a temporary table.
The column aliases following the query_name and the set operators separating
multiple subqueries in the AS clause are valid and required for recursive subquery
factoring. The search_clause and cycle_clause are valid only for recursive subquery
factoring but are not required. See "Recursive Subquery Factoring" on page 19-13.
You can specify this clause in any top-level SELECT statement and in most types of
subqueries. The query name is visible to the main query and to all subsequent
subqueries. For recursive subquery factoring, the query name is even visible to the
subquery that defines the query name itself.
If a subquery_factoring_clause refers to its own
query_name in the subquery that defines it, then the subquery_factoring_clause is
said to be recursive. A recursive subquery_factoring_clause must contain two query
blocks: the first is the anchor member and the second is the recursive member. The
anchor member must appear before the recursive member, and it cannot reference
query_name. The anchor member can be composed of one or more query blocks
combined by the set operators: UNION ALL, UNION, INTERSECT or MINUS. The recursive
member must follow the anchor member and must reference query_name exactly once.
You must combine the recursive member with the anchor member using the UNION ALL
set operator.
Recursive Subquery Factoring

The number of column aliases following WITH query_name and the number of columns
in the SELECT lists of the anchor and recursive query blocks must be the same.
The recursive member cannot contain any of the following elements:
■

The DISTINCT keyword or a GROUP BY clause

■

The model_clause

■

An aggregate function. However, analytic functions are permitted in the select list.
SQL Statements: SAVEPOINT to UPDATE

19-13

SELECT

■

Subqueries that refer to query_name.

■

Outer joins that refer to query_name as the right table.

search_clause
■

■

■
■

■

Use the SEARCH clause to specify an ordering for the rows.

Specify BREADTH FIRST BY if you want sibling rows returned before any child rows
are returned.
Specify DEPTH FIRST BY if you want child rows returned before any siblings rows
are returned.
Sibling rows are ordered by the columns listed after the BY keyword.
The c_alias list following the SEARCH keyword must contain column names from
the column alias list for query_name.
The ordering_column is automatically added to the column list for the query
name. The query that selects from query_name can include an ORDER BY on
ordering_column to return the rows in the order that was specified by the SEARCH
clause.

cycle_clause Use the CYCLE clause to mark cycles in the recursion.
■

■
■

■

■

■

The c_alias list following the CYCLE keyword must contain column names from
the column alias list for query_name. Oracle Database uses these columns to detect
a cycle.
cycle_value and no_cycle_value should be character strings of length 1.
If a cycle is detected, then the cycle mark column specified by cycle_mark_c_
alias for the row causing the cycle is set to the value specified for cycle_value.
The recursion will then stop for this row. That is, it will not look for child rows for
the offending row, but it will continue for other noncyclic rows.
If no cycles are found, then the cycle mark column is set to the default value
specified for no_cycle_value.
The cycle mark column is automatically added to the column list for the query_
name.
A row is considered to form a cycle if one of its ancestor rows has the same values
for the cycle columns.

If you omit the CYCLE clause, then the recursive WITH clause returns an error if cycles
are discovered. In this case, a row forms a cycle if one of its ancestor rows has the same
values for all the columns in the column alias list for query_name that are referenced in
the WHERE clause of the recursive member.
Restrictions on Subquery Factoring This clause is subject to the following
restrictions:
■

■

■
■

You can specify only one subquery_factoring_clause in a single SQL statement.
Any query_name defined in the subquery_factoring_clause can be used in any
subsequent named query block in the subquery_factoring_clause.
In a compound query with set operators, you cannot use the query_name for any of
the component queries, but you can use the query_name in the FROM clause of any
of the component queries.
You cannot specify duplicate names in the column alias list for query_name.
The name used for the ordering_column has to be different from the name used
for cycle_mark_c_alias.

19-14 Oracle Database SQL Language Reference

SELECT

■

The ordering_column and cycle mark column names cannot already be in the
column alias list for query_name.
See Also:
■

Oracle Database Concepts for information about inline views

■

"Subquery Factoring: Example" on page 19-36

■

"Recursive Subquery Factoring: Examples" on page 19-37

hint
Specify a comment that passes instructions to the optimizer on choosing an execution
plan for the statement.
See Also: "Hints" on page 3-74 for the syntax and description of
hints

DISTINCT | UNIQUE
Specify DISTINCT or UNIQUE if you want the database to return only one copy of each
set of duplicate rows selected. These two keywords are synonymous. Duplicate rows
are those with matching values for each expression in the select list.
Restrictions on DISTINCT and UNIQUE Queries These types of queries are subject to

the following restrictions:
■

■

When you specify DISTINCT or UNIQUE, the total number of bytes in all select list
expressions is limited to the size of a data block minus some overhead. This size is
specified by the initialization parameter DB_BLOCK_SIZE.
You cannot specify DISTINCT if the select_list contains LOB columns.

ALL
Specify ALL if you want the database to return all rows selected, including all copies of
duplicates. The default is ALL.

* (all-column wildcard)
Specify the all-column wildcard (asterisk) to select all columns, excluding
pseudocolumns, from all tables, views, or materialized views listed in the FROM clause.
The asterisk can be preceded with a table alias specified in the FROM clause of the same
subquery. The columns are returned in the order indicated by the COLUMN_ID column of
the *_TAB_COLUMNS data dictionary view for the table, view, or materialized view.
If you are selecting from a table rather than from a view or a materialized view, then
columns that have been marked as UNUSED by the ALTER TABLE SET UNUSED statement
are not selected.
ALTER TABLE on page 12-2, "Simple Query Examples" on
page 19-40, and "Selecting from the DUAL Table: Example" on
page 19-57

See Also:

select_list
The select_list lets you specify the columns you want to retrieve from the database.

SQL Statements: SAVEPOINT to UPDATE

19-15

SELECT

query_name
For query_name, specify a name already specified in the subquery_factoring_clause.
You must have specified the subquery_factoring_clause in order to specify query_
name in the select_list. If you specify query_name in the select_list, then you also
must specify query_name in the query_table_expression (FROM clause).
table.* | view.* | materialized view.*
Specify the object name followed by a period and the asterisk to select all columns
from the specified table, view, or materialized view. Oracle Database returns a set of
columns in the order in which the columns were specified when the object was
created. A query that selects rows from two or more tables, views, or materialized
views is a join.
You can use the schema qualifier to select from a table, view, or materialized view in a
schema other than your own. If you omit schema, then the database assumes the table,
view, or materialized view is in your own schema.
See Also:

"Joins" on page 9-11

expr
Specify an expression representing the information you want to select. A column name
in this list can be qualified with schema only if the table, view, or materialized view
containing the column is qualified with schema in the FROM clause. If you specify a
member method of an object type, then you must follow the method name with
parentheses even if the method takes no arguments.
See Also:

"Selecting Sequence Values: Examples" on page 19-57

Specify an alias for the column expression. Oracle Database will use this alias
in the column heading of the result set. The AS keyword is optional. The alias
effectively renames the select list item for the duration of the query. The alias can be
used in the order_by_clause but not other clauses in the query.

c_alias

See Also:
■

■

Oracle Database Data Warehousing Guide for information on using
the expr AS c_alias syntax with the UNION ALL operator in queries
of multiple materialized views
"About SQL Expressions" on page 6-1 for the syntax of expr

Restrictions on the Select List
■

The select list is subject to the following restrictions:

If you also specify a group_by_clause in this statement, then this select list can
contain only the following types of expressions:
–

Constants

–

Aggregate functions and the functions USER, UID, and SYSDATE

–

Expressions identical to those in the group_by_clause. If the group_by_clause
is in a subquery, then all columns in the select list of the subquery must match
the GROUP BY columns in the subquery. If the select list and GROUP BY columns
of a top-level query or of a subquery do not match, then the statement results
in ORA-00979.

–

Expressions involving the preceding expressions that evaluate to the same
value for all rows in a group

19-16 Oracle Database SQL Language Reference

SELECT

■

You can select a rowid from a join view only if the join has one and only one
key-preserved table. The rowid of that table becomes the rowid of the view.
See Also: Oracle Database Administrator's Guide for information on
key-preserved tables

■

If two or more tables have some column names in common, and if you are
specifying a join in the FROM clause, then you must qualify column names with
names of tables or table aliases.

FROM Clause
The FROM clause lets you specify the objects from which data is selected.
query_table_expression
Use the query_table_expression clause to identify a table, view, materialized view,
partition, or subpartition, or to specify a subquery that identifies the objects.
See Also:

"Using Subqueries: Examples" on page 19-50

ONLY The ONLY clause applies only to views. Specify ONLY if the view in the FROM
clause is a view belonging to a hierarchy and you do not want to include rows from
any of its subviews.

flashback_query_clause
Use the flashback_query_clause to retrieve past data from a table, view, or
materialized view.
This clause implements SQL-driven Flashback, which lets you specify a different
system change number or timestamp for each object in the select list. You can also
implement session-level Flashback using the DBMS_FLASHBACK package.
A Flashback Query lets you retrieve a history of changes made to a row. You can
retrieve the corresponding identifier of the transaction that made the change using the
VERSIONS_XID pseudocolumn. You can also retrieve information about the transaction
that resulted in a particular row version by issuing an Oracle Flashback Transaction
Query. You do this by querying the FLASHBACK_TRANSACTION_QUERY data dictionary
view for a particular transaction ID.
AS OF Specify AS OF to retrieve the single version of the rows returned by the query
at a particular change number (SCN) or timestamp. If you specify SCN, then expr must
evaluate to a number. If you specify TIMESTAMP, then expr must evaluate to a
timestamp value. In either case, expr cannot evaluate to NULL. Oracle Database
returns rows as they existed at the specified system change number or time.

Specify VERSIONS to retrieve multiple versions of the rows returned by
the query. Oracle Database returns all committed versions of the rows that existed
between two SCNs or between two timestamp values. The first specified SCN or
timestamp must be earlier than the second specified SCN or timestamp. The rows
returned include deleted and subsequently reinserted versions of the rows.

VERSIONS

■

Specify BETWEEN SCN ... to retrieve the versions of the row that existed between two
SCNs. Both expressions must evaluate to a number and cannot evaluate to NULL.
MINVALUE and MAXVALUE resolve to the SCN of the oldest and most recent data
available, respectively.

SQL Statements: SAVEPOINT to UPDATE

19-17

SELECT

■

Specify BETWEEN TIMESTAMP ... to retrieve the versions of the row that existed
between two timestamps. Both expressions must evaluate to a timestamp value
and cannot evaluate to NULL. MINVALUE and MAXVALUE resolve to the timestamp of
the oldest and most recent data available, respectively.

Oracle Database provides a group of version query pseudocolumns that let you
retrieve additional information about the various row versions. Refer to "Version
Query Pseudocolumns" on page 2-5 for more information.
When both clauses are used together, the AS OF clause determines the SCN or moment
in time from which the database issues the query. The VERSIONS clause determines the
versions of the rows as seen from the AS OF point. The database returns null for a row
version if the transaction started before the first BETWEEN value or ended after the AS OF
point.
Note on Flashback Queries When performing a flashback query, Oracle Database
might not use query optimizations that it would use for other types of queries, which
could have a negative impact on performance. In particular, this occurs when you
specify multiple flashback queries in a hierarchical query.
Restrictions on Flashback Queries

These queries are subject to the following

restrictions:
■

You cannot specify a column expression or a subquery in the expression of the AS
OF clause.

■

You cannot specify the AS OF clause if you have specified the for_update_clause.

■

You cannot use the AS OF clause in the defining query of a materialized view.

■

■

■

You cannot use the VERSIONS clause in flashback queries to temporary or external
tables, or tables that are part of a cluster.
You cannot use the VERSIONS clause in flashback queries to views. However, you
can use the VERSIONS syntax in the defining query of a view.
You cannot specify the flashback_query_clause if you have specified query_name
in the query_table_expression.
See Also:
■

■
■

■

Oracle Database Advanced Application Developer's Guide for more
information on Oracle Flashback Query
"Using Flashback Queries: Example" on page 19-41
Oracle Database Advanced Application Developer's Guide and Oracle
Database PL/SQL Packages and Types Reference for information about
session-level Flashback using the DBMS_FLASHBACK package
Oracle Database Administrator's Guide and to the description of
FLASHBACK_TRANSACTION_QUERY in the Oracle Database Reference for
more information about transaction history

For PARTITION or SUBPARTITION, specify the name or key
value of the partition or subpartition within table from which you want to retrieve
data.

partition_extension_clause

For range- and list-partitioned data, as an alternative to this clause, you can specify a
condition in the WHERE clause that restricts the retrieval to one or more partitions of
table. Oracle Database will interpret the condition and fetch data from only those

19-18 Oracle Database SQL Language Reference

SELECT

partitions. It is not possible to formulate such a WHERE condition for hash-partitioned
data.
See Also: "References to Partitioned Tables and Indexes" on
page 3-119 and "Selecting from a Partition: Example" on page 19-41
dblink For dblink, specify the complete or partial name for a database link to a

remote database where the table, view, or materialized view is located. This database
need not be an Oracle Database.
See Also:
■

■

"References to Objects in Remote Databases" on page 3-117 for
more information on referring to database links
"Distributed Queries" on page 9-16 for more information about
distributed queries and "Using Distributed Queries: Example" on
page 19-56

If you omit dblink, then the database assumes that the table, view, or materialized
view is on the local database.
Restrictions on Database Links Database links are subject to the following
restrictions:
■
■

You cannot query a user-defined type or an object REF on a remote table.
You cannot query columns of type ANYTYPE, ANYDATA, or ANYDATASET from remote
tables.

table | view | materialized view Specify the name of a table, view, or materialized
view from which data is selected.

sample_clause
The sample_clause lets you instruct the database to select from a random sample of
data from the table, rather than from the entire table.
See Also:

"Selecting a Sample: Examples" on page 19-41

BLOCK BLOCK instructs the database to attempt to perform random block sampling

instead of random row sampling.
Block sampling is possible only during full table scans or index fast full scans. If a
more efficient execution path exists, then Oracle Database does not perform block
sampling. If you want to guarantee block sampling for a particular table or index, then
use the FULL or INDEX_FFS hint.
sample_percent For sample_percent, specify the percentage of the total row or block
count to be included in the sample. The value must be in the range .000001 to, but not
including, 100. This percentage indicates the probability of each row, or each cluster of
rows in the case of block sampling, being selected as part of the sample. It does not
mean that the database will retrieve exactly sample_percent of the rows of table.
Caution: The use of statistically incorrect assumptions when using
this feature can lead to incorrect or undesirable results.

SQL Statements: SAVEPOINT to UPDATE

19-19

SELECT

SEED seed_value Specify this clause to instruct the database to attempt to return the
same sample from one execution to the next. The seed_value must be an integer
between 0 and 4294967295. If you omit this clause, then the resulting sample will
change from one execution to the next.
Restrictions on sample_clause
■
■

The following restrictions apply to the SAMPLE clause:

You cannot specify the SAMPLE clause in a subquery in a DML statement.
You can specify the SAMPLE clause in a query on a base table, a container table of a
materialized view, or a view that is key preserving. You cannot specify this clause
on a view that is not key preserving.

The subquery_restriction_clause lets you restrict the
subquery in one of the following ways:

subquery_restriction_clause

WITH READ ONLY

Specify WITH READ ONLY to indicate that the table or view cannot

be updated.
WITH CHECK OPTION Specify WITH CHECK OPTION to indicate that Oracle Database
prohibits any changes to the table or view that would produce rows that are not
included in the subquery. When used in the subquery of a DML statement, you can
specify this clause in a subquery in the FROM clause but not in subquery in the WHERE
clause.
CONSTRAINT constraint Specify the name of the CHECK OPTION constraint. If you
omit this identifier, then Oracle automatically assigns the constraint a name of the
form SYS_Cn, where n is an integer that makes the constraint name unique within the
database.
See Also: "Using the WITH CHECK OPTION Clause: Example" on
page 19-48

table_collection_expression
The table_collection_expression lets you inform Oracle that the value of
collection_expression should be treated as a table for purposes of query and DML
operations. The collection_expression can be a subquery, a column, a function, or a
collection constructor. Regardless of its form, it must return a collection value—that is,
a value whose type is nested table or varray. This process of extracting the elements of
a collection is called collection unnesting.
The optional plus (+) is relevant if you are joining the TABLE collection expression with
the parent table. The + creates an outer join of the two, so that the query returns rows
from the outer table even if the collection expression is null.
In earlier releases of Oracle, when collection_expression
was a subquery, table_collection_expression was expressed as
THE subquery. That usage is now deprecated.
Note:

The collection_expression can reference columns of tables defined to its left in the
FROM clause. This is called left correlation. Left correlation can occur only in table_
collection_expression. Other subqueries cannot contains references to columns
defined outside the subquery.
The optional (+) lets you specify that table_collection_expression should return a
row with all fields set to null if the collection is null or empty. The (+) is valid only if
19-20 Oracle Database SQL Language Reference

SELECT

collection_expression uses left correlation. The result is similar to that of an outer
join.
When you use the (+) syntax in the WHERE clause of a subquery in an UPDATE or DELETE
operation, you must specify two tables in the FROM clause of the subquery. Oracle
Database ignores the outer join syntax unless there is a join in the subquery itself.
See Also:
■
■

"Outer Joins" on page 9-12
"Table Collections: Examples" on page 19-54 and "Collection
Unnesting: Examples" on page 19-54

t_alias
Specify a correlation name, which is alias for the table, view, materialized view, or
subquery for evaluating the query. This alias is required if the select list references any
object type attributes or object type methods. Correlation names are most often used in
a correlated query. Other references to the table, view, or materialized view throughout
the query must refer to this alias.
See Also:

"Using Correlated Subqueries: Examples" on page 19-57

pivot_clause
The pivot_clause lets you write cross-tabulation queries that rotate rows into
columns, aggregating data in the process of the rotation. The output of a pivot
operation typically includes more columns and fewer rows than the starting data set.
The pivot_clause performs the following steps:
1.

The pivot_clause computes the aggregation functions specified at the beginning
of the clause. Aggregation functions must specify a GROUP BY clause to return
multiple values, yet the pivot_clause does not contain an explicit GROUP BY
clause. Instead, the pivot_clause performs an implicit GROUP BY. The implicit
grouping is based on all the columns not referred to in the pivot_clause, along
with the set of values specified in the pivot_in_clause.).

2.

The grouping columns and aggregated values calculated in Step 1 are configured
to produce the following cross-tabular output:
a.

All the implicit grouping columns not referred to in the pivot_clause,
followed by

b.

New columns corresponding to values in the pivot_in_clause Each
aggregated value is transposed to the appropriate new column in the
cross-tabulation. If you specify the XML keyword, then the result is a single new
column that expresses the data as an XML string.

The subclauses of the pivot_clause have the following semantics:
The optional XML keyword generates XML output for the query. The XML
keyword permits the pivot_in_clause to contain either a subquery or the wildcard
keyword ANY. Subqueries and ANY wildcards are useful when the pivot_in_clause
values are not known in advance. With XML output, the values of the pivot column
are evaluated at execution time. You cannot specify XML when you specify explicit
pivot values using expressions in the pivot_in_clause.
XML

When XML output is generated, the aggregate function is applied to each distinct
pivot value, and the database returns a column of XMLType containing an XML string
for all value and measure pairs.

SQL Statements: SAVEPOINT to UPDATE

19-21

SELECT

expr For expr, specify an expression that evaluates to a constant value of a pivot
column. You can optionally provide an alias for each pivot column value. If there is no
alias, the column heading becomes a quoted identifier.

A subquery is used only in conjunction with the XML keyword. When you
specify a subquery, all values found by the subquery are used for pivoting. The output
is not the same cross-tabular format returned by non-XML pivot queries. Instead of
multiple columns specified in the pivot_in_clause, the subquery produces a single
XML string column. The XML string for each row holds aggregated data
corresponding to the implicit GROUP BY value of that row. The XML string for each
output row includes all pivot values found by the subquery, even if there are no
corresponding rows in the input data.
subquery

The subquery must return a list of unique values at the execution time of the pivot
query. If the subquery does not return a unique value, then Oracle Database raises a
run-time error. Use the DISTINCT keyword in the subquery if you are not sure the
query will return unique values.
ANY The ANY keyword is used only in conjunction with the XML keyword. The ANY
keyword acts as a wildcard and is similar in effect to subquery. The output is not the
same cross-tabular format returned by non-XML pivot queries. Instead of multiple
columns specified in the pivot_in_clause, the ANY keyword produces a single XML
string column. The XML string for each row holds aggregated data corresponding to
the implicit GROUP BY value of that row. However, in contrast to the behavior when
you specify subquery, the ANY wildcard produces an XML string for each output row
that includes only the pivot values found in the input data corresponding to that row.
See Also: Oracle Database Data Warehousing Guide for more
information about PIVOT and UNPIVOT and "Using PIVOT and
UNPIVOT: Examples" on page 19-48

unpivot_clause
The unpivot_clause rotates columns into rows.
■

■

■

■

The INCLUDE | EXCLUDE NULLS clause gives you the option of including or
excluding null-valued rows. INCLUDE NULLS causes the unpivot operation to
include null-valued rows; EXCLUDE NULLS eliminates null-values rows from the
return set. If you omit this clause, then the unpivot operation excludes nulls.
For column, specify a name for each output column that will hold measure values,
such as sales_quantity.
In the pivot_for_clause, specify a name for each output column that will hold
descriptor values, such as quarter or product.
In the unpivot_in_clause, specify the input data columns whose names will
become values in the output columns of the pivot_for_clause. These input data
columns have names specifying a category value, such as Q1, Q2, Q3, Q4. The
optional AS clause lets you map the input data column names to the specified
literal values in the output columns.

The unpivot operation turns a set of value columns into one column. Therefore, the
data types of all the value columns must be in the same data type group, such as
numeric or character.
■

If all the value columns are CHAR, then the unpivoted column is CHAR. If any value
column is VARCHAR2, then the unpivoted column is VARCHAR2.

19-22 Oracle Database SQL Language Reference

SELECT

■

If all the value columns are NUMBER, then the unpivoted column is NUMBER. If any
value column is BINARY_DOUBLE, then the unpivoted column is BINARY_DOUBLE. If
no value column is BINARY_DOUBLE but any value column is BINARY_FLOAT, then
the unpivoted column is BINARY_FLOAT.

join_clause
Use the appropriate join_clause syntax to identify tables that are part of a join from
which to select data. The inner_cross_join_clause lets you specify an inner or cross
join. The outer_join_clause lets you specify an outer join.
When you join more than two row sources, you can use parentheses to override
default precedence. For example, the following syntax:
SELECT ... FROM a JOIN (b JOIN c) ...

results in a join of b and c, and then a join of that result set with a.
See Also: "Joins" on page 9-11 for more information on joins, "Using
Join Queries: Examples" on page 19-49, "Using Self Joins: Example" on
page 19-50, and "Using Outer Joins: Examples" on page 19-51

inner_cross_join_clause
Inner joins return only those rows that satisfy the join condition.
INNER Specify INNER to explicitly specify an inner join.
JOIN The JOIN keyword explicitly states that a join is being performed. You can use
this syntax to replace the comma-delimited table expressions used in WHERE clause
joins with FROM clause join syntax.
ON condition Use the ON clause to specify a join condition. Doing so lets you specify
join conditions separate from any search or filter conditions in the WHERE clause.
USING (column) When you are specifying an equijoin of columns that have the same
name in both tables, the USING column clause indicates the columns to be used. You can
use this clause only if the join columns in both tables have the same name. Within this
clause, do not qualify the column name with a table name or table alias.

The CROSS keyword indicates that a cross join is being performed. A cross
join produces the cross-product of two relations and is essentially the same as the
comma-delimited Oracle Database notation.

CROSS

NATURAL The NATURAL keyword indicates that a natural join is being performed.
Refer to NATURAL on page 19-24 for the full semantics of this clause.

outer_join_clause
Outer joins return all rows that satisfy the join condition and also return some or all of
those rows from one table for which no rows from the other satisfy the join condition.
You can specify two types of outer joins: a conventional outer join using the table_
reference syntax on both sides of the join, or a partitioned outer join using the query_
partition_clause on one side or the other. A partitioned outer join is similar to a
conventional outer join except that the join takes place between the outer table and
each partition of the inner table. This type of join lets you selectively make sparse data
more dense along the dimensions of interest. This process is called data densification.

SQL Statements: SAVEPOINT to UPDATE

19-23

SELECT

query_partition_clause The query_partition_clause lets you define a partitioned
outer join. Such a join extends the conventional outer join syntax by applying the
outer join to partitions returned by the query. Oracle Database creates a partition of
rows for each expression you specify in the PARTITION BY clause. The rows in each
query partition have same value for the PARTITION BY expression.

The query_partition_clause can be on either side of the outer join. The result of a
partitioned outer join is a UNION of the outer joins of each of the partitions in the
partitioned result set and the table on the other side of the join. This type of result is
useful for filling gaps in sparse data, which simplifies analytic calculations.
If you omit this clause, then the database treats the entire table expression—everything
specified in table_reference—as a single partition, resulting in a conventional outer
join.
To use the query_partition_clause in an analytic function, use the upper branch of
the syntax (without parentheses). To use this clause in a model query (in the model_
column_clauses) or a partitioned outer join (in the outer_join_clause), use the lower
branch of the syntax (with parentheses).
Restrictions on Partitioned Outer Joins Partitioned outer joins are subject to the

following restrictions:
■

■
■

You can specify the query_partition_clause on either the right or left side of the
join, but not both.
You cannot specify a FULL partitioned outer join.
If you specify the query_partition_clause in an outer join with an ON clause, then
you cannot specify a subquery in the ON condition.
See Also:

"Using Partitioned Outer Joins: Examples" on page 19-52

NATURAL The NATURAL keyword indicates that a natural join is being performed. A
natural join is based on all columns in the two tables that have the same name. It
selects rows from the two tables that have equal values in the relevant columns. If two
columns with the same name do not have compatible data types, then an error is
raised. When specifying columns that are involved in the natural join, do not qualify
the column name with a table name or table alias.

On occasion, the table pairings in natural or cross joins may be ambiguous. For
example, consider the following join syntax:
a NATURAL LEFT JOIN b LEFT JOIN c ON b.c1 = c.c1

This example can be interpreted in either of the following ways:
a NATURAL LEFT JOIN (b LEFT JOIN c ON b.c1 = c.c1)
(a NATURAL LEFT JOIN b) LEFT JOIN c ON b.c1 = c.c1

To avoid this ambiguity, you can use parentheses to specify the pairings of joined
tables. In the absence of such parentheses, the database uses left associativity, pairing
the tables from left to right.
Restriction on Natural Joins You cannot specify a LOB column, columns of ANYTYPE,

ANYDATA, or ANYDATASET, or a collection column as part of a natural join.
outer_join_type The outer_join_type indicates the kind of outer join being
performed:
■

Specify RIGHT to indicate a right outer join.

19-24 Oracle Database SQL Language Reference

SELECT

■
■

■

Specify LEFT to indicate a left outer join.
Specify FULL to indicate a full or two-sided outer join. In addition to the inner join,
rows from both tables that have not been returned in the result of the inner join
will be preserved and extended with nulls.
You can specify the optional OUTER keyword following RIGHT, LEFT, or FULL to
explicitly clarify that an outer join is being performed.

ON condition Use the ON clause to specify a join condition. Doing so lets you specify
join conditions separate from any search or filter conditions in the WHERE clause.

You cannot specify this clause with a

Restriction on the ON condition Clause

NATURAL outer join.
USING column In an outer join with the USING clause, the query returns a single
column which is a coalesce of the two matching columns in the join. The coalesce
functions as follows:
COALESCE (a, b) = a if a NOT NULL, else b.

Therefore:
■

■

■

A left outer join returns all the common column values from the left table in the
FROM clause.
A right outer join returns all the common column values from the right table in the
FROM clause.
A full outer join returns all the common column values from both joined tables.

Restriction on the USING column Clause
■

■

■

Within this clause, do not qualify the column name with a table name or table
alias.
You cannot specify a LOB column or a collection column in the USING column
clause.
You cannot specify this clause with a NATURAL outer join.
See Also:
■

■

■

"Outer Joins" on page 9-12 for additional rules and restrictions
pertaining to outer joins
Oracle Database Data Warehousing Guide for a complete discussion
of partitioned outer joins and data densification
"Using Outer Joins: Examples" on page 19-51

where_clause
The WHERE condition lets you restrict the rows selected to those that satisfy one or more
conditions. For condition, specify any valid SQL condition.
If you omit this clause, then the database returns all rows from the tables, views, or
materialized views in the FROM clause.

SQL Statements: SAVEPOINT to UPDATE

19-25

SELECT

If this clause refers to a DATE column of a partitioned table or
index, then the database performs partition pruning only if:

Note:
■

■

You created the table or index partitions by fully specifying the
year using the TO_DATE function with a 4-digit format mask, and
You specify the date in the where_clause of the query using the
TO_DATE function and either a 2- or 4-digit format mask.

See Also:
■

Chapter 7, "Conditions" for the syntax description of condition

■

"Selecting from a Partition: Example" on page 19-41

hierarchical_query_clause
The hierarchical_query_clause lets you select rows in a hierarchical order.
SELECT statements that contain hierarchical queries can contain the LEVEL
pseudocolumn in the select list. LEVEL returns the value 1 for a root node, 2 for a child
node of a root node, 3 for a grandchild, and so on. The number of levels returned by a
hierarchical query may be limited by available user memory.
Oracle processes hierarchical queries as follows:
■

A join, if present, is evaluated first, whether the join is specified in the FROM clause
or with WHERE clause predicates.

■

The CONNECT BY condition is evaluated.

■

Any remaining WHERE clause predicates are evaluated.

If you specify this clause, then do not specify either ORDER BY or GROUP BY, because they
will destroy the hierarchical order of the CONNECT BY results. If you want to order rows
of siblings of the same parent, then use the ORDER SIBLINGS BY clause.
See Also: "Hierarchical Queries" on page 9-3 for a discussion of
hierarchical queries and "Using the LEVEL Pseudocolumn: Examples"
on page 19-55

START WITH Clause
Specify a condition that identifies the row(s) to be used as the root(s) of a hierarchical
query. The condition can be any condition as described in Chapter 7, "Conditions."
Oracle Database uses as root(s) all rows that satisfy this condition. If you omit this
clause, then the database uses all rows in the table as root rows.
CONNECT BY Clause
Specify a condition that identifies the relationship between parent rows and child rows
of the hierarchy. The condition can be any condition as described in Chapter 7,
"Conditions." However, it must use the PRIOR operator to refer to the parent row.

19-26 Oracle Database SQL Language Reference

SELECT

See Also:
■
■

■

Chapter 2, "Pseudocolumns" for more information on LEVEL
"Hierarchical Queries" on page 9-3 for general information on
hierarchical queries
"Hierarchical Query Examples" on page 19-44

group_by_clause
Specify the GROUP BY clause if you want the database to group the selected rows based
on the value of expr(s) for each row and return a single row of summary information
for each group. If this clause contains CUBE or ROLLUP extensions, then the database
produces superaggregate groupings in addition to the regular groupings.
Expressions in the GROUP BY clause can contain any columns of the tables, views, or
materialized views in the FROM clause, regardless of whether the columns appear in the
select list.
The GROUP BY clause groups rows but does not guarantee the order of the result set. To
order the groupings, use the ORDER BY clause.
If the NLS_SORT parameter has a setting other than BINARY and the NLS_COMP parameter
is set to LINGUISTIC, then expression values are compared linguistically according to
the linguistic definition specified in NLS_SORT to decide if they are equal and therefore
belong to the same group. When character values are compared linguistically for
GROUP BY, they are first transformed to collation keys and then compared like RAW
values. The collation keys are the same values that are returned by the function
NLSSORT and are subject to the same restrictions that are described in "NLSSORT" on
page 5-164. As a result of these restrictions, two values may compare as linguistically
equal and be grouped together if they do not differ in the prefix that was used to
produce the collation key, even if they differ in the rest of the value.
See Also:
■

■

■

Oracle Database Data Warehousing Guide for an expanded
discussion and examples of using SQL grouping syntax for data
aggregation
the GROUP_ID, GROUPING, and GROUPING_ID functions on
page 5-109 for examples
"Using the GROUP BY Clause: Examples" on page 19-42

ROLLUP The ROLLUP operation in the simple_grouping_clause groups the selected
rows based on the values of the first n, n-1, n-2, ... 0 expressions in the GROUP BY
specification, and returns a single row of summary for each group. You can use the
ROLLUP operation to produce subtotal values by using it with the SUM function. When
used with SUM, ROLLUP generates subtotals from the most detailed level to the grand
total. Aggregate functions such as COUNT can be used to produce other kinds of
superaggregates.

For example, given three expressions (n=3) in the ROLLUP clause of the simple_
grouping_clause, the operation results in n+1 = 3+1 = 4 groupings.
Rows grouped on the values of the first n expressions are called regular rows, and the
others are called superaggregate rows.
See Also: Oracle Database Data Warehousing Guide for information on
using ROLLUP with materialized views
SQL Statements: SAVEPOINT to UPDATE

19-27

SELECT

The CUBE operation in the simple_grouping_clause groups the selected rows
based on the values of all possible combinations of expressions in the specification. It
returns a single row of summary information for each group. You can use the CUBE
operation to produce cross-tabulation values.

CUBE

For example, given three expressions (n=3) in the CUBE clause of the simple_grouping_
clause, the operation results in 2n = 23 = 8 groupings. Rows grouped on the values of n
expressions are called regular rows, and the rest are called superaggregate rows.
See Also:
■

■

Oracle Database Data Warehousing Guide for information on using
CUBE with materialized views
"Using the GROUP BY CUBE Clause: Example" on page 19-43

GROUPING SETS GROUPING SETS are a further extension of the GROUP BY clause that
let you specify multiple groupings of data. Doing so facilitates efficient aggregation by
pruning the aggregates you do not need. You specify just the desired groups, and the
database does not need to perform the full set of aggregations generated by CUBE or
ROLLUP. Oracle Database computes all groupings specified in the GROUPING SETS clause
and combines the results of individual groupings with a UNION ALL operation. The
UNION ALL means that the result set can include duplicate rows.

Within the GROUP BY clause, you can combine expressions in various ways:
■

■

To specify composite columns, group columns within parentheses so that the
database treats them as a unit while computing ROLLUP or CUBE operations.
To specify concatenated grouping sets, separate multiple grouping sets, ROLLUP,
and CUBE operations with commas so that the database combines them into a
single GROUP BY clause. The result is a cross-product of groupings from each
grouping set.
See Also: "Using the GROUPING SETS Clause: Example" on
page 19-43

HAVING Clause
Use the HAVING clause to restrict the groups of returned rows to those groups for which
the specified condition is TRUE. If you omit this clause, then the database returns
summary rows for all groups.
Specify GROUP BY and HAVING after the where_clause and hierarchical_query_clause.
If you specify both GROUP BY and HAVING, then they can appear in either order.
See Also:

"Using the HAVING Condition: Example" on page 19-44

Restrictions on the GROUP BY Clause: This clause is subject to the following
restrictions:
■

You cannot specify LOB columns, nested tables, or varrays as part of expr.

■

The expressions can be of any form except scalar subquery expressions.

■

If the group_by_clause references any object type columns, then the query will not
be parallelized.

model_clause
The model_clause lets you view selected rows as a multidimensional array and
randomly access cells within that array. Using the model_clause, you can specify a
19-28 Oracle Database SQL Language Reference

SELECT

series of cell assignments, referred to as rules, that invoke calculations on individual
cells and ranges of cells. These rules operate on the results of a query and do not
update any database tables.
When using the model_clause in a query, the SELECT and ORDER BY clauses must refer
only to those columns defined in the model_column_clauses.
See Also:
■

■

■

The syntax description of expr in "About SQL Expressions" on
page 6-1 and the syntax description of condition in Chapter 7,
"Conditions"
Oracle Database Data Warehousing Guide for an expanded
discussion and examples
"The MODEL clause: Examples" on page 19-45

main_model
The main_model clause defines how the selected rows will be viewed in a
multidimensional array and what rules will operate on which cells in that array.
model_column_clauses
The model_column_clauses define and classify the columns of a query into three
groups: partition columns, dimension columns, and measure columns. If the expr in
these three subclauses is a model column, then the column alias (c_alias) is optional.
If the expr is not a model column, then the column alias is required.
PARTITION BY The PARTITION BY clause specifies the columns that will be used to

divide the selected rows into partitions based on the values of the specified columns.
DIMENSION BY The DIMENSION BY clause specifies the columns that will identify a
row within a partition. The values of the dimension columns, along with those of the
partition columns, serve as array indexes to the measure columns within a row.
MEASURES The MEASURES clause identifies the columns on which the calculations
can be performed. Measure columns in individual rows are treated like cells that you
can reference, by specifying the values for the partition and dimension columns, and
update.
model_column model_column identifies a column to be used in defining the model. A

column alias is required if expr is not a column name. Refer to "Model Expressions" on
page 6-11 for information on model expressions.
cell_reference_options
Use the cell_reference_options clause to specify how null and absent values are
treated in rules and how column uniqueness is constrained.
IGNORE NAV When you specify IGNORE NAV, the database returns the following

values for the null and absent values of the data type specified:
■

Zero for numeric data types

■

01-JAN-2000 for datetime data types

■

An empty string for character data types

■

Null for all other data types

SQL Statements: SAVEPOINT to UPDATE

19-29

SELECT

KEEP NAV When you specify KEEP NAV, the database returns null for both null and
absent cell values. KEEP NAV is the default.

When you specify UNIQUE SINGLE REFERENCE, the
database checks only single-cell references on the right-hand side of the rule for
uniqueness, not the entire query result set.

UNIQUE SINGLE REFERENCE

When you specify UNIQUE DIMENSION, the database checks that
the PARTITION BY and DIMENSION BY columns form a unique key to the query. UNIQUE
DIMENSION is the default.
UNIQUE DIMENSION

model_rules_clause
Use the model_rules_clause to specify the cells to be updated, the rules for updating
those cells, and optionally, how the rules are to be applied and processed.
Each rule represents an assignment and consists of a left-hand side and right-hand
side. The left-hand side of the rule identifies the cells to be updated by the right-hand
side of the rule. The right-hand side of the rule evaluates to the values to be assigned
to the cells specified on the left-hand side of the rule.
UPSERT ALL allows UPSERT behavior for a rule with both positional and
symbolic references on the left-hand side of the rule. When evaluating an UPSERT ALL
rule, Oracle performs the following steps to create a list of cell references to be
upserted:

UPSERT ALL

1.

Find the existing cells that satisfy all the symbolic predicates of the cell reference.

2.

Using just the dimensions that have symbolic references, find the distinct
dimension value combinations of these cells.

3.

Perform a cross product of these value combinations with the dimension values
specified by way of positional references.

Refer to Oracle Database Data Warehousing Guide for more information on the semantics
of UPSERT ALL.
When you specify UPSERT, the database applies the rules to those cells
referenced on the left-hand side of the rule that exist in the multidimensional array,
and inserts new rows for those that do not exist. UPSERT behavior applies only when
positional referencing is used on the left-hand side and a single cell is referenced.
UPSERT is the default. Refer to cell_assignment on page 19-31 for more information on
positional referencing and single-cell references.
UPSERT

UPDATE and UPSERT can be specified for individual rules as well. When either UPDATE or
UPSERT is specified for a specific rule, it takes precedence over the option specified in
the RULES clause.
If an UPSERT ALL, UPSERT, or
UPDATE rule does not contain the appropriate predicates, then the
database may implicitly convert it to a different type of rule:
Notes on UPSERT [ALL] and UPDATE:

■

■

If an UPSERT rule contains an existential predicate, then the rule is
treated as an UPDATE rule.
An UPSERT ALL rule must have at least one existential predicate
and one qualified predicate on its left side. If it has no existential
predicate, then it is treated as an UPSERT rule. If it has no qualified
predicate, then it is treated as an UPDATE rule.

19-30 Oracle Database SQL Language Reference

SELECT

When you specify UPDATE, the database applies the rules to those cells
referenced on the left-hand side of the rule that exist in the multidimensional array. If
the cells do not exist, then the assignment is ignored.

UPDATE

When you specify AUTOMATIC ORDER, the database evaluates the
rules based on their dependency order. In this case, a cell can be assigned a value once
only.

AUTOMATIC ORDER

When you specify SEQUENTIAL ORDER, the database evaluates
the rules in the order they appear. In this case, a cell can be assigned a value more than
once. SEQUENTIAL ORDER is the default.

SEQUENTIAL ORDER

ITERATE ... [UNTIL] Use ITERATE ... [UNTIL] to specify the number of times to cycle
through the rules and, optionally, an early termination condition. The parentheses
around the UNTIL condition are optional.

When you specify ITERATE ... [UNTIL], rules are evaluated in the order in which they
appear. Oracle Database returns an error if both AUTOMATIC ORDER and ITERATE ...
[UNTIL] are specified in the model_rules_clause.
cell_assignment
The cell_assignment clause, which is the left-hand side of the rule, specifies one or
more cells to be updated. When a cell_assignment references a single cell, it is called
a single-cell reference. When more than one cell is referenced, it is called a
multiple-cell reference.
All dimension columns defined in the model_clause must be qualified in the cell_
assignment clause. A dimension can be qualified using either symbolic or positional
referencing.
A symbolic reference qualifies a single dimension column using a Boolean condition
like dimension_column=constant. A positional reference is one where the dimension
column is implied by its position in the DIMENSION BY clause. The only difference
between symbolic references and positional references is in the treatment of nulls.
Using a single-cell symbolic reference such as a[x=null,y=2000], no cells qualify
because x=null evaluates to FALSE. However, using a single-cell positional reference
such as a[null,2000], a cell where x is null and y is 2000 qualifies because null = null
evaluates to TRUE. With single-cell positional referencing, you can reference, update,
and insert cells where dimension columns are null.
You can specify a condition or an expression representing a dimension column value
using either symbolic or positional referencing. condition cannot contain aggregate
functions or the CV function, and condition must reference a single dimension
column. expr cannot contain a subquery. Refer to "Model Expressions" on page 6-11 for
information on model expressions.
single_column_for_loop
The single_column_for_loop clause lets you specify a range of cells to be updated
within a single dimension column.
The IN clause lets you specify the values of the dimension column as either a list of
values or as a subquery. When using subquery, it cannot:
■

Be a correlated query

■

Return more than 10,000 rows

■

Be a query defined in the WITH clause

SQL Statements: SAVEPOINT to UPDATE

19-31

SELECT

The FROM clause lets you specify a range of values for a dimension column with
discrete increments within the range. The FROM clause can only be used for those
columns with a data type for which addition and subtraction is supported. The
INCREMENT and DECREMENT values must be positive.
Optionally, you can specify the LIKE clause within the FROM clause. In the LIKE clause,
pattern is a character string containing a single pattern-matching character %. This
character is replaced during execution with the current incremented or decremented
value in the FROM clause.
If all dimensions other than those used by a FOR loop involve a single-cell reference,
then the expressions can insert new rows. The number of dimension value
combinations generated by FOR loops is counted as part of the 10,000 row limit of the
MODEL clause.
multi_column_for_loop
The multi_column_for_loop clause lets you specify a range of cells to be updated
across multiple dimension columns. The IN clause lets you specify the values of the
dimension columns as either multiple lists of values or as a subquery. When using
subquery, it cannot:
■

Be a correlated query

■

Return more than 10,000 rows

■

Be a query defined in the WITH clause

If all dimensions other than those used by a FOR loop involve a single-cell reference,
then the expressions can insert new rows. The number of dimension value
combinations generated by FOR loops is counted as part of the 10,000 row limit of the
MODEL clause.
See Also: Oracle Database Data Warehousing Guide for more
information about using FOR loops in the MODEL clause

order_by_clause
Use the ORDER BY clause to specify the order in which cells on the left-hand side of the
rule are to be evaluated. The expr must resolve to a dimension or measure column. If
the ORDER BY clause is not specified, then the order defaults to the order of the columns
as specified in the DIMENSION BY clause. See order_by_clause on page 19-33 for more
information.
Restrictions on the order_by_clause

Use of the ORDER BY clause in the model rule is

subject to the following restrictions:
■

■

You cannot specify SIBLINGS, position, or c_alias in the order_by_clause of the
model_clause.
You cannot specify this clause on the left-hand side of the model rule and also
specify a FOR loop on the right-hand side of the rule.

expr
Specify an expression representing the value or values of the cell or cells specified on
the right-hand side of the rule. expr cannot contain a subquery. Refer to "Model
Expressions" on page 6-11 for information on model expressions.

19-32 Oracle Database SQL Language Reference

SELECT

return_rows_clause
The return_rows_clause lets you specify whether to return all rows selected or only
those rows updated by the model rules. ALL is the default.

reference_model
Use the reference_model clause when you need to access multiple arrays from inside
the model_clause. This clause defines a read-only multidimensional array based on
the results of a query.
The subclauses of the reference_model clause have the same semantics as for the
main_model clause. Refer to model_column_clauses on page 19-29 and cell_reference_
options on page 19-29.
Restrictions on the reference_model clause This clause is subject to the following
restrictions:
■

PARTITION BY columns cannot be specified for reference models.

■

The subquery of the reference model cannot refer to columns in an outer subquery.

Set Operators: UNION, UNION ALL, INTERSECT, MINUS
The set operators combine the rows returned by two SELECT statements into a single
result. The number and data types of the columns selected by each component query
must be the same, but the column lengths can be different. The names of the columns
in the result set are the names of the expressions in the select list preceding the set
operator.
If you combine more than two queries with set operators, then the database evaluates
adjacent queries from left to right. The parentheses around the subquery are optional.
You can use them to specify a different order of evaluation.
Refer to "The UNION [ALL], INTERSECT, MINUS Operators" on page 9-8 for
information on these operators, including restrictions on their use.

order_by_clause
Use the ORDER BY clause to order rows returned by the statement. Without an order_
by_clause, no guarantee exists that the same query executed more than once will
retrieve rows in the same order.
The SIBLINGS keyword is valid only if you also specify the hierarchical_
query_clause (CONNECT BY). ORDER SIBLINGS BY preserves any ordering specified in the
hierarchical query clause and then applies the order_by_clause to the siblings of the
hierarchy.

SIBLINGS

expr expr orders rows based on their value for expr. The expression is based on
columns in the select list or columns in the tables, views, or materialized views in the
FROM clause.
position Specify position to order rows based on their value for the expression in
this position of the select list. The position value must be an integer.

You can specify multiple expressions in the order_by_clause. Oracle Database first
sorts rows based on their values for the first expression. Rows with the same value for
the first expression are then sorted based on their values for the second expression,
and so on. The database sorts nulls following all others in ascending order and
preceding all others in descending order. Refer to "Sorting Query Results" on page 9-10
for a discussion of ordering query results.
SQL Statements: SAVEPOINT to UPDATE

19-33

SELECT

ASC | DESC

Specify whether the ordering sequence is ascending or descending. ASC

is the default.
NULLS FIRST | NULLS LAST Specify whether returned rows containing null values
should appear first or last in the ordering sequence.

NULLS LAST is the default for ascending order, and NULLS FIRST is the default for
descending order.
Restrictions on the ORDER BY Clause The following restrictions apply to the ORDER

BY clause:
■

If you have specified the DISTINCT operator in this statement, then this clause
cannot refer to columns unless they appear in the select list.

■

An order_by_clause can contain no more than 255 expressions.

■

You cannot order by a LOB, LONG, or LONG RAW column, nested table, or varray.

■

If you specify a group_by_clause in the same statement, then this order_by_clause
is restricted to the following expressions:
–

Constants

–

Aggregate functions

–

Analytic functions

–

The functions USER, UID, and SYSDATE

–

Expressions identical to those in the group_by_clause

–

Expressions comprising the preceding expressions that evaluate to the same
value for all rows in a group
See Also:

"Using the ORDER BY Clause: Examples" on page 19-45

for_update_clause
The FOR UPDATE clause lets you lock the selected rows so that other users cannot lock or
update the rows until you end your transaction. You can specify this clause only in a
top-level SELECT statement, not in subqueries.
Prior to updating a LOB value, you must lock the row
containing the LOB. One way to lock the row is with an embedded
SELECT ... FOR UPDATE statement. You can do this using one of the
programmatic languages or DBMS_LOB package. For more information
on lock rows before writing to a LOB, see Oracle Database SecureFiles
and Large Objects Developer's Guide.
Note:

Nested table rows are not locked as a result of locking the parent table rows. If you
want the nested table rows to be locked, then you must lock them explicitly.
Restrictions on the FOR UPDATE Clause

This clause is subject to the following

restrictions:
■

You cannot specify this clause with the following other constructs: the DISTINCT
operator, CURSOR expression, set operators, group_by_clause, or aggregate
functions.

19-34 Oracle Database SQL Language Reference

SELECT

■

The tables locked by this clause must all be located on the same database and on
the same database as any LONG columns and sequences referenced in the same
statement.
See Also: "Using the FOR UPDATE Clause: Examples" on
page 19-47

Using the FOR UPDATE Clause on Views In general, this clause is not supported on
views. However, in some cases, a SELECT ... FOR UPDATE query on a view can succeed
without any errors. This occurs when the view has been merged to its containing
query block internally by the query optimizer, and SELECT ... FOR UPDATE succeeds on
the internally transformed query. The examples in this section illustrate when using
the FOR UPDATE clause on a view can succeed or fail.
■

Using the FOR UPDATE clause on merged views
An error can occur when you use the FOR UPDATE clause on a merged view if both
of the following conditions apply:
–

The underlying column of the view is an expression

–

The FOR UPDATE clause applies to a column list

The following statement succeeds because the underlying column of the view is
not an expression:
SELECT employee_id FROM (SELECT * FROM employees)
FOR UPDATE OF employee_id;

The following statement succeeds because, while the underlying column of the
view is an expression, the FOR UPDATE clause does not apply to a column list:
SELECT employee_id FROM (SELECT employee_id+1 AS employee_id FROM employees)
FOR UPDATE;

The following statement fails because the underlying column of the view is an
expression and the FOR UPDATE clause applies to a column list:
SELECT employee_id FROM (SELECT employee_id+1 AS employee_id FROM employees)
FOR UPDATE OF employee_id;
*
Error at line 2:
ORA-01733: virtual column not allowed here
■

Using the FOR UPDATE clause on non-merged views
Since the FOR UPDATE clause is not supported on views, anything that prevents
view merging, such as the NO_MERGE hint, parameters that disallow view merging,
or something in the query structure that prevents view merging, will result in an
ORA-02014 error.
In the following example, the GROUP BY statement prevents view merging, which
causes an error:
SELECT avgsal
FROM (SELECT AVG(salary) AS avgsal FROM employees GROUP BY job_id)
FOR UPDATE;
FROM (SELECT AVG(salary) AS avgsal FROM employees GROUP BY job_id)
*
ERROR at line 2:
ORA-02014: cannot select FOR UPDATE from view with DISTINCT, GROUP BY, etc.

SQL Statements: SAVEPOINT to UPDATE

19-35

SELECT

Due to the complexity of the view merging mechanism, Oracle
recommends against using the FOR UPDATE clause on views.

Note:

OF ... column
Use the OF ... column clause to lock the select rows only for a particular table or view in
a join. The columns in the OF clause only indicate which table or view rows are locked.
The specific columns that you specify are not significant. However, you must specify
an actual column name, not a column alias. If you omit this clause, then the database
locks the selected rows from all the tables in the query.
NOWAIT | WAIT
The NOWAIT and WAIT clauses let you tell the database how to proceed if the SELECT
statement attempts to lock a row that is locked by another user.
■
■

Specify NOWAIT to return control to you immediately if a lock exists.
Specify WAIT to instruct the database to wait integer seconds for the row to
become available and then return control to you.

If you specify neither WAIT nor NOWAIT, then the database waits until the row is
available and then returns the results of the SELECT statement.
SKIP LOCKED is an alternative way to handle a contending transaction
that is locking some rows of interest. Specify SKIP LOCKED to instruct the database to
attempt to lock the rows specified by the WHERE clause and to skip any rows that are
found to be already locked by another transaction. This feature is designed for use in
multiconsumer queue environments, such as Oracle Streams Advanced Queuing. It
enables queue consumers to skip rows that are locked by other consumers and obtain
unlocked rows without waiting for the other consumers to finish. Oracle recommends
that you use the Oracle Streams Advanced Queuing APIs instead of directly using the
SKIP LOCKED functionality. Refer to Oracle Streams Advanced Queuing User's Guide for
more information.
SKIP LOCKED

Note on the WAIT and SKIP LOCKED Clauses

If you specify WAIT or SKIP LOCKED and the table is locked in exclusive mode, then the
database will not return the results of the SELECT statement until the lock on the table
is released. In the case of WAIT, the SELECT FOR UPDATE clause is blocked regardless of
the wait time specified.

Examples
19

The following statement creates the query names
dept_costs and avg_cost for the initial query block containing a join, and then uses
the query names in the body of the main query.
Subquery Factoring: Example

WITH
dept_costs AS (
SELECT department_name, SUM(salary) dept_total
FROM employees e, departments d
WHERE e.department_id = d.department_id
GROUP BY department_name),
avg_cost AS (
SELECT SUM(dept_total)/COUNT(*) avg
FROM dept_costs)
SELECT * FROM dept_costs
WHERE dept_total >
19-36 Oracle Database SQL Language Reference

SELECT

(SELECT avg FROM avg_cost)
ORDER BY department_name;
DEPARTMENT_NAME
DEPT_TOTAL
------------------------------ ---------Sales
304500
Shipping
156400

The following statement shows the
employees who directly or indirectly report to employee 101 and their reporting level.

Recursive Subquery Factoring: Examples

WITH
reports_to_101 (eid, emp_last, mgr_id, reportLevel) AS
(
SELECT employee_id, last_name, manager_id, 0 reportLevel
FROM employees
WHERE employee_id = 101
UNION ALL
SELECT e.employee_id, e.last_name, e.manager_id, reportLevel+1
FROM reports_to_101 r, employees e
WHERE r.eid = e.manager_id
)
SELECT eid, emp_last, mgr_id, reportLevel
FROM reports_to_101
ORDER BY reportLevel, eid;
EID
---------101
108
200
203
204
205
109
110
111
112
113
206

EMP_LAST
MGR_ID REPORTLEVEL
------------------------- ---------- ----------Kochhar
100
0
Greenberg
101
1
Whalen
101
1
Mavris
101
1
Baer
101
1
Higgins
101
1
Faviet
108
2
Chen
108
2
Sciarra
108
2
Urman
108
2
Popp
108
2
Gietz
205
2

The following statement shows employees who directly or indirectly report to
employee 101, their reporting level, and their management chain.
WITH
reports_to_101 (eid, emp_last, mgr_id, reportLevel, mgr_list) AS
(
SELECT employee_id, last_name, manager_id, 0 reportLevel,
CAST(manager_id AS VARCHAR2(2000))
FROM employees
WHERE employee_id = 101
UNION ALL
SELECT e.employee_id, e.last_name, e.manager_id, reportLevel+1,
CAST(mgr_list || ',' || manager_id AS VARCHAR2(2000))
FROM reports_to_101 r, employees e
WHERE r.eid = e.manager_id
)
SELECT eid, emp_last, mgr_id, reportLevel, mgr_list
FROM reports_to_101
ORDER BY reportLevel, eid;

SQL Statements: SAVEPOINT to UPDATE

19-37

SELECT

EID EMP_LAST
MGR_ID REPORTLEVEL MGR_LIST
---------- ------------------------- ---------- ----------- -------101 Kochhar
100
0 100
108 Greenberg
101
1 100,101
200 Whalen
101
1 100,101
203 Mavris
101
1 100,101
204 Baer
101
1 100,101
205 Higgins
101
1 100,101
109 Faviet
108
2 100,101,108
110 Chen
108
2 100,101,108
111 Sciarra
108
2 100,101,108
112 Urman
108
2 100,101,108
113 Popp
108
2 100,101,108
206 Gietz
205
2 100,101,205

The following statement shows the employees who directly or indirectly report to
employee 101 and their reporting level. It stops at reporting level 1.
WITH
reports_to_101 (eid, emp_last, mgr_id, reportLevel) AS
(
SELECT employee_id, last_name, manager_id, 0 reportLevel
FROM employees
WHERE employee_id = 101
UNION ALL
SELECT e.employee_id, e.last_name, e.manager_id, reportLevel+1
FROM reports_to_101 r, employees e
WHERE r.eid = e.manager_id
)
SELECT eid, emp_last, mgr_id, reportLevel
FROM reports_to_101
WHERE reportLevel <= 1
ORDER BY reportLevel, eid;
EID
---------101
108
200
203
204
205

EMP_LAST
MGR_ID REPORTLEVEL
------------------------- ---------- ----------Kochhar
100
0
Greenberg
101
1
Whalen
101
1
Mavris
101
1
Baer
101
1
Higgins
101
1

The following statement shows the entire organization, indenting for each level of
management.
WITH
org_chart (eid, emp_last, mgr_id, reportLevel, salary, job_id) AS
(
SELECT employee_id, last_name, manager_id, 0 reportLevel, salary, job_id
FROM employees
WHERE manager_id is null
UNION ALL
SELECT e.employee_id, e.last_name, e.manager_id,
r.reportLevel+1 reportLevel, e.salary, e.job_id
FROM org_chart r, employees e
WHERE r.eid = e.manager_id
)
SEARCH DEPTH FIRST BY emp_last SET order1
SELECT lpad(' ',2*reportLevel)||emp_last emp_name, eid, mgr_id, salary, job_id
FROM org_chart

19-38 Oracle Database SQL Language Reference

SELECT

ORDER BY order1;
EMP_NAME
EID
MGR_ID
SALARY JOB_ID
-------------------- ---------- ---------- ---------- ---------King
100
24000 AD_PRES
Cambrault
148
100
11000 SA_MAN
Bates
172
148
7300 SA_REP
Bloom
169
148
10000 SA_REP
Fox
170
148
9600 SA_REP
Kumar
173
148
6100 SA_REP
Ozer
168
148
11500 SA_REP
Smith
171
148
7400 SA_REP
De Haan
102
100
17000 AD_VP
Hunold
103
102
9000 IT_PROG
Austin
105
103
4800 IT_PROG
Ernst
104
103
6000 IT_PROG
Lorentz
107
103
4200 IT_PROG
Pataballa
106
103
4800 IT_PROG
Errazuriz
147
100
12000 SA_MAN
Ande
166
147
6400 SA_REP
. . .

The following statement shows the entire organization, indenting for each level of
management, with each level ordered by hire_date. The value of is_cycle is set to Y
for any employee who has the same hire_date as any manager above him in the
management chain.
WITH
dup_hiredate (eid, emp_last, mgr_id, reportLevel, hire_date, job_id) AS
(
SELECT employee_id, last_name, manager_id, 0 reportLevel, hire_date, job_id
FROM employees
WHERE manager_id is null
UNION ALL
SELECT e.employee_id, e.last_name, e.manager_id,
r.reportLevel+1 reportLevel, e.hire_date, e.job_id
FROM dup_hiredate r, employees e
WHERE r.eid = e.manager_id
)
SEARCH DEPTH FIRST BY hire_date SET order1
CYCLE hire_date SET is_cycle TO 'Y' DEFAULT 'N'
SELECT lpad(' ',2*reportLevel)||emp_last emp_name, eid, mgr_id,
hire_date, job_id, is_cycle
FROM dup_hiredate
ORDER BY order1;
EMP_NAME
EID
MGR_ID HIRE_DATE JOB_ID
IS_CYCLE
-------------------- ---------- ---------- --------- ---------- -------King
100
17-JUN-03 AD_PRES
N
De Haan
102
100 13-JAN-01 AD_VP
N
Hunold
103
102 03-JAN-06 IT_PROG
N
Austin
105
103 25-JUN-05 IT_PROG
N
. . .
Kochhar
101
100 21-SEP-05 AD_VP
N
Mavris
203
101 07-JUN-02 HR_REP
N
Baer
204
101 07-JUN-02 PR_REP
N
Higgins
205
101 07-JUN-02 AC_MGR
N
Gietz
206
205 07-JUN-02 AC_ACCOUNT
Y
Greenberg
108
101 17-AUG-02 FI_MGR
N
Faviet
109
108 16-AUG-02 FI_ACCOUNT
N

SQL Statements: SAVEPOINT to UPDATE

19-39

SELECT

Chen

110

108 28-SEP-05 FI_ACCOUNT

N

. . .

The following statement counts the number of employees under each manager.
WITH
emp_count (eid, emp_last, mgr_id, mgrLevel, salary, cnt_employees) AS
(
SELECT employee_id, last_name, manager_id, 0 mgrLevel, salary, 0 cnt_employees
FROM employees
UNION ALL
SELECT e.employee_id, e.last_name, e.manager_id,
r.mgrLevel+1 mgrLevel, e.salary, 1 cnt_employees
FROM emp_count r, employees e
WHERE e.employee_id = r.mgr_id
)
SEARCH DEPTH FIRST BY emp_last SET order1
SELECT emp_last, eid, mgr_id, salary, sum(cnt_employees), max(mgrLevel) mgrLevel
FROM emp_count
GROUP BY emp_last, eid, mgr_id, salary
HAVING max(mgrLevel) > 0
ORDER BY mgr_id NULLS FIRST, emp_last;
EMP_LAST
EID
MGR_ID
SALARY SUM(CNT_EMPLOYEES)
MGRLEVEL
------------------ ---------- ---------- ---------- ------------------ ---------King
100
24000
106
3
Cambrault
148
100
11000
7
2
De Haan
102
100
17000
5
2
Errazuriz
147
100
12000
6
1
Fripp
121
100
8200
8
1
Hartstein
201
100
13000
1
1
Kaufling
122
100
7900
8
1
. . .

Simple Query Examples The following statement selects rows from the employees
table with the department number of 30:
SELECT *
FROM employees
WHERE department_id = 30
ORDER BY last_name;

The following statement selects the name, job, salary and department number of all
employees except purchasing clerks from department number 30:
SELECT last_name, job_id, salary, department_id
FROM employees
WHERE NOT (job_id = 'PU_CLERK' AND department_id = 30)
ORDER BY last_name;

The following statement selects from subqueries in the FROM clause and for each
department returns the total employees and salaries as a decimal value of all the
departments:
SELECT a.department_id "Department",
a.num_emp/b.total_count "%_Employees",
a.sal_sum/b.total_sal "%_Salary"
FROM
(SELECT department_id, COUNT(*) num_emp, SUM(salary) sal_sum
FROM employees
GROUP BY department_id) a,

19-40 Oracle Database SQL Language Reference

SELECT

(SELECT COUNT(*) total_count, SUM(salary) total_sal
FROM employees) b
ORDER BY a.department_id;

Selecting from a Partition: Example You can select rows from a single partition of a
partitioned table by specifying the keyword PARTITION in the FROM clause. This SQL
statement assigns an alias for and retrieves rows from the sales_q2_2000 partition of
the sample table sh.sales:
SELECT * FROM sales PARTITION (sales_q2_2000) s
WHERE s.amount_sold > 1500
ORDER BY cust_id, time_id, channel_id;

The following example selects rows from the oe.orders table for orders earlier than a
specified date:
SELECT * FROM orders
WHERE order_date < TO_DATE('2006-06-15', 'YYYY-MM-DD');

Selecting a Sample: Examples

The following query estimates the number of orders

in the oe.orders table:
SELECT COUNT(*) * 10 FROM orders SAMPLE (10);
COUNT(*)*10
----------70

Because the query returns an estimate, the actual return value may differ from one
query to the next.
SELECT COUNT(*) * 10 FROM orders SAMPLE (10);
COUNT(*)*10
----------80

The following query adds a seed value to the preceding query. Oracle Database always
returns the same estimate given the same seed value:
SELECT COUNT(*) * 10 FROM orders SAMPLE(10) SEED (1);
COUNT(*)*10
----------130
SELECT COUNT(*) * 10 FROM orders SAMPLE(10) SEED(4);
COUNT(*)*10
----------120
SELECT COUNT(*) * 10 FROM orders SAMPLE(10) SEED (1);
COUNT(*)*10
----------130

Using Flashback Queries: Example The following statements show a current value
from the sample table hr.employees and then change the value. The intervals used in

SQL Statements: SAVEPOINT to UPDATE

19-41

SELECT

these examples are very short for demonstration purposes. Time intervals in your own
environment are likely to be larger.
SELECT salary FROM employees
WHERE last_name = 'Chung';
SALARY
---------3800
UPDATE employees SET salary = 4000
WHERE last_name = 'Chung';
1 row updated.
SELECT salary FROM employees
WHERE last_name = 'Chung';
SALARY
---------4000

To learn what the value was before the update, you can use the following Flashback
Query:
SELECT salary FROM employees
AS OF TIMESTAMP (SYSTIMESTAMP - INTERVAL '1' MINUTE)
WHERE last_name = 'Chung';
SALARY
---------3800

To learn what the values were during a particular time period, you can use a version
Flashback Query:
SELECT salary FROM employees
VERSIONS BETWEEN TIMESTAMP
SYSTIMESTAMP - INTERVAL '10' MINUTE AND
SYSTIMESTAMP - INTERVAL '1' MINUTE
WHERE last_name = 'Chung';

To revert to the earlier value, use the Flashback Query as the subquery of another
UPDATE statement:
UPDATE employees SET salary =
(SELECT salary FROM employees
AS OF TIMESTAMP (SYSTIMESTAMP - INTERVAL '2' MINUTE)
WHERE last_name = 'Chung')
WHERE last_name = 'Chung';
1 row updated.
SELECT salary FROM employees
WHERE last_name = 'Chung';
SALARY
---------3800

Using the GROUP BY Clause: Examples To return the minimum and maximum
salaries for each department in the employees table, issue the following statement:
SELECT department_id, MIN(salary), MAX (salary)
19-42 Oracle Database SQL Language Reference

SELECT

FROM employees
GROUP BY department_id
ORDER BY department_id;

To return the minimum and maximum salaries for the clerks in each department, issue
the following statement:
SELECT department_id, MIN(salary), MAX (salary)
FROM employees
WHERE job_id = 'PU_CLERK'
GROUP BY department_id
ORDER BY department_id;

To return the number of employees
and their average yearly salary across all possible combinations of department and job
category, issue the following query on the sample tables hr.employees and
hr.departments:
Using the GROUP BY CUBE Clause: Example

SELECT DECODE(GROUPING(department_name), 1, 'All Departments',
department_name) AS department_name,
DECODE(GROUPING(job_id), 1, 'All Jobs', job_id) AS job_id,
COUNT(*) "Total Empl", AVG(salary) * 12 "Average Sal"
FROM employees e, departments d
WHERE d.department_id = e.department_id
GROUP BY CUBE (department_name, job_id)
ORDER BY department_name, job_id;
DEPARTMENT_NAME
-----------------------------Accounting
Accounting
Accounting
Administration
. . .
Shipping
Shipping

JOB_ID
Total Empl Average Sal
---------- ---------- ----------AC_ACCOUNT
1
99600
AC_MGR
1
144000
All Jobs
2
121800
AD_ASST
1
52800
ST_CLERK
ST_MAN

20
5

33420
87360

Using the GROUPING SETS Clause: Example The following example finds the sum
of sales aggregated for three precisely specified groups:
■

(channel_desc, calendar_month_desc, country_id)

■

(channel_desc, country_id)

■

(calendar_month_desc, country_id)

Without the GROUPING SETS syntax, you would have to write less efficient queries with
more complicated SQL. For example, you could run three separate queries and UNION
them, or run a query with a CUBE(channel_desc, calendar_month_desc, country_
id) operation and filter out five of the eight groups it would generate.
SELECT channel_desc, calendar_month_desc, co.country_id,
TO_CHAR(sum(amount_sold) , '9,999,999,999') SALES$
FROM sales, customers, times, channels, countries co
WHERE sales.time_id=times.time_id
AND sales.cust_id=customers.cust_id
AND sales.channel_id= channels.channel_id
AND customers.country_id = co.country_id
AND channels.channel_desc IN ('Direct Sales', 'Internet')
AND times.calendar_month_desc IN ('2000-09', '2000-10')
AND co.country_iso_code IN ('UK', 'US')
GROUP BY GROUPING SETS(

SQL Statements: SAVEPOINT to UPDATE

19-43

SELECT

(channel_desc, calendar_month_desc, co.country_id),
(channel_desc, co.country_id),
(calendar_month_desc, co.country_id) );
CHANNEL_DESC
-------------------Internet
Direct Sales
Internet
Direct Sales

Internet
Direct Sales

CALENDAR COUNTRY_ID
-------- ---------2000-09
52790
2000-09
52790
2000-10
52790
2000-10
52790
2000-09
52790
2000-10
52790
52790
52790

SALES$
---------124,224
638,201
137,054
682,297
762,425
819,351
261,278
1,320,497

See Also: The functions GROUP_ID, GROUPING, and
GROUPING_ID on page 5-109 for more information on those
functions
Hierarchical Query Examples The following query with a CONNECT BY clause defines
a hierarchical relationship in which the employee_id value of the parent row is equal
to the manager_id value of the child row:
SELECT last_name, employee_id, manager_id FROM employees
CONNECT BY employee_id = manager_id
ORDER BY last_name;

In the following CONNECT BY clause, the PRIOR operator applies only to the employee_id
value. To evaluate this condition, the database evaluates employee_id values for the
parent row and manager_id, salary, and commission_pct values for the child row:
SELECT last_name, employee_id, manager_id FROM employees
CONNECT BY PRIOR employee_id = manager_id
AND salary > commission_pct
ORDER BY last_name;

To qualify as a child row, a row must have a manager_id value equal to the employee_
id value of the parent row and it must have a salary value greater than its
commission_pct value.
To return the minimum and maximum
salaries for the employees in each department whose lowest salary is less than $5,000,
issue the next statement:

Using the HAVING Condition: Example

SELECT department_id, MIN(salary), MAX (salary)
FROM employees
GROUP BY department_id
HAVING MIN(salary) < 5000
ORDER BY department_id;
DEPARTMENT_ID MIN(SALARY) MAX(SALARY)
------------- ----------- ----------10
4400
4400
30
2500
11000
50
2100
8200
60
4200
9000

The following example uses a correlated subquery in a HAVING clause that eliminates
from the result set any departments without managers and managers without
departments:
19-44 Oracle Database SQL Language Reference

SELECT

SELECT department_id, manager_id
FROM employees
GROUP BY department_id, manager_id HAVING (department_id, manager_id) IN
(SELECT department_id, manager_id FROM employees x
WHERE x.department_id = employees.department_id)
ORDER BY department_id;

To select all purchasing clerk records from
employees and order the results by salary in descending order, issue the following
statement:
Using the ORDER BY Clause: Examples

SELECT *
FROM employees
WHERE job_id = 'PU_CLERK'
ORDER BY salary DESC;

To select information from employees ordered first by ascending department number
and then by descending salary, issue the following statement:
SELECT last_name, department_id, salary
FROM employees
ORDER BY department_id ASC, salary DESC, last_name;

To select the same information as the previous SELECT and use the positional ORDER BY
notation, issue the following statement, which orders by ascending department_id,
then descending salary, and finally alphabetically by last_name:
SELECT last_name, department_id, salary
FROM employees
ORDER BY 2 ASC, 3 DESC, 1;

The MODEL clause: Examples The view created below is based on the sample sh
schema and is used by the example that follows.
CREATE OR REPLACE VIEW sales_view_ref AS
SELECT country_name country,
prod_name prod,
calendar_year year,
SUM(amount_sold) sale,
COUNT(amount_sold) cnt
FROM sales,times,customers,countries,products
WHERE sales.time_id = times.time_id
AND sales.prod_id = products.prod_id
AND sales.cust_id = customers.cust_id
AND customers.country_id = countries.country_id
AND ( customers.country_id = 52779
OR customers.country_id = 52776 )
AND ( prod_name = 'Standard Mouse'
OR prod_name = 'Mouse Pad' )
GROUP BY country_name,prod_name,calendar_year;
SELECT country, prod, year, sale
FROM sales_view_ref
ORDER BY country, prod, year;
COUNTRY
---------France
France
France
France

PROD
----------------------------------Mouse Pad
Mouse Pad
Mouse Pad
Mouse Pad

YEAR
-------1998
1999
2000
2001

SALE
--------2509.42
3678.69
3000.72
3269.09

SQL Statements: SAVEPOINT to UPDATE

19-45

SELECT

France
France
France
France
Germany
Germany
Germany
Germany
Germany
Germany
Germany
Germany

Standard Mouse
Standard Mouse
Standard Mouse
Standard Mouse
Mouse Pad
Mouse Pad
Mouse Pad
Mouse Pad
Standard Mouse
Standard Mouse
Standard Mouse
Standard Mouse

1998
1999
2000
2001
1998
1999
2000
2001
1998
1999
2000
2001

2390.83
2280.45
1274.31
2164.54
5827.87
8346.44
7375.46
9535.08
7116.11
6263.14
2637.31
6456.13

16 rows selected.

The next example creates a multidimensional array from sales_view_ref with
columns containing country, product, year, and sales. It also:
■

■

Assigns the sum of the sales of the Mouse Pad for years 1999 and 2000 to the sales
of the Mouse Pad for year 2001, if a row containing sales of the Mouse Pad for year
2001 exists.
Assigns the value of sales of the Standard Mouse for year 2001 to sales of the
Standard Mouse for year 2002, creating a new row if a row containing sales of the
Standard Mouse for year 2002 does not exist.

SELECT country,prod,year,s
FROM sales_view_ref
MODEL
PARTITION BY (country)
DIMENSION BY (prod, year)
MEASURES (sale s)
IGNORE NAV
UNIQUE DIMENSION
RULES UPSERT SEQUENTIAL ORDER
(
s[prod='Mouse Pad', year=2001] =
s['Mouse Pad', 1999] + s['Mouse Pad', 2000],
s['Standard Mouse', 2002] = s['Standard Mouse', 2001]
)
ORDER BY country, prod, year;
COUNTRY
---------France
France
France
France
France
France
France
France
France
Germany
Germany
Germany
Germany
Germany
Germany
Germany
Germany

PROD
----------------------------------Mouse Pad
Mouse Pad
Mouse Pad
Mouse Pad
Standard Mouse
Standard Mouse
Standard Mouse
Standard Mouse
Standard Mouse
Mouse Pad
Mouse Pad
Mouse Pad
Mouse Pad
Standard Mouse
Standard Mouse
Standard Mouse
Standard Mouse

19-46 Oracle Database SQL Language Reference

YEAR
-------1998
1999
2000
2001
1998
1999
2000
2001
2002
1998
1999
2000
2001
1998
1999
2000
2001

SALE
--------2509.42
3678.69
3000.72
6679.41
2390.83
2280.45
1274.31
2164.54
2164.54
5827.87
8346.44
7375.46
15721.9
7116.11
6263.14
2637.31
6456.13

SELECT

Germany

Standard Mouse

2002

6456.13

18 rows selected.

The first rule uses UPDATE behavior because symbolic referencing is used on the
left-hand side of the rule. The rows represented by the left-hand side of the rule exist,
so the measure columns are updated. If the rows did not exist, then no action would
have been taken.
The second rule uses UPSERT behavior because positional referencing is used on the
left-hand side and a single cell is referenced. The rows do not exist, so new rows are
inserted and the related measure columns are updated. If the rows did exist, then the
measure columns would have been updated.
See Also: Oracle Database Data Warehousing Guide for an expanded
discussion and examples

The next example uses the same sales_view_ref view and the analytic function SUM to
calculate a cumulative sum (csum) of sales per country and per year.
SELECT country, year, sale, csum
FROM
(SELECT country, year, SUM(sale) sale
FROM sales_view_ref
GROUP BY country, year
)
MODEL DIMENSION BY (country, year)
MEASURES (sale, 0 csum)
RULES (csum[any, any]=
SUM(sale) OVER (PARTITION BY country
ORDER BY year
ROWS UNBOUNDED PRECEDING)
)
ORDER BY country, year;
COUNTRY
YEAR
SALE
CSUM
--------------- ---------- ---------- ---------France
1998
4900.25
4900.25
France
1999
5959.14
10859.39
France
2000
4275.03
15134.42
France
2001
5433.63
20568.05
Germany
1998
12943.98
12943.98
Germany
1999
14609.58
27553.56
Germany
2000
10012.77
37566.33
Germany
2001
15991.21
53557.54
8 rows selected.

Using the FOR UPDATE Clause: Examples The following statement locks rows in
the employees table with purchasing clerks located in Oxford, which has location_id
2500, and locks rows in the departments table with departments in Oxford that have
purchasing clerks:
SELECT e.employee_id, e.salary, e.commission_pct
FROM employees e, departments d
WHERE job_id = 'SA_REP'
AND e.department_id = d.department_id
AND location_id = 2500
ORDER BY e.employee_id
FOR UPDATE;

SQL Statements: SAVEPOINT to UPDATE

19-47

SELECT

The following statement locks only those rows in the employees table with purchasing
clerks located in Oxford. No rows are locked in the departments table:
SELECT e.employee_id, e.salary, e.commission_pct
FROM employees e JOIN departments d
USING (department_id)
WHERE job_id = 'SA_REP'
AND location_id = 2500
ORDER BY e.employee_id
FOR UPDATE OF e.salary;

The following statement is legal
even though the third value inserted violates the condition of the subquery where_
clause:

Using the WITH CHECK OPTION Clause: Example

INSERT INTO (SELECT department_id, department_name, location_id
FROM departments WHERE location_id < 2000)
VALUES (9999, 'Entertainment', 2500);

However, the following statement is illegal because it contains the WITH CHECK OPTION
clause:
INSERT INTO (SELECT department_id, department_name, location_id
FROM departments WHERE location_id < 2000 WITH CHECK OPTION)
VALUES (9999, 'Entertainment', 2500);
*
ERROR at line 2:
ORA-01402: view WITH CHECK OPTION where-clause violation

Using PIVOT and UNPIVOT: Examples The oe.orders table contains information
about when an order was placed (order_date), how it was place (order_mode), and the
total amount of the order (order_total), as well as other information. The following
example shows how to use the PIVOT clause to pivot order_mode values into columns,
aggregating order_total data in the process, to get yearly totals by order mode:
CREATE TABLE pivot_table AS
SELECT * FROM
(SELECT EXTRACT(YEAR FROM order_date) year, order_mode, order_total FROM orders)
PIVOT
(SUM(order_total) FOR order_mode IN ('direct' AS Store, 'online' AS Internet));
SELECT * FROM pivot_table ORDER BY year;
YEAR
STORE
INTERNET
---------- ---------- ---------2004
5546.6
2006
371895.5
100056.6
2007 1274078.8 1271019.5
2008
252108.3
393349.4

The UNPIVOT clause lets you rotate specified columns so that the input column
headings are output as values of one or more descriptor columns, and the input
column values are output as values of one or more measures columns. The first query
that follows shows that nulls are excluded by default. The second query shows that
you can include nulls using the INCLUDE NULLS clause.
SELECT * FROM pivot_table
UNPIVOT (yearly_total FOR order_mode IN (store AS 'direct',
internet AS 'online'))
ORDER BY year, order_mode;

19-48 Oracle Database SQL Language Reference

SELECT

YEAR
---------2004
2006
2006
2007
2007
2008
2008

ORDER_ YEARLY_TOTAL
------ -----------direct
5546.6
direct
371895.5
online
100056.6
direct
1274078.8
online
1271019.5
direct
252108.3
online
393349.4

7 rows selected.
SELECT * FROM pivot_table
UNPIVOT INCLUDE NULLS
(yearly_total FOR order_mode IN (store AS 'direct', internet AS 'online'))
ORDER BY year, order_mode;
YEAR
---------2004
2004
2006
2006
2007
2007
2008
2008

ORDER_ YEARLY_TOTAL
------ -----------direct
5546.6
online
direct
371895.5
online
100056.6
direct
1274078.8
online
1271019.5
direct
252108.3
online
393349.4

8 rows selected.

Using Join Queries: Examples The following examples show various ways of
joining tables in a query. In the first example, an equijoin returns the name and job of
each employee and the number and name of the department in which the employee
works:
SELECT last_name, job_id, departments.department_id, department_name
FROM employees, departments
WHERE employees.department_id = departments.department_id
ORDER BY last_name, job_id;
LAST_NAME
JOB_ID
DEPARTMENT_ID DEPARTMENT_NAME
------------------- ---------- ------------- ---------------------Abel
SA_REP
80 Sales
Ande
SA_REP
80 Sales
Atkinson
ST_CLERK
50 Shipping
Austin
IT_PROG
60 IT
. . .

You must use a join to return this data because employee names and jobs are stored in
a different table than department names. Oracle Database combines rows of the two
tables according to this join condition:
employees.department_id = departments.department_id

The following equijoin returns the name, job, department number, and department
name of all sales managers:
SELECT last_name, job_id, departments.department_id, department_name
FROM employees, departments
WHERE employees.department_id = departments.department_id

SQL Statements: SAVEPOINT to UPDATE

19-49

SELECT

AND job_id = 'SA_MAN'
ORDER BY last_name;
LAST_NAME
------------------Cambrault
Errazuriz
Partners
Russell
Zlotkey

JOB_ID
DEPARTMENT_ID DEPARTMENT_NAME
---------- ------------- ----------------------SA_MAN
80 Sales
SA_MAN
80 Sales
SA_MAN
80 Sales
SA_MAN
80 Sales
SA_MAN
80 Sales

This query is identical to the preceding example, except that it uses an additional
where_clause condition to return only rows with a job value of 'SA_MAN'.
To determine who works in the same department as
employee 'Lorentz', issue the following statement:

Using Subqueries: Examples

SELECT last_name, department_id FROM employees
WHERE department_id =
(SELECT department_id FROM employees
WHERE last_name = 'Lorentz')
ORDER BY last_name, department_id;

To give all employees in the employees table a 10% raise if they have changed jobs—if
they appear in the job_history table—issue the following statement:
UPDATE employees
SET salary = salary * 1.1
WHERE employee_id IN (SELECT employee_id FROM job_history);

To create a second version of the departments table new_departments, with only three
of the columns of the original table, issue the following statement:
CREATE TABLE new_departments
(department_id, department_name, location_id)
AS SELECT department_id, department_name, location_id
FROM departments;

The following query uses a self join to return the name of
each employee along with the name of the employee's manager. A WHERE clause is
added to shorten the output.

Using Self Joins: Example

SELECT e1.last_name||' works for '||e2.last_name
"Employees and Their Managers"
FROM employees e1, employees e2
WHERE e1.manager_id = e2.employee_id
AND e1.last_name LIKE 'R%'
ORDER BY e1.last_name;
Employees and Their Managers
------------------------------Rajs works for Mourgos
Raphaely works for King
Rogers works for Kaufling
Russell works for King

The join condition for this query uses the aliases e1 and e2 for the sample table
employees:
e1.manager_id = e2.employee_id

19-50 Oracle Database SQL Language Reference

SELECT

Using Outer Joins: Examples The following example shows how a partitioned outer
join fills data gaps in rows to facilitate analytic function specification and reliable
report formatting. The example first creates a small data table to be used in the join:
SELECT d.department_id, e.last_name
FROM departments d LEFT OUTER JOIN employees e
ON d.department_id = e.department_id
ORDER BY d.department_id, e.last_name;

Users familiar with the traditional Oracle Database outer joins syntax will recognize
the same query in this form:
SELECT d.department_id, e.last_name
FROM departments d, employees e
WHERE d.department_id = e.department_id(+)
ORDER BY d.department_id, e.last_name;

Oracle strongly recommends that you use the more flexible FROM clause join syntax
shown in the former example.
The left outer join returns all departments, including those without any employees.
The same statement with a right outer join returns all employees, including those not
yet assigned to a department:
The employee Zeuss was added to the employees table for
these examples, and is not part of the sample data.

Note:

SELECT d.department_id, e.last_name
FROM departments d RIGHT OUTER JOIN employees e
ON d.department_id = e.department_id
ORDER BY d.department_id, e.last_name;
DEPARTMENT_ID
------------. . .
110
110

LAST_NAME
------------------------Gietz
Higgins
Grant
Zeuss

It is not clear from this result whether employees Grant and Zeuss have department_
id NULL, or whether their department_id is not in the departments table. To determine
this requires a full outer join:
SELECT d.department_id as d_dept_id, e.department_id as e_dept_id,
e.last_name
FROM departments d FULL OUTER JOIN employees e
ON d.department_id = e.department_id
ORDER BY d.department_id, e.last_name;
D_DEPT_ID E_DEPT_ID LAST_NAME
---------- ---------- ------------------------. . .
110
110 Gietz
110
110 Higgins
. . .
260
270
999 Zeuss

SQL Statements: SAVEPOINT to UPDATE

19-51

SELECT

Grant

Because the column names in this example are the same in both tables in the join, you
can also use the common column feature by specifying the USING clause of the join
syntax. The output is the same as for the preceding example except that the USING
clause coalesces the two matching columns department_id into a single column
output:
SELECT department_id AS d_e_dept_id, e.last_name
FROM departments d FULL OUTER JOIN employees e
USING (department_id)
ORDER BY department_id, e.last_name;
D_E_DEPT_ID
----------. . .
110
110
. . .
260
270
999

LAST_NAME
------------------------Higgins
Gietz

Zeuss
Grant

Using Partitioned Outer Joins: Examples The following example shows how a
partitioned outer join fills in gaps in rows to facilitate analytic calculation specification
and reliable report formatting. The example first creates and populates a simple table
to be used in the join:
CREATE TABLE inventory (time_id
product
quantity
INSERT
INSERT
INSERT
INSERT

INTO
INTO
INTO
INTO

inventory
inventory
inventory
inventory

VALUES
VALUES
VALUES
VALUES

DATE,
VARCHAR2(10),
NUMBER);

(TO_DATE('01/04/01',
(TO_DATE('06/04/01',
(TO_DATE('01/04/01',
(TO_DATE('04/04/01',

'DD/MM/YY'),
'DD/MM/YY'),
'DD/MM/YY'),
'DD/MM/YY'),

'bottle', 10);
'bottle', 10);
'can', 10);
'can', 10);

SELECT times.time_id, product, quantity FROM inventory
PARTITION BY (product)
RIGHT OUTER JOIN times ON (times.time_id = inventory.time_id)
WHERE times.time_id BETWEEN TO_DATE('01/04/01', 'DD/MM/YY')
AND TO_DATE('06/04/01', 'DD/MM/YY')
ORDER BY 2,1;
TIME_ID
--------01-APR-01
02-APR-01
03-APR-01
04-APR-01
05-APR-01
06-APR-01
01-APR-01
02-APR-01
03-APR-01
04-APR-01
05-APR-01
06-APR-01

PRODUCT
QUANTITY
---------- ---------bottle
10
bottle
bottle
bottle
bottle
bottle
10
can
10
can
can
can
10
can
can

19-52 Oracle Database SQL Language Reference

SELECT

12 rows selected.

The data is now more dense along the time dimension for each partition of the product
dimension. However, each of the newly added rows within each partition is null in the
quantity column. It is more useful to see the nulls replaced by the preceding non-NULL
value in time order. You can achieve this by applying the analytic function LAST_VALUE
on top of the query result:
SELECT time_id, product, LAST_VALUE(quantity IGNORE NULLS)
OVER (PARTITION BY product ORDER BY time_id) quantity
FROM ( SELECT times.time_id, product, quantity
FROM inventory PARTITION BY (product)
RIGHT OUTER JOIN times ON (times.time_id = inventory.time_id)
WHERE times.time_id BETWEEN TO_DATE('01/04/01', 'DD/MM/YY')
AND TO_DATE('06/04/01', 'DD/MM/YY'))
ORDER BY 2,1;
TIME_ID
--------01-APR-01
02-APR-01
03-APR-01
04-APR-01
05-APR-01
06-APR-01
01-APR-01
02-APR-01
03-APR-01
04-APR-01
05-APR-01
06-APR-01

PRODUCT
QUANTITY
---------- ---------bottle
10
bottle
10
bottle
10
bottle
10
bottle
10
bottle
10
can
10
can
10
can
10
can
10
can
10
can
10

12 rows selected.

See Also: Oracle Database Data Warehousing Guide for an expanded
discussion on filling gaps in time series calculations and examples of
usage
Using Antijoins: Example The following example selects a list of employees who are
not in a particular set of departments:
SELECT * FROM employees
WHERE department_id NOT IN
(SELECT department_id FROM departments
WHERE location_id = 1700)
ORDER BY last_name;

Using Semijoins: Example In the following example, only one row needs to be
returned from the departments table, even though many rows in the employees table
might match the subquery. If no index has been defined on the salary column in
employees, then a semijoin can be used to improve query performance.
SELECT * FROM departments
WHERE EXISTS
(SELECT * FROM employees
WHERE departments.department_id = employees.department_id
AND employees.salary > 2500)
ORDER BY department_name;

SQL Statements: SAVEPOINT to UPDATE

19-53

SELECT

Table Collections: Examples You can perform DML operations on nested tables only
if they are defined as columns of a table. Therefore, when the query_table_expr_
clause of an INSERT, DELETE, or UPDATE statement is a table_collection_expression,
the collection expression must be a subquery that uses the TABLE collection expression
to select the nested table column of the table. The examples that follow are based on
the following scenario:

Suppose the database contains a table hr_info with columns department_id,
location_id, and manager_id, and a column of nested table type people which has
last_name, department_id, and salary columns for all the employees of each
respective manager:
CREATE TYPE people_typ AS OBJECT (
last_name
VARCHAR2(25),
department_id NUMBER(4),
salary
NUMBER(8,2));
/
CREATE TYPE people_tab_typ AS TABLE OF people_typ;
/
CREATE TABLE hr_info (
department_id
NUMBER(4),
location_id
NUMBER(4),
manager_id
NUMBER(6),
people
people_tab_typ)
NESTED TABLE people STORE AS people_stor_tab;
INSERT INTO hr_info VALUES (280, 1800, 999, people_tab_typ());

The following example inserts into the people nested table column of the hr_info
table for department 280:
INSERT INTO TABLE(SELECT h.people FROM hr_info h
WHERE h.department_id = 280)
VALUES ('Smith', 280, 1750);

The next example updates the department 280 people nested table:
UPDATE TABLE(SELECT h.people FROM hr_info h
WHERE h.department_id = 280) p
SET p.salary = p.salary + 100;

The next example deletes from the department 280 people nested table:
DELETE TABLE(SELECT h.people FROM hr_info h
WHERE h.department_id = 280) p
WHERE p.salary > 1700;

Collection Unnesting: Examples To select data from a nested table column, use the

TABLE collection expression to treat the nested table as columns of a table. This process
is called collection unnesting.
You could get all the rows from hr_info, which was created in the preceding example,
and all the rows from the people nested table column of hr_info using the following
statement:
SELECT t1.department_id, t2.* FROM hr_info t1, TABLE(t1.people) t2
WHERE t2.department_id = t1.department_id;

Now suppose that people is not a nested table column of hr_info, but is instead a
separate table with columns last_name, department_id, address, hiredate, and

19-54 Oracle Database SQL Language Reference

SELECT

salary. You can extract the same rows as in the preceding example with this
statement:
SELECT t1.department_id, t2.*
FROM hr_info t1, TABLE(CAST(MULTISET(
SELECT t3.last_name, t3.department_id, t3.salary
FROM people t3
WHERE t3.department_id = t1.department_id)
AS people_tab_typ)) t2;

Finally, suppose that people is neither a nested table column of table hr_info nor a
table itself. Instead, you have created a function people_func that extracts from
various sources the name, department, and salary of all employees. You can get the
same information as in the preceding examples with the following query:
SELECT t1.department_id, t2.* FROM hr_info t1, TABLE(CAST
(people_func( ... ) AS people_tab_typ)) t2;

See Also: Oracle Database Object-Relational Developer's Guide for more
examples of collection unnesting.

The following statement returns all
employees in hierarchical order. The root row is defined to be the employee whose job
is AD_VP. The child rows of a parent row are defined to be those who have the
employee number of the parent row as their manager number.

Using the LEVEL Pseudocolumn: Examples

SELECT LPAD(' ',2*(LEVEL-1)) || last_name org_chart,
employee_id, manager_id, job_id
FROM employees
START WITH job_id = 'AD_VP'
CONNECT BY PRIOR employee_id = manager_id;
ORG_CHART
EMPLOYEE_ID MANAGER_ID JOB_ID
------------------ ----------- ---------- ---------Kochhar
101
100 AD_VP
Greenberg
108
101 FI_MGR
Faviet
109
108 FI_ACCOUNT
Chen
110
108 FI_ACCOUNT
Sciarra
111
108 FI_ACCOUNT
Urman
112
108 FI_ACCOUNT
Popp
113
108 FI_ACCOUNT
Whalen
200
101 AD_ASST
Mavris
203
101 HR_REP
Baer
204
101 PR_REP
Higgins
205
101 AC_MGR
Gietz
206
205 AC_ACCOUNT
De Haan
102
100 AD_VP
Hunold
103
102 IT_PROG
Ernst
104
103 IT_PROG
Austin
105
103 IT_PROG
Pataballa
106
103 IT_PROG
Lorentz
107
103 IT_PROG

The following statement is similar to the previous one, except that it does not select
employees with the job FI_MGR.
SELECT LPAD(' ',2*(LEVEL-1)) || last_name org_chart,
employee_id, manager_id, job_id
FROM employees
WHERE job_id != 'FI_MGR'
START WITH job_id = 'AD_VP'
SQL Statements: SAVEPOINT to UPDATE

19-55

SELECT

CONNECT BY PRIOR employee_id = manager_id;
ORG_CHART
EMPLOYEE_ID MANAGER_ID JOB_ID
------------------ ----------- ---------- ---------Kochhar
101
100 AD_VP
Faviet
109
108 FI_ACCOUNT
Chen
110
108 FI_ACCOUNT
Sciarra
111
108 FI_ACCOUNT
Urman
112
108 FI_ACCOUNT
Popp
113
108 FI_ACCOUNT
Whalen
200
101 AD_ASST
Mavris
203
101 HR_REP
Baer
204
101 PR_REP
Higgins
205
101 AC_MGR
Gietz
206
205 AC_ACCOUNT
De Haan
102
100 AD_VP
Hunold
103
102 IT_PROG
Ernst
104
103 IT_PROG
Austin
105
103 IT_PROG
Pataballa
106
103 IT_PROG
Lorentz
107
103 IT_PROG

Oracle Database does not return the manager Greenberg, although it does return
employees who are managed by Greenberg.
The following statement is similar to the first one, except that it uses the LEVEL
pseudocolumn to select only the first two levels of the management hierarchy:
SELECT LPAD(' ',2*(LEVEL-1)) || last_name org_chart,
employee_id, manager_id, job_id
FROM employees
START WITH job_id = 'AD_PRES'
CONNECT BY PRIOR employee_id = manager_id AND LEVEL <= 2;
ORG_CHART
EMPLOYEE_ID MANAGER_ID JOB_ID
------------------ ----------- ---------- ---------King
100
AD_PRES
Kochhar
101
100 AD_VP
De Haan
102
100 AD_VP
Raphaely
114
100 PU_MAN
Weiss
120
100 ST_MAN
Fripp
121
100 ST_MAN
Kaufling
122
100 ST_MAN
Vollman
123
100 ST_MAN
Mourgos
124
100 ST_MAN
Russell
145
100 SA_MAN
Partners
146
100 SA_MAN
Errazuriz
147
100 SA_MAN
Cambrault
148
100 SA_MAN
Zlotkey
149
100 SA_MAN
Hartstein
201
100 MK_MAN

This example shows a query that joins the
departments table on the local database with the employees table on the remote
database:
Using Distributed Queries: Example

SELECT last_name, department_name
FROM employees@remote, departments
WHERE employees.department_id = departments.department_id;

19-56 Oracle Database SQL Language Reference

SELECT

Using Correlated Subqueries: Examples The following examples show the general
syntax of a correlated subquery:
SELECT select_list
FROM table1 t_alias1
WHERE expr operator
(SELECT column_list
FROM table2 t_alias2
WHERE t_alias1.column
operator t_alias2.column);
UPDATE table1 t_alias1
SET column =
(SELECT expr
FROM table2 t_alias2
WHERE t_alias1.column = t_alias2.column);
DELETE FROM table1 t_alias1
WHERE column operator
(SELECT expr
FROM table2 t_alias2
WHERE t_alias1.column = t_alias2.column);

The following statement returns data about employees whose salaries exceed their
department average. The following statement assigns an alias to employees, the table
containing the salary information, and then uses the alias in a correlated subquery:
SELECT department_id, last_name, salary
FROM employees x
WHERE salary > (SELECT AVG(salary)
FROM employees
WHERE x.department_id = department_id)
ORDER BY department_id;

For each row of the employees table, the parent query uses the correlated subquery to
compute the average salary for members of the same department. The correlated
subquery performs the following steps for each row of the employees table:
1.

The department_id of the row is determined.

2.

The department_id is then used to evaluate the parent query.

3.

If the salary in that row is greater than the average salary of the departments of
that row, then the row is returned.

The subquery is evaluated once for each row of the employees table.
Selecting from the DUAL Table: Example
current date:

The following statement returns the

SELECT SYSDATE FROM DUAL;

You could select SYSDATE from the employees table, but the database would return 14
rows of the same SYSDATE, one for every row of the employees table. Selecting from
DUAL is more convenient.
The following statement increments the
employees_seq sequence and returns the new value:
Selecting Sequence Values: Examples
SELECT employees_seq.nextval
FROM DUAL;

SQL Statements: SAVEPOINT to UPDATE

19-57

SELECT

The following statement selects the current value of employees_seq:
SELECT employees_seq.currval
FROM DUAL;

19-58 Oracle Database SQL Language Reference

SET CONSTRAINT[S]

SET CONSTRAINT[S]
Purpose
19

Use the SET CONSTRAINTS statement to specify, for a particular transaction, whether a
deferrable constraint is checked following each DML statement (IMMEDIATE) or when
the transaction is committed (DEFERRED). You can use this statement to set the mode for
a list of constraint names or for ALL constraints.
The SET CONSTRAINTS mode lasts for the duration of the transaction or until another
SET CONSTRAINTS statement resets the mode.
You can also use an ALTER SESSION statement with the SET
CONSTRAINTS clause to set all deferrable constraints. This is equivalent
to making issuing a SET CONSTRAINTS statement at the start of each
transaction in the current session.
Note:

You cannot specify this statement inside of a trigger definition.
SET CONSTRAINTS can be a distributed statement. Existing database links that have
transactions in process are notified when a SET CONSTRAINTS ALL statement is issued,
and new links are notified that it was issued as soon as they start a transaction.

Prerequisites
19

To specify when a deferrable constraint is checked, you must have SELECT privilege on
the table to which the constraint is applied unless the table is in your schema.

Syntax
19

set_constraints::=
,
CONSTRAINT

constraint

IMMEDIATE

SET

;
CONSTRAINTS

ALL

DEFERRED

Semantics
19

constraint
Specify the name of one or more integrity constraints.

ALL
Specify ALL to set all deferrable constraints for this transaction.

IMMEDIATE
Specify IMMEDIATE to cause the specified constraints to be checked immediately on
execution of each constrained DML statement. Oracle Database first checks any
constraints that were deferred earlier in the transaction and then continues
immediately checking constraints of any further statements in that transaction, as long
as all the checked constraints are consistent and no other SET CONSTRAINTS statement is

SQL Statements: SAVEPOINT to UPDATE

19-59

SET CONSTRAINT[S]

issued. If any constraint fails the check, then an error is signaled. At that point, a
COMMIT statement causes the whole transaction to undo.
Making constraints immediate at the end of a transaction is a way of checking whether
COMMIT can succeed. You can avoid unexpected rollbacks by setting constraints to
IMMEDIATE as the last statement in a transaction. If any constraint fails the check, you
can then correct the error before committing the transaction.

DEFERRED
Specify DEFERRED to indicate that the conditions specified by the deferrable constraint
are checked when the transaction is committed.
You can verify the success of deferrable constraints prior to
committing them by issuing a SET CONSTRAINTS ALL IMMEDIATE
statement.

Note:

Examples
19

Setting Constraints: Examples The following statement sets all deferrable
constraints in this transaction to be checked immediately following each DML
statement:
SET CONSTRAINTS ALL IMMEDIATE;

The following statement checks three deferred constraints when the transaction is
committed. This example fails if the constraints were specified to be NOT DEFERRABLE.
SET CONSTRAINTS emp_job_nn, emp_salary_min ,
hr.jhist_dept_fk@remote DEFERRED;

19-60 Oracle Database SQL Language Reference

SET ROLE

SET ROLE
Purpose
19

When a user logs on to Oracle Database, the database enables all privileges granted
explicitly to the user and all privileges in the user's default roles. During the session,
the user or an application can use the SET ROLE statement any number of times to
enable or disable the roles currently enabled for the session.
You cannot enable more than 148 user-defined roles at one time.

Notes:
■

■

■

For most roles, you cannot enable or disable a role unless it was
granted to you either directly or through other roles. However, a
secure application role can be granted and enabled by its
associated PL/SQL package. See the CREATE ROLE semantics for
USING package on page 15-60 and Oracle Database Security Guide
for information about secure application roles.
SET ROLE succeeds only if there are no definer's rights units on the
call stack. If at least one DR unit is on the call stack, then issuing
the SET ROLE command causes ORA-06565. See Oracle Database
PL/SQL Language Reference for more information about definer's
rights units.
To run the SET ROLE command from PL/SQL, you must use
dynamic SQL, preferably the EXECUTE IMMEDIATE statement. See
Oracle Database PL/SQL Language Reference for more information
about this statement.

You can see which roles are currently enabled by examining the SESSION_ROLES data
dictionary view.
See Also:
■
■

■

CREATE ROLE on page 15-59 for information on creating roles
ALTER USER on page 13-6 for information on changing a user's
default roles
Oracle Database Reference for information on the SESSION_ROLES
session parameter

Prerequisites
19

You must already have been granted the roles that you name in the SET ROLE
statement.

SQL Statements: SAVEPOINT to UPDATE

19-61

SET ROLE

Syntax
19

set_role::=
,
IDENTIFIED

BY

password

role
,
SET

ROLE

EXCEPT

role

;

ALL
NONE

Semantics
19

role
Specify one or more roles to be enabled for the current session. All roles not specified
are disabled for the current session or until another SET ROLE statement is issued in the
current session.
In the IDENTIFIED BY password clause, specify the password for a role. If the role has a
password, then you must specify the password to enable the role.
You cannot specify a role identified globally. Global
roles are enabled by default at login, and cannot be reenabled later.

Restriction on Setting Roles

ALL Clause
Specify ALL to enable all roles granted to you for the current session except those
optionally listed in the EXCEPT clause.
Roles listed in the EXCEPT clause must be roles granted directly to you. They cannot be
roles granted to you through other roles.
If you list a role in the EXCEPT clause that has been granted to you both directly and
through another role, then the role remains enabled by virtue of the role to which it
has been granted.
Restrictions on the ALL Clause The following restrictions apply to the ALL clause:
■

■

You cannot specify this clause if you have been directly granted any roles with
passwords. Doing so will result in an ORA-01979 error.
You cannot use this clause to enable a secure application role, which is a role that
can be enabled only by applications using an authorized package. Refer to Oracle
Database Security Guide for information on creating a secure application role and
Oracle Database 2 Day + Security Guide for a tutorial.

NONE
Specify NONE to disable all roles for the current session, including the DEFAULT role.

Examples
19

To enable the role dw_manager identified by the password
warehouse for your current session, issue the following statement:
Setting Roles: Examples

19-62 Oracle Database SQL Language Reference

SET ROLE

SET ROLE dw_manager IDENTIFIED BY warehouse;

To enable all roles granted to you for the current session, issue the following statement:
SET ROLE ALL;

To enable all roles granted to you except dw_manager, issue the following statement:
SET ROLE ALL EXCEPT dw_manager;

To disable all roles granted to you for the current session, issue the following
statement:
SET ROLE NONE;

SQL Statements: SAVEPOINT to UPDATE

19-63

SET TRANSACTION

SET TRANSACTION
Purpose
19

Use the SET TRANSACTION statement to establish the current transaction as read-only or
read/write, establish its isolation level, assign it to a specified rollback segment, or
assign a name to the transaction.
A transaction implicitly begins with any operation that obtains a TX lock:
■

When a statement that modifies data is issued

■

When a SELECT ... FOR UPDATE statement is issued

■

When a transaction is explicitly started with a SET TRANSACTION statement or the
DBMS_TRANSACTION package

Issuing either a COMMIT or ROLLBACK statement explicitly ends the current transaction.
The operations performed by a SET TRANSACTION statement affect only your current
transaction, not other users or other transactions. Your transaction ends whenever you
issue a COMMIT or ROLLBACK statement. Oracle Database implicitly commits the current
transaction before and after executing a data definition language (DDL) statement.
See Also:

COMMIT on page 13-49 and ROLLBACK on page 18-96

Prerequisites
19

If you use a SET TRANSACTION statement, then it must be the first statement in your
transaction. However, a transaction need not have a SET TRANSACTION statement.

Syntax
19

set_transaction::=
ONLY
READ
WRITE
SERIALIZABLE
ISOLATION
TRANSACTION

USE
NAME

string

LEVEL
READ

SET

NAME

ROLLBACK

SEGMENT

COMMITTED
rollback_segment

;

string

Semantics
19

READ ONLY
The READ ONLY clause establishes the current transaction as a read-only transaction.
This clause established transaction-level read consistency.
All subsequent queries in that transaction see only changes that were committed
before the transaction began. Read-only transactions are useful for reports that run
multiple queries against one or more tables while other users update these same
tables.

19-64 Oracle Database SQL Language Reference

SET TRANSACTION

This clause is not supported for the user SYS. Queries by SYS will return changes made
during the transaction even if SYS has set the transaction to be READ ONLY.
Restriction on Read-only Transactions Only the following statements are permitted

in a read-only transaction:
■

Subqueries—SELECT statements without the for_update_clause

■

LOCK TABLE

■

SET ROLE

■

ALTER SESSION

■

ALTER SYSTEM

READ WRITE
Specify READ WRITE to establish the current transaction as a read/write transaction.
This clause establishes statement-level read consistency, which is the default.
Restriction on Read/Write Transactions You cannot toggle between transaction-level
and statement-level read consistency in the same transaction.

ISOLATION LEVEL Clause
Use the ISOLATION LEVEL clause to specify how transactions containing database
modifications are handled.
■

■

The SERIALIZABLE setting specifies serializable transaction isolation mode as
defined in the SQL standard. If a serializable transaction contains data
manipulation language (DML) that attempts to update any resource that may have
been updated in a transaction uncommitted at the start of the serializable
transaction, then the DML statement fails.
The READ COMMITTED setting is the default Oracle Database transaction behavior. If
the transaction contains DML that requires row locks held by another transaction,
then the DML statement waits until the row locks are released.

USE ROLLBACK SEGMENT Clause
This clause is relevant and valid only if you are using rollback
segments for undo. Oracle strongly recommends that you use
automatic undo management to handle undo space. If you follow this
recommendation and run your database in automatic undo mode,
then Oracle Database ignores this clause.

Note:

Specify USE ROLLBACK SEGMENT to assign the current transaction to the specified
rollback segment. This clause also implicitly establishes the transaction as a read/write
transaction.
Parallel DML requires more than one rollback segment. Therefore, if your transaction
contains parallel DML operations, then the database ignores this clause.

NAME Clause
Use the NAME clause to assign a name to the current transaction. This clause is
especially useful in distributed database environments when you must identify and
resolve in-doubt transactions. The string value is limited to 255 bytes.

SQL Statements: SAVEPOINT to UPDATE

19-65

SET TRANSACTION

If you specify a name for a distributed transaction, then when the transaction commits,
the name becomes the commit comment, overriding any comment specified explicitly
in the COMMIT statement.
Oracle Database Concepts for more information about
transaction naming

See Also:

Examples
19

Setting Transactions: Examples The following statements could be run at midnight

of the last day of every month to count the products and quantities on hand in the
Toronto warehouse in the sample Order Entry (oe) schema. This report would not be
affected by any other user who might be adding or removing inventory to a different
warehouse.
COMMIT;
SET TRANSACTION READ ONLY NAME 'Toronto';
SELECT product_id, quantity_on_hand FROM inventories
WHERE warehouse_id = 5
ORDER BY product_id;
COMMIT;

The first COMMIT statement ensures that SET TRANSACTION is the first statement in the
transaction. The last COMMIT statement does not actually make permanent any changes
to the database. It simply ends the read-only transaction.

19-66 Oracle Database SQL Language Reference

TRUNCATE CLUSTER

TRUNCATE CLUSTER
Purpose
19

Caution:

You cannot roll back a TRUNCATE CLUSTER statement.

Use the TRUNCATE CLUSTER statement to remove all rows from a cluster. By default,
Oracle Database also performs the following tasks:
■

■

Deallocates all space used by the removed rows except that specified by the
MINEXTENTS storage parameter
Sets the NEXT storage parameter to the size of the last extent removed from the
segment by the truncation process

Removing rows with the TRUNCATE statement can be more efficient than dropping and
re-creating a cluster. Dropping and re-creating a cluster invalidates dependent objects
of the cluster, requires you to regrant object privileges on the cluster, and requires you
to re-create the indexes and cluster on the table and respecify its storage parameters.
Truncating has none of these effects.
Removing rows with the TRUNCATE CLUSTER statement can be faster than removing all
rows with the DELETE statement, especially if the cluster has numerous indexes and
other dependencies.
See Also:
■

■

DELETE on page 17-26 and DROP CLUSTER on page 17-36 for
information on other ways of dropping data from a cluster
TRUNCATE TABLE on page 19-69 for information on truncating a
table

Prerequisites
19

To truncate a cluster, the cluster must be in your schema or you must have DROP ANY
TABLE system privilege.
"Restrictions on Truncating Tables" on page 19-71

See Also:

Syntax
19

truncate_cluster::=
DROP
STORAGE
schema
TRUNCATE

.

CLUSTER

REUSE
cluster

;

Semantics
19

CLUSTER Clause
Specify the schema and name of the cluster to be truncated. You can truncate only an
indexed cluster, not a hash cluster. If you omit schema, then the database assumes the
cluster is in your own schema.
SQL Statements: SAVEPOINT to UPDATE

19-67

TRUNCATE CLUSTER

When you truncate a cluster, the database also automatically deletes all data in the
indexes of the cluster tables.

STORAGE Clauses
The STORAGE clauses let you determine what happens to the space freed by the
truncated rows. The DROP STORAGE clause and REUSE STORAGE clause also apply to the
space freed by the data deleted from associated indexes.
Specify DROP STORAGE to deallocate all space from the deleted rows
from the cluster except the space allocated by the MINEXTENTS parameter of the cluster.
This space can subsequently be used by other objects in the tablespace. Oracle
Database also sets the NEXT storage parameter to the size of the last extent removed
from the segment in the truncation process. This is the default.
DROP STORAGE

Specify REUSE STORAGE to retain the space from the deleted rows
allocated to the cluster. Storage values are not reset to the values when the table or
cluster was created. This space can subsequently be used only by new data in the
cluster resulting from insert or update operations. This clause leaves storage
parameters at their current settings.

REUSE STORAGE

If you have specified more than one free list for the object you are truncating, then the
REUSE STORAGE clause also removes any mapping of free lists to instances and resets the
high-water mark to the beginning of the first extent.

Examples
19

Truncating a Cluster: Example The following statement removes all rows from all
tables in the personnel cluster, but leaves the freed space allocated to the tables:
TRUNCATE CLUSTER personnel REUSE STORAGE;

The preceding statement also removes all data from all indexes on the tables in the
personnel cluster.

19-68 Oracle Database SQL Language Reference

TRUNCATE TABLE

TRUNCATE TABLE
Purpose
19

Caution: You cannot roll back a TRUNCATE TABLE statement, nor can
you use a FLASHBACK TABLE statement to retrieve the contents of a table
that has been truncated.

Use the TRUNCATE TABLE statement to remove all rows from a table. By default, Oracle
Database also performs the following tasks:
■

■

Deallocates all space used by the removed rows except that specified by the
MINEXTENTS storage parameter
Sets the NEXT storage parameter to the size of the last extent removed from the
segment by the truncation process

Removing rows with the TRUNCATE TABLE statement can be more efficient than
dropping and re-creating a table. Dropping and re-creating a table invalidates
dependent objects of the table, requires you to regrant object privileges on the table,
and requires you to re-create the indexes, integrity constraints, and triggers on the
table and respecify its storage parameters. Truncating has none of these effects.
Removing rows with the TRUNCATE TABLE statement can be faster than removing all
rows with the DELETE statement, especially if the table has numerous triggers, indexes,
and other dependencies.
See Also:
■

■

DELETE on page 17-26 and DROP TABLE on page 18-5 for
information on other ways of removing data from a table
TRUNCATE CLUSTER on page 19-67 for information on
truncating a cluster

Prerequisites
19

To truncate a table, the table must be in your schema or you must have the DROP ANY
TABLE system privilege.
See Also:

"Restrictions on Truncating Tables" on page 19-71

SQL Statements: SAVEPOINT to UPDATE

19-69

TRUNCATE TABLE

Syntax
19

truncate_table::=
PRESERVE
MATERIALIZED
schema
TRUNCATE

.

TABLE

VIEW

LOG

PURGE
table

ALL
DROP
STORAGE
REUSE
;

Semantics
19

TABLE Clause
Specify the schema and name of the table to be truncated. This table cannot be part of a
cluster. If you omit schema, then Oracle Database assumes the table is in your own
schema.
■

■

■

■

■

You can truncate index-organized tables and temporary tables. When you truncate
a temporary table, only the rows created during the current session are removed.
Oracle Database changes the NEXT storage parameter of table to be the size of the
last extent deleted from the segment in the process of truncation.
Oracle Database also automatically truncates and resets any existing UNUSABLE
indicators for the following indexes on table: range and hash partitions of local
indexes and subpartitions of local indexes.
If table is not empty, then the database marks UNUSABLE all nonpartitioned
indexes and all partitions of global partitioned indexes on the table. However,
when the table is truncated, the index is also truncated, and a new high water
mark is calculated for the index segment. This operation is equivalent to creating a
new segment for the index. Therefore, at the end of the truncate operation, the
indexes are once again USABLE.
For a domain index, this statement invokes the appropriate truncate routine to
truncate the domain index data.
See Also: Oracle Database Data Cartridge Developer's Guide for more
information on domain indexes

■

■

If a regular or index-organized table contains LOB columns, then all LOB data and
LOB index segments are truncated.
If table is partitioned, then all partitions or subpartitions, as well as the LOB data
and LOB index segments for each partition or subpartition, are truncated.

19-70 Oracle Database SQL Language Reference

TRUNCATE TABLE

When you truncate a table, Oracle Database automatically
removes all data in the table's indexes and any materialized view
direct-path INSERT information held in association with the table. This
information is independent of any materialized view log. If this
direct-path INSERT information is removed, then an incremental
refresh of the materialized view may lose data.
Note:

■

All cursors are invalidated.

Restrictions on Truncating Tables This statement is subject to the following
restrictions:
■

You cannot roll back a TRUNCATE TABLE statement.

■

You cannot flash back to the state of the table before the truncate operation.

■

■

■

■

You cannot individually truncate a table that is part of a cluster. You must either
truncate the cluster, delete all rows from the table, or drop and re-create the table.
You cannot truncate the parent table of an enabled foreign key constraint. You
must disable the constraint before truncating the table. An exception is that you
can truncate the table if the integrity constraint is self-referential.
If a domain index is defined on table, then neither the index nor any index
partitions can be marked IN_PROGRESS.
You cannot truncate the parent table of a reference-partitioned table. You must first
drop the reference-partitioned child table.

MATERIALIZED VIEW LOG Clause
The MATERIALIZED VIEW LOG clause lets you specify whether a materialized view log
defined on the table is to be preserved or purged when the table is truncated. This
clause permits materialized view master tables to be reorganized through export or
import without affecting the ability of primary key materialized views defined on the
master to be fast refreshed. To support continued fast refresh of primary key
materialized views, the materialized view log must record primary key information.
The keyword SNAPSHOT is supported in place of MATERIALIZED
VIEW for backward compatibility.
Note:

PRESERVE Specify PRESERVE if any materialized view log should be preserved when
the master table is truncated. This is the default.
PURGE Specify PURGE if any materialized view log should be purged when the
master table is truncated.

Oracle Database Advanced Replication for more information
about materialized view logs and the TRUNCATE statement

See Also:

STORAGE Clauses
The STORAGE clauses let you determine what happens to the space freed by the
truncated rows. The DROP STORAGE clause, DROP ALL STORAGE clause, and REUSE STORAGE
clause also apply to the space freed by the data deleted from associated indexes.

SQL Statements: SAVEPOINT to UPDATE

19-71

TRUNCATE TABLE

Specify DROP STORAGE to deallocate all space from the deleted rows
from the table except the space allocated by the MINEXTENTS parameter of the table.
This space can subsequently be used by other objects in the tablespace. Oracle
Database also sets the NEXT storage parameter to the size of the last extent removed
from the segment in the truncation process. This setting, which is the default, is useful
for small and medium-sized objects. The extent management in locally managed
tablespace is very fast in these cases, so there is no need to reserve space.
DROP STORAGE

DROP ALL STORAGE Specify DROP ALL STORAGE to deallocate all space from the

deleted rows from the table, including the space allocated by the MINEXTENTS
parameter. All segments for the table, as well as all segments for its dependent objects,
will be deallocated.
Restrictions on DROP ALL STORAGE This clause is subject to the same restrictions
as described in "Restrictions on Deferred Segment Creation" on page 16-33.
Note: The DROP ALL STORAGE clause is available starting with Oracle
Database 11g Release 2 (11.2.0.2).

Specify REUSE STORAGE to retain the space from the deleted rows
allocated to the table. Storage values are not reset to the values when the table was
created. This space can subsequently be used only by new data in the table resulting
from insert or update operations. This clause leaves storage parameters at their current
settings.

REUSE STORAGE

This setting is useful as an alternative to deleting all rows of a very large table—when
the number of rows is very large, the table entails many thousands of extents, and
when data is to be reinserted in the future.
This clause is not valid for temporary tables. A session becomes unbound from the
temporary table when the table is truncated, so the storage is automatically dropped.
If you have specified more than one free list for the object you are truncating, then the
REUSE STORAGE clause also removes any mapping of free lists to instances and resets the
high-water mark to the beginning of the first extent.

Examples
19

Truncating a Table: Example The following statement removes all rows from a
hypothetical copy of the sample table hr.employees and returns the freed space to the
tablespace containing employees:
TRUNCATE TABLE employees_demo;

The preceding statement also removes all data from all indexes on employees and
returns the freed space to the tablespaces containing them.
The following
statements are examples of TRUNCATE statements that preserve materialized view logs:

Preserving Materialized View Logs After Truncate: Example
TRUNCATE TABLE sales_demo PRESERVE MATERIALIZED VIEW LOG;
TRUNCATE TABLE orders_demo;

19-72 Oracle Database SQL Language Reference

UPDATE

UPDATE
Purpose
19

Use the UPDATE statement to change existing values in a table or in the base table of a
view or the master table of a materialized view.

Prerequisites
19

For you to update values in a table, the table must be in your own schema or you must
have the UPDATE object privilege on the table.
For you to update values in the base table of a view:
■
■

You must have the UPDATE object privilege on the view, and
Whoever owns the schema containing the view must have the UPDATE object
privilege on the base table.

The UPDATE ANY TABLE system privilege also allows you to update values in any table
or in the base table of any view.
You must also have the SELECT object privilege on the object you want to update if:
■
■

The object is on a remote database or
The SQL92_SECURITY initialization parameter is set to TRUE and the UPDATE
operation references table columns, such as the columns in a where_clause.

Syntax
19

update::=
hint

t_alias

dml_table_expression_clause

UPDATE
ONLY

(

dml_table_expression_clause

where_clause

returning_clause

)

error_logging_clause

update_set_clause

;

(DML_table_expression_clause::= on page 19-73, update_set_clause::= on page 19-74,
where_clause::= on page 19-74, returning_clause::= on page 19-74, error_logging_clause::=
on page 19-75)
DML_table_expression_clause::=
partition_extension_clause
@
schema

dblink

table

.

@

view

dblink

materialized view
subquery_restriction_clause
(

subquery

)

table_collection_expression

SQL Statements: SAVEPOINT to UPDATE

19-73

UPDATE

(partition_extension_clause::= on page 19-74, subquery::= on page 19-5--part of SELECT,
subquery_restriction_clause::= on page 19-74, table_collection_expression::= on page 19-74)
partition_extension_clause::=
(

partition

)

PARTITION

,
FOR

(

(

partition_key_value

subpartition

)

)

SUBPARTITION

,
FOR

(

subpartition_key_value

)

subquery_restriction_clause::=

READ

CONSTRAINT

ONLY

constraint

WITH
CHECK

OPTION

table_collection_expression::=

TABLE

(

collection_expression

(

+

)

(

subquery

)

update_set_clause::=
,
,
(

column

)

=

)

expr
column

=

SET

(

subquery

)

DEFAULT
expr
VALUE

(

t_alias

)

=
(

subquery

where_clause::=
WHERE

condition

returning_clause::=
,

,

RETURN
expr

INTO

data_item

RETURNING

19-74 Oracle Database SQL Language Reference

)

UPDATE

error_logging_clause::=
schema
INTO
LOG

.
table

(

simple_expression

)

ERRORS

integer
REJECT

LIMIT
UNLIMITED

Semantics
19

hint
Specify a comment that passes instructions to the optimizer on choosing an execution
plan for the statement.
You can place a parallel hint immediately after the UPDATE keyword to parallelize both
the underlying scan and UPDATE operations.
See Also:
■
■

"Hints" on page 3-74 for the syntax and description of hints
Oracle Database Concepts for detailed information about parallel
execution

DML_table_expression_clause
The ONLY clause applies only to views. Specify ONLY syntax if the view in the UPDATE
clause is a view that belongs to a hierarchy and you do not want to update rows from
any of its subviews.
See Also: "Restrictions on the DML_table_expression_clause" on
page 19-77 and "Updating a Table: Examples" on page 19-80

schema
Specify the schema containing the object to be updated. If you omit schema, then the
database assumes the object is in your own schema.
table | view | materialized_view |subquery
Specify the name of the table, view, materialized view, or the columns returned by a
subquery to be updated. Issuing an UPDATE statement against a table fires any UPDATE
triggers associated with the table.
■

If you specify view, then the database updates the base table of the view. You
cannot update a view except with INSTEAD OF triggers if the defining query of the
view contains one of the following constructs:
A set operator
A DISTINCT operator
An aggregate or analytic function
A GROUP BY, ORDER BY, MODEL, CONNECT BY, or START WITH clause
A collection expression in a SELECT list
A subquery in a SELECT list
A subquery designated WITH READ ONLY
A recursive WITH clause

SQL Statements: SAVEPOINT to UPDATE

19-75

UPDATE

Joins, with some exceptions, as documented in Oracle Database Administrator's
Guide
■
■

■

■

You cannot update more than one base table through a view.
In addition, if the view was created with the WITH CHECK OPTION, then you can
update the view only if the resulting data satisfies the view's defining query.
If table or the base table of view contains one or more domain index columns,
then this statement executes the appropriate indextype update routine.
You cannot update rows in a read-only materialized view. If you update rows in a
writable materialized view, then the database updates the rows from the
underlying container table. However, the updates are overwritten at the next
refresh operation. If you update rows in an updatable materialized view that is
part of a materialized view group, then the database also updates the
corresponding rows in the master table.
See Also:
■

■

Oracle Database Data Cartridge Developer's Guide for more
information on the indextype update routines
CREATE MATERIALIZED VIEW on page 15-4 for information on
creating updatable materialized views

partition_extension_clause
Specify the name or partition key value of the partition or subpartition within table
targeted for updates. You need not specify the partition name when updating values in
a partitioned table. However in some cases specifying the partition name can be more
efficient than a complicated where_clause.
See Also: "References to Partitioned Tables and Indexes" on
page 3-119 and "Updating a Partition: Example" on page 19-81

dblink
Specify a complete or partial name of a database link to a remote database where the
object is located. You can use a database link to update a remote object only if you are
using Oracle Database distributed functionality.
If you omit dblink, then the database assumes the object is on the local database.
See Also: "References to Objects in Remote Databases" on page 3-117
for information on referring to database links

subquery_restriction_clause
Use the subquery_restriction_clause to restrict the subquery in one of the following
ways:
WITH READ ONLY

Specify WITH READ ONLY to indicate that the table or view cannot

be updated.
WITH CHECK OPTION Specify WITH CHECK OPTION to indicate that Oracle Database
prohibits any changes to the table or view that would produce rows that are not
included in the subquery. When used in the subquery of a DML statement, you can
specify this clause in a subquery in the FROM clause but not in subquery in the WHERE
clause.

19-76 Oracle Database SQL Language Reference

UPDATE

CONSTRAINT constraint Specify the name of the CHECK OPTION constraint. If you
omit this identifier, then Oracle automatically assigns the constraint a name of the
form SYS_Cn, where n is an integer that makes the constraint name unique within the
database.
See Also: "Using the WITH CHECK OPTION Clause: Example" on
page 19-48

table_collection_expression
The table_collection_expression lets you inform Oracle that the value of
collection_expression should be treated as a table for purposes of query and DML
operations. The collection_expression can be a subquery, a column, a function, or a
collection constructor. Regardless of its form, it must return a collection value—that is,
a value whose type is nested table or varray. This process of extracting the elements of
a collection is called collection unnesting.
The optional plus (+) is relevant if you are joining the TABLE collection expression with
the parent table. The + creates an outer join of the two, so that the query returns rows
from the outer table even if the collection expression is null.
In earlier releases of Oracle, when collection_expression
was a subquery, table_collection_expression was expressed as
THE subquery. That usage is now deprecated.
Note:

You can use a table_collection_expression to update rows in one table based on
rows from another table. For example, you could roll up four quarterly sales tables
into a yearly sales table.
t_alias
Specify a correlation name (alias) for the table, view, or subquery to be referenced
elsewhere in the statement. This alias is required if the DML_table_expression_clause
references any object type attributes or object type methods.
See Also:

"Correlated Update: Example" on page 19-81

Restrictions on the DML_table_expression_clause

This clause is subject to the

following restrictions:
■

■

■

■

You cannot execute this statement if table or the base table of view contains any
domain indexes marked IN_PROGRESS or FAILED.
You cannot insert into a partition if any affected index partitions are marked
UNUSABLE.
You cannot specify the order_by_clause in the subquery of the DML_table_
expression_clause.
If you specify an index, index partition, or index subpartition that has been
marked UNUSABLE, then the UPDATE statement will fail unless the SKIP_UNUSABLE_
INDEXES session parameter has been set to TRUE.
ALTER SESSION on page 11-45 for information on the
SKIP_UNUSABLE_INDEXES session parameter
See Also:

update_set_clause
The update_set_clause lets you set column values.
SQL Statements: SAVEPOINT to UPDATE

19-77

UPDATE

column
Specify the name of a column of the object that is to be updated. If you omit a column
of the table from the update_set_clause, then the value of that column remains
unchanged.
If column refers to a LOB object attribute, then you must first initialize it with a value
of empty or null. You cannot update it with a literal. Also, if you are updating a LOB
value using some method other than a direct UPDATE SQL statement, then you must
first lock the row containing the LOB. See for_update_clause on page 19-34 for more
information.
If column is a virtual column, you cannot specify it here. Rather, you must update the
values from which the virtual column is derived.
If column is part of the partitioning key of a partitioned table, then UPDATE will fail if
you change a value in the column that would move the row to a different partition or
subpartition, unless you enable row movement. Refer to the row_movement_clause of
CREATE TABLE on page 16-6 or ALTER TABLE on page 12-2.
In addition, if column is part of the partitioning key of a list-partitioned table, then
UPDATE will fail if you specify a value for the column that does not already exist in the
partition_key_value list of one of the partitions.
subquery
Specify a subquery that returns exactly one row for each row updated.
■

■

■
■

If you specify only one column in the update_set_clause, then the subquery can
return only one value.
If you specify multiple columns in the update_set_clause, then the subquery
must return as many values as you have specified columns.
If the subquery returns no rows, then the column is assigned a null.
If this subquery refers to remote objects, then the UPDATE operation can run in
parallel as long as the reference does not loop back to an object on the local
database. However, if the subquery in the DML_table_expression_clause refers to
any remote objects, then the UPDATE operation will run serially without
notification.

You can use the flashback_query_clause within the subquery to update table with
past data. Refer to the flashback_query_clause of SELECT on page 19-17 for more
information on this clause.
See Also:
■
■

SELECT on page 19-4 and "Using Subqueries" on page 9-14
parallel_clause on page 16-68 in the CREATE TABLE
documentation

expr
Specify an expression that resolves to the new value assigned to the corresponding
column.
Chapter 6, "Expressions" for the syntax of expr and
"Updating an Object Table: Example" on page 19-81

See Also:

19-78 Oracle Database SQL Language Reference

UPDATE

DEFAULT Specify DEFAULT to set the column to the value previously specified as the

default value for the column. If no default value for the corresponding column has
been specified, then the database sets the column to null.
Restriction on Updating to Default Values You cannot specify DEFAULT if you are

updating a view.
VALUE Clause
The VALUE clause lets you specify the entire row of an object table.
Restriction on the VALUE clause

You can specify this clause only for an object table.

If you insert string literals into a RAW column, then during
subsequent queries, Oracle Database will perform a full table scan
rather than using any index that might exist on the RAW column.

Note:

See Also:

"Updating an Object Table: Example" on page 19-81

where_clause
The where_clause lets you restrict the rows updated to those for which the specified
condition is true. If you omit this clause, then the database updates all rows in the
table or view. Refer to Chapter 7, "Conditions" for the syntax of condition.
The where_clause determines the rows in which values are updated. If you do not
specify the where_clause, then all rows are updated. For each row that satisfies the
where_clause, the columns to the left of the equality operator (=) in the update_set_
clause are set to the values of the corresponding expressions to the right of the
operator. The expressions are evaluated as the row is updated.

returning_clause
The returning clause retrieves the rows affected by a DML statement. You can specify
this clause for tables and materialized views and for views with a single base table.
When operating on a single row, a DML statement with a returning_clause can
retrieve column expressions using the affected row, rowid, and REFs to the affected
row and store them in host variables or PL/SQL variables.
When operating on multiple rows, a DML statement with the returning_clause stores
values from expressions, rowids, and REFs involving the affected rows in bind arrays.
expr

Each item in the expr list must be a valid expression syntax.

The INTO clause indicates that the values of the changed rows are to be stored
in the variable(s) specified in data_item list.

INTO

Each data_item is a host variable or PL/SQL variable that stores the
retrieved expr value.

data_item

For each expression in the RETURNING list, you must specify a corresponding
type-compatible PL/SQL variable or host variable in the INTO list.
Restrictions The following restrictions apply to the RETURNING clause:
■

The expr is restricted as follows:

SQL Statements: SAVEPOINT to UPDATE

19-79

UPDATE

■

–

For UPDATE and DELETE statements each expr must be a simple expression or a
single-set aggregate function expression. You cannot combine simple
expressions and single-set aggregate function expressions in the same
returning_clause. For INSERT statements, each expr must be a simple
expression. Aggregate functions are not supported in an INSERT statement
RETURNING clause.

–

Single-set aggregate function expressions cannot include the DISTINCT
keyword.

If the expr list contains a primary key column or other NOT NULL column, then the
update statement fails if the table has a BEFORE UPDATE trigger defined on it.

■

You cannot specify the returning_clause for a multitable insert.

■

You cannot use this clause with parallel DML or with remote objects.

■

You cannot retrieve LONG types with this clause.

■

You cannot specify this clause for a view on which an INSTEAD OF trigger has been
defined.
Oracle Database PL/SQL Language Reference for information
on using the BULK COLLECT clause to return multiple values to
collection variables

See Also:

error_logging_clause
The error_logging_clause has the same behavior in an UPDATE statement as it does in an
INSERT statement. Refer to the INSERT statement error_logging_clause on page 18-65 for
more information.
See Also: "Inserting Into a Table with Error Logging: Example" on
page 18-66

Examples
19

The following statement gives null commissions to all
employees with the job SH_CLERK:

Updating a Table: Examples

UPDATE employees
SET commission_pct = NULL
WHERE job_id = 'SH_CLERK';

The following statement promotes Douglas Grant to manager of Department 20 with a
$1,000 raise:
UPDATE employees SET
job_id = 'SA_MAN', salary = salary + 1000, department_id = 120
WHERE first_name||' '||last_name = 'Douglas Grant';

The following statement increases the salary of an employee in the employees table on
the remote database:
UPDATE employees@remote
SET salary = salary*1.1
WHERE last_name = 'Baer';

The next example shows the following syntactic constructs of the UPDATE statement:
■

Both forms of the update_set_clause together in a single statement

19-80 Oracle Database SQL Language Reference

UPDATE

■

A correlated subquery

■

A where_clause to limit the updated rows

UPDATE employees a
SET department_id =
(SELECT department_id
FROM departments
WHERE location_id = '2100'),
(salary, commission_pct) =
(SELECT 1.1*AVG(salary), 1.5*AVG(commission_pct)
FROM employees b
WHERE a.department_id = b.department_id)
WHERE department_id IN
(SELECT department_id
FROM departments
WHERE location_id = 2900
OR location_id = 2700);

The preceding UPDATE statement performs the following operations:
■

■

■
■

Updates only those employees who work in Geneva or Munich (locations 2900
and 2700)
Sets department_id for these employees to the department_id corresponding to
Bombay (location_id 2100)
Sets each employee's salary to 1.1 times the average salary of their department
Sets each employee's commission to 1.5 times the average commission of their
department

Updating a Partition: Example The following example updates values in a single
partition of the sales table:
UPDATE sales PARTITION (sales_q1_1999) s
SET s.promo_id = 494
WHERE amount_sold > 1000;

The following statement creates two object
tables, people_demo1 and people_demo2, of the people_typ object created in Table
Collections: Examples on page 19-54. The example shows how to update a row of
people_demo1 by selecting a row from people_demo2:
Updating an Object Table: Example

CREATE TABLE people_demo1 OF people_typ;
CREATE TABLE people_demo2 OF people_typ;
UPDATE people_demo1 p SET VALUE(p) =
(SELECT VALUE(q) FROM people_demo2 q
WHERE p.department_id = q.department_id)
WHERE p.department_id = 10;

The example uses the VALUE object reference function in both the SET clause and the
subquery.
Correlated Update: Example For an example that uses a correlated subquery to
update nested table rows, refer to "Table Collections: Examples" on page 19-54.

The following example
returns values from the updated row and stores the result in PL/SQL variables bnd1,
bnd2, bnd3:
Using the RETURNING Clause During UPDATE: Example

SQL Statements: SAVEPOINT to UPDATE

19-81

UPDATE

UPDATE employees
SET job_id ='SA_MAN', salary = salary + 1000, department_id = 140
WHERE last_name = 'Jones'
RETURNING salary*0.25, last_name, department_id
INTO :bnd1, :bnd2, :bnd3;

The following example shows that you can specify a single-set aggregate function in
the expression of the returning clause:
UPDATE employees
SET salary = salary * 1.1
WHERE department_id = 100
RETURNING SUM(salary) INTO :bnd1;

19-82 Oracle Database SQL Language Reference

A
A

How to Read Syntax Diagrams

This appendix describes how to read syntax diagrams.

Graphic Syntax Diagrams
Syntax diagrams are drawings that illustrate valid SQL syntax. To read a diagram,
trace it from left to right, in the direction shown by the arrows.
Commands and other keywords appear in UPPERCASE inside rectangles. Type them
exactly as shown in the rectangles. Parameters appear in lowercase inside ovals.
Variables are used for the parameters. Punctuation, operators, delimiters, and
terminators appear inside circles.
If the syntax diagram has more than one path, then you can choose any path. For
example, in the following syntax you can specify either NOPARALLEL or PARALLEL:
parallel_clause::=
NOPARALLEL
integer
PARALLEL

If you have the choice of more than one keyword, operator, or parameter, then your
options appear in a vertical list. For example, in the following syntax diagram, you can
specify one or more of the four parameters in the stack:
physical_attributes_clause::=
PCTFREE

integer

PCTUSED

integer

INITRANS

integer

storage_clause

The following table shows parameters that appear in the syntax diagrams and
provides examples of the values you might substitute for them in your statements:

How to Read Syntax Diagrams

A-1

Graphic Syntax Diagrams

Table A–1

Syntax Parameters

Parameter

Description

Examples

table

The substitution value must be the name of an
object of the type specified by the parameter. For a
list of all types of objects, see the section, "Schema
Objects" on page 3-109.

employees

c

The substitution value must be a single character
from your database character set.

T
s

'text'

The substitution value must be a text string in single 'Employee records'
quotation marks. See the syntax description of 'text'
in "Text Literals" on page 3-45.

char

The substitution value must be an expression of
data type CHAR or VARCHAR2 or a character literal in
single quotation marks.

last_name
'Smith'

condition

The substitution value must be a condition that
last_name >'A'
evaluates to TRUE or FALSE. See the syntax
description of condition in Chapter 7, "Conditions".

date

The substitution value must be a date constant or an TO_DATE(
expression of DATE data type.
'01-Jan-2002',

d

'DD-MON-YYYY')
expr

The substitution value can be an expression of any
data type as defined in the syntax description of
expr in "About SQL Expressions" on page 6-1.

integer

The substitution value must be an integer as defined 72
by the syntax description of integer in "Integer
Literals" on page 3-47.

number

The substitution value must be an expression of
AVG(salary)
NUMBER data type or a number constant as defined in
15 * 7
the syntax description of number in "Numeric
Literals" on page 3-47.

m
n

salary + 1000

raw

The substitution value must be an expression of
data type RAW.

HEXTORAW('7D')

subquery

The substitution value must be a SELECT statement
that will be used in another SQL statement. See
SELECT on page 19-4.

SELECT last_name
FROM employees

db_name

The substitution value must be the name of a
sales_db
nondefault database in an embedded SQL program.

db_string

The substitution value must be the database
—
identification string for an Oracle Net database
connection. For details, see the user's guide for your
specific Oracle Net protocol.

Required Keywords and Parameters
Required keywords and parameters can appear singly or in a vertical list of
alternatives. Single required keywords and parameters appear on the main path,
which is the horizontal line you are currently traveling. In the following example,
library_name is a required parameter:
drop_library::=
DROP

LIBRARY

library_name

A-2 Oracle Database SQL Language Reference

;

Graphic Syntax Diagrams

If there is a library named HQ_LIB, then, according to the diagram, the following
statement is valid:
DROP LIBRARY hq_lib;

If multiple keywords or parameters appear in a vertical list that intersects the main
path, then one of them is required. You must choose one of the keywords or
parameters, but not necessarily the one that appears on the main path. In the following
example, you must choose one of the two settings:
key_compression::=
integer
COMPRESS
NOCOMPRESS

Optional Keywords and Parameters
If keywords and parameters appear in a vertical list above the main path, then they are
optional. In the following example, instead of traveling down a vertical line, you can
continue along the main path:
deallocate_unused_clause::=
KEEP
DEALLOCATE

size_clause

UNUSED

size_clause::=
K
M
G
T
P
E
integer

According to the diagrams, all of the following statements are valid:
DEALLOCATE
DEALLOCATE
DEALLOCATE
DEALLOCATE

UNUSED;
UNUSED KEEP 1000;
UNUSED KEEP 10G;
UNUSED 8T;

Syntax Loops
Loops let you repeat the syntax within them as many times as you like. In the
following example, after choosing one value expression, you can go back repeatedly to
choose another, separated by commas.

How to Read Syntax Diagrams

A-3

Graphic Syntax Diagrams

query_partition_clause::=
,
expr
PARTITION

BY

,
(

expr

)

Multipart Diagrams
Read a multipart diagram as if all the main paths were joined end to end. The
following example is a three-part diagram:
alter_java::=
SOURCE
ALTER

schema

.

JAVA

object_name
CLASS

,
RESOLVER

(

(

schema_name

match_string

)

)

–

COMPILE
RESOLVE

;

invoker_rights_clause

According to the diagram, the following statement is valid:
ALTER JAVA SOURCE jsource_1 COMPILE;

Database Objects
The names of Oracle identifiers, such as tables and columns, must not exceed 30
characters in length. The first character must be a letter, but the rest can be any
combination of letters, numerals, dollar signs ($), pound signs (#), and underscores (_).
However, if an Oracle identifier is enclosed by double quotation marks ("), then it can
contain any combination of legal characters, including spaces but excluding quotation
marks. Oracle identifiers are not case sensitive except within double quotation marks.

See Also: "Database Object Naming Rules" on page 3-111 for more
information

A-4 Oracle Database SQL Language Reference

B
Automatic and Manual Locking Mechanisms
During SQL Operations

B

This appendix describes mechanisms that lock data either automatically or as specified
by the user during SQL statements. For a general discussion of locking mechanisms in
the context of data concurrency and consistency, see Oracle Database Concepts.
This appendix contains the following sections:
■

Automatic Locks in DML Operations

■

Automatic Locks in DDL Operations

■

Manual Data Locking

Automatic Locks in DML Operations
The purpose of a DML lock, also called a data lock, is to guarantee the integrity of data
being accessed concurrently by multiple users. For example, a DML lock can prevent
multiple customers from buying the last copy of a book available from an online
bookseller. DML locks prevent destructive interference of simultaneous conflicting
DML or DDL operations.
DML statements automatically acquire locks at both the table level and the row level.
In the sections that follow, the acronym in parentheses after each type of lock or lock
mode is the abbreviation used in the Locks Monitor of Oracle Enterprise Manager.
Enterprise Manager might display "TM" for any table lock, rather than indicate the
mode of table lock (such as RS or SRX).
The types of row and table locks are summarized here. For a more complete discussion
of the types of row and table locks, see Oracle Database Concepts.
Row Locks (TX) A row lock, also called a TX lock, is a lock on a single row of a table.
A transaction acquires a row lock for each row modified by one of the following
statements: INSERT, UPDATE, DELETE, MERGE, and SELECT ... FOR UPDATE. The row lock
exists until the transaction commits or rolls back.

When a transaction obtains a row lock for a row, the transaction also acquires a table
lock for the table in which the row resides. The table lock prevents conflicting DDL
operations that would override data changes in a current transaction.
Table Locks (TM) A transaction automatically acquires a table lock (TM lock) when a
table is modified with the following statements: INSERT, UPDATE, DELETE, MERGE, and
SELECT ... FOR UPDATE. These DML operations require table locks to reserve DML access
to the table on behalf of a transaction and to prevent DDL operations that would
conflict with the transaction. You can explicitly obtain a table lock using the LOCK

Automatic and Manual Locking Mechanisms During SQL Operations

B-1

Automatic Locks in DML Operations

TABLE statement, as described in "Manual Data Locking" on page B-5.
A table lock can be held in any of the following modes:
■

■

■

■

■

A row share lock (RS), also called a subshare table lock (SS), indicates that the
transaction holding the lock on the table has locked rows in the table and intends
to update them. An SS lock is the least restrictive mode of table lock, offering the
highest degree of concurrency for a table.
A row exclusive lock (RX), also called a subexclusive table lock (SX), indicates
that the transaction holding the lock has updated table rows or issued SELECT ...
FOR UPDATE. An SX lock allows other transactions to query, insert, update, delete,
or lock rows concurrently in the same table. Therefore, SX locks allow multiple
transactions to obtain simultaneous SX and SS locks for the same table.
A share table lock (S) held by one transaction allows other transactions to query
the table (without using SELECT ... FOR UPDATE) but allows updates only if a single
transaction holds the share table lock. Multiple transactions may hold a share table
lock concurrently, so holding this lock is not sufficient to ensure that a transaction
can modify the table.
A share row exclusive table lock (SRX), also called a share-subexclusive table
lock (SSX), is more restrictive than a share table lock. Only one transaction at a
time can acquire an SSX lock on a given table. An SSX lock held by a transaction
allows other transactions to query the table (except for SELECT ... FOR UPDATE) but
not to update the table.
An exclusive table lock (X) is the most restrictive mode of table lock, allowing the
transaction that holds the lock exclusive write access to the table. Only one
transaction can obtain an X lock for a table.
See Also:

"Manual Data Locking" on page B-5

Locks in DML Operations
Oracle Database automatically obtains row-level and table-level locks on behalf of
DML operations. The type of operation determines the locking behavior. Table B–1
summarizes the information in this section.
The implicit SX locks shown for the DML statements in
Table B–1 can sometimes be exclusive (X) locks for a short time owing
to side effects from constraints.

Note:

Table B–1

Summary of Locks Obtained by DML Statements

SQL Statement

Row
Locks

Table
Lock
Mode

RS

RX

S

SRX X

SELECT ... FROM table...

—

none

Y

Y

Y

Y

Y

INSERT INTO table ...

Yes

SX

Y

Y

N

N

N

UPDATE table ...

Yes

SX

Y1

Y1

N

N

N

MERGE INTO table ...

Yes

SX

Y

Y

N

N

N

DELETE FROM table ...

Yes

SX

Y1

Y1

N

N

N

SELECT ... FROM table FOR UPDATE OF ...

Yes

SX

Y1

Y1

N

N

N

LOCK TABLE table IN ...

—

B-2 Oracle Database SQL Language Reference

Automatic Locks in DML Operations

Table B–1 (Cont.) Summary of Locks Obtained by DML Statements
Table
Lock
Mode

RS

RX

S

SRX X

ROW SHARE MODE

SS

Y

Y

Y

Y

N

ROW EXCLUSIVE MODE

SX

Y

Y

N

N

N

SHARE MODE

S

Y

N

Y

N

N

SHARE ROW EXCLUSIVE MODE

SSX

Y

N

N

N

N

EXCLUSIVE MODE

X

N

N

N

N

N

Row
Locks

SQL Statement

1

Yes, if no conflicting row locks are held by another transaction. Otherwise, waits occur.

Locks When Rows Are Queried
A query can be explicit, as in the SELECT statement, or implicit, as in most INSERT,
MERGE, UPDATE, and DELETE statements. The only DML statement that does not
necessarily include a query component is an INSERT statement with a VALUES clause.
Because queries only read data, they are the SQL statements least likely to interfere
with other SQL statements.
The following characteristics apply to a query without the FOR UPDATE clause:
■

■

The query acquires no data locks. Therefore, other transactions can query and
update a table being queried, including the specific rows being queried. Because
queries without the FOR UPDATE clause do not acquire any data locks to block other
operations, such queries are often referred to as nonblocking queries.
The query does not have to wait for any data locks to be released. Therefore, the
query can always proceed. An exception to this rule is that queries may have to
wait for data locks in some very specific cases of pending distributed transactions.

Locks When Rows Are Modified
Some databases use a lock manager to maintain a list of locks in memory. Oracle
Database, in contrast, stores lock information in the data block that contains the locked
row. Each row lock affects only a single row.
Oracle Database uses a queuing mechanism for acquisition of row locks. If a
transaction requires a row lock, and if the row is not already locked, then the
transaction acquires a lock in the row's data block. The transaction itself has an entry
in the interested transaction list (ITL) section of the block header. Each row modified
by this transaction points to a copy of the transaction ID stored in the ITL. Thus, 100
rows in the same block modified by a single transaction require 100 row locks, but all
100 rows reference a single transaction ID.
When a transaction ends, the transaction ID remains in the ITL section of the data
block header. If a new transaction wants to modify a row, then it uses the transaction
ID to determine whether the lock is active. If the lock is active, then the session of the
new transaction asks to be notified when the lock is released; otherwise, the new
transaction acquires the lock.
The characteristics of INSERT, UPDATE, DELETE, and SELECT ... FOR UPDATE statements are
as follows:
■

A transaction containing a DML statement acquires exclusive row locks on the
rows modified by the statement. Therefore, other transactions cannot update or
delete the locked rows until the locking transaction either commits or rolls back.

Automatic and Manual Locking Mechanisms During SQL Operations

B-3

Automatic Locks in DDL Operations

■

■

In addition to these row locks, a transaction containing a DML statement that
modifies data also requires at least a subexclusive table lock (SX) on the table that
contains the affected rows. If the transaction already holds an S, SRX, or X table
lock for the table, which are more restrictive than an SX lock, then the SX lock is
not needed and is not acquired. If the containing transaction already holds only an
SS lock, however, then Oracle Database automatically converts the SS lock to an SX
lock.
A transaction that contains a DML statement does not require row locks on any
rows selected by a subquery or an implicit query.
In the following sample UPDATE statement, the SELECT statement in parentheses is a
subquery, whereas the WHERE a > 5 clause is an implicit query:
UPDATE t SET x = ( SELECT y FROM t2 WHERE t2.z = t.z ) WHERE a > 5;

A subquery or implicit query inside a DML statement is guaranteed to be
consistent as of the start of the query and does not see the effects of the DML
statement of which it forms a part.
■

A query in a transaction can see the changes made by previous DML statements in
the same transaction, but not the uncommitted changes of other transactions.
Oracle Database Concepts for information on locks in
foreign keys

See Also:

Automatic Locks in DDL Operations
A data dictionary (DDL) lock protects the definition of a schema object while it is
acted upon or referred to by an ongoing DDL operation. For example, when a user
creates a procedure, Oracle Database automatically acquires DDL locks for all schema
objects referenced in the procedure definition. The DDL locks prevent these objects
from being altered or dropped before procedure compilation is complete.
Oracle Database acquires a DDL lock automatically on behalf of any DDL transaction
requiring it. Users cannot explicitly request DDL locks. Only individual schema objects
that are modified or referenced are locked during DDL operations. The whole data
dictionary is never locked.
DDL operations also acquire DML locks on the schema object to be modified.

Exclusive DDL Locks
An exclusive DDL lock prevents other session from obtaining a DDL or DML lock.
Most DDL operations require exclusive DDL locks to prevent destructive interference
with other DDL operations that might modify or reference the same schema object. For
example, a DROP TABLE operation is not allowed to drop a table while an ALTER TABLE
operation is adding a column to it, and vice versa. However, a query against the table
is not blocked.
Exclusive DDL locks last for the duration of DDL statement execution and automatic
commit. During the acquisition of an exclusive DDL lock, if another DDL lock is
already held on the schema object by another operation, then the acquisition waits
until the older DDL lock is released and then proceeds.

Share DDL Locks
A share DDL lock for a resource prevents destructive interference with conflicting
DDL operations, but allows data concurrency for similar DDL operations.
B-4 Oracle Database SQL Language Reference

Manual Data Locking

For example, when a CREATE PROCEDURE statement is run, the containing transaction
acquires share DDL locks for all referenced tables. Other transactions can concurrently
create procedures that reference the same tables and acquire concurrent share DDL
locks on the same tables, but no transaction can acquire an exclusive DDL lock on any
referenced table.
A share DDL lock lasts for the duration of DDL statement execution and automatic
commit. Thus, a transaction holding a share DDL lock is guaranteed that the definition
of the referenced schema object is constant for the duration of the transaction.

Breakable Parse Locks
A parse lock is held by a SQL statement or PL/SQL program unit for each schema
object that it references. Parse locks are acquired so that the associated shared SQL area
can be invalidated if a referenced object is altered or dropped. A parse lock is called a
breakable parse lock because it does not disallow any DDL operation and can be
broken to allow conflicting DDL operations.
A parse lock is acquired in the shared pool during the parse phase of SQL statement
execution. The lock is held as long as the shared SQL area for that statement remains in
the shared pool.

Manual Data Locking
Oracle Database always performs locking automatically to ensure data concurrency,
data integrity, and statement-level read consistency. However, you can override the
Oracle default locking mechanisms. This can be useful in situations such as the
following:
■

■

When your application requires consistent data for the duration of the transaction,
not reflecting changes by other transactions, you can achieve transaction-level read
consistency by using explicit locking, read-only transactions, serializable
transactions, or by overriding default locking.
When your application requires that a transaction have exclusive access to a
resource so that the transaction does not have to wait for other transactions to
complete, you can explicitly lock the data for the duration of the transaction.

You can override automatic locking at two levels:
■

Transaction. You can override transaction-level locking with the following SQL
statements:
–

SET TRANSACTION ISOLATION LEVEL

–

LOCK TABLE

–

SELECT ... FOR UPDATE

Locks acquired by these statements are released after the transaction commits or
rolls back.
■

Session. A session can set the required transaction isolate level with an ALTER
SESSION SET ISOLATION LEVEL statement.

Automatic and Manual Locking Mechanisms During SQL Operations

B-5

Manual Data Locking

When overriding Oracle default locking, the database
administrator or application developer should ensure that data
integrity is guaranteed, data concurrency is acceptable, and deadlocks
are not possible or, if possible, are appropriately handled. For more
information on these criteria, see Oracle Database Concepts.

Note:

B-6 Oracle Database SQL Language Reference

C
C

Oracle and Standard SQL

This appendix discusses Oracle's conformance with the SQL:2008 standards. The
mandatory portion of SQL:2008 is known as Core SQL:2008 and is found in SQL:2008
Part 2 (Foundation) and Part 11 (Schemata). The Foundation features are analyzed in
Annex F of Part 2 in the table "Feature taxonomy and definition for mandatory
features of SQL/Foundation". The Schemata features are analyzed in Annex E of Part
11 in the table "Feature taxonomy and definition for mandatory features of
SQL/Schemata".
This appendix declares Oracle's conformance to the SQL standards established by the
American National Standards Institute (ANSI) and the International Organization for
Standardization (ISO). (The ANSI and ISO SQL standards are identical.)
This appendix contains the following sections:
■

ANSI Standards

■

ISO Standards

■

Oracle Compliance To Core SQL:2008

■

Oracle Support for Optional Features of SQL/Foundation:2008

■

Oracle Compliance with SQL/CLI:2008

■

Oracle Compliance with SQL/PSM:2008

■

Oracle Compliance with SQL/MED:2008

■

Oracle Compliance with SQL/OLB:2008

■

Oracle Compliance with SQL/JRT:2008

■

Oracle Compliance with SQL/XML:2008

■

Oracle Compliance with FIPS 127-2

■

Oracle Extensions to Standard SQL

■

Oracle Compliance with Older Standards

■

Character Set Support

ANSI Standards
The following documents of the American National Standards Institute (ANSI) relate
to SQL:
■

ANSI/ISO/IEC 9075-1:2008, Information technology—Database
languages—SQL—Part 1: Framework (SQL/Framework)

Oracle and Standard SQL C-1

ISO Standards

■

■

■

■

■

■

■

■

ANSI/ISO/IEC 9075-2:2008, Information technology—Database
languages—SQL—Part 2: Foundation (SQL/Foundation)
ANSI/ISO/IEC 9075-3:2008, Information technology—Database
languages—SQL—Part 3: Call-Level Interface (SQL/CLI)
ANSI/ISO/IEC 9075-4:2008, Information technology—Database
languages—SQL—Part 4: Persistent Stored Modules (SQL/PSM)
ANSI/ISO/IEC 9075-9:2008, Information technology—Database
languages—SQL—Part 9: Management of External Data (SQL/MED)
ANSI/ISO/IEC 9075-10:2008, Information technology—Database
languages—SQL—Part 10: Object Language Bindings (SQL/OLB)
ANSI/ISO/IEC 9075-11:2008, Information technology—Database
languages—SQL—Part 11: Information and Definition Schemas (SQL/Schemata)
ANSI/ISO/IEC 9075-13:2008, Information technology—Database
languages—SQL—Part 13: SQL Routines and Types using the Java Programming
Language (SQL/JRT)
ANSI/ISO/IEC 9075-14:2008, Information technology—Database
languages—SQL—Part 14: XML-Related Specifications (SQL/XML)

These standards are identical to the corresponding ISO standards listed in the next
section.
You can obtain a copy of ANSI standards from this address:
American National Standards Institute
25 West 43rd Street, fourth floor
New York, NY 10036 USA
Telephone: +1.212.642.4900
Fax: +1.212.398.0023
Web site: http://www.ansi.org/
You can also obtain the standards from their Web site:
http://webstore.ansi.org/default.aspx

A subset of ANSI standards, including the SQL standard, are INCITS standards. You
can obtain these from the InterNational Committee for Information Technology
Standards (INCITS) at:
http://www.incits.org/

ISO Standards
The following documents of the International Organization for Standardization (ISO)
relate to SQL:
■

■

■

■

ISO/IEC 9075-1:2008, Information technology—Database languages—SQL—Part
1: Framework (SQL/Framework)
ISO/IEC 9075-2:2008, Information technology—Database languages—SQL—Part
2: Foundation (SQL/Foundation)
ISO/IEC 9075-3:2008, Information technology—Database languages—SQL—Part
3: Call-Level Interface (SQL/CLI)
ISO/IEC 9075-4:2008, Information technology—Database languages—SQL—Part
4: Persistent Stored Modules (SQL/PSM)

C-2 Oracle Database SQL Language Reference

Oracle Compliance To Core SQL:2008

■

■

■

■

■

ISO/IEC 9075-9:2008, Information technology—Database languages—SQL—Part
9: Management of External Data (SQL/MED)
ISO/IEC 9075-10:2008, Information technology—Database languages—SQL—Part
10: Object Language Bindings (SQL/OLB)
ISO/IEC 9075-11:2008, Information technology—Database languages—SQL—Part
11: Information and Definition Schemas (SQL/Schemata)
ISO/IEC 9075-13:2008, Information technology—Database languages—SQL—Part
13: SQL Routines and Types using the Java Programming Language (SQL/JRT)
ISO/IEC 9075-14:2008, Information technology—Database languages—SQL—Part
14: XML-Related Specifications (SQL/XML)

You can obtain a copy of ISO standards from this address:
International Organization for Standardization
1, ch. de la Voie-Creuse
Case postale 56
CH-1211, Geneva 20, Switzerland
Phone: +41.22.749.0111
Fax: +41.22.733.3430
Web site: http://www.iso.org/
or from their Web store:
http://www.iso.org/iso/store.htm

Oracle Compliance To Core SQL:2008
The ANSI and ISO SQL standards require conformance claims to state the type of
conformance and the implemented facilities. The minimum claim of conformance is
called Core SQL:2008 and is defined in Part 2, SQL/Foundation, and Part 11,
SQL/Schemata, of the standard. The following products provide full or partial
conformance with Core SQL:2008 as described in the tables that follow:
■

Oracle Database server

■

Pro*C/C++, release 9.2.0

■

Pro*COBOL, release 9.2.0

■

Pro*Fortran, release 1.8.77

■

SQL Module for Ada (Mod*Ada), release 9.2.0

■

Pro*COBOL 1.8, release 1.8.77

■

Pro*PL/I, release 1.6.28

■

OTT (Oracle Type Translator)

The SQL standards conformance features can be used either as a guide to portability,
or as a guide to functionality. From the standpoint of portability, the user is interested
in conformance to both the precise syntax and semantics of the standard feature. From
the standpoint of functionality, the user is less concerned about the precise syntax and
more concerned with issues of semantics. The tables in this appendix use the following
terms regarding support for standard syntax and semantics:
■
■

Full Support: The feature is supported with standard syntax and semantics.
Partial Support: Some, but not all, of the standard syntax is supported; whatever is
supported has standard semantics.
Oracle and Standard SQL C-3

Oracle Compliance To Core SQL:2008

■

■

■

Enhanced Supported: The standard semantics is supported, but gives functionality
that differs from the standard by enhancing it.
Equivalent Support: The standard semantics is supported using non-standard
syntax.
Similar Support: Neither the standard's syntax nor semantics are supported
precisely, but similar functionality is provided.

Oracle's support for the features of Core SQL:2008 is listed in Table C–1:
Table C–1

Oracle Support of Core SQL:2008 Features

Feature ID,
Feature

Support

E011, Numeric
data types

Oracle fully supports this feature.

E021, Character
data types

Oracle fully supports these subfeatures:
■

E021-01, CHARACTER data type

■

E021-07, Character concatenation

■

E021-08, UPPER and LOWER functions

■

E021-09, TRIM function

■

E021-10, Implicit casting among character data types

Oracle partially supports these subfeatures:
■

■

■

E021-02, CHARACTER VARYING data type (Oracle does not distinguish a
zero-length VARCHAR string from NULL)
E021-03, Character literals (Oracle regards the zero-length literal '' as
being null)
E021-12, Character comparison (Oracle's rules for padding the shorter
of two strings to be compared differs from the standard)

Oracle has equivalent functionality for these subfeatures:

E031, Identifiers

■

E021-04, CHARACTER_LENGTH function: use LENGTH function instead

■

E021-05, OCTET_LENGTH function: use LENGTHB function instead

■

E021-06, SUBSTRING function: use SUBSTR function instead

■

E021-11, POSITION function: use INSTR function instead

Oracle supports this feature, with the following exceptions:
■

■

■

Oracle does not support the escape sequence to permit a double quote
within a quoted identifier
A non-quoted identifier may not be equivalent to an Oracle reserved
word (the list of Oracle reserved words differs from the standard's list)
A column name may not be ROWID, even as a quoted identifier

Oracle extends this feature as follows:
■

An identifier may be up to 30 characters long

■

A non-quoted identifier may have dollar sign ($) or pound sign (#)

C-4 Oracle Database SQL Language Reference

Oracle Compliance To Core SQL:2008

Table C–1 (Cont.) Oracle Support of Core SQL:2008 Features
Feature ID,
Feature
E051, Basic
query
specification

Support
Oracle fully supports the following subfeatures:
■

E051-01, SELECT DISTINCT

■

E051-02, GROUP BY clause

■

E051-04, GROUP BY can contain columns not in SELECT list

■

E051-05, SELECT list items can be renamed

■

E051-06, HAVING clause

■

E051-07, Qualified * in SELECT list

Oracle partially supports the following subfeatures:
■

E051-08, Correlation names in FROM clause (Oracle supports correlation
names, but not the optional AS keyword)

Oracle has equivalent functionality for the following subfeature:
■

E051-09, Rename columns in the FROM clause (column names can be
renamed in a subquery in the FROM clause)

E061, Basic
predicates and
search
conditions

Oracle fully supports this feature, except that Oracle comparison of
character strings differs from the standard as follows: In the standard, two
character strings of unequal length are compared by either padding the
shorter string with spaces or a fictitious character that is less than all actual
characters. The decision on padding is made on the basis of the character
set. In Oracle, the decision is based on whether the comparands are of fixed
or varying length.

E071, Basic
query
expressions

Oracle fully supports the following subfeatures:
■

E071-01, UNION DISTINCT table operator

■

E071-02, UNION ALL table operator

■

■

E071-05, Columns combined by table operators need not have exactly
the same type
E071-06, table operators in subqueries

Oracle has equivalent functionality for the following subfeature:
■

E071-03, EXCEPT DISTINCT table operator: Use MINUS instead of EXCEPT
DISTINCT

E081, Basic
privileges

Oracle fully supports all subfeatures of this feature, except E081-09, USAGE
privileges. In the standard, the USAGE privilege permits the user to use
domains, collations, character sets, transliterations, user-defined types and
sequence generators. Oracle does not support domains or transliterations.
No privileges are required to access collations and character sets. The Oracle
privilege to use a user-defined type is EXECUTE. The Oracle privilege to use a
sequence type is SELECT.

E091, Set
functions

Oracle fully supports this feature.

E101, Basic data
manipulation

Oracle fully supports this feature.

E111, Single row
SELECT
statement

Oracle fully supports this feature.

Oracle and Standard SQL C-5

Oracle Compliance To Core SQL:2008

Table C–1 (Cont.) Oracle Support of Core SQL:2008 Features
Feature ID,
Feature
E121, Basic
cursor support

Support
Oracle fully supports the following subfeatures:
■

E121-02, ORDER BY columns need not be in SELECT list

■

E121-03, Value expressions in ORDER BY clause

■

E121-04, OPEN statement

■

E121-06, Positioned UPDATE statement

■

E121-07, Positioned DELETE statement

■

E121-08, CLOSE statement

Oracle provides partial support for the following subfeatures:
■

■

E121-01, DECLARE CURSOR - fully supported, except for the FOR READ ONLY
syntax
E121-10 FETCH statement, implicit NEXT - fully supported, except for the
noise word FROM

Oracle provides enhanced support for the following subfeature:
■

E121-17, WITH HOLD cursors (in the standard, a cursor is not held
through a ROLLBACK, but Oracle does hold through ROLLBACK)

E131, Null value Oracle fully supports this feature, with this exception: In Oracle, a null of
support
character type is indistinguishable from a zero-length character string.
E141, Basic
integrity
constraints

Oracle fully supports this feature.

E151,
Transaction
support

Oracle fully supports this feature.

E152, Basic SET
TRANSACTION
statement

Oracle fully supports this feature.

E153, Updatable
queries with
subqueries

Oracle fully supports this feature.

E161, SQL
Oracle fully supports this feature.
comments using
leading double
minus
E171, SQLSTATE Oracle fully supports this feature.
support
E182, Module
language

Oracle supports this feature for Ada only.

C-6 Oracle Database SQL Language Reference

Oracle Compliance To Core SQL:2008

Table C–1 (Cont.) Oracle Support of Core SQL:2008 Features
Feature ID,
Feature
F021, Basic
information
schema

Support
Oracle does not have any of the views in this feature. However, Oracle
makes the same information available in other metadata views:
■

Instead of TABLES, use ALL_TABLES.

■

Instead of COLUMNS, use ALL_TAB_COLUMNS.

■

Instead of VIEWS, use ALL_VIEWS.
However, Oracle's ALL_VIEWS does not display whether a user view was
defined WITH CHECK OPTION or if it is updatable. To see whether a view
has WITH CHECK OPTION, use ALL_CONSTRAINTS, with TABLE_NAME equal to
the view name and look for CONSTRAINT_TYPE equal to 'V'.

■

Instead of TABLE_CONSTRAINTS, REFERENTIAL_CONSTRAINTS, and CHECK_
CONSTRAINTS, use ALL_CONSTRAINTS.
However, Oracle's ALL_CONSTRAINTS does not display whether a
constraint is deferrable or initially deferred.

F031, Basic
schema
manipulation

Oracle fully supports these subfeatures:
■

F031-01, CREATE TABLE statement to create persistent base tables

■

F031-02, CREATE VIEW statement

■

F031-03, GRANT statement

Oracle partially supports this subfeature:
■

F031-04, ALTER TABLE statement: ADD COLUMN clause (Oracle does not
support the optional keyword COLUMN in this syntax. Also, Oracle
requires the column definition to be enclosed in parentheses, unlike the
standard.)

Oracle does not support these subfeatures (because Oracle does not support
the keyword RESTRICT):
■

F031-13, DROP TABLE statement: RESTRICT clause

■

F031-16, DROP VIEW statement: RESTRICT clause

■

F031-19, REVOKE statement: RESTRICT clause

(Oracle DROP commands enhance the standard by invalidating dependent
objects, so that they can be subsequently revalidated without user action,
rather than either cascading all drops to dependent objects or prohibiting a
drop if there is a dependent object.)
F041, Basic
joined table

Oracle fully supports this feature.

F051, Basic date
and time

Oracle fully supports this feature, except the following subfeatures are not
supported:
■

F051-02, TIME data type

■

F051-07, LOCALTIME

F081, UNION and
EXCEPT in views

Oracle fully supports UNION in views. The equivalent in Oracle of the
standard's EXCEPT is called MINUS, which is fully supported in views.

F131, Grouped
operations

Oracle fully supports this feature.

F181, Multiple
module support

Oracle fully supports this feature.

F201, CAST
function

Oracle fully supports this feature.

F221, Explicit
defaults

Oracle fully supports this feature.

Oracle and Standard SQL C-7

Oracle Compliance To Core SQL:2008

Table C–1 (Cont.) Oracle Support of Core SQL:2008 Features
Feature ID,
Feature

Support

F261, CASE
expressions

Oracle fully supports this feature.

F311, Schema
definition
statement

Oracle fully supports this feature.

F471, Scalar
Oracle fully supports this feature.
subquery values
F481, Expanded
null predicate

Oracle fully supports this feature.

F501, Feature
and
conformance
views

Oracle does not support this feature.

F812, Basic
flagging

Oracle has a flagger, but it flags SQL-92 compliance rather than SQL:2008
compliance.

S011, Distinct
types

Distinct types are strongly typed scalar types. A distinct type can be
emulated in Oracle using an object type with only one attribute. The
standard's Information Schema view called USER_DEFINED_TYPES is
equivalent to Oracle's metadata view ALL_TYPES.

T321, Basic
SQL-invoked
routines

Oracle fully supports these subfeatures:
■

T321-03, function invocation

■

T321-04, CALL statement

Oracle supports these subfeatures with syntactic differences:
■

T321-01, user-defined functions with no overloading

■

T321-02, user-defined procedures with no overloading

The Oracle syntax for CREATE FUNCTION and CREATE PROCEDURE differs from
the standard as follows:
■

■
■

■

■

In the standard, the mode of a parameter (IN, OUT, or INOUT) comes
before the parameter name, whereas in Oracle it comes after the
parameter name.
The standard uses INOUT, whereas Oracle uses IN OUT.
Oracle requires either IS or AS after the return type and before the
definition of the routine body, while the standard lacks these keywords.
If the routine body is in C (for example), then the standard uses the
keywords LANGUAGE C EXTERNAL NAME to name the routine, whereas
Oracle uses LANGUAGE C NAME.
If the routine body is in SQL, then Oracle uses its proprietary
procedural extension called PL/SQL.

Oracle supports the following subfeature in PL/SQL but not in Oracle SQL:
■

T321-05, RETURN statement

Oracle provides equivalent functionality for the following subfeatures:
■
■

T631, IN
predicate with
one list element

T321-06, ROUTINES view: Use the ALL PROCEDURES metadata view.
T321-07, PARAMETERS view: Use the ALL_ARGUMENTS and ALL_METHOD_
PARAMS metadata views.

Oracle fully supports this feature.

C-8 Oracle Database SQL Language Reference

Oracle Support for Optional Features of SQL/Foundation:2008

Oracle Support for Optional Features of SQL/Foundation:2008
Oracle's support for optional features of SQL/Foundation:2008 is listed in Table C–2:
Table C–2

Oracle Support for Optional Features of SQL/Foundation:2008

Feature ID, Feature

Support

B012, Embedded C

Oracle fully supports this feature.

B013, Embedded COBOL

Oracle fully supports this feature.

B014, Embedded Fortran

Oracle fully supports this feature.

B021, Direct SQL

Oracle fully supports this feature, as SQL*Plus.

B031, Basic dynamic SQL

Oracle supports dynamic SQL in two styles, documented in the
embedded language manuals as "Oracle dynamic SQL" and
"ANSI dynamic SQL."
ANSI dynamic SQL is an implementation of the standard, with
the following restrictions:
■
■

■

■

Oracle supports a subset of the descriptor items.
For , Oracle only supports .
For , Oracle only supports .
Dynamic parameters are indicated by a colon followed by
an identifier rather than a question mark.

Oracle dynamic SQL is similar to standard dynamic SQL, with
the following modifications:
■

■

■

B032, Extended dynamic
SQL

Parameters are indicated by a colon followed by an
identifier, instead of a question mark.
Oracle's DESCRIBE SELECT LIST FOR statement replaces the
standard's DESCRIBE OUTPUT.
Oracle provides DECLARE STATEMENT if you want to declare a
cursor using a dynamic SQL statement physically prior to
the PREPARE statement that prepares the dynamic SQL
statement.

In ANSI dynamic SQL, Oracle only implements the ability to
declare global statements and global cursors from this feature;
the rest of the feature is not supported.
In Oracle dynamic SQL, Oracle's DESCRIBE BIND VARIABLES is
equivalent to the standard's DESCRIBE INPUT; the rest of this
feature is not supported.

B111, Module language Ada Oracle fully supports this feature.
B122, Routine language C

Oracle supports external routines written in C, though Oracle
does not support the standard syntax for creating such routines.

B128, Routine language SQL Oracle supports routines written in PL/SQL, which is Oracle's
equivalent to the standard procedural language SQL/PSM.
F032, CASCADE drop
behavior

In Oracle, a DROP command invalidates all of the dropped
object's dependent objects. Invalidated objects are effectively
unusable until the dropped object is redefined in such a way to
allow successful recompilation of the invalidated object.

F033, ALTER TABLE
statement: DROP COLUMN
clause

Oracle provides a DROP COLUMN clause, but without the RESTRICT
or CASCADE options found in the standard.

Oracle and Standard SQL C-9

Oracle Support for Optional Features of SQL/Foundation:2008

Table C–2 (Cont.) Oracle Support for Optional Features of SQL/Foundation:2008
Feature ID, Feature

Support

F034, Extended REVOKE
statement

Oracle supports the following parts of this feature:
■

■

F034-01, REVOKE statement performed by other than the
owner of a schema object
F034-03, REVOKE statement to revoke a privilege that the
grantee has WITH GRANT OPTION

Oracle provides equivalent functionality for the following parts
of this feature:
■

CASCADE: In Oracle, a REVOKE invalidates all dependent
objects, which become effectively unusable until the
metadata is changed through subsequent CREATE and GRANT
commands enabling the invalidated object to be successfully
recompiled.

F052, Intervals and datetime Oracle only supports the INTERVAL YEAR TO MONTH and INTERVAL
arithmetic
DAY TO SECOND data types.
F111, Isolations levels other
than SERIALIZABLE

In addition to SERIALIZABLE, Oracle supports the READ
COMMITTED isolation level.

F121, Basic diagnostics
management

Much of the functionality of this feature is provided through the
SQLCA in embedded languages.

F191, Referential delete
actions

Oracle supports ON DELETE CASCADE and ON DELETE SET NULL.

F200, TRUNCATE TABLE

Oracle fully supports this feature, and extends it by permitting
truncation of a table that references itself in a referential integrity
constraint.

F231, Privilege tables

Oracle makes this information available in the following
metadata views:
■

Instead of TABLE_PRIVILEGES, use ALL_TAB_PRIVS.

■

Instead of COLUMN_PRIVILEGES, use ALL_COL_PRIVS.

■

Oracle does not support USAGE privileges so there is no
equivalent to USAGE_PRIVILEGES.

F281, LIKE enhancements

Oracle fully supports this feature.

F291, UNIQUE predicate

The IS A SET condition may be used to test whether a multiset is
a set; that is, each row is unique. Thus, the equivalent of
UNIQUE 
is
CAST (
 AS MULTISET) IS A SET

F302, INTERSECT table
operator

Oracle supports INTERSECT but not INTERSECT ALL. Syntactically,
Oracle differs from the standard in that UNION, INTERSECT, and
MINUS have the same precedence.

F312, MERGE statement

The Oracle MERGE statement is almost the same as the standard,
with these exceptions:
■

■

■

F341, Usage tables

Oracle does not support the optional AS keyword before a
table alias.
Oracle does not support the ability to rename columns of
the table specified in the USING clause with a parenthesized
list of column names following the table alias.
Oracle does not support the .

Oracle makes this information available in the views ALL_
DEPENDENCIES, DBA_DEPENDENCIES, and USER_DEPENDENCIES.

C-10 Oracle Database SQL Language Reference

Oracle Support for Optional Features of SQL/Foundation:2008

Table C–2 (Cont.) Oracle Support for Optional Features of SQL/Foundation:2008
Feature ID, Feature

Support

F381, Extended schema
manipulation

Oracle fully supports the following element of this feature:
■

Oracle supports the standard syntax to add a table
constraint using ALTER TABLE.

Oracle partially supports the following element of this feature:
■

Oracle supports the standard syntax to drop a table
constraint, except that Oracle does not support RESTRICT.

Oracle provides equivalent functionality for the following
element of this feature:
■

To alter the default value of a column, use the MODIFY option
of ALTER TABLE.

Oracle does not support the following parts of this feature:
■

DROP SCHEMA statement

■

ALTER ROUTINE statement

F382, Alter column data
type

Oracle supports this functionality, though with non-standard
syntax.

F391, Long identifiers

Oracle supports identifiers up to 30 characters in length.

F393, Unicode escapes in
literals

The Oracle UNISTR function supports numeric escape sequences
for all Unicode characters.

F394, Optional normal form
specification

This feature adds the keywords NFC, NFD, NFKC, and NKD to the
NORMALIZE function and the IS NORMAL predicate. Without these
keywords, NFC is the default (see Feature T061, UCS support).
Oracle supports all four normalization forms, with nonstandard
syntax, as follows:
■

For NFC, use COMPOSE

■

For NFD, use DECOMPOSE with the CANONICAL option

■

For NFKD, use DECOMPOSE with the COMPATIBILITY option

■

For NFKC, use DECOMPOSE with the CANONICAL option
followed by COMPOSE

Oracle does not support the IS NORMAL predicate.
F401, Extended joined table

Oracle supports FULL outer joins, CROSS joins, and NATURAL joins.

F402, Named column joins
for LOBs, arrays and
multisets

Oracle supports named column joins for columns whose
declared type is nested table. Oracle does not support named
column joins for LOBs or arrays.

F403, Partitioned join tables

Oracle supports this feature, except with FULL outer joins.

F411, Time zone
specification

Oracle fully supports TIMESTAMP WITH TIME ZONE, but does not
support TIME WITH TIME ZONE.

F421, National character

Oracle fully supports this feature.

F441, Extended set function
support

Oracle supports the following parts of this feature:
■

■
■

F442, Mixed column
references in set functions

The ability in the WHERE clause to reference a column that is
defined using an aggregate, either in a view or an inline
view
COUNT without DISTINCT of an expression
Aggregates that reference columns that are outer references
with respect to the aggregating query

Oracle fully supports this feature.

Oracle and Standard SQL

C-11

Oracle Support for Optional Features of SQL/Foundation:2008

Table C–2 (Cont.) Oracle Support for Optional Features of SQL/Foundation:2008
Feature ID, Feature

Support

F461, Named character sets

Oracle supports many character sets with Oracle-defined names.
Oracle does not support any other aspect of this feature.

F491, Constraint
management

Oracle fully supports this feature.

F531, Temporary tables

Oracle supports GLOBAL TEMPORARY tables.

F555, Enhanced seconds
precision

Oracle provides enhanced support for this feature, supporting
up to 9 places after the decimal point.

F561, Full value expressions Oracle fully supports this feature.
F571, Truth value tests

Oracle's LNNVL function is equivalent to the standard's IS NOT
TRUE predicate.

F591, Derived tables

Oracle supports , with the exception of:
■

■

Oracle does not support the optional AS keyword before a
table alias.
Oracle does not support .

F641, Row and table
constructors

In Oracle, a row constructor may be used in an equality or
inequality comparison with another row constructor or with a
subquery. Oracle does not support anything else in this feature.

F690, Collation support

Oracle's NLSSORT function may be used to change the collation of
character expressions.

F693, SQL-sessions and
client module collations

To set a session collation, use ALTER SESSION SET NLS_COMP =
'LINGUISTIC' and also set NLS_SORT to your desired collation.
Oracle does not support client module collations.

F695, Translation support

The Oracle CONVERT function can convert between the database
character set and the national character set. For other character
sets, store the data in the RAW data type and use the PL/SQL
package function UTL_RAW.CONVERT. Oracle does not provide the
ability to add or drop character set conversions.

F721, Deferrable constraints

Oracle fully supports this feature.

F731, INSERT column
privileges

Oracle fully supports this feature.

F761, Session management

Oracle provides the following equivalents for elements of this
feature:
■

■

■

The equivalent to the standard's SET SESSION
CHARACTERISTICS AS TRANSACTION SERIALIZABLE is ALTER
SESSION SET ISOLATION_LEVEL = SERIALIZABLE.
The equivalent to the standard's SET SCHEMA is ALTER
SESSION SET CURRENT_SCHEMA.
The equivalent to the standard's SET COLLATION is ALTER
SESSION SET NLS_SORT.

F771, Connection
management

Oracle's CONNECT statement provides the same functionality as
the standard's CONNECT statement, though with different syntax.
Instead of using the standard's SET CONNECTION, Oracle provides
the AT clause to indicate which connection a SQL statement
should be performed on. Oracle embedded languages let you
disconnect from a connection by using the RELEASE option of
either COMMIT or ROLLBACK.

F781, Self-referencing
operations

Oracle fully supports this feature.

F801, Full set function

Oracle fully supports this feature.

C-12 Oracle Database SQL Language Reference

Oracle Support for Optional Features of SQL/Foundation:2008

Table C–2 (Cont.) Oracle Support for Optional Features of SQL/Foundation:2008
Feature ID, Feature

Support

F831, Full cursor update

Oracle supports the combination of FOR UPDATE and ORDER BY
clauses in a query.

F841, LIKE_REGEX predicate

Oracle's equivalent is REGEXP_LIKE. Oracle's pattern syntax lacks
some of the features of the standard's. Oracle's match parameter
has the same capabilities as the standard's, though with a few
spelling differences.

F842, OCCURRENCES_REGEX
function

Oracle's equivalent is REGEXP_COUNT. Oracle's pattern syntax
lacks some of the features of the standard's. Oracle's match
parameter has the same capabilities as the standard's, though
with a few spelling differences.

F843, POSITION_REGEX
function

Oracle's equivalent is REGEXP_INSTR. Oracle's pattern syntax
lacks some of the features of the standard's. Oracle's match
parameter has the same capabilities as the standard's, though
with a few spelling differences.

F844, SUBSTRING_REGEX
function

Oracle's equivalent is REGEXP_SUBSTR. Oracle's pattern syntax
lacks some of the features of the standard's. Oracle's match
parameter has the same capabilities as the standard's, though
with a few spelling differences.

F845, TRANSLATE_REGEX
function

Oracle's equivalent is REGEXP_REPLACE. Oracle's pattern syntax
lacks some of the features of the standard's. Oracle's match
parameter has the same capabilities as the standard's, though
with a few spelling differences.

S023, Basic structured types

Oracle's object types are equivalent to structured types in the
standard.

S024, Enhanced structured
types

Oracle's syntax is non-standard, but provides equivalents for the
following:
■

NOT INSTANTIABLE

■

STATIC methods

■

■

RELATIVE, MAP, and STATE orderings. The keyword in Oracle
for RELATIVE orderings is ORDER. There is no keyword for
STATE orderings (this is the default, if no other ordering is
defined). Unlike the standard, Oracle does not support
EQUALS ONLY on non-STATE orderings. (See also Feature S251,
User-defined orderings.)
SELF AS RESULT in the signature of constructor methods

S025, Final structured types

Oracle's final object types are equivalent to final structured types
in the standard.

S026, Self-referencing
structured types

In Oracle, an object type OT may have a reference that references
OT.

S041, Basic reference types

Oracle's reference types are equivalent to reference types in the
standard. To dereference a reference, dot notation is used,
instead of -> as in the standard.

S043, Enhanced reference
types

Oracle supports the following elements of this feature:
■
■

■
■

DEREF operator to return the object referenced by a reference
SCOPE clause as a constraint on columns of tables or
materialized views
Adding and dropping the scope of a column
References that are either system-generated or derived from
the primary key (but not from any other list of columns, nor
from a list of attributes of the type)

Oracle and Standard SQL

C-13

Oracle Support for Optional Features of SQL/Foundation:2008

Table C–2 (Cont.) Oracle Support for Optional Features of SQL/Foundation:2008
Feature ID, Feature

Support

S051, Create table of type

Oracle's object tables are equivalent to tables of structured type
in the standard.

S081, Subtables

Oracle supports hierarchies of object views, but not of object
base tables. To emulate a hierarchy of base tables, create a
hierarchy of views on those base tables.

S091, Array types

Oracle VARRAY types are equivalent to array types in the
standard. However, Oracle does not support storage of arrays of
LOBs. To access a single element of an array using a subscript,
you must use PL/SQL. Oracle supports the following aspects of
this feature with nonstandard syntax:
■

■

To construct an instance of varray type, including an empty
array, use the varray type constructor.
To unnest a varray in the FROM clause, use the TABLE
operator.

S092, Arrays of user-defined Oracle supports VARRAYs of object types.
types
S094, Arrays of reference
types

Oracle supports VARRAYs of references.

S095, Array constructors by
query

Oracle supports this using CAST (MULTISET (SELECT ...) AS
varray_type). The ability to order the elements of the array
using ORDER BY is not supported.

S097, Array element
assignment

In PL/SQL, you can assign to array elements, using syntax that
is similar to the standard (SQL/PSM).

S111, ONLY in query
expressions

Oracle supports the ONLY clause for view hierarchies; Oracle
does not support hierarchies of base tables.

S151, Type predicate

Oracle fully supports this feature.

S161, Subtype treatment

Supported, with a minor syntactic difference: The standard
requires parentheses around the referenced type's name; Oracle
does not support parentheses in this position.

S162, Subtype treatment for
references

The standard requires parentheses around the referenced type's
name; Oracle does not support parentheses in this position.

S201, SQL-invoked routines
on arrays

PL/SQL provides the ability to pass arrays as parameters and
return arrays as the result of functions. Procedures and functions
written in C may pass arrays and return arrays as the result of
functions using the Oracle Type Translator (OTT).

S202, SQL-invoked routines
on multisets

A PL/SQL routine may have nested tables as parameters, and
may return a nested table. Routines written in C may pass arrays
and return arrays as the result of functions using the Oracle Type
Translator.

S232, Array locators

Oracle Type Translator supports descriptors for arrays, which
achieve the same purpose as locators.

S233, Multiset locators

Oracle supports locators for nested tables.

S241, Transform functions

The Oracle Type Translator provides the same capability as
transforms.

C-14 Oracle Database SQL Language Reference

Oracle Support for Optional Features of SQL/Foundation:2008

Table C–2 (Cont.) Oracle Support for Optional Features of SQL/Foundation:2008
Feature ID, Feature

Support

S251, User-defined
orderings

Oracle's object type ordering capabilities correspond to the
standard's capabilities as follows:
■

■

■

■

Oracle's MAP ordering corresponds to the standard's ORDER
FULL BY MAP ordering.
Oracle's ORDER ordering corresponds to the standard's ORDER
FULL BY RELATIVE ordering.
If an Oracle object type has neither MAP nor ORDER declared,
then this corresponds to EQUALS ONLY BY STATE in the
standard.
Oracle does not have unordered object types; you can alter
the ordering but you cannot drop it.

S271, Basic multiset support Multisets in the standard are supported as nested table types in
Oracle. The Oracle nested table data type based on a scalar type
ST is equivalent, in standard terminology, to a multiset of rows
having a single field of type ST and named column_value. The
Oracle nested table type based on an object type is equivalent to
a multiset of structured type in the standard.
Oracle supports the following elements of this feature on nested
tables using the same syntax as the standard has for multisets:
■

The CARDINALITY function

■

The SET function

■

The MEMBER predicate

■

The IS A SET predicate

■

The COLLECT aggregate

All other aspects of this feature are supported with non-standard
syntax, as follows:
■

■

■

■

■

To create an empty multiset, denoted MULTISET[] in the
standard, use an empty constructor of the nested table type.
To obtain the sole element of a multiset with one element,
denoted ELEMENT () in the
standard, use a scalar subquery to select the single element
from the nested table.
To construct a multiset by enumeration, use the constructor
of the nested table type.
To construct a multiset by query, use CAST with a multiset
argument, casting to the nested table type.
To unnest a multiset, use the TABLE operator in the FROM
clause.

S272, Multisets of
user-defined types

Oracle's nested table type permits a multiset of structured types.
Oracle does not have distinct types, so a multiset of distinct
types is not supported

S274, Multisets of reference
types

A nested table type can have one or more columns of reference
type.

S275, Advanced multiset
support

Oracle supports the following elements of this feature on nested
tables using the same syntax as the standard has for multisets:
■

The MULTISET UNION, MULTISET INTERSECTION, and MULTISET
EXCEPT operators

■

The SUBMULTISET predicate

■

= and <> predicates

Oracle does not support the FUSION or INTERSECTION aggregates.

Oracle and Standard SQL

C-15

Oracle Support for Optional Features of SQL/Foundation:2008

Table C–2 (Cont.) Oracle Support for Optional Features of SQL/Foundation:2008
Feature ID, Feature

Support

S281, Nested collection
types

Oracle permits nesting of its collection types (varray and nested
table).

T041, Basic LOB data type
support

Oracle supports the following aspects of this feature:
■

The keywords BLOB, CLOB, and NCLOB

■

Concatenation, UPPER, LOWER, and TRIM on CLOBs

Oracle provides equivalent support for the following aspects of
this feature:
■

Use INSTR instead of POSITION.

■

Use LENGTH instead of CHAR_LENGTH.

■

Use SUBSTR instead of SUBSTRING.

Oracle does not support the following aspects of this feature:
■

■
■

■

T042, Extended LOB
support

The keywords BINARY LARGE OBJECT, CHARACTER LARGE
OBJECT, and NATIONAL CHARACTER LARGE OBJECT as synonyms
for BLOB, CLOB, and NCLOB, respectively

The ability to specify an upper bound on the length of a
LOB or CLOB
Concatenation of BLOBs

Oracle fully supports the following element of this feature:
■

TRIM function on a CLOB argument

Oracle provides equivalent functionality for the following
elements of this feature:
■
■

BLOB and CLOB substring, supported using SUBSTR
SIMILAR predicate, supported using REGEXPR_LIKE to
perform pattern matching with a Perl-like syntax

The following elements of this feature are not supported:
■

Comparison predicates with BLOB or CLOB operands

■

CAST with a BLOB or CLOB operand

■

■

OVERLAY (This may be emulated using SUBSTR and string
concatenation.)
LIKE predicate with BLOB or CLOB operands

T051, Row types

Oracle object types can be used in place of the standard's row
types.

T061, UCS support

Oracle provides equivalent functionality for the following
elements of this feature:
■

■

Oracle supports the keyword CHAR instead of CHARACTERS,
and BYTE instead of OCTETS, in a character data type
declaration.
The Oracle COMPOSE function is equivalent to the standard's
NORMALIZE function.

Oracle does not support the IS NORMALIZED predicate.
T071, BIGINT data type

On many implementations, BIGINT refers to a binary integer
type with 64 bits, which supports almost 19 decimal digits. The
Oracle NUMBER type supports 39 decimal digits.

T111, Updatable joins,
unions and columns

Oracle's updatable join views are a subset of the standard's
updatable join capabilities.

C-16 Oracle Database SQL Language Reference

Oracle Support for Optional Features of SQL/Foundation:2008

Table C–2 (Cont.) Oracle Support for Optional Features of SQL/Foundation:2008
Feature ID, Feature

Support

T121, WITH (excluding
RECURSIVE) in query
expression

Oracle fully supports this feature.

T122, WITH (excluding
RECURSIVE) in subquery

Oracle fully supports this feature.

T131, Recursive query

Oracle supports the use of a WITH clause element that references
itself, but without the RECURSIVE keyword. Alternatively,
Oracle's START WITH and CONNECT BY clauses can be used to
perform many recursive queries.

T132, Recursive query in
subquery

Oracle supports the use of a WITH clause element that references
itself, but without the RECURSIVE keyword. Alternatively,
Oracle's START WITH and CONNECT BY clauses can be used to
perform many recursive queries.

T141, SIMILAR predicate

Oracle provides REGEXP_LIKE for pattern matching with a
Perl-like syntax.

T172, AS subquery clause in
table definition

Oracle's AS subquery feature of CREATE TABLE has substantially
the same functionality as the standard, though there are some
syntactic differences.

T175, Generated columns

Oracle supports this feature, with the following restrictions:
■
■

Generated columns are not supported in temporary tables.
The data type of a generated column may not be LOB or
XML.

T176, Sequence generator
support

Oracle's sequences have the same capabilities as the standard's,
though with different syntax.

T201, Comparable data
types for referential
constraints

Oracle fully supports this feature.

Oracle and Standard SQL

C-17

Oracle Support for Optional Features of SQL/Foundation:2008

Table C–2 (Cont.) Oracle Support for Optional Features of SQL/Foundation:2008
Feature ID, Feature

Support

T211, Basic trigger capability Oracle's triggers differ from the standard as follows:
■

■

■

■

■

■

■

■

Oracle does not provide the optional syntax FOR EACH
STATEMENT for the default case, the statement trigger.
Oracle does not support OLD TABLE and NEW TABLE; the
transition tables specified in the standard (the multiset of
before and after images of affected rows) are not available.
The trigger body is written in PL/SQL, which is
functionally equivalent to the standard's procedural
language PSM, but not the same.
In the trigger body, the new and old transition variables are
referenced beginning with a colon.
Oracle's row triggers are executed as the row is processed,
instead of buffering them and executing all of them after
processing all rows. The standard's semantics are
deterministic, but Oracle's in-flight row triggers are more
performant.
Oracle's before-row and before-statement triggers can
perform DML statements, which is forbidden in the
standard. However, Oracle's after-row statements cannot
perform DML, while it is permitted in the standard.
When multiple triggers apply, the standard says they are
executed in order of definition. In Oracle the execution
order is nondeterministic.
Oracle uses the system privileges CREATE TRIGGER and
CREATE ANY TRIGGER to regulate creation of triggers, instead
of the standard's TRIGGER privilege, which is a table
privilege.

T212, Enhanced trigger
capability

This feature permits statements triggers, which Oracle supports,
as described for feature T211, Basic trigger capability.

T213, INSTEAD OF triggers

Oracle supports INSTEAD OF triggers on views, with syntax and
semantics agreeing with the standard except as noted for feature
211, Basic trigger capability.

T241, START TRANSACTION
statement

Oracle's SET TRANSACTION statement starts a transaction making
it equivalent to the standard's START TRANSACTION rather than the
standard's SET TRANSACTION. Oracle's READ ONLY transactions are
at SERIALIZABLE isolation level.

T271, Savepoints

Oracle supports this feature, except:
■

Oracle does not support RELEASE SAVEPOINT.

■

Oracle does not support savepoint levels.

T285, Enhanced derived
column names

This feature pertains only to derived columns in a SELECT list
with no column alias and consisting of a SQL parameter
reference. In that case, the column name defaults to the
parameter name, the same as in the standard.

T322, Overloading of
SQL-invoked functions and
procedures

Oracle supports overloading of functions and procedures.
However, the rules for handling certain data type combinations
are not the same as the standard. For example, the standard
permits the coexistence of two functions of the same name
differing only in the numeric types of the arguments, whereas
Oracle does not permit this.

T323, Explicit security for
external routines

The Oracle syntax AUTHID { CURRENT USER | DEFINER } when used
when creating an external function, procedure, or package is
equivalent to the standard's EXTERNAL SECURITY { DEFINER |
INVOKER }.

C-18 Oracle Database SQL Language Reference

Oracle Support for Optional Features of SQL/Foundation:2008

Table C–2 (Cont.) Oracle Support for Optional Features of SQL/Foundation:2008
Feature ID, Feature

Support

T324, Explicit security for
SQL routines

Oracle's syntax AUTHID { CURRENT USER | DEFINER } when used
when creating a PL/SQL function, procedure, or package is
equivalent to the standard's SQL SECURITY { DEFINER | INVOKER }.

T325, Qualified SQL
parameter reference

PL/SQL supports the use of a routine name to qualify a
parameter name.

T326, Table functions

Oracle provides equivalents for the following elements of this
feature:
■

■

■

■

 is supported using
CAST (MULTISET () AS )

 is supported using the TABLE
operator in the FROM clause with a varray or nested table as
the argument
 is equivalent to an Oracle
expression resulting in a varray or nested table
 is equivalent to a PL/SQL function
that returns a nested table

T331, Basic roles

Oracle supports this feature, except for REVOKE ADMIN OPTION FOR
.

T351, Bracketed comments

Oracle fully supports this feature.

T431, Extended grouping
capabilities

Oracle fully supports this feature.

T432, Nested and
Oracle supports concatenated GROUPING SETS, but not nested
concatenated GROUPING SETS GROUPING SETS.
T433, Multiargument
function GROUPING

The Oracle GROUP_ID function can be used to conveniently
distinguish groups in a grouped query, serving the same
purpose as the standard multiargument GROUPING function.

T441, ABS and MOD functions

Oracle supports the ABS function. Oracle's MOD function is similar
to the standard, though the behavior is different if the two
arguments are of opposite sign.

T471, Result sets return
value

PL/SQL ref cursors provide all the functionality of the
standard's result set cursors.

T491, LATERAL derived
tables

The Oracle TABLE operator in the FROM clause is equivalent to the
LATERAL operator in the standard, but only for collection
expressions.

T501, Enhanced EXISTS
predicate

Oracle fully supports this feature.

T571, Array-returning
external SQL-invoked
function

Oracle table functions returning a varray can be defined in
external programming languages. When declaring such
functions in SQL, use the CREATE FUNCTION command with the
PIPELINED USING clause.

T572, Multiset-returning
external SQL-invoked
function

Oracle table functions returning a nested table can be defined in
external programming languages. When declaring such
functions in SQL, use the CREATE FUNCTION command with the
PIPELINED USING clause. In the body of the function, use the
OCITable interface. The function must be invoked within the
TABLE operator in the FROM clause.

T581, Regular expressions
substring functions

Oracle provides the REGEXP_SUBSTR function to perform
substring operations using regular expression matching.

Oracle and Standard SQL

C-19

Oracle Compliance with SQL/CLI:2008

Table C–2 (Cont.) Oracle Support for Optional Features of SQL/Foundation:2008
Feature ID, Feature

Support

T591, UNIQUE constraints of
possibly null columns

Oracle permits a UNIQUE constraint on one or more nullable
columns. If the UNIQUE constraint is on a single column, then the
semantics are the same as the standard (the constraint permits
any number of rows that are null in the designated column). If
the UNIQUE constraint is on two or more columns, then the
semantics are nonstandard. Oracle permits any number of rows
that are null in all the designated columns. Unlike the standard,
if a row is non-null in at least one of the designated columns,
then another row having the same values in the non-null
columns of the constraint is a constraint violation and not
permitted.

T611, Elementary OLAP
operations

Oracle fully supports this feature.

T612, Advanced OLAP
operations

Oracle supports the following elements of this feature: PERCENT_
RANK, CUME_DIST, WIDTH_BUCKET, hypothetical set functions,
PERCENTILE_CONT, and PERCENTILE_DISC.
Oracle does not support the following elements of this feature:
■

window names

■

ROW_NUMBER without an ORDER BY clause

T613, Sampling

Oracle uses the keyword SAMPLE instead of the standard's
keyword, TABLESAMPLE. Oracle uses the keyword BLOCK instead
of the standard's keyword, SYSTEM. Oracle uses the absence of
the keyword BLOCK to indicate a Bernoulli sampling of rows,
indicated in the standard by the keyword BERNOULLI.

T621, Enhanced numeric
functions

Oracle fully supports this feature, except for the alternate
spelling CEILING of the CEIL function.

T641, Multiple column
assignment

The standard syntax to assign to multiple columns is supported
if the assignment source is a subquery.

T652, SQL-dynamic
statements in SQL routines.

PL/SQL supports dynamic SQL.

T654, SQL-dynamic
statements in external
routines

Oracle supports dynamic SQL in embedded C, which may be
used to create an external routine.

T655, Cyclically dependent
routines

PL/SQL supports recursion.

Oracle Compliance with SQL/CLI:2008
The Oracle ODBC driver conforms to SQL/CLI:2008.

Oracle Compliance with SQL/PSM:2008
Oracle PL/SQL provides functionality equivalent to SQL/PSM:2008, with minor
syntactic differences, such as the spelling or arrangement of keywords.

Oracle Compliance with SQL/MED:2008
Oracle does not comply with SQL/MED:2008.

C-20 Oracle Database SQL Language Reference

Oracle Compliance with SQL/XML:2008

Oracle Compliance with SQL/OLB:2008
Oracle SQLJ conforms to SQL/OLB:1999 and not yet to SQL/OLB 2008.

Oracle Compliance with SQL/JRT:2008
Oracle fully supports stored routines and SQL types implemented in Java(TM). Oracle
provides equivalent support for the creation and maintenance of such types and
procedures. Oracle's capabilities are in general a superset of the functionality defined
by the standard.

Oracle Compliance with SQL/XML:2008
The XML data type in the standard is XML. The Oracle equivalent data type is XMLType.
A feature of the standard is considered to be fully supported if the only difference
between Oracle and the standard is the spelling of the data type name.
Table C–3 describes Oracle's support for the features of SQL/XML:2008.
Table C–3

Oracle Support for Features of SQL/XML:2008

Feature ID, Feature

Support

X010, XML type

Oracle fully supports this feature.

X011, Arrays of XML types

Oracle supports this feature using named array types (in the
standard, array types are anonymous)

X012, Multisets of XML type

The Oracle equivalent of a multiset of XML type is a nested
table with a single column of XML type.

X013, Distinct types of XML

A distinct type can be emulated using an object type with a
single attribute.

X014, Attributes of XML type

In Oracle, attributes of object types may be of type XMLType, but
the syntax for creating object types is nonstandard.

X016, Persistent XML values

Oracle fully supports this feature.

X020, XML Concatenation

Oracle fully supports this feature.

X025, XMLCast

Oracle provides equivalents for the following elements of this
feature:
■

■

To cast from XML to a scalar type, use EXTRACTVALUE. If the
XML value is typed, then the result is in the nearest analog
to the XML type, otherwise the result type is
VARCHAR(4000). Use CAST to convert to any other scalar
type.
To cast from a scalar type to XML, pass the scalar value in
to XMLQuery and insert it in a document constructor.

Since Oracle has only one XML type, there is no need to cast
from XML to XML.
X031, XMLElement

Oracle fully supports this feature.

X032, XMLForest

Oracle fully supports this feature.

X034, XMLAgg

Oracle fully supports this feature.

X035, XMLAgg: ORDER BY option Oracle fully supports this feature.
X036, XMLComment

Oracle fully supports this feature.

X036, XMLPi

Oracle fully supports this feature.

Oracle and Standard SQL

C-21

Oracle Compliance with SQL/XML:2008

Table C–3 (Cont.) Oracle Support for Features of SQL/XML:2008
Feature ID, Feature

Support

X038, XMLText

The Oracle XMLCData function may be used to create a text
node.

X040, Basic table mapping

Oracle table mappings are available through a Java interface
and through a package. Oracle table mappings have been
generalized to map queries and not just tables. To map only a
table: SELECT * FROM table_name. This provides support for the
following elements of this feature:
■

X041, Basic table mapping: null absent

■

X042, Basic table mapping: null as nil

■

X043, Basic table mapping: table as forest

■

X044, Basic table mapping: table as element

■

X045, Basic table mapping: with target namespace

■

X046, Basic table mapping: data mapping

■

X047, Basic table mapping: metadata mapping

■

X049, Basic table mapping: hex encoding

Oracle does not support the following element of this feature:
■

X048, Basic table mapping: base64 encoding

X060, XMLParse: Character
string input and CONTENT
option

Oracle does not support the {PRESERVE | STRIP} WHITESPACE
syntax. The behavior is always STRIP WHITESPACE.

X061, XMLParse: Character
string input and DOCUMENT
option

Oracle does not support the {PRESERVE | STRIP} WHITESPACE
syntax. The behavior is always STRIP WHITESPACE.

X070, XMLSerialize:
Character string serialization
and CONTENT option

Oracle fully supports this feature.

X071, XMLSerialize:
Character string serialization
and DOCUMENT option

Oracle fully supports this feature.

X072, XMLSerialize:
Character string serialization

Oracle fully supports this feature.

X076, XMLSerialize: VERSION
option

Use XMLRoot to set the XML version prior to serialization.

X080, Namespaces in XML
publishing

In the Oracle implementation of XMLElement, XMLAttributes
are used to define namespaces (XMLNamespaces is not
implemented). However, XMLAttributes is not supported for
XMLForest.

X086, XML namespace
declarations in XMLTable

Oracle fully supports this feature.

X090, XML document
predicate

In Oracle, you can test whether an XML value is a document by
using the ISFRAGMENT method.

X096, XMLExists

Use EXISTSNODE to evaluate an XPath, returning 1 if a node is
found, 0 if not. XQuery expressions other than XPath
expressions are not supported. Also, Oracle supports XPath 1.0
expressions (not XPath 2,0, which is a subset of XQuery).

X120, XML parameters in
SQL routines

Oracle fully supports this feature.

C-22 Oracle Database SQL Language Reference

Oracle Compliance with SQL/XML:2008

Table C–3 (Cont.) Oracle Support for Features of SQL/XML:2008
Feature ID, Feature

Support

X121, XML parameters in
external routines

Oracle supports XML values passed to external routines using
a non-standard interface.

X141, IS VALID predicate: data The XMLISVALID method is equivalent to the IS VALID
drive case
predicate, and supports the data-driven case.
X142, IS VALID predicate:
ACCORDING TO clause

The XMLISVALID method is equivalent to the IS VALID
predicate, and includes the equivalent of the ACCORDING TO
clause.

X143, IS VALID predicate:
ELEMENT clause

The XMLISVALID method is equivalent to the IS VALID
predicate, and includes the equivalent of the ELEMENT clause.

X144, IS VALID predicate:
schema location

The XMLISVALID method is equivalent to the IS VALID
predicate, and supports the specification of a schema location
for a registered XML Schema.

X145, IS VALID predicate
outside check constraints

The XMLISVALID method is equivalent to the IS VALID
predicate, and may be used outside check constraints.

X151, IS VALID predicate with The XMLISVALID method is equivalent to the IS VALID
predicate, and performs validation equivalent to the DOCUMENT
DOCUMENT option
clause. (XMLISVALID does not support "content" validation.)
X156, IS VALID predicate:
optional NAMESPACE with
ELEMENT clause

The XMLISVALID method is equivalent to the IS VALID
predicate, and may be used to validate against an element in
any namespace.

X157, IS VALID predicate: NO
NAMESPACE with ELEMENT
clause

The XMLISVALID method is equivalent to the IS VALID
predicate, and may be used to validate against an element in
the "no name" namespace.

X160, Basic Information
Schema for registered XML
Schemas

The Oracle static data dictionary view ALL_XML_SCHEMAS
provides a list of the registered XML schemas that are
accessible to the current user. The ALL_XML_SCHEMAS.SCHEMA_
URL column corresponds to the standard XML_SCHEMAS.XML_
SCHEMA_LOCATION column. The target namespace of the
registered XML Schemas can be learned by examining ALL_
XML_SCHEMAS.SCHEMA. Oracle has no equivalents for the other
columns of the standard's XML_SCHEMAS.

X161, Advanced Information
Schema for registered XML
Schemas

Oracle does not have static data dictionary views
corresponding to XML_SCHEMA_NAMESPACES and XML_SCHEMA_
ELEMENTS in the standard. However, all the information about
registered XML Schemas may be learned by examining the
actual XML Schema, which is found in the ALL_XML_
SCHEMAS.SCHEMA column. This may also be examined to learn
whether a registered XML Schema is nondeterministic, and
which of its namespaces and elements are nondeterministic.

X191, XML(DOCUMENT
(XMLSCHEMA)) type

Oracle does not support this syntax. However, a column of a
table can be constrained by a registered XML Schema, in which
case all values of the column will be of
XML(DOCUMENT(XMLSCHEMA)) type.

Oracle and Standard SQL

C-23

Oracle Compliance with SQL/XML:2008

Table C–3 (Cont.) Oracle Support for Features of SQL/XML:2008
Feature ID, Feature

Support

X200, XMLQuery

Oracle fully supports the following elements of this feature:
■

X201, XMLQuery: RETURNING CONTENT

■

X203, XMLQuery: passing a context item

■

X204, XMLQuery: initializing an XQuery variable

Oracle does not support the following elements of this feature:
■

X202, XMLQuery: RETURNING SEQUENCE

■

{ NULL | EMPTY } ON EMPTY syntax

■

Mandatory BY { REF | VALUE } in the PASSING clause (Oracle
supports only value semantics)

X221, XML passing
mechanism BY VALUE

Oracle supports only value semantics, but does not support the
explicit BY VALUE clause.

X232, XML(CONTENT(ANY)) type

Oracle does not support this syntax as a type modifier, but the
Oracle XMLType supports this data type for transient values.
Persistent values are of type XML(DOCUMENT(ANY)), which is a
subset of XML(CONTENT(ANY)).

X241, RETURNING CONTENT in
XML publishing

Oracle does not support this syntax. In Oracle, the behavior of
the publishing functions (XMLAgg, XMLComment, XMLConcat,
XMLElement, XMLForest, and XMLPi) is always RETURNING
CONTENT.

X251, Persistent XML values
of XML(DOCUMENT(UNTYPED))
type

Oracle fully supports this feature.

X252, Persistent values of
type XML(DOCUMENT(ANY))

Oracle fully supports this feature.

X256, Persistent values of
XML(DOCUMENT(XMLSCHEMA))
type

Oracle fully supports this feature.

X260, XML type, ELEMENT
clause

Oracle does not support this syntax. However, a column of a
table may be constrained by a top-level element in a registered
XML Schema.

X262, XML type, optional
NAMESPACE with ELEMENT
clause

Oracle does not support this syntax. However, a column of a
table may be constrained by a top-level element in a
namespace other than the target namespace of a registered
XML Schema.

X263, XML type: NO
NAMESPACE with ELEMENT
clause

Oracle does not support this syntax. However, a column of a
table may be constrained by a top-level element in the "no
name" namespace of a registered XML Schema.

X264, XML type: schema
location

Oracle does not support this syntax. However, a column of a
table may be constrained by a registered XML Schema that is
identified by a schema location.

X271, XMLValidate: data
driven case

The SCHEMAVALIDATE method is equivalent to XMLValidate, and
supports the data-driven case.

X272, XMLValidate:
ACCORDING TO clause

The SCHEMAVALIDATE method is equivalent to XMLValidate, and
may be used to specify a particular registered XML Schema.

X273, XMLValidate: ELEMENT
clause

The SCHEMAVALIDATE method is equivalent to XMLValidate, and
may be used to specify a particular element of a particular
registered XML Schema.

C-24 Oracle Database SQL Language Reference

Oracle Compliance with FIPS 127-2

Table C–3 (Cont.) Oracle Support for Features of SQL/XML:2008
Feature ID, Feature

Support

X274, XMLValidate: schema
location

The SCHEMAVALIDATE method is equivalent to XMLValidate, and
may be used to specify a particular registered XML Schema by
its schema location URL.

X281, XMLValidate with
DOCUMENT option

The SCHEMAVALIDATE method is equivalent to XMLValidate.
SCHEMAVALIDATE performs validation only of XML documents
(not content).

X285, XMLValidate: optional
NAMESPACE with ELEMENT
clause

The SCHEMAVALIDATE method is equivalent to XMLValidate, and
may be used to specify a particular element in a namespace
other than the target namespace of a particular registered XML
Schema.

X286, XMLValidate: NO
NAMESPACE with ELEMENT
clause

The SCHEMAVALIDATE method is equivalent to XMLValidate, and
may be used to specify a particular element in the "no name"
namespace of a particular registered XML Schema.

X300, XMLTable

Oracle does not support reverse axes in the column path
expressions. Aside from that restriction, Oracle fully supports
the following elements of this feature:
■

X086, XML namespace declarations in XMLTable

■

X302, XMLTable with ordinality column

■

X303, XMLTable: column default option

■

X304, XMLTable: passing a context item

■

X305, XMLTable: initializing an XQuery variable

Oracle does not support the following elements of this feature:
■
■

X301, XMLTable: derived column list option
Mandatory BY {REF | VALUE} in the PASSING clause. Oracle
supports only BY VALUE semantics currently.

Oracle Compliance with FIPS 127-2
Oracle complied fully with last Federal Information Processing Standard (FIPS), which
was FIPS PUB 127-2. That standard is no longer published. However, for users whose
applications depend on information about the sizes of some database constructs that
were defined in FIPS 127-2, the details of our compliance are listed in Table C–4.
Table C–4

Sizing for Database Constructs

Database Constructs

FIPS

Oracle Database

Length of an identifier (in bytes)

18

30

Length of CHARACTER data type (in bytes)

240

2,000

Decimal precision of NUMERIC data type

15

38

Decimal precision of DECIMAL data type

15

38

Decimal precision of INTEGER data type

9

38

Decimal precision of SMALLINT data type

4

38

Binary precision of FLOAT data type

20

126

Binary precision of REAL data type

20

63

Binary precision of DOUBLE PRECISION data type

30

126

Columns in a table

100

1,000

Oracle and Standard SQL

C-25

Oracle Extensions to Standard SQL

Table C–4 (Cont.) Sizing for Database Constructs
Database Constructs

FIPS

Oracle Database

Values in an INSERT statement

100

1,000

SET clauses in an UPDATE statement (Note 1)

20

1,000

Length of a row (Note2, Note 3)

2,000

2,000,000

Columns in a UNIQUE constraint

6

32

Length of a UNIQUE constraint (Note 2)

120

(Note 4)

Length of foreign key column list (Note 2)

120

(Note 4)

Columns in a GROUP BY clause

6

255 (Note 5)

Length of GROUP BY column list

120

(Note 5)

Sort specifications in ORDER BY clause

6

255 (Note 5)

Length of ORDER BY column list

120

(Note 5)

Columns in a referential integrity constraint

6

32

Tables referenced in a SQL statement

15

No limit

Cursors simultaneously open

10

(Note 6)

Items in a SELECT list

100

1,000

Note 1: The number of SET clauses in an UPDATE statement refers to the number items
separated by commas following the SET keyword.
Note 2: The FIPS PUB defines the length of a collection of columns to be the sum of:
twice the number of columns, the length of each character column in bytes, decimal
precision plus 1 of each exact numeric column, binary precision divided by 4 plus 1 of
each approximate numeric column.
Note 3: The Oracle limit for the maximum row length is based on the maximum length
of a row containing a LONG value of length 2 gigabytes and 999 VARCHAR2 values, each
of length 4000 bytes: 2(254) + 231 + (999(4000)).
Note 4: The Oracle limit for a UNIQUE key is half the size of an Oracle data block
(specified by the initialization parameter DB_BLOCK_SIZE) minus some overhead.
Note 5: Oracle places no limit on the number of columns in a GROUP BY clause or the
number of sort specifications in an ORDER BY clause. However, the sum of the sizes of
all the expressions in either a GROUP BY clause or an ORDER BY clause is limited to the
size of an Oracle data block (specified by the initialization parameter DB_BLOCK_SIZE)
minus some overhead.
Note 6: The Oracle limit for the number of cursors simultaneously opened is specified
by the initialization parameter OPEN_CURSORS. The maximum value of this parameter
depends on the memory available on your operating system and exceeds 100 in all
cases.

Oracle Extensions to Standard SQL
Oracle supports numerous features that extend beyond standard SQL. If you are
concerned with the portability of your applications to other implementations of SQL,
then use Oracle's FIPS Flagger to help identify the use of Oracle extensions to Entry
SQL-92 in your embedded SQL programs. The FIPS Flagger is part of the Oracle
precompilers and the SQL*Module compiler. The FIPS Flagger can also be enabled in
SQL*Plus by using ALTER SESSION SET FLAGGER = ENTRY. While SQL-92 has been
C-26 Oracle Database SQL Language Reference

Character Set Support

superseded by SQL:2008, there has been no conformance testing authority for any
version of SQL since SQL-92; hence, Entry SQL-92 offers you the most assurance of
portability.
Pro*COBOL Programmer's Guide and Pro*C/C++
Programmer's Guide for information on how to use the FIPS Flagger

See Also:

Oracle Compliance with Older Standards
This release of Oracle Database conforms to SQL:2008, the most recent edition of the
SQL standard, as itemized in preceding sections of this appendix. Oracle does not
formally claim that this release of the database conforms to SQL-92—and in particular,
to SQL-92 Entry Level—or to SQL:1999, because those standards have been
superseded by SQL:2008. Some, mostly minor, changes between editions of the SQL
standard—both between SQL-92 and SQL:1999 and between SQL:1999 and
SQL:2008—might affect applications. The SQL standard, or a reference discussing that
standard, can be consulted to determine the details of any incompatibilities that have
been introduced. One important source is Annex E of SQL/Foundation:1999 and
SQL/Foundation:2008.
In some cases, this release of Oracle Database might continue to recognize constructs
from older editions of SQL. Such recognition is often allowed as a valid vendor
extension. It is the general policy of Oracle to keep incompatibilities between versions
of the database as few as possible. This policy extends to retention of older forms
when that is feasible. In any case, the differences between older SQL and SQL:2008 (as
noted above) are relatively inconsequential.

Character Set Support
Oracle supports most national, international, and vendor-specific encoded character
set standards. A complete list of character sets supported by Oracle appears in Oracle
Database Globalization Support Guide.
Unicode is a universal encoded character set that lets you store information from any
language using a single character set. Unicode is required by modern standards such
as XML, Java, JavaScript, and LDAP. Unicode is compliant with ISO/IEC standard
10646. You can obtain a copy of ISO/IEC standard 10646 from this address:
International Organization for Standardization
1 Rue de Varembé
Case postale 56
CH-1211, Geneva 20, Switzerland
Phone: +41.22.749.0111
Fax: +41.22.733.3430
Web site: http://www.iso.ch/
Oracle Database complies fully with Unicode 4.0, the fourth and most recent version of
the Unicode standard. For up-to-date information on this standard, visit the Web site
of the Unicode Consortium:
http://www.unicode.org

Oracle uses UTF-8 (8-bit) encoding by way of three database character sets, two for
ASCII-based platforms (UTF8 and AL32UTF8) and one for EBCDIC platforms (UTFE).
If you prefer to implement Unicode support incrementally, then you can store Unicode
data in either the UTF-16 or UTF-8 encoding form, in the national character set, for the
SQL NCHAR data types (NCHAR, NVARCHAR2, and NCLOB).

Oracle and Standard SQL

C-27

Character Set Support

See Also: Oracle Database Globalization Support Guide for details on
Oracle character set support

C-28 Oracle Database SQL Language Reference

D
D

Oracle Regular Expression Support

Oracle's implementation of regular expressions conforms with the IEEE Portable
Operating System Interface (POSIX) regular expression standard and to the Unicode
Regular Expression Guidelines of the Unicode Consortium.
This appendix contains the following sections:
■

Multilingual Regular Expression Syntax

■

Regular Expression Operator Multilingual Enhancements

■

Perl-influenced Extensions in Oracle Regular Expressions

Multilingual Regular Expression Syntax
Table D–1 lists the full set of operators defined in the POSIX standard Extended
Regular Expression (ERE) syntax. Oracle follows the exact syntax and matching
semantics for these operators as defined in the POSIX standard for matching ASCII
(English language) data. For more complete descriptions of the operators, examples of
their use, and Oracle multilingual enhancements of the operators, refer to Oracle
Database Advanced Application Developer's Guide. Notes following the table provide
more complete descriptions of the operators and their functions, as well as Oracle
multilingual enhancements of the operators. Table D–2 summarizes Oracle support for
and multilingual enhancement of the POSIX operators.
Table D–1

Regular Expression Operators and Metasymbols

Operator

Description

\

The backslash character can have four different meanings depending on
the context. It can:
■

Stand for itself

■

Quote the next character

■

Introduce an operator

■

Do nothing

*

Matches zero or more occurrences

+

Matches one or more occurrences

?

Matches zero or one occurrence

|

Alternation operator for specifying alternative matches

^

Matches the beginning of a string by default. In multiline mode, it matches
the beginning of any line anywhere within the source string.

Oracle Regular Expression Support

D-1

Regular Expression Operator Multilingual Enhancements

Table D–1 (Cont.) Regular Expression Operators and Metasymbols
Operator

Description

$

Matches the end of a string by default. In multiline mode, it matches the
end of any line anywhere within the source string.

.

Matches any character in the supported character set except NULL

[]

Bracket expression for specifying a matching list that should match any
one of the expressions represented in the list. A non-matching list
expression begins with a circumflex (^) and specifies a list that matches
any character except for the expressions represented in the list.
To specify a right bracket (]) in the bracket expression, place it first in the
list (after the initial circumflex (^), if any).
To specify a hyphen in the bracket expression, place it first in the list (after
the initial circumflex (^), if any), last in the list, or as an ending range point
in a range expression.

()

Grouping expression, treated as a single subexpression

{m}

Matches exactly m times

{m,}

Matches at least m times

{m,n}

Matches at least m times but no more than n times

\n

The backreference expression (n is a digit between 1 and 9) matches the nth
subexpression enclosed between '(' and ')' preceding the \n

[..]

Specifies one collation element, and can be a multicharacter element (for
example, [.ch.] in Spanish)

[: :]

Specifies character classes (for example, [:alpha:]). It matches any character
within the character class.

[==]

Specifies equivalence classes. For example, [=a=] matches all characters
having base letter 'a'.

Regular Expression Operator Multilingual Enhancements
When applied to multilingual data, Oracle's implementation of the POSIX operators
extends beyond the matching capabilities specified in the POSIX standard. Table D–2
shows the relationship of the operators in the context of the POSIX standard.
■
■

■

The first column lists the supported operators.
The second and third columns indicate whether the POSIX standard (Basic
Regular Expression—BRE and Extended Regular Expression—ERE, respectively)
defines the operator
The fourth column indicates whether Oracle's implementation extends the
operator's semantics for handling multilingual data.

Oracle lets you enter multibyte characters directly, if you have a direct input method,
or you can use functions to compose the multibyte characters. You cannot use the
Unicode hexadecimal encoding value of the form '\xxxx'. Oracle evaluates the
characters based on the byte values used to encode the character, not the graphical
representation of the character. All accented characters are considered word characters.
Table D–2

POSIX and Multilingual Operator Relationships

Operator

POSIX BRE syntax

POSIX ERE Syntax

Multilingual
Enhancement

\

Yes

Yes

—

D-2 Oracle Database SQL Language Reference

Perl-influenced Extensions in Oracle Regular Expressions

Table D–2 (Cont.) POSIX and Multilingual Operator Relationships
Operator

POSIX BRE syntax

POSIX ERE Syntax

Multilingual
Enhancement

*

Yes

Yes

—

+

--

Yes

—

?

—

Yes

—

|

—

Yes

—

^

Yes

Yes

Yes

$

Yes

Yes

Yes

.

Yes

Yes

Yes

[]

Yes

Yes

Yes

()

Yes

Yes

—

{m}

Yes

Yes

—

{m,}

Yes

Yes

—

{m,n}

Yes

Yes

—

\n

Yes

Yes

Yes

[..]

Yes

Yes

Yes

[::]

Yes

Yes

Yes

[==]

Yes

Yes

Yes

Perl-influenced Extensions in Oracle Regular Expressions
Oracle Database regular expression functions and conditions accept a number of
Perl-influenced operators that are in common use, although not part of the POSIX
standard. Table D–3 lists those operators. For more complete descriptions with
examples, refer to Oracle Database Advanced Application Developer's Guide.
Table D–3

Perl-influenced Operators in Oracle Regular Expressions

Operator

Description

\d

A digit character.

\D

A nondigit character.

\w

A word character.

\W

A nonword character.

\s

A whitespace character.

\S

A non-whitespace character.

\A

Matches only at the beginning of a string, or before a newline
character at the end of a string.

\Z

Matches only at the end of a string.

*?

Matches the preceding pattern element 0 or more times
(nongreedy).

Oracle Regular Expression Support

D-3

Perl-influenced Extensions in Oracle Regular Expressions

Table D–3 (Cont.) Perl-influenced Operators in Oracle Regular Expressions
Operator

Description

+?

Matches the preceding pattern element 1 or more times
(nongreedy).

??

Matches the preceding pattern element 0 or 1 time (nongreedy).

{n}?

Matches the preceding pattern element exactly n times
(nongreedy).

{n,}?

Matches the preceding pattern element at least n times
(nongreedy).

{n,m}?

Matches the preceding pattern element at least n but not more
than m times (nongreedy).

D-4 Oracle Database SQL Language Reference

E
E

Oracle SQL Reserved Words and Keywords
This appendix contains the following sections:
■

Oracle SQL Reserved Words

■

Oracle SQL Keywords

Oracle SQL Reserved Words
This section lists Oracle SQL reserved words. You cannot use Oracle SQL reserved
words as nonquoted identifiers. Quoted identifiers can be reserved words, although
this is not recommended.
In addition to the following reserved words, Oracle uses
system-generated names beginning with "SYS_" for implicitly
generated schema objects and subobjects. Oracle discourages you
from using this prefix in the names you explicitly provide to your
schema objects and subobjects to avoid possible conflict in name
resolution.

Note:

The V$RESERVED_WORDS data dictionary view provides additional information on each
reserved word, including whether it is always reserved or is reserved only for
particular uses. Refer to Oracle Database Reference for more information.
Words followed by an asterisk (*) are also ANSI reserved words.
ACCESS
ADD
ALL *
ALTER *
AND *
ANY *
AS *
ASC
AUDIT
BETWEEN *
BY *
CHAR *
CHECK *
CLUSTER
COLUMN *
COLUMN_VALUE (See Note 1 at the end of this list)
COMMENT

Oracle SQL Reserved Words and Keywords E-1

Oracle SQL Reserved Words

COMPRESS
CONNECT *
CREATE *
CURRENT *
DATE *
DECIMAL *
DEFAULT *
DELETE *
DESC
DISTINCT *
DROP *
ELSE *
EXCLUSIVE
EXISTS *
FILE
FLOAT *
FOR *
FROM *
GRANT *
GROUP *
HAVING *
IDENTIFIED
IMMEDIATE
IN *
INCREMENT
INDEX
INITIAL
INSERT *
INTEGER *
INTERSECT *
INTO *
IS *
LEVEL
LIKE *
LOCK
LONG
MAXEXTENTS
MINUS
MLSLABEL
MODE
MODIFY
NESTED_TABLE_ID (See Note 1 at the end of this list)
NOAUDIT
NOCOMPRESS
NOT *
NOWAIT
NULL *
NUMBER
OF *
OFFLINE
ON *
ONLINE
OPTION
OR *
ORDER *

E-2 Oracle Database SQL Language Reference

Oracle SQL Keywords

PCTFREE
PRIOR
PRIVILEGES
PUBLIC
RAW
RENAME
RESOURCE
REVOKE *
ROW *
ROWID (See Note 2 at the end of this list)
ROWNUM
ROWS *
SELECT *
SESSION
SET *
SHARE
SIZE
SMALLINT *
START *
SUCCESSFUL
SYNONYM
SYSDATE
TABLE *
THEN *
TO *
TRIGGER *
UID
UNION *
UNIQUE *
UPDATE *
USER *
VALIDATE
VALUES *
VARCHAR *
VARCHAR2
VIEW
WHENEVER *
WHERE *
WITH *
Note 1: This keyword is only reserved for use as an attribute name.
Note 2: You cannot use the uppercase word ROWID, either quoted or nonquoted, as a
column name. However, you can use the uppercase word as a quoted identifier that is
not a column name, and you can use the word with one or more lowercase letters (for
example, "Rowid" or "rowid") as any quoted identifier, including a column name.

Oracle SQL Keywords
Oracle SQL keywords are not reserved. However, Oracle uses them internally in
specific ways. Therefore, if you use these words as names for objects and object parts,
then your SQL statements may be more difficult to read and may lead to unpredictable
results.
You can obtain a list of keywords by querying the V$RESERVED_WORDS data dictionary
view. All keywords in the view that are not listed as always reserved or reserved for a
Oracle SQL Reserved Words and Keywords E-3

Oracle SQL Keywords

specific use are Oracle SQL keywords. Refer to Oracle Database Reference for more
information.

E-4 Oracle Database SQL Language Reference

F
F

Extended Examples

The body of the SQL Language Reference contains examples for almost every reference
topic. This appendix contains lengthy examples that are not appropriate in the context
of a single SQL statement. These examples are intended to provide uninterrupted the
series of steps that you would use to take advantage of particular Oracle functionality.
They do not replace the syntax diagrams and semantics found for each individual SQL
statement in the body of the reference. Use the cross-references provided to access
additional information, such as privileges required and restrictions, as well as syntax.
This appendix contains the following sections:
■

Using Extensible Indexing

■

Using XML in SQL Statements

Using Extensible Indexing
This section provides examples of the steps entailed in a simple but realistic extensible
indexing scenario.
Suppose you want to rank the salaries in the HR.employees table and then find those
that rank between 10 and 20. You could use the DENSE_RANK function, as follows:
SELECT last_name, salary FROM
(SELECT last_name, DENSE_RANK() OVER
(ORDER BY salary DESC) rank_val, salary FROM employees)
WHERE rank_val BETWEEN 10 AND 20;

See Also:

DENSE_RANK on page 5-82

This nested query is somewhat complex, and it requires a full scan of the employees
table as well as a sort. An alternative would be to use extensible indexing to achieve
the same goal. The resulting query will be simpler. The query will require only an
index scan and a table access by rowid, and will therefore perform much more
efficiently.
The first step is to create the implementation type position_im, including method
headers for index definition, maintenance, and creation. Most of the type body uses
PL/SQL, which is shown in italics.
The type must created with the AUTHID CURRENT_USER clause because of the EXECUTE
IMMEDIATE statement inside the function ODCIINDEXCREATE(). By default that function
runs with the definer rights. When the function is called in the subsequent creation of
the domain index, the invoker does not have the same rights.

Extended Examples F-1

Using Extensible Indexing

See Also:
■

■

CREATE TYPE on page 17-3 and CREATE TYPE BODY on
page 17-5
Oracle Database Data Cartridge Developer's Guide for complete
information on the ODCI routines in this statement

CREATE OR REPLACE TYPE position_im AUTHID CURRENT_USER AS OBJECT
(
curnum NUMBER,
howmany NUMBER,
lower_bound NUMBER,
upper_bound NUMBER,
/* lower_bound and upper_bound are used for the
index-based functional implementation */
STATIC FUNCTION ODCIGETINTERFACES(ifclist OUT SYS.ODCIOBJECTLIST) RETURN NUMBER,
STATIC FUNCTION ODCIINDEXCREATE
(ia SYS.ODCIINDEXINFO, parms VARCHAR2, env SYS.ODCIEnv) RETURN NUMBER,
STATIC FUNCTION ODCIINDEXTRUNCATE (ia SYS.ODCIINDEXINFO,
env SYS.ODCIEnv) RETURN NUMBER,
STATIC FUNCTION ODCIINDEXDROP(ia SYS.ODCIINDEXINFO,
env SYS.ODCIEnv) RETURN NUMBER,
STATIC FUNCTION ODCIINDEXINSERT(ia SYS.ODCIINDEXINFO, rid ROWID,
newval NUMBER, env SYS.ODCIEnv) RETURN NUMBER,
STATIC FUNCTION ODCIINDEXDELETE(ia SYS.ODCIINDEXINFO, rid ROWID, oldval NUMBER,
env SYS.ODCIEnv) RETURN NUMBER,
STATIC FUNCTION ODCIINDEXUPDATE(ia SYS.ODCIINDEXINFO, rid ROWID, oldval NUMBER,
newval NUMBER, env SYS.ODCIEnv) RETURN NUMBER,
STATIC FUNCTION ODCIINDEXSTART(SCTX IN OUT position_im, ia SYS.ODCIINDEXINFO,
op SYS.ODCIPREDINFO, qi SYS.ODCIQUERYINFO,
strt NUMBER, stop NUMBER, lower_pos NUMBER,
upper_pos NUMBER, env SYS.ODCIEnv) RETURN NUMBER,
MEMBER FUNCTION ODCIINDEXFETCH(SELF IN OUT position_im, nrows NUMBER,
rids OUT SYS.ODCIRIDLIST, env SYS.ODCIEnv)
RETURN NUMBER,
MEMBER FUNCTION ODCIINDEXCLOSE(env SYS.ODCIEnv) RETURN NUMBER
);
/
CREATE OR REPLACE TYPE BODY position_im
IS
STATIC FUNCTION ODCIGETINTERFACES(ifclist OUT SYS.ODCIOBJECTLIST)
RETURN NUMBER IS
BEGIN
ifclist := SYS.ODCIOBJECTLIST(SYS.ODCIOBJECT('SYS','ODCIINDEX2'));
RETURN ODCICONST.SUCCESS;
END ODCIGETINTERFACES;
STATIC FUNCTION ODCIINDEXCREATE (ia SYS.ODCIINDEXINFO, parms VARCHAR2, env SYS.ODCIEnv) RETURN
NUMBER
IS
stmt
VARCHAR2(2000);
BEGIN
/* Construct the SQL statement */
stmt := 'Create Table ' || ia.INDEXSCHEMA || '.' || ia.INDEXNAME ||
'_STORAGE_TAB' || '(col_val, base_rowid, constraint pk PRIMARY KEY ' ||
'(col_val, base_rowid)) ORGANIZATION INDEX AS SELECT ' ||
ia.INDEXCOLS(1).COLNAME || ', ROWID FROM ' ||
ia.INDEXCOLS(1).TABLESCHEMA || '.' || ia.INDEXCOLS(1).TABLENAME;
EXECUTE IMMEDIATE stmt;

F-2 Oracle Database SQL Language Reference

Using Extensible Indexing

RETURN ODCICONST.SUCCESS;
END;
STATIC FUNCTION ODCIINDEXDROP(ia SYS.ODCIINDEXINFO, env SYS.ODCIEnv) RETURN NUMBER IS
stmt VARCHAR2(2000);
BEGIN
/* Construct the SQL statement */
stmt := 'DROP TABLE ' || ia.INDEXSCHEMA || '.' || ia.INDEXNAME ||
'_STORAGE_TAB';
/* Execute the statement */
EXECUTE IMMEDIATE stmt;
RETURN ODCICONST.SUCCESS;
END;
STATIC FUNCTION ODCIINDEXTRUNCATE(ia SYS.ODCIINDEXINFO, env SYS.ODCIEnv) RETURN NUMBER IS
stmt VARCHAR2(2000);
BEGIN
/* Construct the SQL statement */
stmt := 'TRUNCATE TABLE ' || ia.INDEXSCHEMA || '.' || ia.INDEXNAME || '_STORAGE_TAB';
EXECUTE IMMEDIATE stmt;
RETURN ODCICONST.SUCCESS;
END;
STATIC FUNCTION ODCIINDEXINSERT(ia SYS.ODCIINDEXINFO, rid ROWID,
newval NUMBER, env SYS.ODCIEnv) RETURN NUMBER IS
stmt VARCHAR2(2000);
BEGIN
/* Construct the SQL statement */
stmt := 'INSERT INTO ' || ia.INDEXSCHEMA || '.' || ia.INDEXNAME ||
'_STORAGE_TAB VALUES (''' || newval || ''' , ''' || rid || ''' )';
/* Execute the SQL statement */
EXECUTE IMMEDIATE stmt;
RETURN ODCICONST.SUCCESS;
END;
STATIC FUNCTION ODCIINDEXDELETE(ia SYS.ODCIINDEXINFO, rid ROWID, oldval NUMBER,
env SYS.ODCIEnv)
RETURN NUMBER IS
stmt VARCHAR2(2000);
BEGIN
/* Construct the SQL statement */
stmt := 'DELETE FROM ' || ia.INDEXSCHEMA || '.' || ia.INDEXNAME ||
'_STORAGE_TAB WHERE col_val = ''' || oldval || ''' AND base_rowid = ''' || rid || '''';
/* Execute the statement */
EXECUTE IMMEDIATE stmt;
RETURN ODCICONST.SUCCESS;
END;
STATIC FUNCTION ODCIINDEXUPDATE(ia SYS.ODCIINDEXINFO, rid ROWID, oldval NUMBER,
newval NUMBER, env SYS.ODCIEnv) RETURN NUMBER IS
stmt VARCHAR2(2000);
BEGIN
/* Construct the SQL statement */
stmt := 'UPDATE ' || ia.INDEXSCHEMA || '.' || ia.INDEXNAME ||
'_STORAGE_TAB SET col_val = ''' || newval || ''' WHERE f2 = '''|| rid ||'''';
/* Execute the statement */
EXECUTE IMMEDIATE stmt;
RETURN ODCICONST.SUCCESS;
END;
STATIC FUNCTION ODCIINDEXSTART(SCTX IN OUT position_im, ia SYS.ODCIINDEXINFO,
op SYS.ODCIPREDINFO, qi SYS.ODCIQUERYINFO,
strt NUMBER, stop NUMBER, lower_pos NUMBER,
upper_pos NUMBER, env SYS.ODCIEnv) RETURN NUMBER IS

Extended Examples F-3

Using Extensible Indexing

rid
storage_tab_name
lower_bound_stmt
upper_bound_stmt
range_query_stmt
lower_bound
upper_bound
cnum
nrows

VARCHAR2(5072);
VARCHAR2(65);
VARCHAR2(2000);
VARCHAR2(2000);
VARCHAR2(2000);
NUMBER;
NUMBER;
INTEGER;
INTEGER;

BEGIN
/* Take care of some error cases.
The only predicates in which position operator can appear are
op() = 1
OR
op() = 0
OR
op() between 0 and 1
*/
IF (((strt != 1) AND (strt != 0)) OR
((stop != 1) AND (stop != 0)) OR
((strt = 1) AND (stop = 0))) THEN
RAISE_APPLICATION_ERROR(-20101,
'incorrect predicate for position_between operator');
END IF;
IF (lower_pos > upper_pos) THEN
RAISE_APPLICATION_ERROR(-20101, 'Upper Position must be greater than or
equal to Lower Position');
END IF;
IF (lower_pos <= 0) THEN
RAISE_APPLICATION_ERROR(-20101, 'Both Positions must be greater than zero');
END IF;
storage_tab_name := ia.INDEXSCHEMA || '.' || ia.INDEXNAME ||
'_STORAGE_TAB';
upper_bound_stmt := 'Select MIN(col_val) FROM (Select /*+ INDEX_DESC(' ||
storage_tab_name || ') */ DISTINCT ' ||
'col_val FROM ' || storage_tab_name || ' ORDER BY ' ||
'col_val DESC) WHERE rownum <= ' || lower_pos;
EXECUTE IMMEDIATE upper_bound_stmt INTO upper_bound;
IF (lower_pos != upper_pos) THEN
lower_bound_stmt := 'Select MIN(col_val) FROM (Select /*+ INDEX_DESC(' ||
storage_tab_name || ') */ DISTINCT ' ||
'col_val FROM ' || storage_tab_name ||
' WHERE col_val < ' || upper_bound || ' ORDER BY ' ||
'col_val DESC) WHERE rownum <= ' ||
(upper_pos - lower_pos);
EXECUTE IMMEDIATE lower_bound_stmt INTO lower_bound;
ELSE
lower_bound := upper_bound;
END IF;
IF (lower_bound IS NULL) THEN
lower_bound := upper_bound;
END IF;
range_query_stmt := 'Select base_rowid FROM ' || storage_tab_name ||
' WHERE col_val BETWEEN ' || lower_bound || ' AND ' ||
upper_bound;
cnum := DBMS_SQL.OPEN_CURSOR;
DBMS_SQL.PARSE(cnum, range_query_stmt, DBMS_SQL.NATIVE);
/* set context as the cursor number */
SCTX := position_im(cnum, 0, 0, 0);
/* return success */
RETURN ODCICONST.SUCCESS;

F-4 Oracle Database SQL Language Reference

Using Extensible Indexing

END;
MEMBER FUNCTION ODCIINDEXFETCH(SELF IN OUT position_im, nrows NUMBER,
rids OUT SYS.ODCIRIDLIST, env SYS.ODCIEnv)
RETURN NUMBER IS
cnum
INTEGER;
rid_tab DBMS_SQL.Varchar2_table;
rlist
SYS.ODCIRIDLIST := SYS.ODCIRIDLIST();
i
INTEGER;
d
INTEGER;
BEGIN
cnum := SELF.curnum;
IF self.howmany = 0 THEN
dbms_sql.define_array(cnum, 1, rid_tab, nrows, 1);
d := DBMS_SQL.EXECUTE(cnum);
END IF;
d := DBMS_SQL.FETCH_ROWS(cnum);
IF d = nrows THEN
rlist.extend(d);
ELSE
rlist.extend(d+1);
END IF;
DBMS_SQL.COLUMN_VALUE(cnum, 1, rid_tab);
for i in 1..d loop
rlist(i) := rid_tab(i+SELF.howmany);
end loop;
SELF.howmany := SELF.howmany + d;
rids := rlist;
RETURN ODCICONST.SUCCESS;
END;
MEMBER FUNCTION ODCIINDEXCLOSE(env SYS.ODCIEnv) RETURN NUMBER IS
cnum INTEGER;
BEGIN
cnum := SELF.curnum;
DBMS_SQL.CLOSE_CURSOR(cnum);
RETURN ODCICONST.SUCCESS;
END;
END;
/

The next step is to create the functional implementation function_for_position_
between for the operator that will be associated with the indextype. (The PL/SQL
blocks are shown in parentheses.)
This function is for use with an index-based function evaluation. Therefore, it takes an
index context and scan context as parameters.
See Also:
■

■

Oracle Database Data Cartridge Developer's Guide for information on
creating index-based functional implementation
CREATE FUNCTION on page 14-58 and Oracle Database PL/SQL
Language Reference

CREATE OR REPLACE FUNCTION function_for_position_between
(col NUMBER, lower_pos NUMBER, upper_pos NUMBER,
indexctx IN SYS.ODCIIndexCtx,
scanctx IN OUT position_im,
scanflg IN NUMBER)
RETURN NUMBER AS
rid
ROWID;
storage_tab_name VARCHAR2(65);
Extended Examples F-5

Using Extensible Indexing

lower_bound_stmt VARCHAR2(2000);
upper_bound_stmt VARCHAR2(2000);
col_val_stmt
VARCHAR2(2000);
lower_bound
NUMBER;
upper_bound
NUMBER;
column_value
NUMBER;
BEGIN
IF (indexctx.IndexInfo IS NOT NULL) THEN
storage_tab_name := indexctx.IndexInfo.INDEXSCHEMA || '.' ||
indexctx.IndexInfo.INDEXNAME || '_STORAGE_TAB';
IF (scanctx IS NULL) THEN
/* This is the first call. Open a cursor for future calls.
First, do some error checking
*/
IF (lower_pos > upper_pos) THEN
RAISE_APPLICATION_ERROR(-20101,
'Upper Position must be greater than or equal to Lower Position');
END IF;
IF (lower_pos <= 0) THEN
RAISE_APPLICATION_ERROR(-20101,
'Both Positions must be greater than zero');
END IF;
/* Obtain the upper and lower value bounds for the range we're interested in.
*/
upper_bound_stmt := 'Select MIN(col_val) FROM (Select /*+ INDEX_DESC(' ||
storage_tab_name || ') */ DISTINCT ' ||
'col_val FROM ' || storage_tab_name || ' ORDER BY ' ||
'col_val DESC) WHERE rownum <= ' || lower_pos;
EXECUTE IMMEDIATE upper_bound_stmt INTO upper_bound;
IF (lower_pos != upper_pos) THEN
lower_bound_stmt := 'Select MIN(col_val) FROM (Select /*+ INDEX_DESC(' ||
storage_tab_name || ') */ DISTINCT ' ||
'col_val FROM ' || storage_tab_name ||
' WHERE col_val < ' || upper_bound || ' ORDER BY ' ||
'col_val DESC) WHERE rownum <= ' ||
(upper_pos - lower_pos);
EXECUTE IMMEDIATE lower_bound_stmt INTO lower_bound;
ELSE
lower_bound := upper_bound;
END IF;
IF (lower_bound IS NULL) THEN
lower_bound := upper_bound;
END IF;
/* Store the lower and upper bounds for future function invocations for
the positions.
*/
scanctx := position_im(0, 0, lower_bound, upper_bound);
END IF;
/* Fetch the column value corresponding to the rowid, and see if it falls
within the determined range.
*/
col_val_stmt := 'Select col_val FROM ' || storage_tab_name ||
' WHERE base_rowid = ''' || indexctx.Rid || '''';
EXECUTE IMMEDIATE col_val_stmt INTO column_value;
IF (column_value <= scanctx.upper_bound AND
column_value >= scanctx.lower_bound AND
scanflg = ODCICONST.RegularCall) THEN
RETURN 1;
ELSE
RETURN 0;

F-6 Oracle Database SQL Language Reference

Using Extensible Indexing

END IF;
ELSE
RAISE_APPLICATION_ERROR(-20101, 'A column that has a domain index of' ||
'Position indextype must be the first argument');
END IF;
END;
/

Next, create the position_between operator, which uses the function_for_position_
between function. The operator takes an indexed NUMBER column as the first argument,
followed by a NUMBER lower and upper bound as the second and third arguments.
See Also:

CREATE OPERATOR on page 15-35

CREATE OR REPLACE OPERATOR position_between
BINDING (NUMBER, NUMBER, NUMBER) RETURN NUMBER
WITH INDEX CONTEXT, SCAN CONTEXT position_im
USING function_for_position_between;

In this CREATE OPERATOR statement, the WITH INDEX CONTEXT, SCAN CONTEXT position_im
clause is included so that the index context and scan context are passed in to the
functional evaluation, which is index based.
Now create the position_indextype indextype for the position_operator:
See Also:

CREATE INDEXTYPE on page 14-87

CREATE INDEXTYPE position_indextype
FOR position_between(NUMBER, NUMBER, NUMBER)
USING position_im;

The operator position_between uses an index-based functional implementation.
Therefore, a domain index must be defined on the referenced column so that the index
information can be passed into the functional evaluation. So the final step is to create
the domain index salary_index using the position_indextype indextype:
See Also:

CREATE INDEX on page 14-60

CREATE INDEX salary_index ON employees(salary)
INDEXTYPE IS position_indextype;

Now you can use the position_between operator function to rewrite the original
query as follows:
SELECT last_name, salary FROM employees
WHERE position_between(salary, 10, 20)=1
ORDER BY salary DESC, last_name;
LAST_NAME
SALARY
------------------------- ---------Tucker
10000
King
10000
Baer
10000
Bloom
10000
Fox
9600
Bernstein
9500
Sully
9500
Greene
9500
Hunold
9000

Extended Examples F-7

Using XML in SQL Statements

Faviet
McEwen
Hall
Hutton
Taylor
Livingston
Gietz
Chen
Fripp
Weiss
Olsen
Smith
Kaufling

9000
9000
9000
8800
8600
8400
8300
8200
8200
8000
8000
8000
7900

Using XML in SQL Statements
This section describes some of the ways you can use XMLType data in the database.

XMLType Tables
The sample schema oe contains a table warehouses, which contains an XMLType column
warehouse_spec. Suppose you want to create a separate table with the warehouse_spec
information. The following example creates a very simple XMLType table with one CLOB
column:
CREATE TABLE xwarehouses OF XMLTYPE
XMLTYPE STORE AS CLOB;

You can insert into such a table using XMLType syntax, as shown in the next statement.
(The data inserted in this example corresponds to the data in the warehouse_spec
column of the sample table oe.warehouses where warehouse_id = 1.)
INSERT INTO xwarehouses VALUES
(xmltype('

1
Southlake, Texas
Owned
25000
2
Rear load
true
N
Street
10
'));

Oracle XML DB Developer's Guide for information on
XMLType and its member methods
See Also:

You can query this table with the following statement:
SELECT e.getClobVal() FROM xwarehouses e;

CLOB columns are subject to all of the restrictions on LOB columns. To avoid these
restrictions, create an XMLSchema-based table. The XMLSchema maps the XML
elements to their object-relational equivalents. The following example registers an
XMLSchema locally. The XMLSchema (xwarhouses.xsd) reflects the same structure as
the xwarehouses table. (XMLSchema declarations use PL/SQL and the DBMS_
XMLSCHEMA package, so the example is shown in italics.)
F-8 Oracle Database SQL Language Reference

Using XML in SQL Statements

See Also: Oracle XML DB Developer's Guide for information on
creating XMLSchemas
begin
dbms_xmlschema.registerSchema(
'http://www.example.com/xwarehouses.xsd',
'


















',
TRUE, TRUE, FALSE, FALSE);
end;
/

type
type
type
type
type
type
type
type
type
type

=
=
=
=
=
=
=
=
=
=

"positiveInteger"/>
"string"/>
"who:RentalType"/>
"positiveInteger"/>
"positiveInteger"/>
"string"/>
"boolean"/>
"boolean"/>
"who:ParkingType"/>
"positiveInteger"/>

Now you can create an XMLSchema-based table, as shown in the following example:
CREATE TABLE xwarehouses OF XMLTYPE
XMLSCHEMA "http://www.example.com/xwarehouses.xsd"
ELEMENT "Warehouse";

By default, Oracle stores this as an object-relational table. Therefore, you can insert into
it as shown in the example that follows. (The data inserted in this example
corresponds to the data in the warehouse_spec column of the sample table
oe.warehouses where warehouse_id = 1.)
INSERT INTO xwarehouses VALUES(
xmltype.createxml('

1
Southlake, Texas
Owned
25000
2
Rear load
true
false
Street
10
'));
...

You can define constraints on an XMLSchema-based table. To do so, you use the
XMLDATA pseudocolumn to refer to the appropriate attribute within the Warehouse XML
element:
ALTER TABLE xwarehouses ADD (PRIMARY KEY(XMLDATA."WarehouseId"));

Because the data in xwarehouses is stored object relationally, Oracle rewrites queries to
this XMLType table to go to the underlying storage when possible. Therefore the
following queries would use the index created by the primary key constraint in the
preceding example:
SELECT * FROM xwarehouses x
WHERE EXISTSNODE(VALUE(x), '/Warehouse[WarehouseId="1"]',
'xmlns:who="http://www.example.com/xwarehouses.xsd"') = 1;
SELECT * FROM xwarehouses x
WHERE EXTRACTVALUE(VALUE(x), '/Warehouse/WarehouseId',
'xmlns:who="http://www.example.com/xwarehouses.xsd"') = 1;

You can also explicitly create indexes on XMLSchema-based tables, which greatly
enhance the performance of subsequent queries. You can create object-relational views
on XMLType tables, and you can create XMLType views on object-relational tables.
See Also:
■

XMLDATA Pseudocolumn on page 2-11 for information on the
XMLDATA pseudocolumn

■

"Creating an XMLType View: Example" on page 17-25

■

Creating an Index on an XMLType Table: Example on page 14-81

XMLType Columns
The sample table oe.warehouses was created with a warehouse_spec column of type
XMLType. The examples in this section create a shortened form of the oe.warehouses
table, using two different types of storage.
The first example creates a table with an XMLType table stored as a CLOB. This table does
not require an XMLSchema, so the content structure is not predetermined:
CREATE TABLE xwarehouses (
warehouse_id
NUMBER,
warehouse_spec
XMLTYPE)
XMLTYPE warehouse_spec STORE AS CLOB
(TABLESPACE example
STORAGE (INITIAL 6144)
CHUNK 4000

F-10

Oracle Database SQL Language Reference

Using XML in SQL Statements

NOCACHE LOGGING);

The following example creates a similar table, but stores the XMLType data in an
object-relational XMLType column whose structure is determined by the specified
XMLSchema:
CREATE TABLE xwarehouses (
warehouse_id
NUMBER,
warehouse_spec XMLTYPE)
XMLTYPE warehouse_spec STORE AS OBJECT RELATIONAL
XMLSCHEMA "http://www.example.com/xwarehouses.xsd"
ELEMENT "Warehouse";

Extended Examples

F-11

Using XML in SQL Statements

F-12

Oracle Database SQL Language Reference

Index
Symbols
% (percent) used with LIKE operator, 7-16
+ (plus sign) in Oracle Automatic Storage
Management filenames, 8-32

Numerics
20th century, 3-65
21st century, 3-65
3GL functions and procedures, calling,

15-2

A
ABORT LOGICAL STANDBY clause
of ALTER DATABASE, 10-37, 10-38
ABS function, 5-18
ACCESSED GLOBALLY clause
of CREATE CONTEXT, 14-10
ACCOUNT LOCK clause
of ALTER USER. See CREATE USER
of CREATE USER, 17-12
ACCOUNT UNLOCK clause
of ALTER USER. See CREATE USER
of CREATE USER, 17-12
ACOS function, 5-19
ACTIVATE STANDBY DATABASE clause
of ALTER DATABASE, 10-34
AD and A.D. datetime format elements, 3-65
ADD clause
of ALTER DIMENSION, 10-49
of ALTER INDEXTYPE, 10-98
of ALTER TABLE, 12-42
of ALTER VIEW, 13-15
ADD DATAFILE clause
of ALTER TABLESPACE, 12-96
ADD LOGFILE clause
of ALTER DATABASE, 10-13
ADD LOGFILE GROUP clause
of ALTER DATABASE, 10-30
ADD LOGFILE INSTANCE clause
of ALTER DATABASE, 10-30
ADD LOGFILE MEMBER clause
of ALTER DATABASE, 10-13, 10-31
ADD LOGFILE THREAD clause
of ALTER DATABASE, 10-30

ADD OVERFLOW clause
of ALTER TABLE, 12-40
ADD PARTITION clause
of ALTER TABLE, 12-64, 12-65
ADD PRIMARY KEY clause
of ALTER MATERIALIZED VIEW LOG, 11-22
ADD ROWID clause
of ALTER MATERIALIZED VIEW, 11-22
ADD SUPPLEMENTAL LOG DATA clause
of ALTER DATABASE, 10-32
ADD SUPPLEMENTAL LOG GROUP clause
of ALTER TABLE, 12-35
ADD TEMPFILE clause
of ALTER TABLESPACE, 12-96
ADD VALUES clause
of ALTER TABLE ... MODIFY PARTITION, 12-61
ADD_MONTHS function, 5-20
adding a constraint to a table, 12-55
ADMINISTER ANY SQL TUNING SET system
privilege, 18-40
ADMINISTER DATABASE TRIGGER system
privilege, 18-46
ADMINISTER SQL MANAGEMENT OBJECT system
privilege, 18-40, 18-44
ADMINISTER SQL TUNING SET system
privilege, 18-40
ADVISE clause
of ALTER SESSION, 11-46
ADVISOR system privilege, 18-40
aggregate functions, 5-10
alias
for a column, 9-2
for an expressions in a view query, 17-18
specifying in queries and subqueries, 19-21
ALL clause
of SELECT, 19-15
of SET CONSTRAINTS, 19-59
of SET ROLE, 19-62
ALL EXCEPT clause
of SET ROLE, 19-62
ALL operator, 7-5
ALL PRIVILEGES clause
of GRANT, 18-37
of REVOKE, 18-91
ALL PRIVILEGES shortcut
of AUDIT, 13-32

Index-1

ALL shortcut
of AUDIT, 13-32
ALL_COL_COMMENTS data dictionary
view, 13-47
ALL_INDEXTYPE_COMMENTS data dictionary
view, 13-47
ALL_MVIEW_COMMENTS data dictionary
view, 13-47
ALL_OPERATOR_COMMENTS data dictionary
view, 13-47
ALL_ROWS hint, 3-79
ALL_TAB_COMMENTS data dictionary view, 13-47
all-column wildcard, 19-15
ALLOCATE EXTENT clause
of ALTER CLUSTER, 10-6, 10-7
of ALTER INDEX, 10-84
of ALTER MATERIALIZED VIEW, 11-7
of ALTER TABLE, 12-36
ALLOW CORRUPTION clause
of ALTER DATABASE ... RECOVER, 10-22
ALTER ANY CLUSTER system privilege, 18-40
ALTER ANY CUBE DIMENSION system
privilege, 18-43
ALTER ANY CUBE system privilege, 18-43
ALTER ANY DIMENSION system privilege, 18-41
ALTER ANY INDEX system privilege, 18-41
ALTER ANY INDEXTYPE system privilege, 18-41
ALTER ANY LIBRARY system privilege, 18-42
ALTER ANY MATERIALIZED VIEW system
privilege, 18-42
ALTER ANY MINING MODEL system
privilege, 18-43
ALTER ANY OPERATOR system privilege, 18-44
ALTER ANY OUTLINE system privilege, 18-44
ALTER ANY PROCEDURE system privilege, 18-44
ALTER ANY ROLE system privilege, 18-44
ALTER ANY SEQUENCE system privilege, 18-45
ALTER ANY SQL PROFILE system privilege, 18-40
ALTER ANY TABLE system privilege, 18-45
ALTER ANY TRIGGER system privilege, 18-46
ALTER ANY TYPE system privilege, 18-46
ALTER CLUSTER statement, 10-5
ALTER DATABASE LINK system privilege, 18-41
ALTER DATABASE statement, 10-8
ALTER DATABASE system privilege, 18-40
ALTER DIMENSION statement, 10-48
ALTER DISKGROUP statement, 10-51
ALTER FLASHBACK ARCHIVE statement, 10-74
ALTER FUNCTION statement, 10-77
ALTER INDEX statement, 10-78
ALTER INDEXTYPE statement, 10-97
ALTER JAVA CLASS statement, 10-100
ALTER JAVA SOURCE statement, 10-100
ALTER LIBRARY statement, 11-2
ALTER MATERIALIZED VIEW LOG
statement, 11-18
ALTER MATERIALIZED VIEW statement, 11-3
ALTER object privilege
on a mining model, 18-49
on a sequence, 18-50

Index-2

on a table, 18-51
on an OLAP cube, 18-50
on an OLAP cube dimension, 18-50
ALTER OPERATOR statement, 11-25
ALTER OUTLINE statement, 11-28
ALTER PACKAGE statement, 11-30
ALTER PROCEDURE statement, 11-31
ALTER PROFILE statement, 11-32
ALTER PROFILE system privilege, 18-44
ALTER PUBLIC DATABASE LINK system
privilege, 18-41
ALTER RESOURCE COST statement, 11-35
ALTER RESOURCE COST system privilege, 18-45
ALTER ROLE statement, 11-38
ALTER ROLLBACK SEGMENT statement, 11-40
ALTER ROLLBACK SEGMENT system
privilege, 18-44
ALTER SEQUENCE statement, 11-43
ALTER SESSION statement, 11-45
ALTER SESSION system privilege, 18-45
ALTER SNAPSHOT LOG. See ALTER
MATERIALIZED VIEW LOG
ALTER SNAPSHOT. See ALTER MATERIALIZED
VIEW
ALTER SYSTEM statement, 11-58
ALTER SYSTEM system privilege, 18-40
ALTER TABLE statement, 12-2
ALTER TABLESPACE statement, 12-90
ALTER TABLESPACE system privilege, 18-46
ALTER TRIGGER statement, 13-2
ALTER TYPE statement, 13-4
ALTER USER statement, 13-6
ALTER USER system privilege, 18-46
ALTER VIEW statement, 13-14
alter_external_table clause
of ALTER TABLE, 12-18
AM and A.M. datetime format elements, 3-65
American National Standards Institute (ANSI), C-1
data types, 3-28
conversion to Oracle data types, 3-28
standards, 1-1, C-1
supported data types, 3-4
analytic functions, 5-11
ANALYZE ANY system privilege, 18-47
ANALYZE CLUSTER statement, 13-17
ANALYZE INDEX statement, 13-17
ANALYZE TABLE statement, 13-17
ANCILLARY TO clause
of CREATE OPERATOR, 15-36
AND condition, 7-8, 7-9
AND DATAFILES clause
of DROP TABLESPACE, 18-11
ANSI. See American National Standards Institute
(ANSI)
antijoins, 9-14
ANY operator, 7-5
APPEND hint, 3-80
APPEND_VALUES hint, 3-80
APPENDCHILDXML function, 5-21
application servers

allowing connection as user, 13-10
applications
allowing connection as user, 13-10
securing, 14-9
validating, 14-9
ARCHIVE LOG clause
of ALTER SYSTEM, 11-61
archive mode
specifying, 14-25
archived redo logs
location, 10-20
ARCHIVELOG clause
of ALTER DATABASE, 10-13, 10-28
of CREATE CONTROLFILE, 14-16
of CREATE DATABASE, 14-25
arguments
of operators, 4-1
arithmetic
with DATE values, 3-20
arithmetic operators, 4-3
AS source_char clause
of CREATE JAVA, 14-94
AS subquery clause
of CREATE MATERIALIZED VIEW, 15-21
of CREATE TABLE, 16-67
of CREATE VIEW, 17-20
ASC clause
of CREATE INDEX, 14-71
ASCII character set, 3-38
ASCII function, 5-22
ASCIISTR function, 5-23
ASIN function, 5-24
ASM_DISKGROUPS initialization parameter
setting with ALTER SYSTEM, 10-71, 14-43
ASM_DISKSTRING initialization parameter
setting with ALTER SYSTEM, 14-46
ASSOCIATE STATISTICS statement, 13-25
asterisk
all-column wildcard in queries, 19-15
asynchronous commit, 13-50
ATAN function, 5-25
ATAN2 function, 5-26
ATTRIBUTE clause
of ALTER DIMENSION, 10-49
of CREATE DIMENSION, 14-37, 14-39
attributes
adding to a dimension, 10-49
dropping from a dimension, 10-49
maximum number of in object type, 16-26
of dimensions, defining, 14-39
of disk groups, 10-68, 14-46
AUDIT ANY system privilege, 18-47
AUDIT statement
locks, B-4
AUDIT SYSTEM system privilege, 18-40
auditing
options
for SQL statements, 13-37
policies
value-based, 13-29

SQL statements, 13-30
by a proxy, 13-30
by a user, 13-30
on a directory, 13-30
on a schema, 13-30
stopping, 18-79
system privileges, 13-30
AUTHENTICATED BY clause
of CREATE DATABASE LINK, 14-34
AUTHENTICATED clause
of ALTER USER, 13-11
AUTHENTICATION REQUIRED clause
of ALTER USER, 13-11
AUTHID CURRENT_USER clause
of ALTER JAVA, 10-101
of CREATE JAVA, 14-92, 14-93
AUTHID DEFINER clause
of ALTER JAVA, 10-101
of CREATE JAVA, 14-92, 14-93
AUTOALLOCATE clause
of CREATE TABLESPACE, 16-90
AUTOEXTEND clause
of ALTER DATABASE, 10-12
of CREATE DATABASE, 14-21
automatic segment-space management, 16-91
automatic undo mode, 11-40, 14-29
AVG function, 5-27

B
BACKUP ANY TABLE system privilege, 18-45
BACKUP CONTROLFILE clause
of ALTER DATABASE, 10-14, 10-33
backups, 10-40
basic table compression, 16-35
BC and B.C. datetime format elements, 3-65
BECOME USER system privilege, 18-47
BEGIN BACKUP clause
of ALTER DATABASE, 10-25
of ALTER TABLESPACE, 12-95
BETWEEN condition, 7-22
BFILE
data type, 3-25
locators, 3-25
BFILENAME function, 5-29
BIN_TO_NUM function, 5-30
binary large objects. See BLOB
binary operators, 4-2
binary XML format, 16-49
binary XML storage, 16-50
BINDING clause
of CREATE OPERATOR, 15-35
bindings
adding to an operator, 11-26
dropping from an operator, 11-27
bit vectors
converting to numbers, 5-30
BITAND function, 5-32
BITMAP clause
of CREATE INDEX, 14-67

Index-3

bitmap indexes, 14-67
creating join indexes, 14-61
blank padding
specifying in format models, 3-67
suppressing, 3-67
BLOB data type, 3-26
transactional support, 3-26
BLOCKSIZE clause
of CREATE TABLESPACE, 16-88
bottom-N reporting, 5-82, 5-209, 5-236
buffer cache
flushing, 11-65
BUFFER_POOL parameter
of STORAGE clause, 8-55
BUILD DEFERRED clause
of CREATE MATERIALIZED VIEW, 15-16
BUILD IMMEDIATE clause
of CREATE MATERIALIZED VIEW, 15-16
BY ACCESS clause
of AUDIT, 13-34
BY SESSION clause
of AUDIT, 13-34
BYTE character semantics, 3-9
BYTE length semantics, 12-47

C
CACHE clause
of ALTER MATERIALIZED VIEW, 11-11
of ALTER MATERIALIZED VIEW LOG, 11-22
of ALTER TABLE, 16-62
of CREATE CLUSTER, 14-7
of CREATE MATERIALIZED VIEW, 15-15
of CREATE MATERIALIZED VIEW LOG, 15-30
CACHE hint, 3-81
CACHE parameter
of ALTER SEQUENCE. See CREATE
SEQUENCE, 11-43
of CREATE SEQUENCE, 15-69
CACHE READS clause
of ALTER TABLE, 12-45
of CREATE TABLE, 16-62
cached cursors
execution plan for, 18-20
call spec. See call specifications
call specifications
in procedures, 15-48
CALL statement, 13-42
calls
limiting CPU time for, 15-52
limiting data blocks read, 15-53
CARDINALITY function, 5-34
Cartesian products, 9-12
CASCADE clause
of CREATE TABLE, 16-65
of DROP PROFILE, 17-65
of DROP USER, 18-16
CASCADE CONSTRAINTS clause
of DROP CLUSTER, 17-37
of DROP TABLE, 18-7

Index-4

of DROP TABLESPACE, 18-11
of DROP VIEW, 18-18
of REVOKE, 18-91
CASE expressions, 6-5
searched, 6-5
simple, 6-5
CAST function, 5-35
CATSEARCH condition, 7-2
CATSEARCH operator, 4-1
CEIL function, 5-38
chained rows
listing, 13-22
of clusters, 13-20
CHANGE CATEGORY clause
of ALTER OUTLINE, 11-29
CHANGE NOTIFICATION system privilege, 18-47
CHANGE_DUPKEY_ERROR_INDEX hint, 3-81
CHAR character semantics, 3-9
CHAR data type, 3-9
converting to VARCHAR2, 3-57
CHAR length semantics, 12-47
character functions
returning character values, 5-4
returning number values, 5-5
character large objects. See CLOB
character length semantics, 12-47
character literal. See text
character set
changing, 10-38
character set functions, 5-5
CHARACTER SET parameter
of CREATE CONTROLFILE, 14-17
of CREATE DATABASE, 14-23
character sets
common, 3-37
database, specifying, 14-23
multibyte characters, 3-112
specifying for database, 14-23
character strings
comparison rules, 3-37
exact matching, 3-67
fixed-length, 3-9
national character set, 3-9
variable-length, 3-10, 3-14
zero-length, 3-9
CHARTOROWID function, 5-39
CHECK clause
of constraints, 8-12
of CREATE TABLE, 16-30
check constraints, 8-12
CHECK DATAFILES clause
of ALTER SYSTEM, 11-63
CHECKPOINT clause
of ALTER SYSTEM, 11-63
checkpoints
forcing, 11-63
CHR function, 5-40
CHUNK clause
of ALTER TABLE, 12-45
of CREATE TABLE, 16-44

CLEAR LOGFILE clause
of ALTER DATABASE, 10-13, 10-29
CLOB data type, 3-26
transactional support, 3-26
clone databases
mounting, 10-18
CLOSE DATABASE LINK clause
of ALTER SESSION, 11-46
CLUSTER clause
of ANALYZE, 13-20
of CREATE INDEX, 14-68
of CREATE TABLE, 16-41
of TRUNCATE, 19-67
CLUSTER hint, 3-82
CLUSTER_ID function, 5-42
CLUSTER_PROBABILITY function, 5-44
CLUSTER_SET function, 5-46
clusters
assigning tables to, 16-41
caching retrieved blocks, 14-7
cluster indexes, 14-68
collecting statistics on, 13-20
creating, 14-2
deallocating unused extents, 10-6
degree of parallelism
changing, 10-6, 10-7
when creating, 14-6
dropping tables, 17-36
extents, allocating, 10-6, 10-7
granting system privileges for, 18-40
hash, 14-5
single-table, 14-6
sorted, 14-4, 16-27
indexed, 14-5
key values
allocating space for, 14-5
modifying space for, 10-6
migrated and chained rows in, 13-20, 13-22
modifying, 10-5
physical attributes
changing, 10-5
specifying, 14-4
releasing unused space, 10-7
removing from the database, 17-36
SQL examples, 17-37
storage attributes
changing, 10-5
storage characteristics
changing, 10-6
tablespace in which created, 14-5
validating structure, 13-21
COALESCE clause
for partitions, 12-65
of ALTER INDEX, 10-91
of ALTER TABLE, 12-40, 12-41, 12-60
of ALTER TABLESPACE, 12-94
COALESCE function, 5-48
as a variety of CASE expression, 5-48
COALESCE SUBPARTITION clause
of ALTER TABLE, 12-60

COLLECT function, 5-50
collection functions, 5-7
collection types
multilevel, 16-48
collections
inserting rows into, 18-60
modifying, 12-52
modifying retrieval method, 12-10
nested tables, 3-31
testing for empty, 7-12
treating as a table, 18-60, 19-20, 19-75, 19-77
unnesting, 19-20
examples, 19-54
varrays, 3-30
collection-typed values
converting to data types, 5-35
column expressions, 6-6
column REF constraints, 8-12
of CREATE TABLE, 16-30
column values
unpivoting into rows, 19-22
COLUMN_VALUE pseudocolumn, 2-6
columns
adding, 12-42
aliases for, 9-2
altering storage, 12-44
associating statistics types with, 13-25
basing an index on, 14-69
comments on, 13-47
creating comments about, 13-46
defining, 16-6
disassociating statistics types from, 17-34
dropping from a table, 12-49
LOB
storage attributes, 12-45
maximum number of, 16-26
modifying existing, 12-46
parent-child relationships between, 14-36
properties, altering, 12-11, 12-44
qualifying names of, 9-2
REF
describing, 8-12
renaming, 12-52
restricting values for, 8-4
specifying
as primary key, 8-9
constraints on, 16-30
default values, 16-27
storage properties, 16-42
substitutable, identifying type, 5-289
virtual
adding to a table, 12-43
creating, 16-29
modifying, 12-43
COLUMNS clause
of ASSOCIATE STATISTICS, 13-25, 13-26
of DISASSOCIATE STATISTICS, 17-34
COMMENT ANY MINING MODEL system
privilege, 18-43
COMMENT ANY TABLE system privilege, 18-47

Index-5

COMMENT clause
of COMMIT, 13-50
COMMENT statement, 13-46
comments, 3-72
adding to objects, 13-46
associating with a transaction, 13-51
dropping from objects, 13-46
in SQL statements, 3-73
on indextypes, 13-47
on operators, 13-47
on schema objects, 3-74
on table columns, 13-47
on tables, 13-47
removing from the data dictionary, 13-46
specifying, 3-73
viewing, 13-47
commit
asynchronous, 13-50
automatic, 13-49
COMMIT IN PROCEDURE clause
of ALTER SESSION, 11-46
COMMIT statement, 13-49
COMMIT TO SWITCHOVER clause
of ALTER DATABASE, 10-36
comparison conditions, 7-4
comparison functions, 5-6
comparison semantics
of character strings, 3-37
COMPILE clause
of ALTER DIMENSION, 10-50
of ALTER JAVA SOURCE, 10-101
of ALTER MATERIALIZED VIEW, 11-15
of ALTER VIEW, 13-16
of CREATE JAVA, 14-92
COMPOSE function, 5-51
composite foreign keys, 8-10
composite partitioning
range-list, 12-60, 16-59
when creating a table, 16-22, 16-56
composite primary keys, 8-9
composite range partitions, 16-56
COMPOSITE_LIMIT parameter
of ALTER PROFILE, 11-32
of CREATE PROFILE, 15-53
compound conditions, 7-21
compound expressions, 6-4
COMPRESS clause
of ALTER INDEX ... REBUILD, 10-88
of CREATE TABLE, 16-39
compression
of index keys, 10-81
of tables, 16-34
of tablespaces, 16-34
CONCAT function, 5-52
concatenation operator, 4-4
conditions
BETWEEN, 7-22
comparison, 7-4
compound, 7-21
EXISTS, 7-20, 7-22

Index-6

floating-point, 7-7
group comparison, 7-6
IN, 7-23
in SQL syntax, 7-1
interval, 7-22
IS [NOT] EMPTY, 7-12
IS ANY, 7-9
IS OF type, 7-25
IS PRESENT, 7-10
LIKE, 7-14
logical, 7-8
MEMBER, 7-13
membership, 7-13, 7-23
model, 7-9
multiset, 7-11
null, 7-19
pattern matching, 7-14
range, 7-22
REGEXP_LIKE, 7-18
SET, 7-12
simple comparison, 7-5
SUBMULTISET, 7-13
UNDER_PATH, 7-21
XML, 7-20
CONNECT BY clause
of queries and subqueries, 19-26
of SELECT, 9-4, 19-26
CONNECT clause
of SELECT and subqueries, 19-10
CONNECT TO clause
of CREATE DATABASE LINK, 14-33
CONNECT_BY_ISCYCLE pseudocolumn, 2-1
CONNECT_BY_ISLEAF pseudocolumn, 2-2
CONNECT_BY_ROOT operator, 4-5
CONNECT_TIME parameter
of ALTER PROFILE, 11-32
of ALTER RESOURCE COST, 11-35
connection qualifier, 3-118
CONSIDER FRESH clause
of ALTER MATERIALIZED VIEW, 11-15
constant values. See literals
CONSTRAINT(S) session parameter, 11-51
constraints
adding to a table, 12-55
altering, 12-11
check, 8-12
checking
at end of transaction, 8-14
at start of transaction, 8-15
at the end of each DML statement, 8-14
column REF, 8-12
deferrable, 8-14, 19-59
enforcing, 11-51
defining, 8-4, 16-6
for a table, 16-30
on a column, 16-30
disabling, 16-64
after table creation, 12-78
cascading, 16-65
during table creation, 16-24

dropping, 12-11, 12-55, 18-11
enabling, 16-64, 16-65
after table creation, 12-78
during table creation, 16-24
foreign key, 8-10
modifying existing, 12-55
on views
dropping, 13-15, 18-18
partitioning referential, 12-55, 16-60
primary key, 8-9
attributes of index, 8-17
enabling, 16-65
referential integrity, 8-10
renaming, 12-55
restrictions, 8-8
setting state for a transaction, 19-59
storing rows in violation, 12-73
table REF, 8-12
unique
attributes of index, 8-17
enabling, 16-65
CONTAINS condition, 7-2
CONTAINS operator, 4-1
context namespaces
accessible to instance, 14-10
associating with package, 14-9
initializing using OCI, 14-10
initializing using the LDAP directory, 14-10
removing from the database, 17-38
contexts
creating namespaces for, 14-9
granting system privileges for, 18-40
control file clauses
of ALTER DATABASE, 10-14
control files
allowing reuse, 14-14, 14-22
backing up, 10-33
force logging mode, 14-17
re-creating, 14-12
standby, creating, 10-33
CONTROLFILE REUSE clause
of CREATE DATABASE, 14-22
conversion
functions, 5-6
rules, string to date, 3-69
CONVERT function, 5-53
CORR function, 5-55
CORR_K function, 5-59
CORR_S function, 5-58
correlated subqueries, 9-14
correlation functions
Kendall’s tau-b, 5-57
Pearson’s, 5-55
Spearman’s rho, 5-57
correlation names
for base tables of indexes, 14-69
in DELETE, 17-31
in SELECT, 19-21
COS function, 5-60
COSH function, 5-61

COUNT function, 5-62
COVAR_POP function, 5-64
COVAR_SAMP function, 5-66
CPU_PER_CALL parameter
of ALTER PROFILE, 11-32
of CREATE PROFILE, 15-52
CPU_PER_SESSION parameter
of ALTER PROFILE, 11-32
of ALTER RESOURCE COST, 11-35
of CREATE PROFILE, 15-52
CREATE ANY CLUSTER system privilege, 18-40
CREATE ANY CONTEXT system privilege, 18-40
CREATE ANY CUBE BUILD PROCESS system
privilege, 18-44
CREATE ANY CUBE DIMENSION system
privilege, 18-43
CREATE ANY CUBE system privilege, 18-43
CREATE ANY DIMENSION system privilege, 18-41
CREATE ANY DIRECTORY system privilege, 18-41
CREATE ANY INDEX system privilege, 18-41
CREATE ANY INDEXTYPE system privilege, 18-41
CREATE ANY JOB system privilege, 18-42
CREATE ANY LIBRARY system privilege, 18-42
CREATE ANY MATERIALIZED VIEW system
privilege, 18-42
CREATE ANY MEASURE FOLDER system
privilege, 18-43
CREATE ANY MINING MODEL system
privilege, 18-43
CREATE ANY OPERATOR system privilege, 18-44
CREATE ANY OUTLINE system privilege, 18-44
CREATE ANY PROCEDURE system
privilege, 18-44
CREATE ANY SEQUENCE system privilege, 18-45
CREATE ANY SQL PROFILE system
privilege, 18-40
CREATE ANY SYNONYM system privilege, 18-45
CREATE ANY TABLE system privilege, 18-45
CREATE ANY TRIGGER system privilege, 18-46
CREATE ANY TYPE system privilege, 18-46
CREATE ANY VIEW system privilege, 18-47
CREATE CLUSTER statement, 14-2
CREATE CLUSTER system privilege, 18-40
CREATE CONTEXT statement, 14-9
CREATE CONTROLFILE statement, 14-12
CREATE CUBE BUILD PROCESS system
privilege, 18-43
CREATE CUBE DIMENSION system
privilege, 18-43
CREATE CUBE system privilege, 18-43
CREATE DATABASE LINK statement, 14-31
CREATE DATABASE LINK system privilege, 18-41
CREATE DATABASE statement, 14-19
CREATE DATAFILE clause
of ALTER DATABASE, 10-11, 10-26
CREATE DIMENSION statement, 14-36
CREATE DIMENSION system privilege, 18-41
CREATE DIRECTORY statement, 14-41
CREATE DISKGROUP statement, 14-43
CREATE EXTERNAL JOB system privilege, 18-42

Index-7

CREATE FLASHBACK ARCHIVE statement, 14-55
CREATE FUNCTION statement, 14-58
CREATE INDEX statement, 14-60
CREATE INDEXTYPE statement, 14-87
CREATE INDEXTYPE system privilege, 18-41
CREATE JAVA statement, 14-91
CREATE JOB system privilege, 18-42
CREATE LIBRARY statement, 15-2
CREATE LIBRARY system privilege, 18-42
CREATE MATERIALIZED VIEW LOG
statement, 15-27
CREATE MATERIALIZED VIEW statement, 15-4
CREATE MATERIALIZED VIEW system
privilege, 18-42
CREATE MEASURE FOLDER system
privilege, 18-43
CREATE MINING MODEL system privilege, 18-43
CREATE OPERATOR statement, 15-35
CREATE OPERATOR system privilege, 18-44
CREATE OUTLINE statement, 15-38
CREATE PACKAGE BODY statement, 15-44
CREATE PACKAGE statement, 15-42
locks, B-4
CREATE PFILE statement, 15-46
CREATE PROCEDURE statement, 15-48
locks, B-4
CREATE PROCEDURE system privilege, 18-44
CREATE PROFILE statement, 15-50
CREATE PROFILE system privilege, 18-44
CREATE PUBLIC DATABASE LINK system
privilege, 18-41
CREATE PUBLIC SYNONYM system
privilege, 18-45
CREATE RESTORE POINT statement, 15-56
CREATE ROLE statement, 15-59
CREATE ROLE system privilege, 18-44
CREATE ROLLBACK SEGMENT statement, 15-62
CREATE ROLLBACK SEGMENT system
privilege, 18-44
CREATE SCHEMA statement, 15-65
CREATE SEQUENCE statement, 15-67
CREATE SEQUENCE system privilege, 18-45
CREATE SESSION system privilege, 18-45
CREATE SPFILE statement, 15-71
CREATE STANDBY CONTROLFILE clause
of ALTER DATABASE, 10-14, 10-33
CREATE SYNONYM statement, 16-2
CREATE SYNONYM system privilege, 18-45
CREATE TABLE statement, 16-6
CREATE TABLE system privilege, 18-45
CREATE TABLESPACE statement, 16-83
CREATE TABLESPACE system privilege, 18-46
CREATE TRIGGER statement, 16-98
CREATE TRIGGER system privilege, 18-46
CREATE TYPE BODY statement, 17-5
CREATE TYPE statement, 17-3
CREATE TYPE system privilege, 18-46
CREATE USER statement, 17-7
CREATE USER system privilege, 18-46
CREATE VIEW statement, 17-14

Index-8

CREATE VIEW system privilege, 18-47
cross joins, 19-23
CUBE clause
of SELECT statements, 19-28
CUBE_TABLE function, 5-67
cubes
extracting data, 5-67
CUME_DIST function, 5-69
cumulative distributions, 5-69
currency
group separators, 3-58
currency symbol
ISO, 3-58
local, 3-58
union, 3-59
CURRENT_DATE function, 5-71
CURRENT_SCHEMA session parameter, 11-51
CURRENT_TIMESTAMP function, 5-72
CURRENT_USER clause
of CREATE DATABASE LINK, 14-33
CURRVAL pseudocolumn, 2-3, 15-67
CURSOR expressions, 6-7
CURSOR_SHARING_EXACT hint, 3-82
cursors
cached, 18-20
CV function, 5-73
CYCLE parameter
of ALTER SEQUENCE. See CREATE
SEQUENCE, 11-43
of CREATE SEQUENCE, 15-69

D
data
aggregation
composite columns of GROUP BY, 19-28
concatenated grouping sets of GROUP
BY, 19-28
grouping sets, 19-28
analyzing a subset, 5-178
caching frequently used, 16-62
independence, 16-2
integrity checking on input, 3-11
locks on, B-1
pivoting, 19-21
retrieving, 9-1
specifying as temporary, 16-25
undo
preserving, 12-100, 16-94
unpivoting, 19-22
data cartridge functions, 5-17
data conversion, 3-40
between character data types, 3-42
implicit
disadvantages, 3-40
implicit versus explicit, 3-40
when performed implicitly, 3-40, 3-42
when specified explicitly, 3-43
data definition language
locks, B-4

data definition language (DDL)
statements, 10-2
and implicit commit, 10-2
causing recompilation, 10-2
PL/SQL support, 10-2
statements requiring exclusive access, 10-2
data dictionary
adding comments to, 13-46
locks, B-4
data files
bringing online, 10-27
changing size of, 10-27
creating new, 10-26
defining for a tablespace, 16-84, 16-85, 16-86
defining for the database, 14-21
designing media recovery, 10-19
dropping, 12-96
enabling autoextend, 8-35
end online backup of, 10-27, 12-96
extending automatically, 8-35
putting online, 10-27
recover damaged, 10-19
recovering, 10-21
re-creating lost or damaged, 10-26
renaming, 10-25
resizing, 10-27
reusing, 8-35
size of, 8-34
specifying, 8-29
for a tablespace, 16-87
for database, 14-26
system generated, 10-26
taking offline, 10-27
temporary
shrinking, 12-97
data manipulation language (DML)
allowing during indexing, 10-86
operations
during index creation, 14-73
during index rebuild, 12-77
restricting, 11-67
parallelizing, 16-63
retrieving rows affected by, 17-31, 18-62, 19-79
statements, 10-2
PL/SQL support, 10-3
data mining functions, 5-7
data redaction
granting system privileges for, 18-40
data types, 3-1
"Any" types, 3-31
ANSI-supported, 3-4
BFILE, 3-25
BLOB, 3-26
CHAR, 3-9
character, 3-8
CLOB, 3-26
comparison rules, 3-36
converting to collection-typed values, 5-35
converting to other data types, 5-35
DATE, 3-17

datetime, 3-16
interval, 3-16
INTERVAL DAY TO SECOND, 3-19
INTERVAL YEAR TO MONTH, 3-19
length semantics, 3-9
LONG, 3-14
LONG RAW, 3-23
media types, 3-35
NCHAR, 3-9
NCLOB, 3-26
NUMBER, 3-10
NVARCHAR2, 3-9
Oracle-supplied types, 3-31
RAW, 3-23
ROWID, 3-27
SDO_TOPO_GEOMETRY, 3-34
spatial types, 3-34
TIMESTAMP, 3-18
TIMESTAMP WITH LOCAL TIME ZONE, 3-19
TIMESTAMP WITH TIME ZONE, 3-18
UROWID, 3-28
user-defined, 3-29
VARCHAR, 3-10
VARCHAR2, 3-10
XML types, 3-32
database links, 9-16
altering, 10-46
closing, 11-46
creating, 3-117, 14-31
creating synonyms with, 16-4
current user, 14-33
granting system privileges for, 18-41
naming, 3-117
public, 14-32
dropping, 17-40
referring to, 3-118
removing from the database, 17-40
shared, 14-32
syntax, 3-118
updating passwords, 10-46
username and password, 3-118
database objects
dropping, 18-16
nonschema, 3-110
schema, 3-109
Database Smart Flash Cache, 8-55
database triggers. See triggers
databases
accounts
creating, 17-7
allowing changes to, 11-46
allowing generation of redo logs, 10-18
allowing reuse of control files, 14-22
allowing unlimited resources to users, 15-52
archive mode, specifying, 14-25
beginning backup of, 10-25
blocks
specifying size, 16-88
cancel-based recovery, 10-21
terminating, 10-22

Index-9

change-based recovery, 10-21
changing characteristics, 14-12
changing global name, 10-40
changing name, 14-12, 14-14
character set, specifying, 14-23
committing to standby status, 10-36
connect strings, 3-118
consistency-based recovery, 10-21
controlling use, 10-42
create script for, 10-33
creating, 14-19
data files
modifying, 10-25
specifying, 14-26
default edition, setting, 10-38
designing media recovery, 10-19
dropping, 17-39
ending backup of, 10-25
erasing all data from, 14-19
flashing back, 18-24
granting system privileges for, 18-40
in FLASHBACK mode, 10-41
in FORCE LOGGING mode, 10-29, 14-17, 14-25
instances of, 14-23
limiting resources for users, 15-50
log files
modifying, 10-28
specifying, 14-24
managed recovery, 10-11
modifying, 10-8
mounting, 10-18, 14-19
moving a subset to a different database, 12-72
namespaces, 3-113
naming, 10-18
national character set, specifying, 14-23
no-data-loss mode, 10-35
online
adding log files, 10-30
opening, 10-18, 14-19
prepare to re-create, 10-33
preventing changes to, 10-42
protection mode of, 10-35
quiesced state, 11-67
read-only, 10-18
read/write, 10-18
reconstructing damaged, 10-19
recovering, 10-19, 10-21
recovery
allowing corrupt blocks, 10-22
testing, 10-22
with backup control file, 10-21
re-creating control file for, 14-12
remote
accessing, 9-16
authenticating users to, 14-34
connecting to, 14-33
inserting into, 18-60
service name of, 14-34
table locks on, 18-72
resetting

Index-10

to an earlier version, 10-8
restoring earlier version of, 10-41, 12-100, 16-92
restricting users to read-only transactions, 10-19
resuming activity, 11-66
returning to a past time, 18-24
standby
adding log files, 10-30
suspending activity, 11-66
system user passwords, 14-22
temp files
modifying, 10-25
time zone
determining, 5-76
setting, valid values for, 10-41, 14-30
time-based recovery, 10-21
upgrading, 10-8
DATAFILE clause
of CREATE DATABASE, 14-26
DATAFILE clauses
of ALTER DATABASE, 10-11, 10-27
DATAFILE OFFLINE clause
of ALTER DATABASE, 10-27
DATAFILE ONLINE clause
of ALTER DATABASE, 10-27
DATAFILE RESIZE clause
of ALTER DATABASE, 10-27
datafiles
dropping, 18-11
online backup of, 12-95
online, updating information on, 11-63
DATAOBJ_TO_PARTITION function, 5-75
DATE columns
converting to datetime columns, 12-46
DATE data type, 3-17
julian, 3-17
date format models, 3-60, 3-62
long, 3-62
punctuation in, 3-61
short, 3-62
text in, 3-61
date functions, 5-5
dates
arithmetic, 3-20
comparison rules, 3-37
datetime arithmetic, 3-20
boundary cases, 11-51
calculating daylight saving time, 3-22
datetime columns
creating from DATE columns, 12-46
datetime data types, 3-16
daylight saving time, 3-22
datetime expressions, 6-8
datetime field
extracting from a datetime or interval value, 5-91
datetime format elements, 3-60
and Globalization Support, 3-65
capitalization, 3-61
ISO standard, 3-65
RR, 3-65
suffixes, 3-66

datetime functions, 5-5
datetime literals, 3-50
DAY datetime format element, 3-65
daylight saving time, 3-22
boundary cases, 3-22
going into or coming out of effect, 3-22
DB2 data types, 3-28
restrictions on, 3-29
DBA_2PC_PENDING data dictionary view, 11-46
DBA_COL_COMMENTS data dictionary
view, 13-47
DBA_INDEXTYPE_COMMENTS data dictionary
view, 13-47
DBA_MVIEW_COMMENTS data dictionary
view, 13-47
DBA_OPERATOR_COMMENTS data dictionary
view, 13-47
DBA_ROLLBACK_SEGS data dictionary
view, 17-68
DBA_TAB_COMMENTS data dictionary
view, 13-47
DBMS_ROWID package
and extended rowids, 3-27
DBTIMEZONE function, 5-76
DDL. See data definition language (DDL)
DEALLOCATE UNUSED clause
of ALTER CLUSTER, 10-6, 10-7
of ALTER INDEX, 10-79
of ALTER TABLE, 12-36
DEBUG ANY PROCEDURE system privilege, 18-41
DEBUG object privilege
on a function, procedure, or package, 18-50
on a table, 18-51
on a view, 18-51
on an object type, 18-49
debugging
granting system privileges for, 18-41
decimal characters, 3-48
specifying, 3-58
DECODE function, 5-77
decoding functions, 5-9
DECOMPOSE function, 5-79
DEFAULT clause
of ALTER TABLE, 12-43
of ALTER TABLESPACE, 12-93
of CREATE TABLE, 16-27, 16-30
of CREATE TABLESPACE, 16-90
DEFAULT COST clause
of ASSOCIATE STATISTICS, 13-26, 13-27
default index, suppressing, 15-16
DEFAULT profile
assigning to users, 17-65
DEFAULT ROLE clause
of ALTER USER, 13-9
DEFAULT SELECTIVITY clause
of ASSOCIATE STATISTICS, 13-26, 13-27
default tablespace, 14-27
DEFAULT TABLESPACE clause
of ALTER DATABASE, 10-39
of ALTER USER, 13-9

of ALTER USER. See CREATE USER
of CREATE USER, 17-10
default tablespaces
specifying for a user, 13-9
DEFAULT TEMPORARY TABLESPACE clause
of ALTER DATABASE, 10-39
of CREATE DATABASE, 14-21
DEFERRABLE clause
of constraints, 8-14
deferrable constraints, 19-59
DEFERRED clause
of SET CONSTRAINTS, 19-60
DELETE ANY CUBE DIMENSION system
privilege, 18-43
DELETE ANY MEASURE FOLDER system
privilege, 18-43
DELETE ANY TABLE system privilege, 18-45
DELETE object privilege
on a table, 18-51
on a view, 18-51
on an OLAP cube dimension, 18-50
on an OLAP measures folder, 18-50
DELETE statement, 17-26
error logging, 17-32
DELETE STATISTICS clause
of ANALYZE, 13-23
DELETEXML function, 5-80
DENSE_RANK function, 5-82
DEPTH function, 5-84
DEREF function, 5-85
DESC clause
of CREATE INDEX, 14-71
dictionaries
granting system privileges for, 18-41
dimensional objects
extracting data, 5-67
dimensions
attributes
adding, 10-49
changing, 10-48
defining, 14-39
dropping, 10-49
compiling invalidated, 10-50
creating, 14-36
defining levels, 14-36
examples, 14-39
extracting data, 5-67
granting system privileges for, 18-41
hierarchies
adding, 10-49
changing, 10-48
defining, 14-38
dropping, 10-49
levels
adding, 10-49
defining, 14-37
dropping, 10-49
parent-child hierarchy, 14-37
removing from the database, 17-41
directories. See directory objects

Index-11

directory objects
as aliases for operating system directories, 14-41
auditing, 13-33, 13-34
creating, 14-41
granting system privileges for, 18-41
redefining, 14-42
removing from the database, 17-42
direct-path INSERT, 3-80, 18-54
DISABLE ALL TRIGGERS clause
of ALTER TABLE, 12-79
DISABLE clause
of ALTER INDEX, 10-90
of CREATE TABLE, 16-64
DISABLE DISTRIBUTED RECOVERY clause
of ALTER SYSTEM, 11-64
DISABLE NOVALIDATE constraint state, 8-16
DISABLE PARALLEL DML clause
of ALTER SESSION, 11-46
DISABLE QUERY REWRITE clause
of ALTER MATERIALIZED VIEW, 11-14
of CREATE MATERIALIZED VIEW, 15-21
DISABLE RESTRICTED SESSION clause
of ALTER SYSTEM, 11-68
DISABLE RESUMABLE clause
of ALTER SESSION, 11-48
DISABLE ROW MOVEMENT clause
of ALTER TABLE, 12-38
of CREATE TABLE, 16-16, 16-65
DISABLE STORAGE IN ROW clause
of ALTER TABLE, 12-45
of CREATE TABLE, 16-44
DISABLE TABLE LOCK clause
of ALTER TABLE, 12-79
DISABLE VALIDATE constraint state, 8-16
DISASSOCIATE STATISTICS statement, 17-34
DISCONNECT SESSION clause
of ALTER SYSTEM, 11-63
disk group files
changing permission settings, 10-70
Intelligent Data Placement, 10-65
setting owner or user group, 10-70
disk groups
altering, 10-51
creating, 14-43
a tablespace in, 16-87
failure groups, 10-59, 14-45
files in, 8-32
dropping, 17-43
managing Oracle ADVM volumes, 10-67
rebalancing, 10-62
setting attributes, 10-68, 14-46
specifying files in, 8-32
specifying files in control files, 14-15
diskgroup
Oracle ASM, 15-72
disks
bringing online, 10-62
QUORUM, 14-45
regions, 10-65
REGULAR, 14-45

Index-12

taking offline, 10-61
dispatcher processes
creating additional, 11-73
terminating, 11-73
DISTINCT clause
of SELECT, 19-15
distinct queries, 19-15
distributed queries, 9-16
restrictions on, 9-16
distribution
hints for, 3-101
DML. See data manipulation language (DML)
domain indexes, 14-60, 14-77, 14-87
and LONG columns, 12-47
associating statistics types with, 13-25
creating, prerequisites, 14-77
determining user-defined CPU and I/O
costs, 18-20
disassociating statistics types from, 17-34, 17-50
example, F-1
invoking drop routines for, 18-6
local partitioned, 14-78
modifying, 10-89
parallelizing creation of, 14-78
rebuilding, 10-86
removing from the database, 17-50
system managed, 14-89
domain_index_clause
of CREATE INDEX, 14-63
DOWNGRADE clause
of ALTER DATABASE, 10-19
DRIVING_SITE hint, 3-82
DROP ANY CLUSTER system privilege, 18-40
DROP ANY CONTEXT system privilege, 18-40
DROP ANY CUBE BUILD PROCESS system
privilege, 18-44
DROP ANY CUBE DIMENSION system
privilege, 18-43
DROP ANY CUBE system privilege, 18-43
DROP ANY DIMENSION system privilege, 18-41
DROP ANY DIRECTORY system privilege, 18-41
DROP ANY INDEX system privilege, 18-41
DROP ANY INDEXTYPE system privilege, 18-41
DROP ANY LIBRARY system privilege, 18-42
DROP ANY MATERIALIZED VIEW system
privilege, 18-42
DROP ANY MEASURE FOLDER system
privilege, 18-43
DROP ANY MINING MODEL system
privilege, 18-43
DROP ANY OPERATOR system privilege, 18-44
DROP ANY OUTLINE system privilege, 18-44
DROP ANY PROCEDURE system privilege, 18-44
DROP ANY ROLE system privilege, 18-44
DROP ANY SEQUENCE system privilege, 18-45
DROP ANY SQL PROFILE system privilege, 18-40
DROP ANY SYNONYM system privilege, 18-45
DROP ANY TABLE system privilege, 18-45
DROP ANY TRIGGER system privilege, 18-46
DROP ANY TYPE system privilege, 18-46

DROP ANY VIEW system privilege, 18-47
DROP clause
of ALTER DIMENSION, 10-49
of ALTER INDEXTYPE, 10-98
DROP CLUSTER statement, 17-36
DROP COLUMN clause
of ALTER TABLE, 12-49
DROP CONSTRAINT clause
of ALTER TABLE, 12-55
DROP constraint clause
of ALTER VIEW, 13-15
DROP CONTEXT statement, 17-38
DROP DATABASE LINK statement, 17-40
DROP DATABASE statement, 17-39
DROP DIMENSION statement, 17-41
DROP DIRECTORY statement, 17-42
DROP DISKGROUP statement, 17-43
DROP FLASHBACK ARCHIVE statement, 17-47
DROP FUNCTION statement, 17-48
DROP INDEX statement, 17-50
DROP INDEXTYPE statement, 17-52
DROP JAVA statement, 17-53
DROP LIBRARY statement, 17-54
DROP LOGFILE clause
of ALTER DATABASE, 10-13, 10-31
DROP LOGFILE MEMBER clause
of ALTER DATABASE, 10-13, 10-31
DROP MATERIALIZED VIEW LOG
statement, 17-57
DROP MATERIALIZED VIEW statement, 17-55
DROP OPERATOR statement, 17-59
DROP OUTLINE statement, 17-60
DROP PACKAGE BODY statement, 17-62
DROP PACKAGE statement, 17-62
DROP PARTITION clause
of ALTER INDEX, 10-94
of ALTER TABLE, 12-66
DROP PRIMARY constraint clause
of ALTER TABLE, 12-55
DROP PROCEDURE statement, 17-64
DROP PROFILE statement, 17-65
DROP PROFILE system privilege, 18-44
DROP PUBLIC DATABASE LINK system
privilege, 18-41
DROP PUBLIC SYNONYM system privilege, 18-45
DROP RESTORE POINT statement, 17-66
DROP ROLE statement, 17-67
DROP ROLLBACK SEGMENT statement, 17-68
DROP ROLLBACK SEGMENT system
privilege, 18-44
DROP SEQUENCE statement, 18-2
DROP SUPPLEMENTAL LOG DATA clause
of ALTER DATABASE, 10-33
DROP SUPPLEMENTAL LOG GROUP clause
of ALTER TABLE, 12-35
DROP SYNONYM statement, 18-3
DROP TABLE statement, 18-5
DROP TABLESPACE statement, 18-9
DROP TABLESPACE system privilege, 18-46
DROP TRIGGER statement, 18-12

DROP TYPE BODY statement, 18-15
DROP TYPE statement, 18-13
DROP UNIQUE constraint clause
of ALTER TABLE, 12-55
DROP USER statement, 18-16
DROP USER system privilege, 18-46
DROP VALUES clause
of ALTER TABLE ... MODIFY PARTITION, 12-61
DROP VIEW statement, 18-18
DUAL dummy table, 3-112, 9-16
DUMP function, 5-86
DY datetime format element, 3-65
DYNAMIC_SAMPLING hint, 3-83

E
EBCDIC character set, 3-38
editioning views, 17-17
editions
comments on, 13-47
creating, 14-51
dropping, 17-45
granting system privileges for, 18-41
setting default for database, 10-38
setting for a session, 11-49
embedded SQL, 10-3
precompiler support, 10-3
EMPTY_BLOB function, 5-88
EMPTY_CLOB function, 5-88
ENABLE ALL TRIGGERS clause
of ALTER TABLE, 12-79
ENABLE clause
of ALTER INDEX, 10-90
of ALTER TRIGGER, 13-3
of CREATE TABLE, 16-64
ENABLE DISTRIBUTED RECOVERY clause
of ALTER SYSTEM, 11-64
ENABLE NOVALIDATE constraint state, 8-16
ENABLE PARALLEL DML clause
of ALTER SESSION, 11-46
ENABLE QUERY REWRITE clause
of ALTER MATERIALIZED VIEW, 11-14
of CREATE MATERIALIZED VIEW, 15-21
ENABLE RESTRICTED SESSION clause
of ALTER SYSTEM, 11-68
ENABLE RESUMABLE clause
of ALTER SESSION, 11-48
ENABLE ROW MOVEMENT clause
of ALTER TABLE, 12-38
of CREATE TABLE, 16-16, 16-65
ENABLE STORAGE IN ROW clause
of ALTER TABLE, 12-45
of CREATE TABLE, 16-44
ENABLE TABLE LOCK clause
of ALTER TABLE, 12-79
ENABLE VALIDATE constraint state, 8-16
encoding functions, 5-9
encryption, 16-27
of tablespaces, 8-56, 16-89
encryption keys

Index-13

generating, 11-69
END BACKUP clause
of ALTER DATABASE, 10-25
of ALTER DATABASE ... DATAFILE, 10-27
of ALTER TABLESPACE, 12-96
enterprise users
allowing connection as database users, 13-10
environment functions, 5-9
equality test, 7-5
equijoins, 9-11
defining for a dimension, 14-38
equivalency tests, 7-24
error logging
of DELETE operations, 17-32
of INSERT operations, 18-65
of MERGE operations, 18-77
ERROR_ON_OVERLAP_TIME session
parameter, 11-51
EVALUATE operator, 4-1
EXCEPTIONS INTO clause
of ALTER TABLE, 12-73
EXCHANGE PARTITION clause
of ALTER TABLE, 12-27, 12-72
EXCHANGE SUBPARTITION clause
of ALTER TABLE, 12-27, 12-72
exchanging partitions
restrictions on, 12-73
EXCLUDING NEW VALUES clause
of ALTER MATERIALIZED VIEW LOG, 11-23
of CREATE MATERIALIZED VIEW LOG, 15-32
EXCLUSIVE lock mode, 18-73
exclusive locks
row locks (TX), B-1
table locks (TM), B-1
EXECUTE ANY CLASS system privilege, 18-42
EXECUTE ANY INDEXTYPE system
privilege, 18-42
EXECUTE ANY LIBRARY system privilege, 18-42
EXECUTE ANY OPERATOR system privilege, 18-44
EXECUTE ANY PROCEDURE system
privilege, 18-44
EXECUTE ANY PROGRAM system privilege, 18-42
EXECUTE ANY TYPE system privilege, 18-46
EXECUTE object privilege
on a directory, 18-49
on a library, 18-49
on an object type, 18-49
on an operator, 18-50
execution plans
determining, 18-20
dropping outlines for, 17-60
saving, 15-38
EXEMPT ACCESS POLICY system privilege, 18-47
EXEMPT REDACTION POLICY system
privilege, 18-40
EXISTS condition, 7-20, 7-22
EXISTSNODE function, 5-89
EXP function, 5-90
EXPLAIN PLAN statement, 18-20
explicit data conversion, 3-40, 3-43

Index-14

expressions
CASE, 6-5
changing declared type of, 5-327
column, 6-6
comparing, 5-77
compound, 6-4
computing with the DUAL table, 9-16
CURSOR, 6-7
datetime, 6-8
in SQL syntax, 6-1
interval, 6-10
lists of, 6-16
model, 6-11
object access, 6-13
placeholder, 6-14
scalar subqueries as, 6-14
simple, 6-3
type constructor, 6-14
extended rowids
base 64, 3-27
not directly available, 3-27
extensible indexing
example, F-1
EXTENT MANAGEMENT clause
of CREATE DATABASE, 14-21
of CREATE TABLESPACE, 16-85, 16-90
EXTENT MANAGEMENT DICTIONARY clause
of CREATE TABLESPACE, 16-91
EXTENT MANAGEMENT LOCAL clause
of CREATE DATABASE, 14-26
extents
allocating for partitions, 12-36
allocating for subpartitions, 12-36
allocating for tables, 12-36
restricting access by instances, 10-84
specifying maximum number for an object, 8-53
specifying number allocated upon object
creation, 8-52
specifying the first for an object, 8-51
specifying the percentage of size increase, 8-52
specifying the second for an object, 8-51
external functions, 14-58, 15-48
external LOBs, 3-24
external procedures, 15-48
external tables, 16-37
access drivers, 16-40
altering, 12-56
creating, 16-40
ORACLE_DATAPUMP access driver, 16-40
ORACLE_LOADER access driver, 16-40
restrictions on, 16-40
external users, 15-60, 17-9
EXTRACT (datetime) function, 5-91
EXTRACT (XML) function, 5-94
EXTRACTVALUE function, 5-95

F
FACT hint, 3-83
FAILED_LOGIN_ATTEMPTS parameter

of ALTER PROFILE, 11-33
of CREATE PROFILE, 15-53
failure groups
creating for a disk group, 10-59, 14-45
FEATURE_ID function, 5-96
FEATURE_SET function, 5-97
FEATURE_VALUE function, 5-99
files
specifying as a redo log file group, 8-29
specifying as data files, 8-29
specifying as temp files, 8-29
FIPS
compliance, C-25
flagging, 11-52
FIRST function, 5-101
FIRST_ROWS(n) hint, 3-83
FIRST_VALUE function, 5-103
FLAGGER session parameter, 11-52
flash cache, 8-55
FLASH_CACHE parameter
of STORAGE clause, 8-55
FLASHBACK ANY TABLE system privilege, 18-42,
18-45, 18-47
FLASHBACK ARCHIVE ADMINISTER system
privilege, 18-41
FLASHBACK ARCHIVE object privilege, 18-49
flashback data archives
creating, 14-55
dropping, 17-47
modifying, 10-74
privileges for, 18-41
specifying for a table, 12-39, 16-66
FLASHBACK DATABASE statement, 18-24
flashback queries, 19-17
pseudocolumns for, 2-5
using with inserts, 18-58, 19-78
FLASHBACK TABLE statement, 18-27
floating-point conditions, 7-7
floating-point numbers, 3-12
converting to, 5-297, 5-299
handling NaN, 5-154
FLOOR function, 5-105
FLUSH BUFFER_CACHE clause
of ALTER SYSTEM, 11-65
FLUSH GLOBAL CONTEXT clause
of ALTER SYSTEM, 11-65
FLUSH REDO clause
of ALTER SYSTEM, 11-65
FLUSH SHARED_POOL clause
of ALTER SYSTEM, 11-64
FM format model modifier, 3-67
FOR clause
of CREATE INDEXTYPE, 14-88
of EXPLAIN PLAN, 18-21, 18-27
FOR UPDATE clause
of CREATE MATERIALIZED VIEW, 15-21
of SELECT, 19-13, 19-34
FORCE ANY TRANSACTION system
privilege, 18-47
FORCE clause

of COMMIT, 13-51
of CREATE VIEW, 17-16
of DISASSOCIATE STATISTICS, 17-35
of DROP INDEX, 17-51
of DROP INDEXTYPE, 17-52
of DROP OPERATOR, 17-59
of DROP TYPE, 18-14
of REVOKE, 18-91
of ROLLBACK, 18-27, 18-97
FORCE LOGGING clause
of ALTER DATABASE, 10-29
of ALTER TABLESPACE, 12-98
of CREATE CONTROLFILE, 14-17
of CREATE DATABASE, 14-25
of CREATE TABLESPACE, 16-89
FORCE PARALLEL DML clause
of ALTER SESSION, 11-46
FORCE TRANSACTION system privilege, 18-47
foreign key constraints, 8-10
foreign tables
rowids of, 3-28
format models, 3-56
changing the return format, 3-69
date, 3-60
changing, 3-60
default format, 3-60
format elements, 3-60
maximum length, 3-60
modifiers, 3-67
number, 3-57
number, elements of, 3-57
specifying, 3-69
XML, 3-70
formats
for dates and numbers. See format models
of return values from the database, 3-56
of values stored in the database, 3-56
free lists
specifying for a table, partition, cluster, or
index, 8-54
specifying for LOBs, 16-45
FREELIST GROUPS parameter
of STORAGE clause, 8-54
FREELISTS parameter
of STORAGE clause, 8-53
FREEPOOLS parameter
of LOB storage, 16-45
FROM clause
of queries, 9-12
FROM COLUMNS clause
of DISASSOCIATE STATISTICS, 17-35
FROM FUNCTIONS clause
of DISASSOCIATE STATISTICS, 17-35
FROM INDEXES clause
of DISASSOCIATE STATISTICS, 17-35
FROM INDEXTYPES clause
of DISASSOCIATE STATISTICS, 17-35
FROM PACKAGES clause
of DISASSOCIATE STATISTICS, 17-35
FROM TYPES clause

Index-15

of DISASSOCIATE STATISTICS, 17-35
FROM_TZ function, 5-106
FULL hint, 3-84
full outer joins, 19-24
function expressions
built-in, 6-10
user-defined, 6-10
function-based indexes, 14-60
creating, 14-69
disabling, 10-86, 10-90
enabling, 10-86, 10-90
refreshing, 10-37, 10-38
functions
See also SQL functions
3GL,calling, 15-2
associating statistics types with, 13-25
avoiding run-time compilation, 10-77
built_in
as expressions, 6-10
calling, 13-42
changing the declaration of, 14-59
changing the definition of, 14-59
defining an index on, 14-69
disassociating statistics types from, 17-34
executing, 13-42
external, 14-58, 15-48
inverse distribution, 5-182, 5-185
issuing COMMIT or ROLLBACK
statements, 11-46
linear regression, 5-226
naming rules, 3-114
OLAP, 5-17
recompiling invalid, 10-77
re-creating, 14-59, 14-92
removing from the database, 17-48
statistics, assigning default cost, 13-26
statistics, defining default selectivity, 13-26
stored, 14-58
storing return value of, 13-44
synonyms for, 16-2
user-defined, 5-380
as expressions, 6-10
XML, 5-8
FUNCTIONS clause
of ASSOCIATE STATISTICS, 13-25, 13-26
of DISASSOCIATE STATISTICS, 17-34
FX format model modifier, 3-67

of CREATE INDEX, 14-64, 14-74
GLOBAL QUERY REWRITE system privilege, 18-42
GLOBAL TEMPORARY clause
of CREATE TABLE, 16-25
global users, 15-60, 17-10
GLOBAL_TOPIC_ENABLED system
parameter, 11-73
globally partitioned indexes, 14-74, 14-75
GRANT ANY OBJECT PRIVILEGE system
privilege, 18-48
GRANT ANY PRIVILEGE system privilege, 18-48
GRANT ANY ROLE system privilege, 18-44
GRANT CONNECT THROUGH clause
of ALTER USER, 13-7, 13-10
GRANT statement
locks, B-4
GRAPHIC data type
DB2, 3-29
SQL/DS, 3-29
greater than or equal to tests, 7-5
greater than tests, 7-5
GREATEST function, 5-107
GROUP BY clause
CUBE extension, 19-28
identifying duplicate groupings, 5-109
of SELECT and subqueries, 19-10, 19-27
ROLLUP extension of, 19-27
group comparison conditions, 7-6
group separator
specifying, 3-58
GROUP_ID function, 5-109
GROUPING function, 5-110
grouping sets, 19-28
GROUPING SETS clause
of SELECT and subqueries, 19-28
GROUPING_ID function, 5-111
groupings
filtering out duplicate, 5-109
GUARD ALL clause
of ALTER DATABASE, 10-42
GUARD clause
of ALTER DATABASE, 10-42
overriding, 11-46
GUARD NONE clause
of ALTER DATABASE, 10-42
GUARD STANDBY clause
of ALTER DATABASE, 10-42

G

H

general comparison functions, 5-6
general recovery clause
of ALTER DATABASE, 10-10, 10-19
geoimaging, 3-34
global application context information
flushing, 11-65
global indexes. See indexes, globally partitioned
GLOBAL PARTITION BY HASH clause
of CREATE INDEX, 14-74
GLOBAL PARTITION BY RANGE clause

hash clusters
creating, 14-5
single-table, creating, 14-6
specifying hash function for, 14-6
HASH hint, 3-84
HASH IS clause
of CREATE CLUSTER, 14-6
hash partitioning clause
of CREATE TABLE, 16-24, 16-55
hash partitions

Index-16

adding, 12-64
coalescing, 12-60
HASHKEYS clause
of CREATE CLUSTER, 14-5
HAVING condition
of GROUP BY clause, 19-28
heap-organized tables
creating, 16-6
hexadecimal value
returning, 3-59
HEXTORAW function, 5-112
hierarchical functions, 5-7
hierarchical queries, 9-3, 19-26
child rows, 2-2, 9-4
illustrated, 2-2
leaf rows, 2-2
operators in, 4-5
CONNECT_BY_ROOT, 4-5
PRIOR, 4-5
ordering, 19-33
parent rows, 2-2, 9-4
pseudocolumns in, 2-1
CONNECT_BY_ISCYCLE, 2-1
CONNECT_BY_ISLEAF, 2-2
LEVEL, 2-2
retrieving root and node values, 5-278
hierarchical query clause
of SELECT and subqueries, 19-10
hierarchies
adding to a dimension, 10-49
dropping from a dimension, 10-49
of dimensions, defining, 14-38
HIERARCHY clause
of CREATE DIMENSION, 14-37, 14-38
high water mark
of clusters, 10-7
of indexes, 10-84
of tables, 12-36, 13-19
hints, 9-2
ALL_ROWS, 3-79
APPEND, 3-80
APPEND_VALUES, 3-80
CACHE, 3-81
CLUSTER, 3-82
CURSOR_SHARING_EXACT, 3-82
DRIVING_SITE, 3-82
DYNAMIC_SAMPLING, 3-83
FACT, 3-83
FIRST_ROWS(n), 3-83
FULL, 3-84
HASH, 3-84
in SQL statements, 3-74
INDEX, 3-85
INDEX_ASC, 3-86
INDEX_COMBINE, 3-86
INDEX_DESC, 3-87
INDEX_FFS, 3-87
INDEX_JOIN, 3-87
INDEX_SS, 3-88
INDEX_SS_ASC, 3-88

INDEX_SS_DESC, 3-89
LEADING, 3-89
location syntax, 3-75
MERGE, 3-89
MODEL_MIN_ANALYSIS, 3-90
MONITOR, 3-90
NO_EXPAND, 3-91
NO_FACT, 3-91
NO_INDEX, 3-92
NO_INDEX_FFS, 3-92
NO_INDEX_SS, 3-92
NO_MERGE, 3-93
NO_MONITOR, 3-93
NO_PARALLEL, 3-93
NO_PARALLEL_INDEX, 3-94
NO_PUSH_PRED, 3-94
NO_PUSH_SUBQ, 3-95
NO_PX_JOIN_FILTER, 3-95
NO_QUERY_TRANSFORMATION, 3-95
NO_RESULT_CACHE, 3-95
NO_REWRITE, 3-95
NO_STAR_TRANSFORMATION, 3-96
NO_STATEMENT_QUEUING, 3-96
NO_UNNEST, 3-96
NO_USE_HASH, 3-96
NO_USE_MERGE, 3-97
NO_USE_NL, 3-97
NO_XML_QUERY_REWRITE, 3-97
NO_XMLINDEX_REWRITE, 3-98
NOAPPEND, 3-91
NOCACHE, 3-91
NOPARALLEL, 3-93
NOPARALLEL_INDEX, 3-94
NOREWRITE, 3-95
OPT_PARAM, 3-98
ORDERED, 3-98
PARALLEL, 3-98
PARALLEL_INDEX, 3-101
passing to the optimizer, 19-73
PQ_DISTRIBUTE, 3-101
PUSH_PRED, 3-104
PUSH_SUBQ, 3-104
PX_JOIN_FILTER, 3-104
QB_NAME, 3-104
REWRITE, 3-106
specifying a query block, 3-75
STAR_TRANSFORMATION, 3-106
STATEMENT_QUEUING, 3-107
syntax, 3-77
UNNEST, 3-107
USE_CONCAT, 3-107
USE_HASH, 3-108
USE_MERGE, 3-108
USE_NL, 3-108
USE_NL_WITH_INDEX, 3-109
histograms
creating equiwidth, 5-347
Hybrid Columnar Compression, 16-35

Index-17

I
IDENTIFIED BY clause
of ALTER ROLE. See CREATE ROLE
of CREATE DATABASE LINK, 14-34
of SET ROLE, 19-62
IDENTIFIED EXTERNALLY clause
of ALTER ROLE. See CREATE ROLE
of ALTER USER. See CREATE USER
of CREATE ROLE, 15-60
of CREATE USER, 17-9
IDENTIFIED GLOBALLY clause
of ALTER ROLE. See CREATE ROLE
of CREATE ROLE, 15-60
of CREATE USER, 17-10
identifier functions, 5-9
IDLE_TIME parameter
of ALTER PROFILE, 11-32
IEEE754
floating-point arithmetic, 3-13
Oracle conformance with, 3-13
IGNORE_ROW_ON_DUPKEY_INDEX hint, 3-85
IMMEDIATE clause
of SET CONSTRAINTS, 19-59
implicit data conversion, 3-40, 3-42
IN conditions, 7-23
INCLUDING CONTENTS clause
of DROP TABLESPACE, 18-10
INCLUDING DATAFILES clause
of ALTER DATABASE TEMPFILE DROP
clause, 10-28
INCLUDING NEW VALUES clause
of ALTER MATERIALIZED VIEW LOG, 11-23
of CREATE MATERIALIZED VIEW LOG, 15-32
INCLUDING TABLES clause
of DROP CLUSTER, 17-36
incomplete object types, 17-3
creating, 17-3
INCREMENT BY clause
of ALTER SEQUENCE. See CREATE SEQUENCE
INCREMENT BY parameter
of CREATE SEQUENCE, 15-68
incremental
and block change tracking, 10-40
INDEX clause
of ANALYZE, 13-20
of CREATE CLUSTER, 14-5
INDEX hint, 3-85
index keys
compression, 10-81
INDEX object privilege
on a table, 18-51
index partitions
creating subpartitions, 14-66
index subpartitions, 14-66
INDEX_ASC hint, 3-86
INDEX_COMBINE hint, 3-86
INDEX_DESC hint, 3-87
INDEX_FFS hint, 3-87
INDEX_JOIN hint, 3-87
INDEX_SS hint, 3-88
Index-18

INDEX_SS_ASC hint, 3-88
INDEX_SS_DESC hint, 3-89
indexed clusters
creating, 14-5
indexes, 10-85
allocating new extents for, 10-84
application-specific, 14-87
ascending, 14-71
based on indextypes, 14-77
bitmap, 14-67
bitmap join, 14-79
B-tree, 14-60
changing attributes, 10-85
changing parallelism of, 10-85
collecting statistics on, 13-20
on composite-partitioned tables, 14-76
creating, 14-60
creating as unusable, 14-79
creating on a cluster, 14-61
creating on a table, 14-61
deallocating unused space from, 10-84
descending, 14-71
and query rewrite, 14-71
as function-based indexes, 14-71
direct-path inserts, logging, 10-85
domain, 14-60, 14-77, 14-87
domain, example, F-1
dropping index partitions, 17-50
examples, 14-80
full fast scans, 3-87
function-based, 14-60
creating, 14-69
global partitioned, creating, 14-64
globally partitioned, 14-74, 14-75
updating, 12-76
granting system privileges for, 18-41
on hash-partitioned tables, 14-76
invisible to the optimizer, 10-91, 14-73
join, bitmap, 14-79
key compression of, 10-88
key compression, enabling, 10-86
keys, eliminating repetition, 10-86
local domain, 14-78
locally partitioned, 14-75
logging rebuild operations, 10-86
marking as UNUSABLE, 10-90
merging block contents, 10-86
merging contents of index blocks, 10-91
merging contents of index partition blocks, 10-93
modifying attributes, 10-86
moving, 10-86
on clusters, 14-68
on composite-partitioned tables, creating, 14-66
on hash-partitioned tables
creating, 14-66
on index-organized tables, 14-68
on list-partitioned tables
creating, 14-65
on nested table storage tables, 14-68
on partitioned tables, 14-68

on range-partitioned tables, creating, 14-65
on scalar typed object attributes, 14-68
on table columns, 14-68
on XMLType tables, 14-81
online, 14-73
parallelizing creation of, 14-74
partitioned, 3-119, 14-60
user-defined, 14-74
partitioning, 14-74
partitions, 14-74
adding hash, 10-92
adding new, 10-94
changing default attributes, 10-92
changing physical attributes, 10-85
changing storage characteristics, 10-92
coalescing hash partitions, 10-94
deallocating unused space from, 10-84
dropping, 10-94
marking UNUSABLE, 10-94, 12-74
modifying the real characteristics, 10-93
preventing use of, 10-90
rebuilding, 10-86
rebuilding unusable, 12-74
re-creating, 10-86
removing, 10-92
renaming, 10-93
specifying tablespace for, 10-86, 10-88
splitting, 10-92, 10-94
preventing use of, 10-90
purging from the recycle bin, 18-83
on range-partitioned tables, 14-76
rebuilding, 10-86
rebuilding while online, 10-88
re-creating, 10-86
removing from the database, 17-50
renaming, 10-86, 10-91
reverse, 10-86, 10-87, 10-89, 14-73
specifying tablespace for, 10-86, 10-88
statistics on usage, 10-91
subpartitions
allocating extents for, 10-94
changing default attributes, 10-92
changing physical attributes, 10-85
changing storage characteristics, 10-92
deallocating unused space from, 10-84, 10-94
marking UNUSABLE, 10-94
modifying, 10-86
moving, 10-86
preventing use of, 10-90
rebuilding, 10-86
re-creating, 10-86
renaming, 10-93
specifying tablespace for, 10-86, 10-88
tablespace containing, 14-72
unique, 14-67
unsorted, 14-72
used to enforce constraints, 12-56, 16-65
validating structure, 13-21
INDEXES clause
of ASSOCIATE STATISTICS, 13-25, 13-26

of DISASSOCIATE STATISTICS, 17-34
index-organized tables
bitmap indexes on, creating, 16-38
creating, 16-6
mapping tables, 12-77
creating, 16-38
moving, 12-62
merging contents of index blocks, 12-41
modifying, 12-39, 12-41
moving, 12-77
overflow segments
specifying storage, 12-40, 16-56
partitioned, updating secondary indexes, 10-93
PCT_ACCESS_DIRECT statistics, 13-19
primary key indexes
coalescing, 12-40
rebuilding, 12-76
rowids of, 3-28
secondary indexes, updating, 10-92
INDEXTYPE clause
of CREATE INDEX, 14-63, 14-77
indextypes
adding operators, 10-97
altering, 10-97
associating statistics types with, 13-25
changing implementation type, 10-97
comments on, 13-47
creating, 14-87
disassociating statistics types from, 17-34, 17-52
drop routines, invoking, 17-50
granting system privileges for, 18-41
indexes based on, 14-77
instances, 14-60
removing from the database, 17-52
INDEXTYPES clause
of ASSOCIATE STATISTICS, 13-25, 13-26
of DISASSOCIATE STATISTICS, 17-34
in-doubt transactions
forcing, 13-51
forcing commit of, 13-51
forcing rollback, 18-27, 18-97
rolling back, 18-96
inequality test, 7-5
INITCAP function, 5-113
INITIAL parameter
of STORAGE clause, 8-51
initialization parameters
changing session settings, 11-49
setting using ALTER SESSION, 11-50
INITIALIZED EXTERNALLY clause
of CREATE CONTEXT, 14-10
INITIALIZED GLOBALLY clause
of CREATE CONTEXT, 14-10
INITIALLY DEFERRED clause
of constraints, 8-15
INITIALLY IMMEDIATE clause
of constraints, 8-15
INITRANS parameter
of ALTER CLUSTER, 10-6
of ALTER INDEX, 10-85

Index-19

of ALTER MATERIALIZED VIEW LOG, 11-19
of ALTER TABLE, 12-34
of CREATE INDEX. See CREATE TABLE
of CREATE MATERIALIZED VIEW LOG. See
CREATE TABLE
of CREATE MATERIALIZED VIEW. See CREATE
TABLE
of CREATE TABLE, 8-46
inline constraints
of ALTER TABLE, 12-43
of CREATE TABLE, 16-30
inline views, 9-14
inner joins, 9-12, 19-24
inner-N reporting, 5-236
INSERT
direct-path versus conventional, 18-54
INSERT ANY CUBE DIMENSION system
privilege, 18-43
INSERT ANY MEASURE FOLDER system
privilege, 18-43
INSERT ANY TABLE system privilege, 18-45
INSERT clause
of MERGE, 18-75
INSERT object privilege
on a table, 18-51
on a view, 18-51
on an OLAP cube dimension, 18-50
on an OLAP measures folder, 18-50
INSERT statement, 18-54
append, 3-80
error logging, 18-65
INSERTCHILDXML function, 5-114
INSERTCHILDXMLAFTER function, 5-116
INSERTCHILDXMLBEFORE function, 5-117
inserts
and simultaneous update, 18-74
conditional, 18-63
conventional, 18-54
direct-path, 18-54
multitable, 18-63
examples, 18-68
single-table, 18-58
using MERGE, 18-75
INSERTXMLAFTER function, 5-118
INSERTXMLBEFORE function, 5-119
instance recovery
continue after interruption, 10-19
INSTANCE session parameter, 11-52
instances
making index extents available to, 10-84
setting parameters for, 11-71
INSTR function, 5-120
INSTR2 function, 5-120
INSTR4 function, 5-120
INSTRB function, 5-120
INSTRC function, 5-120
integers
generating unique, 15-67
in SQL syntax, 3-47
precision of, 3-47

Index-20

syntax of, 3-47
integrity constraints. See constraints
internal LOBs, 3-24
International Organization for Standardization
(ISO), C-1
standards, 1-1, C-1
INTERSECT set operator, 4-5, 19-33
interval
arithmetic, 3-20
data types, 3-16
literals, 3-53
interval conditions, 7-22
INTERVAL DAY TO SECOND data type, 3-19
INTERVAL expressions, 6-10
interval partitioning, 12-58, 16-52
changing the interval, 12-58
INTERVAL YEAR TO MONTH data type, 3-19
INTO clause
of EXPLAIN PLAN, 18-21
of INSERT, 18-58
INVALIDATE GLOBAL INDEXES clause
of ALTER TABLE, 12-76
inverse distribution functions, 5-182, 5-185
invoker rights
altering for a Java class, 10-101
defining for a Java class, 14-92, 14-93
IS [NOT] EMPTY conditions, 7-12
IS ANY condition, 7-9
IS NOT NULL operator, 7-20
IS NULL operator, 7-20
IS OF type condition, 7-25
IS PRESENT condition, 7-10
ISO. See International Organization for
Standardization (ISO)
ISOLATION_LEVEL session parameter, 11-52
ITERATION_NUMBER function, 5-122

J
Java
class
creating, 14-91, 14-92
dropping, 17-53
resolving, 10-100, 14-92
Java source schema object
creating, 14-92
resource
creating, 14-91, 14-92
dropping, 17-53
schema object
name resolution of, 14-94
source
compiling, 10-100, 14-92
creating, 14-91
dropping, 17-53
job scheduler object privileges, 18-42
JOIN clause
of CREATE DIMENSION, 14-37
JOIN KEY clause
of ALTER DIMENSION, 10-49

of CREATE DIMENSION, 14-38
join views
example, 17-23
making updatable, 17-21
modifying, 17-30, 18-59, 19-75
joins, 9-11
antijoins, 9-14
conditions
defining, 9-11
cross, 19-23
equijoins, 9-11
full outer, 19-24
inner, 9-12, 19-24
left outer, 19-24
natural, 19-24
outer, 9-12
and data densification, 9-12
on grouped tables, 9-12
restrictions, 9-13
parallel, 3-101
right outer, 19-24
self, 9-12
semijoins, 9-14
without join conditions, 9-12
Julian dates, 3-17

K
KEEP keyword
of FIRST function, 5-101
of LAST function, 5-101
with aggregate functions, 5-10
key compression, 16-39
definition, 10-88
disabling, 10-88, 14-72
enabling, 10-86
of index rebuild, 12-78
of indexes
disabling, 10-88
of index-organized tables, 16-39
key-preserved tables, 17-21
keywords, 3-112
in object names, 3-112
optional, A-3
required, A-2
KILL SESSION clause
of ALTER SYSTEM, 11-64

L
LAG function, 5-124
large object functions, 5-7
large objects. See LOB data types
LAST function, 5-126
LAST_DAY function, 5-127
LAST_VALUE function, 5-128
LEAD function, 5-131
LEADING hint, 3-89
LEAST function, 5-133
left outer joins, 19-24

LENGTH function, 5-135
LENGTH2 function, 5-135
LENGTH4 function, 5-135
LENGTHB function, 5-135
LENGTHC function, 5-135
less than tests, 7-5
LEVEL clause
of ALTER DIMENSION, 10-48
of CREATE DIMENSION, 14-36, 14-37
level columns
specifying default values, 16-30
LEVEL pseudocolumn, 2-2, 19-26
levels
adding to a dimension, 10-49
dropping from a dimension, 10-49
of dimensions, defining, 14-37
libraries
creating, 15-2
granting system privileges for, 18-42
re-creating, 15-3
removing from the database, 17-54
library units. See Java schema objects
LIKE conditions, 7-14
linear regression functions, 5-226
LIST CHAINED ROWS clause
of ANALYZE, 13-22
list partitioning
adding default partition, 12-65
adding partitions, 12-65
adding values, 12-61
creating a default partition, 16-54
creating partitions, 16-54
dropping values, 12-61
merging default with nondefault partitions, 12-70
splitting default partition, 12-68
list subpartitions
adding, 12-60
LISTAGG function, 5-136
listeners
registering, 11-71
literals, 3-45
datetime, 3-50
interval, 3-53
LN function, 5-138
LNNVL function, 5-139
LOB columns
adding, 12-42
compressing, 16-46
creating from LONG columns, 3-14, 12-47
deduplication, 16-45
defining properties
for materialized views, 15-9
encrypting, 16-46
modifying, 12-46
modifying storage, 12-45
restricted in joins, 9-11
restrictions on, 3-25
storage characteristics of materialized
views, 11-10
LOB data types, 3-24

Index-21

LOB storage clause
for partitions, 12-45
of ALTER MATERIALIZED VIEW, 11-5, 11-10
of ALTER TABLE, 12-14, 12-45
of CREATE MATERIALIZED VIEW, 15-9, 15-10,
15-14, 15-15
of CREATE TABLE, 16-13, 16-42
LOBs
attributes, initializing, 3-25
columns
difference from LONG and LONG RAW, 3-24
populating, 3-25
external, 3-24
internal, 3-24
locators, 3-24
logging attribute, 16-34
modifying physical attributes, 12-53
number of bytes manipulated in, 16-44
saving old versions, 16-44
saving values in a cache, 12-45, 16-62
specifying directories for, 14-41
storage
attributes, 16-42
characteristics, 8-46
in-line, 16-43
tablespace for
defining, 16-34
LOCAL clause
of CREATE INDEX, 14-65, 14-75
local users, 15-60, 17-9
locale independent, 3-62
locally managed tablespaces
altering, 12-93
storage attributes, 8-51
locally partitioned indexes, 14-75
LOCALTIMESTAMP function, 5-140
location transparency, 16-2
LOCK ANY TABLE system privilege, 18-45
LOCK TABLE statement, 18-71
locking, overriding automatic, 18-71
locks
data, B-1
dictionary, B-4
row (TX), B-1
table (TM), B-1
locks. See table locks
log data
collection during update operations, 10-32
log file clauses
of ALTER DATABASE, 10-13
log files
adding, 10-28
dropping, 10-28
modifying, 10-28
registering, 10-36
renaming, 10-25
specifying for the database, 14-24
LOG function, 5-141
log groups
adding, 12-35

Index-22

dropping, 12-35
LOGFILE clause
OF CREATE DATABASE, 14-24
LOGFILE GROUP clause
of CREATE CONTROLFILE, 14-15
logging
and redo log size, 8-39
specifying minimal, 8-39
supplemental
dropping, 10-33
supplemental, adding log groups, 12-35
supplemental, dropping log groups, 12-35
LOGGING clause
of ALTER INDEX, 10-85
of ALTER MATERIALIZED VIEW, 11-11
of ALTER MATERIALIZED VIEW LOG, 11-21
of ALTER TABLE, 12-35
of ALTER TABLESPACE, 12-98
of CREATE MATERIALIZED VIEW, 15-14
of CREATE MATERIALIZED VIEW LOG, 15-30
of CREATE TABLE, 16-34
of CREATE TABLESPACE, 16-89
logical conditions, 7-8
logical standby database
aborting, 10-37, 10-38
activating, 10-34
stopping, 10-37, 10-38
LOGICAL_READS_PER_CALL parameter
of ALTER PROFILE, 11-32
LOGICAL_READS_PER_SESSION parameter
of ALTER PROFILE, 11-32
of ALTER RESOURCE COST, 11-35
LogMiner
supplemental logging, 12-35, 16-31
LONG columns
and domain indexes, 12-47
converting to LOB, 3-14, 12-47
restrictions on, 3-15
to store text strings, 3-14
to store view definitions, 3-14
where referenced from, 3-14
LONG data type, 3-14
in triggers, 3-15
LONG RAW data type, 3-23
converting from CHAR data, 3-24
LONG VARGRAPHIC data type
DB2, 3-29
SQL/DS, 3-29
LOWER function, 5-142
LPAD function, 5-143
LTRIM function, 5-144

M
MAKE_REF function, 5-145
MANAGE SCHEDULER system privilege, 18-42
MANAGE TABLESPACE system privilege, 18-46
managed recovery
of database, 10-11
managed standby recovery

as background process, 10-23
create a logical standby from the physical
standby, 10-24
overriding delays, 10-23
returning control during, 10-23, 10-24
terminating existing, 10-23, 10-24
MANAGED STANDBY RECOVERY clause
of ALTER DATABASE, 10-22
MAPPING TABLE clause
of ALTER TABLE, 12-62, 12-77
mapping tables
of index-organized tables, 12-77, 16-38
modifying, 12-41
master databases, 15-4
master tables, 15-4
MATCHES condition, 7-2
MATCHES operator, 4-1
materialized join views, 15-27
materialized view logs, 15-27
creating, 15-27
excluding new values from, 11-23
logging changes to, 11-21
object ID based, 11-22
parallelizing creation, 15-30
partition attributes, changing, 11-21
partitioned, 15-31
physical attributes
changing, 11-21
specifying, 15-30
purging, 11-23, 15-33
removing from the database, 17-57
required for fast refresh, 15-27
rowid based, 11-22
saving new values in, 11-23
saving old values in, 15-32
storage attributes
specifying, 15-30
materialized views, 15-16
allowing update of, 15-21
changing from rowid-based to
primary-key-based, 11-13
changing to primary-key-based, 11-22
complete refresh, 11-12, 15-17
compression of, 11-10, 15-14
constraints on, 8-17
creating, 15-4
creating comments about, 13-46
for data warehousing, 15-4
degree of parallelism, 11-10, 11-21
during creation, 15-16
enabling and disabling query rewrite, 15-21
examples, 15-23, 15-33
fast refresh, 11-12, 15-17
forced refresh, 11-13
granting system privileges for, 18-42
index characteristics
changing, 11-10
indexes that maintain, 15-16
join, 15-27
LOB storage attributes, 11-10

logging changes to, 11-11
master table, dropping, 17-56
object type, creating, 15-12
partitions, 11-10
compression of, 11-10, 15-14
physical attributes, 15-14
changing, 11-9
primary key, 15-18
recording values in master table, 11-22
query rewrite
eligibility for, 8-17
enabling and disabling, 11-14
re-creating during refresh, 11-12
refresh, 10-37
after DML on master table, 11-13, 15-18
mode, changing, 11-12
on next COMMIT, 11-13, 15-17
using trusted constraints, 15-20
refresh, time, changing, 11-12
refreshing, 10-38
removing from the database, 17-55
for replication, 15-4
restricting scope of, 15-12
retrieving data from, 19-4
revalidating, 11-15
rowid, 15-19
rowid values
recording in master table, 11-22
saving blocks in a cache, 11-11
storage attributes, 15-14
changing, 11-9
subquery, 15-21
suppressing creation of default index, 15-16
synonyms for, 16-2
when to populate, 15-16
MAX function, 5-146
MAXDATAFILES parameter
of CREATE CONTROLFILE, 14-16
of CREATE DATABASE, 14-23
MAXEXTENTS parameter
of STORAGE clause, 8-53
MAXINSTANCES parameter
of CREATE CONTROLFILE, 14-16
OF CREATE DATABASE, 14-23
MAXLOGFILES parameter
of CREATE CONTROLFILE, 14-16
of CREATE DATABASE, 14-25
MAXLOGHISTORY parameter
of CREATE CONTROLFILE, 14-16
of CREATE DATABASE, 14-25
MAXLOGMEMBERS parameter
of CREATE CONTROLFILE, 14-16
of CREATE DATABASE, 14-25
MAXSIZE clause
of ALTER DATABASE, 10-12
MAXTRANS parameter
of physical_attributes_clause, 8-46
MAXVALUE parameter
of ALTER SEQUENCE. See CREATE SEQUENCE
of CREATE SEQUENCE, 15-69

Index-23

media recovery
avoid on startup, 10-27
designing, 10-19
disabling, 10-25
from specified redo logs, 10-19
of database, 10-19
of datafiles, 10-19
of standby database, 10-19
of tablespaces, 10-19
performing ongoing, 10-22
preparing for, 10-28
restrictions, 10-20
sustained standby recovery, 10-22
MEDIAN function, 5-148
median values, 5-185
MEMBER conditions, 7-13
membership conditions, 7-13, 7-23
MERGE ANY VIEW system privilege, 18-47
MERGE hint, 3-89
MERGE PARTITIONS clause
of ALTER TABLE, 12-70
MERGE statement, 18-74
deletes during, 18-76
error logging, 18-77
inserts during, 18-76
updates during, 18-76
MERGE VIEW object privilege on a view, 18-51
merge_insert_clause
of MERGE, 18-76
migrated rows
listing, 13-22
of clusters, 13-20
MIN function, 5-150
MINEXTENTS parameter
of STORAGE clause, 8-52
MINIMIZE RECORDS PER BLOCK clause
of ALTER TABLE, 12-38
MINIMUM EXTENT clause
of ALTER TABLESPACE, 12-94
of CREATE TABLESPACE, 16-88
mining models
auditing, 13-33
comments on, 13-47
MINUS set operator, 4-5, 19-33
MINVALUE parameter
of ALTER SEQUENCE. See CREATE SEQUENCE
of CREATE SEQUENCE, 15-69
MOD function, 5-152
MODE clause
of LOCK TABLE, 18-72
MODEL clause
of SELECT, 19-11, 19-28
model conditions, 7-9
IS ANY, 7-9
IS PRESENT, 7-10
model expression, 6-11
model functions, 5-17
MODEL_MIN_ANALYSIS hint, 3-90
MODIFY clause
of ALTER TABLE, 12-46

Index-24

MODIFY CONSTRAINT clause
of ALTER TABLE, 12-11, 12-55
of ALTER VIEW, 13-15
MODIFY DEFAULT ATTRIBUTES clause
of ALTER INDEX, 10-82, 10-92
of ALTER TABLE, 12-57
MODIFY LOB storage clause
of ALTER MATERIALIZED VIEW, 11-6, 11-10
of ALTER TABLE, 12-53
MODIFY NESTED TABLE clause
of ALTER TABLE, 12-10, 12-52
MODIFY PARTITION clause
of ALTER INDEX, 10-93
of ALTER MATERIALIZED VIEW, 11-10
of ALTER TABLE, 12-59
MODIFY scoped_table_ref_constraint clause
of ALTER MATERIALIZED VIEW, 11-11
MODIFY SUBPARTITION clause
of ALTER INDEX, 10-83, 10-94
MODIFY VARRAY clause
of ALTER TABLE, 12-17, 12-54
MON datetime format element, 3-65
MONITOR hint, 3-90
MONITORING USAGE clause
of ALTER INDEX, 10-91
MONTH datetime format element, 3-65
MONTHS_BETWEEN function, 5-153
MOUNT clause
of ALTER DATABASE, 10-18
MOVE clause
of ALTER TABLE, 12-31, 12-76
MOVE ONLINE clause
of ALTER TABLE, 12-77
MOVE SUBPARTITION clause
of ALTER TABLE, 12-63
MTS. See shared server
multilevel collections, 16-48
multiset conditions, 7-11
MULTISET EXCEPT operator, 4-6
MULTISET INTERSECT operator, 4-7
MULTISET keyword
of CAST function, 5-35
multiset operators, 4-6
MULTISET EXCEPT, 4-6
MULTISET INTERSECT, 4-7
MULTISET UNION, 4-8
MULTISET UNION operator, 4-8
multitable inserts, 18-63
conditional, 18-63
examples, 18-68
unconditional, 18-63
multi-threaded server. See shared server

N
NAME clause
of SET TRANSACTION, 19-65
NAMED clause
of CREATE JAVA, 14-93
namespaces

and object naming rules, 3-112
database, 3-113
for nonschema objects, 3-113
for schema objects, 3-112
NANVL function, 5-154
national character set
changing, 10-38
fixed versus variable width, 3-10
multibyte character data, 3-26
multibyte character sets, 3-9, 3-10
variable-length strings, 3-9
NATIONAL CHARACTER SET parameter
of CREATE DATABASE, 14-23
natural joins, 19-24
NCHAR data type, 3-9
NCHR function, 5-155
NCLOB data type, 3-26
transactional support of, 3-26
nested subqueries, 9-14
NESTED TABLE clause
of ALTER TABLE, 12-12, 12-44
of CREATE TABLE, 16-12, 16-48
nested tables, 3-31, 5-188, 5-245, 7-12
changing returned value, 12-52
combining, 4-6
compared with varrays, 3-39
comparison rules, 3-39
creating, 17-3
creating from existing columns, 5-50
defining as index-organized tables, 12-44
determining hierarchy, 7-13
dropping the body of, 18-15
dropping the specification of, 18-13
in materialized views, 15-9, 15-10
indexing columns of, 14-69
modifying, 12-52
modifying column properties, 12-12
multilevel, 16-48
partitioned nested table columns, 12-68
storage characteristics of, 12-44, 16-48
NEW_TIME function, 5-156
NEXT clause
of ALTER MATERIALIZED VIEW ...
REFRESH, 11-13
NEXT parameter
of STORAGE clause, 8-51
NEXT_DAY function, 5-157
NEXTVAL pseudocolumn, 2-3, 15-67
NLS_CHARSET_DECL_LEN function, 5-158
NLS_CHARSET_ID function, 5-159
NLS_CHARSET_NAME function, 5-160
NLS_DATE_LANGUAGE initialization
parameter, 3-65
NLS_INITCAP function, 5-161
NLS_LANGUAGE initialization parameter, 3-65,
9-10
NLS_LENGTH_SEMANTICS initialization parameter
overriding, 3-9
NLS_LOWER function, 5-162
NLS_SORT initialization parameter, 9-10

NLS_TERRITORY initialization parameter, 3-65
NLS_UPPER function, 5-163
NLSSORT function, 5-164
NO FORCE LOGGING clause
of ALTER DATABASE, 10-29
of ALTER TABLESPACE, 12-98
NO_EXPAND hint, 3-91
NO_FACT hint, 3-91
NO_INDEX hint, 3-92
NO_INDEX_FFS hint, 3-92
NO_INDEX_SS hint, 3-92
NO_MERGE hint, 3-93
NO_MONITOR hint, 3-93
NO_PARALLEL hint, 3-93
NO_PARALLEL_INDEX, 3-94
NO_PUSH_PRED hint, 3-94
NO_PUSH_SUBQ hint, 3-95
NO_PX_JOIN_FILTER hint, 3-95
NO_QUERY_TRANSFORMATION hint, 3-95
NO_RESULT_CACHE hint, 3-95
NO_REWRITE hint, 3-95
NO_STAR_TRANSFORMATION hint, 3-96
NO_STATEMENT_QUEUING hint, 3-96
NO_UNNEST hint, 3-96
NO_USE_HASH hint, 3-96
NO_USE_MERGE hint, 3-97
NO_USE_NL hint, 3-97
NO_XML_QUERY_REWRITE hint, 3-97
NO_XMLINDEX_REWRITE hint, 3-98
NOAPPEND hint, 3-91
NOARCHIVELOG clause
of ALTER DATABASE, 10-13, 10-28
of CREATE CONTROLFILE, 14-16
OF CREATE DATABASE, 10-19, 14-25
NOAUDIT statement, 18-79
locks, B-4
NOCACHE clause
of ALTER CLUSTER, 10-7
of ALTER MATERIALIZED VIEW, 11-11
of ALTER MATERIALIZED VIEW LOG, 11-22
of ALTER SEQUENCE. See CREATE SEQUENCE
of ALTER TABLE, 16-62
of CREATE CLUSTER, 14-7
of CREATE MATERIALIZED VIEW, 15-15
of CREATE MATERIALIZED VIEW LOG, 15-30
of CREATE SEQUENCE, 15-70
NOCACHE hint, 3-91
NOCOMPRESS clause
of ALTER INDEX ... REBUILD, 10-88
of CREATE INDEX, 14-72
of CREATE TABLE, 16-39
NOCYCLE parameter
of ALTER SEQUENCE. See CREATE
SEQUENCE, 11-43
of CREATE SEQUENCE, 15-69
NOFORCE clause
of CREATE JAVA, 14-92
of CREATE VIEW, 17-17
NOLOGGING mode
and force logging mode, 8-38

Index-25

for nonpartitioned objects, 8-38
for partitioned objects, 8-38
NOMAXVALUE parameter
of ALTER SEQUENCE. See CREATE SEQUENCE
of CREATE SEQUENCE, 15-69
NOMINIMIZE RECORDS PER BLOCK clause
of ALTER TABLE, 12-38
NOMINVALUE parameter
of ALTER SEQUENCE. See CREATE
SEQUENCE, 11-43
of CREATE SEQUENCE, 15-69
NOMONITORING USAGE clause
of ALTER INDEX, 10-91
NONE clause
of SET ROLE, 19-62
nonempty subsets of, 5-188
nonequivalency tests, 7-24
nonschema objects
list of, 3-110
namespaces, 3-113
NOORDER parameter
of ALTER SEQUENCE. See CREATE
SEQUENCE, 11-43
of CREATE SEQUENCE, 15-70
NOPARALLEL clause
of CREATE INDEX, 8-42, 16-64
NOPARALLEL hint, 3-93
NOPARALLEL_INDEX hint, 3-94
NORELY clause
of constraints, 8-17
NORESETLOGS clause
of CREATE CONTROLFILE, 14-15
NOREVERSE parameter
of ALTER INDEX ... REBUILD, 10-87, 10-89
NOREWRITE hint, 3-95
NOROWDEPENDENCIES clause
of CREATE CLUSTER, 14-7
of CREATE TABLE, 16-64
NOSORT clause
of ALTER INDEX, 14-72
NOT condition, 7-8, 7-9
NOT DEFERRABLE clause
of constraints, 8-14
NOT IDENTIFIED clause
of ALTER ROLE. See CREATE ROLE
of CREATE ROLE, 15-60
NOT IN subqueries
converting to NOT EXISTS subqueries, 5-139
NOT NULL clause
of CREATE TABLE, 16-30
NOWAIT clause
of LOCK TABLE, 18-73
NTH_VALUE function, 5-167
NTILE function, 5-169
null, 3-71
difference from zero, 3-71
in conditions, 3-72
table of, 3-72
in functions, 5-2
with comparison conditions, 3-72

Index-26

null conditions, 7-19
NULLIF function, 5-170
as a form of CASE expression, 5-170
NULL-related functions, 5-9
NUMBER data type, 3-10
converting to VARCHAR2, 3-57
precision, 3-10
scale, 3-10
number format models, 3-57
number functions, 5-3
numbers
comparison rules, 3-36
floating-point, 3-10, 3-12
in SQL syntax, 3-47
precision of, 3-48
spelling out, 3-66
syntax of, 3-47
numeric functions, 5-3
numeric precedence, 3-14
NUMTODSINTERVAL function, 5-171
NUMTOYMINTERVAL function, 5-172
NVARCHAR2 data type, 3-9
NVL function, 5-173
NVL2 function, 5-174

O
object access expressions, 6-13
OBJECT IDENTIFIER clause
of CREATE TABLE, 16-69
object identifiers
contained in REFs, 3-30
of object views, 17-19
primary key, 16-69
specifying, 16-69
specifying an index on, 16-69
system-generated, 16-69
object instances
types of, 7-25
object privileges
FLASHBACK ARCHIVE, 18-49
granting, 15-59
multiple, 15-65
on specific columns, 18-38
on a database object
revoking, 18-92
revoking, 18-89
from a role, 18-87, 18-91
from a user, 18-87, 18-90
from PUBLIC, 18-91
object reference functions, 5-17
object tables
adding rows to, 18-54
as part of hierarchy, 16-68
creating, 16-7, 16-68
querying, 16-68
system-generated column name, 16-68, 16-70,
17-19, 17-21
updating to latest version, 12-37
upgrading, 12-37

object type columns
defining properties
for materialized views, 15-9
in a type hierarchy, 16-42
membership in hierarchy, 12-44
modifying properties
for tables, 12-12, 12-44
substitutability, 12-44
object type materialized views
creating, 15-12
object types, 3-30
associating statistics types with, 13-25
attributes, 3-121
in a type hierarchy, 16-42
membership in hierarchy, 12-44
substitutability, 12-44
bodies
creating, 17-5
re-creating, 17-6
comparison rules, 3-39
MAP function, 3-39
ORDER function, 3-39
components of, 3-30
creating, 17-3
defining member methods of, 17-5
disassociating statistics types from, 17-34, 18-13
dropping the body of, 18-15
dropping the specification of, 18-13
granting system privileges for, 18-46
identifiers, 2-7
incomplete, 17-3
methods, 3-121
privileges on subtypes, 18-38
references to. See REFs
statistics types, 13-25
values of, 2-8
object views, 17-18
base tables
adding rows, 18-54
creating, 17-19
creating subviews, 17-19
defining, 17-14
querying, 17-18
OBJECT_ID pseudocolumn, 2-7, 16-68, 16-70, 17-19,
17-21
OBJECT_VALUE pseudocolumn, 2-8
objects. See object types or database objects
ODCIIndexInsert method
indextype support of, 10-98, 14-89
OF clause
of CREATE VIEW, 17-19
OFFLINE clause
of ALTER TABLESPACE, 12-98
of CREATE TABLESPACE, 16-90
OIDINDEX clause
of CREATE TABLE, 16-69
OIDs. See object identifiers
OLAP functions, 5-17
OLTP table compression, 16-35
ON clause

of CREATE OUTLINE, 15-40
ON COMMIT clause
of CREATE TABLE, 16-32
ON COMMIT REFRESH object privilege
on a materialized view, 18-49
ON COMMIT REFRESH system privilege, 18-42
ON DEFAULT clause
of AUDIT, 13-33
of NOAUDIT, 18-81
ON DELETE CASCADE clause
of constraints, 8-11
ON DELETE SET NULL clause
of constraints, 8-11
ON DIRECTORY clause
of AUDIT, 13-33, 13-34
of NOAUDIT, 18-81
ON MINING MODEL clause
of AUDIT, 13-33
ON object clause
of NOAUDIT, 18-81
of REVOKE, 18-92
ON PREBUILT TABLE clause
of CREATE MATERIALIZED VIEW, 15-13
online backup
of tablespaces, ending, 12-96
ONLINE clause
of ALTER TABLESPACE, 12-98
of CREATE INDEX, 14-73
of CREATE TABLESPACE, 16-90
online indexes, 14-73
rebuilding, 12-77
ONLINE parameter
of ALTER INDEX ... REBUILD, 10-88
online redo logs
reinitializing, 10-29
OPEN clause
of ALTER DATABASE, 10-18
OPEN READ ONLY clause
of ALTER DATABASE, 10-19
OPEN READ WRITE clause
of ALTER DATABASE, 10-18
operands, 4-1
operating system files
dropping, 18-11
removing, 10-28
operators, 4-1
adding to indextypes, 10-98
altering, 11-25
arithmetic, 4-3
binary, 4-2
comments on, 13-47
concatenation, 4-4
CONNECT_BY_ROOT, 4-5
dropping from indextypes, 10-98
granting system privileges for, 18-44
MULTISET EXCEPT, 4-6
MULTISET INTERSECT, 4-7
MULTISET UNION, 4-8
precedence, 4-2
PRIOR, 4-5

Index-27

set, 4-5, 19-33
specifying implementation of, 15-35
unary, 4-2
user-defined, 4-9
binding to a function, 11-26, 15-36
compiling, 11-25
creating, 15-35
dropping, 17-59
how bindings are implemented, 15-36
implementation type, 15-37
return type of binding, 15-36
OPT_PARAM hint, 3-98
OPTIMAL parameter
of STORAGE clause, 8-54
OR condition, 7-8, 7-9
OR REPLACE clause
of CREATE CONTEXT, 14-9
of CREATE DIRECTORY, 14-42
of CREATE FUNCTION, 14-59, 14-92
of CREATE LIBRARY, 15-3
of CREATE OUTLINE, 15-39
of CREATE PACKAGE, 15-43
of CREATE PACKAGE BODY, 15-45
of CREATE PROCEDURE, 15-49
of CREATE TRIGGER, 16-98
of CREATE TYPE, 17-4
of CREATE TYPE BODY, 17-6
of CREATE VIEW, 17-16
ORA_DST_AFFECTED function, 5-175
ORA_DST_CONVERT function, 5-176
ORA_DST_ERROR function, 5-177
ORA_HASH function, 5-178
ORA_ROWSCN pseudocolumn, 2-8
Oracle ADVM volumes, 10-67
Oracle Automatic Storage Management
migrating nodes in a cluster, 11-66
Oracle Call Interface, 1-3
Oracle Expression Filter
conditions, 7-2
operators, 4-1
Oracle reserved words, E-1
Oracle Text
built-in conditions, 7-2
CATSEARCH, 7-2
CONTAINS, 7-2
creating domain indexes, 14-78
MATCHES, 7-2
operators, 4-1
CATSEARCH, 4-1
CONTAINS, 4-1
MATCHES, 4-1
SCORE, 4-1
Oracle Tools
support of SQL, 1-3
ORDAudio data type, 3-35
ORDDicom data type, 3-35
ORDDoc data type, 3-35
ORDER BY clause
of queries, 9-10
of SELECT, 9-10, 19-13, 19-33

Index-28

with ROWNUM, 2-10
ORDER clause
of ALTER SEQUENCE. See CREATE SEQUENCE
ORDER parameter
of CREATE SEQUENCE, 15-70
ORDER SIBLINGS BY clause
of SELECT, 19-33
ORDERED hint, 3-98
ORDImage data type, 3-35
ORDImageSignature data type, 3-35
ordinal numbers
specifying, 3-66
spelling out, 3-66
ORDVideo data type, 3-35
ORGANIZATION EXTERNAL clause
of CREATE TABLE, 16-37, 16-40
ORGANIZATION HEAP clause
of CREATE TABLE, 16-37
ORGANIZATION INDEX clause
of CREATE TABLE, 16-37
outer joins, 9-12
restrictions, 9-13
outlines
assign to a different category, 11-29
assigning to a different category, 11-28, 11-30
copying, 15-39
creating, 15-38
creating on statements, 15-40
dropping from the database, 17-60
enabling and disabling dynamically, 15-39
for use by current session, 15-39
for use by PUBLIC, 15-39
granting system privileges for, 18-44
private, use by the optimizer, 11-54
rebuilding, 11-28, 11-30
recompiling, 11-29
renaming, 11-28, 11-29, 11-30
replacing, 15-39
storing groups of, 15-40
use by the optimizer, 11-72
use to generate execution plans, 11-54
used to generate execution plans, 15-38
out-of-line constraints
of CREATE TABLE, 16-30
OVER clause
of analytic functions, 5-12
OVERFLOW clause
of ALTER INDEX, 10-83
of ALTER TABLE, 12-40
of CREATE TABLE, 16-39

P
package bodies
creating, 15-44
re-creating, 15-45
removing from the database, 17-62
packaged procedures
dropping, 17-64
packages

associating statistics types with, 13-25
creating, 15-42
disassociating statistics types from, 17-34, 17-62
redefining, 15-43
removing from the database, 17-62
synonyms for, 16-2
PACKAGES clause
of ASSOCIATE STATISTICS, 13-25, 13-26
of DISASSOCIATE STATISTICS, 17-34
PARALLEL clause
of ALTER CLUSTER, 10-6, 10-7
of ALTER INDEX, 10-85
of ALTER MATERIALIZED VIEW, 11-7, 11-10
of ALTER MATERIALIZED VIEW LOG, 11-19,
11-21
of ALTER TABLE, 12-76
of CREATE CLUSTER, 14-6
of CREATE INDEX, 14-74
of CREATE MATERIALIZED VIEW, 15-12, 15-16
of CREATE MATERIALIZED VIEW LOG, 15-29,
15-30
of CREATE TABLE, 16-24, 16-63
parallel execution, 8-41
hints, 3-98
of DDL statements, 11-46
of DML statements, 11-46
PARALLEL hint, 3-98
PARALLEL_INDEX hint, 3-101
parameter files
creating, 15-46
from memory, 15-47
parameters
in syntax
optional, A-3
required, A-2
PARAMETERS clause
of CREATE INDEX, 14-78
PARTITION ... LOB storage clause
of ALTER TABLE, 12-45
PARTITION BY HASH clause
of CREATE TABLE, 16-18, 16-55
PARTITION BY LIST clause
of CREATE TABLE, 16-18, 16-54
PARTITION BY RANGE clause
of CREATE TABLE, 16-18, 16-52
PARTITION BY REFERENCE clause
of CREATE TABLE, 16-20, 16-60
PARTITION clause
of ANALYZE, 13-19
of CREATE INDEX, 14-75
of CREATE TABLE, 16-53
of DELETE, 17-29
of INSERT, 18-59
of LOCK TABLE, 18-72
of UPDATE, 19-76
partitioned indexes, 3-119, 14-60, 14-75
local, creating, 14-65
user-defined, 14-74
partitioned index-organized tables
secondary indexes, updating, 10-93

partitioned tables, 3-119
partition-extended table names
in DML statements, 3-120
restrictions on, 3-120
syntax, 3-120
partitioning
by hash, 16-18, 16-55
by list, 16-18, 16-54
by range, 16-18, 16-52
by reference, 16-20, 16-60
clauses
of ALTER INDEX, 10-81
of ALTER TABLE, 12-57
interval, 16-52
of materialized view logs, 11-21, 15-31
of materialized views, 11-10, 15-7, 15-15
range with interval partitions, 16-52
referential constraint, 12-55, 16-60
system, 16-61
partitions
adding, 12-57
adding rows to, 18-54
allocating extents for, 12-36
based on literal values, 16-54
composite
specifying, 16-56
converting into nonpartitioned tables, 12-72
deallocating unused space from, 12-36
dropping, 12-66
exchanging with tables, 12-27
extents
allocating for an index, 10-84
hash
adding, 12-64
coalescing, 12-65
specifying, 16-55
index, 14-74
inserting rows into, 18-59
list, adding, 12-65
LOB storage characteristics of, 12-45
locking, 18-71
logging attribute, 16-34
logging insert operations, 12-35
merging, 12-70
modifying, 12-57, 12-59
physical attributes
changing, 12-34
range
adding, 12-64
specifying, 16-52
removing rows from, 12-67, 17-29
renaming, 12-67
revising values in, 19-76
splitting, 12-68
storage characteristics, 8-46
tablespace for
defining, 16-34
PASSWORD EXPIRE clause
of ALTER USER. See CREATE USER
of CREATE USER, 17-12

Index-29

PASSWORD_GRACE_TIME parameter
of ALTER PROFILE, 11-33
of CREATE PROFILE, 15-54
PASSWORD_LIFE_TIME parameter
of ALTER PROFILE, 11-33
of CREATE PROFILE, 15-53
PASSWORD_LOCK_TIME parameter
of ALTER PROFILE, 11-33
of CREATE PROFILE, 15-54
PASSWORD_REUSE_MAX parameter
of ALTER PROFILE, 11-33
of CREATE PROFILE, 15-53
PASSWORD_REUSE_TIME parameter
of ALTER PROFILE, 11-33
of CREATE PROFILE, 15-53
PASSWORD_VERIFY_FUNCTION parameter
of ALTER PROFILE, 11-33
of CREATE PROFILE, 15-54
passwords
expiration of, 17-12
grace period, 15-53
guaranteeing complexity, 15-53
limiting use and reuse, 15-53
locking, 15-53
making unavailable, 15-53
parameters
of CREATE PROFILE, 15-51
special characters in, 17-9
PATH function, 5-179
PATH_VIEW, 7-20, 7-21
pattern-matching conditions, 7-14
PCT_ACCESS_DIRECT statistics
for index-organized tables, 13-19
PCTFREE parameter
of ALTER CLUSTER, 10-6
of ALTER INDEX, 10-85
of ALTER MATERIALIZED VIEW LOG, 11-19
of ALTER TABLE, 12-34
of CREATE MATERIALIZED VIEW LOG. See
CREATE TABLE.
of CREATE MATERIALIZED VIEW. See CREATE
TABLE.
of CREATE TABLE, 8-45
PCTINCREASE parameter
of STORAGE clause, 8-52
PCTTHRESHOLD parameter
of CREATE TABLE, 16-38
PCTUSED parameter
of ALTER CLUSTER, 10-6
of ALTER INDEX, 10-85
of ALTER MATERIALIZED VIEW LOG, 11-19
of ALTER TABLE, 12-34
of CREATE INDEX. See CREATE TABLE
of CREATE MATERIALIZED VIEW LOG. See
CREATE TABLE.
of CREATE MATERIALIZED VIEW. See CREATE
TABLE.
of CREATE TABLE, 8-45
PCTVERSION parameter
of LOB storage, 16-44

Index-30

of LOB storage clause, 12-53
PERCENT_RANK function, 5-180
PERCENTILE_CONT function, 5-182
PERCENTILE_DISC function, 5-185
PERMANENT clause
of ALTER TABLESPACE, 12-100
physical attributes clause
of ALTER CLUSTER, 10-5
of ALTER INDEX, 10-85
of ALTER MATERIALIZED VIEW LOG, 11-19
of ALTER TABLE, 12-34
of CREATE CLUSTER, 14-3
of CREATE MATERIALIZED VIEW, 15-8
of CREATE TABLE, 16-16, 16-33
physical standby database
activating, 10-34
converting to snapshot standby database, 10-38
pivot operations, 19-21
examples, 19-48
syntax, 19-7
placeholder expressions, 6-14
plan management
granting system privileges for, 18-44
plan stability, 15-38
PLAN_TABLE sample table, 18-20
P.M. datetime format element, 3-65
PM datetime format element, 3-65
POSIX regular expression standard, D-1
POWER function, 5-187
POWERMULTISET function, 5-188
POWERMULTISET_BY_CARDINALITY
function, 5-189
PQ_DISTRIBUTE hint, 3-101
precedence
of conditions, 7-3
of numbers, 3-14
of operators, 4-2
precision
number of digits of, 3-48
of NUMBER data type, 3-10
precompilers, 1-3
predefined roles, 18-36
PREDICTION function, 5-191
PREDICTION_BOUNDS function, 5-193
PREDICTION_COST function, 5-195
PREDICTION_DETAILS function, 5-197
PREDICTION_PROBABILITY function, 5-199
PREDICTION_SET function, 5-201
PREPARE TO SWITCHOVER clause
of ALTER DATABASE, 10-36
PRESENTNNV function, 5-204
PRESENTV function, 5-206
pretty-printing of XML output, 5-373
PREVIOUS function, 5-208
primary database
converting to physical standby database, 10-38
PRIMARY KEY clause
of constraints, 8-9
of CREATE TABLE, 16-30
primary key constraints, 8-9

enabling, 16-65
index on, 16-65
primary keys
generating values for, 15-67
PRIOR clause
of hierarchical queries, 9-3
PRIOR operator, 4-5
PRIVATE clause
of CREATE OUTLINE, 15-39
private outlines
use by the optimizer, 11-54
PRIVATE_SGA parameter
of ALTER PROFILE, 11-32
of ALTER RESOURCE COST, 11-36
privileges
on subtypes of object types, 18-38
revoking from a grantee, 18-88
See also system privileges or object privileges
procedures
3GL,calling, 15-2
calling, 13-42
creating, 15-48
executing, 13-42
external, 15-48
granting system privileges for, 18-44
invalidating local objects dependent on, 17-64
issuing COMMIT or ROLLBACK
statements, 11-46
naming rules, 3-114
recompiling, 11-31
re-creating, 15-49
removing from the database, 17-64
synonyms for, 16-2
PROFILE clause
of ALTER USER. See CREATE USER
of CREATE USER, 17-11
profiles
adding resource limits, 11-32
assigning to a user, 17-11
changing resource limits, 11-32
creating, 15-50
examples, 15-54
deassigning from users, 17-65
dropping resource limits, 11-32
granting system privileges for, 18-44
modifying, examples, 11-33
removing from the database, 17-65
proxy clause
of ALTER USER, 13-7, 13-10
proxy users
allowing connection as database users, 13-10
pseudocolumns, 2-1
COLUMN_VALUE, 2-6
CONNECT_BY_ISCYCLE, 2-1
CONNECT_BY_ISLEAF, 2-2
CURRVAL, 2-3
flashback queries, 2-5
in hierarchical queries, 2-1
LEVEL, 2-2
NEXTVAL, 2-3

OBJECT_ID, 2-7, 16-68, 16-70, 17-19, 17-21
OBJECT_VALUE, 2-8
ORA_ROWSCN, 2-8
ROWID, 2-9
ROWNUM, 2-10
version queries, 2-5
XMLDATA, 2-11
PUBLIC clause
of CREATE OUTLINE, 15-39
of CREATE SYNONYM, 16-3
of DROP DATABASE LINK, 17-40
public database links
dropping, 17-40
public synonyms, 16-3
dropping, 18-3
PURGE statement, 18-83
PUSH_PRED hint, 3-104
PUSH_SUBQ hint, 3-104
PX_JOIN_FILTER hint, 3-104

Q
QB_NAME hint, 3-104
queries, 9-1, 19-4
comments in, 9-2
compound, 9-10
correlated
left correlation, 19-20
default locking of, B-3
defined, 9-1
distributed, 9-16
grouping returned rows on a value, 19-27
hierarchical. See hierarchical queries
hierarchical, ordering, 19-33
hints in, 9-2
join, 9-11, 19-23
locking rows during, 19-34
multiple versions of rows, 19-17
of past data, 19-17
ordering returned rows, 19-33
outer joins in, 19-20
referencing multiple tables, 9-11
select lists of, 9-2
selecting all columns, 19-15
selecting from a random sample of rows, 19-19
sorting results, 9-10
syntax, 9-1
top-level, 9-1
top-N, 2-10
query rewrite
and dimensions, 14-36
defined, 19-4
QUERY REWRITE object privilege
on a materialized view, 18-49
QUERY REWRITE system privilege, 18-42
QUIESCE RESTRICTED clause
of ALTER SYSTEM, 11-67
QUOTA clause
of ALTER USER. See CREATE USER
of CREATE USER, 17-11

Index-31

R
range conditions, 7-22
range partitioning
converting to interval partitioning, 12-58
range partitions
adding, 12-64
creating, 16-52
values of, 16-53
RANK function, 5-209
RATIO_TO_REPORT function, 5-211
RAW data type, 3-23
converting from CHAR data, 3-24
RAWTOHEX function, 5-212
RAWTONHEX function, 5-213
READ object privilege
on a materialized directory, 18-48
READ ONLY clause
of ALTER TABLESPACE, 12-99
of ALTER VIEW, 13-16
READ WRITE clause
of ALTER TABLESPACE, 12-99
of ALTER VIEW, 13-16
REBUILD clause
of ALTER INDEX, 10-86
of ALTER OUTLINE, 11-29
REBUILD PARTITION clause
of ALTER INDEX, 10-87
REBUILD SUBPARTITION clause
of ALTER INDEX, 10-87
REBUILD UNUSABLE LOCAL INDEXES clause
of ALTER TABLE, 12-74
RECOVER AUTOMATIC clause
of ALTER DATABASE, 10-20
RECOVER CANCEL clause
of ALTER DATABASE, 10-10, 10-22
RECOVER clause
of ALTER DATABASE, 10-19
RECOVER CONTINUE clause
of ALTER DATABASE, 10-10, 10-22
RECOVER DATABASE clause
of ALTER DATABASE, 10-10, 10-21
RECOVER DATAFILE clause
of ALTER DATABASE, 10-10, 10-21
RECOVER LOGFILE clause
of ALTER DATABASE, 10-10, 10-22
RECOVER MANAGED STANDBY DATABASE
clause
of ALTER DATABASE, 10-11
RECOVER TABLESPACE clause
of ALTER DATABASE, 10-10, 10-21
RECOVERABLE, 10-86, 16-37
See also LOGGING clause
recovery
discarding data, 10-18
distributed, enabling, 11-64
instance, continue after interruption, 10-19
media, designing, 10-19
media, performing ongoing, 10-22
of database, 10-10
recovery clauses
Index-32

of ALTER DATABASE, 10-9
recursive subquery factoring, 19-13
recycle bin
purging objects from, 18-83
redo log files
specifying, 8-29
specifying for a control file, 14-13
redo logs, 10-18
adding, 10-28, 10-30
applying to logical standby database, 10-37
archive location, 11-62
automatic archiving, 11-61
automatic name generation, 10-19, 10-20
clearing, 10-28
dropping, 10-28, 10-31
enabling and disabling thread, 10-28
manual archiving, 11-61
all, 11-62
by group number, 11-62
by SCN, 11-61
current, 11-62
next, 11-62
with sequence numbers, 11-61
members
adding to existing groups, 10-31
dropping, 10-31
renaming, 10-25
remove changes from, 10-18
reusing, 8-35
size of, 8-34
specifying, 8-29, 14-24
for media recovery, 10-22
specifying archive mode, 14-25
switching groups, 11-66
REF columns
rescoping, 11-11
specifying, 16-30
specifying from table or column, 16-30
REF constraints
defining scope, for materialized views, 11-9
of ALTER TABLE, 12-43
REF function, 5-214
reference partitioning, 16-60
reference-partitioned tables, 12-57
maintenance operations, 12-74
REFERENCES clause
of CREATE TABLE, 16-30
REFERENCES object privilege
on a table, 18-51
on a view, 18-51
referential integrity constraints, 8-10
REFRESH clause
of ALTER MATERIALIZED VIEW, 11-9, 11-12
of CREATE MATERIALIZED VIEW, 15-8
REFRESH COMPLETE clause
of ALTER MATERIALIZED VIEW, 11-12
of CREATE MATERIALIZED VIEW, 15-16
REFRESH FAST clause
of ALTER MATERIALIZED VIEW, 11-12
of CREATE MATERIALIZED VIEW, 15-16

REFRESH FORCE clause
of ALTER MATERIALIZED VIEW, 11-13
of CREATE MATERIALIZED VIEW, 15-16
REFRESH ON COMMIT clause
of ALTER MATERIALIZED VIEW, 11-13
of CREATE MATERIALIZED VIEW, 15-16
REFRESH ON DEMAND clause
of ALTER MATERIALIZED VIEW, 11-13
of CREATE MATERIALIZED VIEW, 15-16
REFs, 3-30, 8-12
as containers for object identifiers, 3-30
dangling, 13-21
updating, 13-21
validating, 13-21
REFTOHEX function, 5-215
REGEXP_COUNT function, 5-216
REGEXP_INSTR function, 5-218
REGEXP_LIKE condition, 7-18
REGEXP_REPLACE function, 5-221
REGEXP_SUBSTR function, 5-224
REGISTER clause
of ALTER SYSTEM, 11-71
REGISTER LOGFILE clause
of ALTER DATABASE, 10-36
REGR_AVGX function, 5-226
REGR_AVGY function, 5-226
REGR_COUNT function, 5-226
REGR_INTERCEPT function, 5-226
REGR_R2 function, 5-226
REGR_SLOPE function, 5-226
REGR_SXX function, 5-226
REGR_SXY function, 5-226
REGR_SYY function, 5-226
regular expressions
multilingual syntax, D-1
operators, multilingual enhancements, D-2
Oracle support of, D-1
Perl-influenced operators, D-3
subexpressions, 5-219, 5-224
relational tables
creating, 16-7, 16-25
RELY clause
of constraints, 8-17
REMAINDER function, 5-231
RENAME clause
of ALTER INDEX, 10-91
of ALTER OUTLINE, 11-29
of ALTER TABLE, 12-39
of ALTER TABLESPACE, 12-94
of ALTER TRIGGER, 13-3
RENAME CONSTRAINT clause
of ALTER TABLE, 12-55
RENAME DATAFILE clause
of ALTER TABLESPACE, 12-96
RENAME FILE clause
of ALTER DATABASE, 10-9, 10-25
RENAME GLOBAL_NAME clause
of ALTER DATABASE, 10-40
RENAME PARTITION clause
of ALTER INDEX, 10-93

of ALTER TABLE, 12-67
RENAME statement, 18-85
RENAME SUBPARTITION clause
of ALTER INDEX, 10-93
of ALTER TABLE, 12-67
REPLACE function, 5-232
replication
row-level dependency tracking, 14-7, 16-64
reserved words, 3-111, E-1
RESET COMPATIBILITY clause
of ALTER DATABASE, 10-8
reset sequence of, 10-18
RESETLOGS parameter
of CREATE CONTROLFILE, 14-15
RESOLVE clause
of ALTER JAVA CLASS, 10-101
of CREATE JAVA, 14-92
RESOLVER clause
of ALTER JAVA CLASS, 10-101
of ALTER JAVA SOURCE, 10-101
of CREATE JAVA, 14-94
Resource Manager, 11-67
resource parameters
of CREATE PROFILE, 15-51
RESOURCE_VIEW, 7-20, 7-21
response time
optimizing, 3-83
restore points
guaranteed, 15-57
preserved, 15-57
using
to flash back a table, 18-29
to flashback the database, 18-25
RESTRICTED SESSION system privilege, 18-41,
18-45
result cache, 16-63
RESULT_CACHE hint, 3-105
resumable space allocation, 11-48
RESUMABLE system privilege, 18-48
RESUME clause
of ALTER SYSTEM, 11-66
RETENTION parameter
of LOB storage, 16-44
RETRY_ON_ROW_CHANGE hint, 3-105
RETURN clause
of CREATE OPERATOR, 15-36
RETURNING clause
of DELETE, 17-31
of INSERT, 18-56
of UPDATE, 19-74, 19-75, 19-79
REUSE clause
of CREATE CONTROLFILE, 14-14
of file specifications, 8-35
REVERSE clause
of CREATE INDEX, 14-73
reverse indexes, 14-73
REVERSE parameter
of ALTER INDEX ... REBUILD, 10-87, 10-89
REVOKE CONNECT THROUGH clause
of ALTER USER, 13-7, 13-10

Index-33

REVOKE statement, 18-87
locks, B-4
REWRITE hint, 3-106
right outer joins, 19-24
roles, 18-36
authorization
by a password, 15-60
by an external service, 15-60
by the database, 15-60
by the enterprise directory service, 15-60
changing, 11-38
creating, 15-59
disabling
for the current session, 19-61, 19-62
enabling
for the current session, 19-61, 19-62
granting, 18-33, 18-36
system privileges for, 18-44
to a user, 18-36
to another role, 18-36
to PUBLIC, 18-36
identifying by password, 15-60
identifying externally, 15-60
identifying through enterprise directory
service, 15-60
identifying using a package, 15-60
removing from the database, 17-67
revoking, 18-87
from another role, 17-67, 18-90
from PUBLIC, 18-90
from users, 17-67, 18-89
rollback segments
removing from the database, 17-68
specifying optimal size of, 8-54
rollback segments granting
system privileges for, 18-44
ROLLBACK statement, 18-96
rollback undo, 11-40, 14-29
ROLLUP clause
of SELECT statements, 19-27
ROUND (date) function, 5-233
format models, 5-379
ROUND (number) function, 5-234
routines
calling, 13-42
executing, 13-42
row constructor, 6-16
ROW EXCLUSIVE lock mode, 18-73
row locking, B-1
ROW SHARE lock mode, 18-72
row value constructor, 6-16
row values
pivoting into columns, 19-21
ROW_NUMBER function, 5-236
ROWDEPENDENCIES clause
of CREATE CLUSTER, 14-7
of CREATE TABLE, 16-64
ROWID data type, 3-27
ROWID pseudocolumn, 2-9, 3-27, 3-28
rowids, 3-27

Index-34

description of, 3-27
extended
base 64, 3-27
not directly available, 3-27
nonphysical, 3-28
of foreign tables, 3-28
of index-organized tables, 3-28
uses for, 2-9
ROWIDTOCHAR function, 5-238
ROWIDTONCHAR function, 5-239
row-level dependency tracking, 14-7, 16-64
row-level locking, B-1
ROWNUM pseudocolumn, 2-10
rows
adding to a table, 18-54
allowing movement of between partitions, 16-16
inserting
into partitions, 18-59
into remote databases, 18-60
into subpartitions, 18-59
locking, B-1
locks on, B-1
movement between partitions, 16-65
removing
from a cluster, 19-67, 19-69
from a table, 19-67, 19-69
from partitions and subpartitions, 17-29
from tables and views, 17-26
selecting in hierarchical order, 9-3
specifying constraints on, 8-12
storing if in violation of constraints, 12-73
RPAD function, 5-240
RR datetime format element, 3-65
RTRIM function, 5-241
run-time compilation
avoiding, 11-31, 13-14

S
SAMPLE clause
of SELECT, 19-19
of SELECT and subqueries, 19-8
SAVEPOINT statement, 19-2
savepoints
erasing, 13-49
rolling back to, 18-96
specifying, 19-2
scalar subqueries, 6-14
scale
greater than precision, 3-11
of NUMBER data type, 3-10
SCHEMA clause
of CREATE JAVA, 14-93
schema objects, 3-109
defining default buffer pool for, 8-55
dropping, 18-16
in other schemas, 3-117
list of, 3-109
name resolution, 3-116
namespaces, 3-112

naming
examples, 3-114
guidelines, 3-115
rules, 3-111
object types, 3-30
on remote databases, 3-117
partitioned indexes, 3-119
partitioned tables, 3-119
parts of, 3-110
protecting location, 16-2
protecting owner, 16-2
providing alternate names for, 16-2
reauthorizing, 10-2
recompiling, 10-2
referring to, 3-115, 11-51
remote, accessing, 14-31
validating structure, 13-21
schemas
changing for a session, 11-51
creating, 15-65
definition of, 3-109
scientific notation, 3-58
SCN_TO_TIMESTAMP function, 5-242
SCOPE FOR clause
of ALTER MATERIALIZED VIEW, 11-9
of CREATE MATERIALIZED VIEW, 15-12
SCORE operator, 4-1
SDO_GEOMETRY data type, 3-34
SDO_GEORASTER data type, 3-34
SDO_TOPO_GEOMETRY data type, 3-34
security
enforcing, 16-98
security clauses
of ALTER SYSTEM, 11-68
segment attributes clause
of CREATE TABLE, 16-16
SEGMENT MANAGEMENT clause
of CREATE TABLESPACE, 16-91
segments
space management
automatic, 16-91
manual, 16-91
using bitmaps, 16-91
using free lists, 16-91
table
compacting, 10-84, 11-11, 11-21, 12-36
SELECT ANY CUBE DIMENSION system
privilege, 18-43
SELECT ANY CUBE system privilege, 18-43
SELECT ANY DICTIONARY system
privilege, 18-48
SELECT ANY MINING MODEL system
privilege, 18-43
SELECT ANY SEQUENCE system privilege, 18-45
SELECT ANY TABLE system privilege, 18-45
select lists, 9-2
ordering, 9-10
SELECT object privilege
granting on a view, 18-37
on a materialized view, 18-49

on a mining model, 18-49
on a sequence, 18-50
on a table, 18-51
on a view, 18-51
on an OLAP cube, 18-50
on an OLAP cube dimension, 18-50
SELECT statement, 9-1, 19-4
self joins, 9-12
semijoins, 9-14
sequences, 2-3, 15-67
accessing values of, 15-67
changing
the increment value, 11-43
creating, 15-67
creating without limit, 15-68
granting system privileges for, 18-45
guarantee consecutive values, 15-70
how to use, 2-4
increment value, setting, 15-68
incrementing, 15-67
initial value, setting, 15-69
maximum value
eliminating, 11-43
setting, 15-69
setting or changing, 11-43
minimum value
eliminating, 11-43
setting, 15-69
setting or changing, 11-43
number of cached values, changing, 11-43
ordering values, 11-43
preallocating values, 15-69
recycling values, 11-43
removing from the database, 18-2
renaming, 18-85
restarting, 18-2
at a different number, 11-43
at a predefined limit, 15-68
values, 15-69
reusing, 15-67
stopping at a predefined limit, 15-68
synonyms for, 16-2
where to use, 2-3
server parameter files
creating, 15-71
from memory, 15-73
server wallet
keys, 11-68
service name
of remote database, 14-34
session control statements, 10-3
PL/SQL support of, 10-3
session locks
releasing, 11-64
session parameters
changing settings, 11-51
INSTANCE, 11-52
SESSION_ROLES view, 19-61
sessions
calculating resource cost limits, 11-35

Index-35

changing resource cost limits, 11-35
disconnecting, 11-63
granting system privileges for, 18-45
limiting CPU time, 11-35
limiting data block reads, 11-35
limiting inactive periods, 11-32
limiting private SGA space, 11-36
limiting resource costs, 11-35
limiting total elapsed time, 11-35
limiting total resources, 11-32
modifying characteristics of, 11-49
restricting, 11-67
restricting to privileged users, 11-68
switching to a different instance, 11-52
terminating, 11-64
terminating across instances, 11-64
time zone setting, 11-53
SESSIONS_PER_USER parameter
of ALTER PROFILE, 11-32
SESSIONTIMEZONE function, 5-244
SET clause
of ALTER SESSION, 11-49
of ALTER SYSTEM, 11-71
SET conditions, 7-12
SET CONSTRAINT(S) statement, 19-59
SET DANGLING TO NULL clause
of ANALYZE, 13-21
SET DATABASE clause
of CREATE CONTROLFILE, 14-14
SET ENCRYPTION KEY clause
of ALTER SYSTEM, 11-69
SET ENCRYPTION WALLET clause
of ALTER SYSTEM, 11-68
SET function, 5-245
set operators, 4-5, 19-33
INTERSECT, 4-5
MINUS, 4-5
UNION, 4-5
UNION ALL, 4-5
SET ROLE statement, 19-61
SET STANDBY DATABASE clause
of ALTER DATABASE, 10-35
SET STATEMENT_ID clause
of EXPLAIN PLAN, 18-21
SET TIME_ZONE clause
of ALTER DATABASE, 10-17, 10-41
of ALTER SESSION, 11-53
of CREATE DATABASE, 14-22
SET TRANSACTION statement, 19-64
SET UNUSED clause
of ALTER TABLE, 12-49
SGA. See system global area (SGA)
SHARE ROW EXCLUSIVE lock mode, 18-73
SHARE UPDATE lock mode, 18-73
SHARED clause
of CREATE DATABASE LINK, 14-32
shared pool
flushing, 11-64
shared server
processes

Index-36

creating additional, 11-73
terminating, 11-73
system parameters, 11-73
short-circuit evaluation
DECODE function, 5-77
SHRINK SPACE clause
of ALTER INDEX, 10-84
of ALTER MATERIALIZED VIEW, 11-11
of ALTER MATERIALIZED VIEW LOG, 11-21
of ALTER TABLE, 12-36
SHUTDOWN clause
of ALTER SYSTEM, 11-70
SI_AverageColor data type, 3-35
SI_Color data type, 3-35
SI_ColorHistogram data type, 3-35
SI_FeatureList data type, 3-36
SI_PositionalColorHistogram data type, 3-36
SI_StillImage data type, 3-36
SI_Texture data type, 3-36
siblings
ordering in a hierarchical query, 19-33
SIGN function, 5-246
simple comparison conditions, 7-5
simple expressions, 6-3
SIN function, 5-247
SINGLE TABLE clause
of CREATE CLUSTER, 14-6
single-row functions, 5-3
single-table insert, 18-58
SINH function, 5-248
SIZE clause
of ALTER CLUSTER, 10-6
of CREATE CLUSTER, 14-5
of file specifications, 8-34
SOME operator, 7-5
SOUNDEX function, 5-249
SP datetime format element suffix, 3-66
special characters
in passwords, 15-54
spelled numbers
specifying, 3-66
spfile
Oracle ASM, 15-72
SPLIT PARTITION clause
of ALTER INDEX, 10-94
of ALTER TABLE, 12-68
SPTH datetime format element suffix, 3-66
SQL Developer, 1-3
SQL functions, 5-2
ABS, 5-18
ACOS, 5-19
ADD_MONTHS, 5-20
aggregate functions, 5-10
analytic functions, 5-11
APPENDCHILDXML, 5-21
applied to LOB columns, 5-2
ASCII, 5-22
ASCIISTR, 5-23
ASIN, 5-24
ATAN, 5-25

ATAN2, 5-26
AVG, 5-27
BFILENAME, 5-29
BIN_TO_NUM, 5-30
BITAND, 5-32
CARDINALITY, 5-34
CAST, 5-35
CEIL, 5-38
character functions
returning character values, 5-4
returning number values, 5-5
character set functions, 5-5
CHARTOROWID, 5-39
CHR, 5-40
CLUSTER_ID, 5-42
CLUSTER_PROBABILITY, 5-44
CLUSTER_SET, 5-46
COALESCE, 5-48
COLLECT, 5-50
collection functions, 5-7
COMPOSE, 5-51
CONCAT, 5-52
conversion functions, 5-6
CONVERT, 5-53
CORR, 5-55
CORR_K, 5-59
CORR_S, 5-58
COS, 5-60
COSH, 5-61
COUNT, 5-62
COVAR_POP, 5-64
COVAR_SAMP, 5-66
CUBE_TABLE, 5-67
CUME_DIST, 5-69
CURRRENT_DATE, 5-71
CURRRENT_TIMESTAMP, 5-72
CV, 5-73
data cartridge functions, 5-17
data mining functions, 5-7
DATAOBJ_TO_PARTITION, 5-75
datetime functions, 5-5
DBTIMEZONE, 5-76
DECODE, 5-77
DECOMPOSE, 5-79
DELETEXML, 5-80
DENSE_RANK, 5-82
DEPTH, 5-84
DEREF, 5-85
DUMP, 5-86
EMPTY_BLOB, 5-88
EMPTY_CLOB, 5-88
encoding and decoding functions, 5-9
environment and identifier functions, 5-9
EXISTSNODE, 5-89
EXP, 5-90
EXTRACT (datetime), 5-91
EXTRACT (XML), 5-94
EXTRACTVALUE, 5-95
FEATURE_ID, 5-96
FEATURE_SET, 5-97

FEATURE_VALUE, 5-99
FIRST, 5-101
FIRST_VALUE, 5-103
FLOOR, 5-105
FROM_TZ, 5-106
general comparison functions, 5-6
GREATEST, 5-107
GROUP_ID, 5-109
GROUPING, 5-110
GROUPING_ID, 5-111
HEXTORAW, 5-112
hierarchical functions, 5-7
INITCAP, 5-113
INSERTCHILDXML, 5-114
INSERTCHILDXMLAFTER, 5-116
INSERTCHILDXMLBEFORE, 5-117
INSERTXMLAFTER, 5-118
INSERTXMLBEFORE, 5-119
INSTR, 5-120
INSTR2, 5-120
INSTR4, 5-120
INSTRB, 5-120
INSTRC, 5-120
ITERATION_NUMBER, 5-122
LAG, 5-124
large object functions, 5-7
LAST, 5-126
LAST_DAY, 5-127
LAST_VALUE, 5-128
LEAD, 5-131
LEAST, 5-133
LENGTH, 5-135
LENGTH2, 5-135
LENGTH4, 5-135
LENGTHB, 5-135
LENGTHC, 5-135
linear regression, 5-226
LISTAGG, 5-136
LN, 5-138
LNNVL, 5-139
LOCALTIMESTAMP, 5-140
LOG, 5-141
LOWER, 5-142
LPAD, 5-143
LTRIM, 5-144
MAKE_REF, 5-145
MAX, 5-146
MEDIAN, 5-148
MIN, 5-150
MOD, 5-152
model functions, 5-17
MONTHS_BETWEEN, 5-153
NANVL, 5-154
NCHR, 5-155
NEW_TIME, 5-156
NEXT_DAY, 5-157
NLS_CHARSET_DECL_LEN, 5-158
NLS_CHARSET_ID, 5-159
NLS_CHARSET_NAME, 5-160
NLS_INITCAP, 5-161

Index-37

NLS_LOWER, 5-162
NLS_UPPER, 5-163
NLSSORT, 5-164
NTH_VALUE, 5-167
NTILE, 5-169
NULLIF, 5-170
NULL-related functions, 5-9
numeric functions, 5-3
NUMTODSINTERVAL, 5-171
NUMTOYMINTERVAL, 5-172
NVL, 5-173
NVL2, 5-174
object reference functiions, 5-17
OLAP functions, 5-17
ORA_DST_AFFECTED, 5-175
ORA_DST_CONVERT, 5-176
ORA_DST_ERROR, 5-177
ORA_HASH, 5-178
PATH, 5-179
PERCENT_RANK, 5-180
PERCENTILE_CONT, 5-182
PERCENTILE_DISC, 5-185
POWER, 5-187
POWERMULTISET, 5-188
POWERMULTISET_BY_CARDINALITY, 5-189
PREDICTION, 5-191
PREDICTION_BOUNDS, 5-193
PREDICTION_COST, 5-195
PREDICTION_DETAILS, 5-197
PREDICTION_PROBABILITY, 5-199
PREDICTION_SET, 5-201
PRESENTNNV, 5-204
PRESENTV, 5-206
PREVIOUS, 5-208
RANK, 5-209
RATIO_TO_REPORT, 5-211
RAWTOHEX, 5-212
RAWTONHEX, 5-213
REF, 5-214
REFTOHEX, 5-215
REGEXP_COUNT, 5-216
REGEXP_INSTR, 5-218
REGEXP_REPLACE, 5-221
REGEXP_SUBSTR, 5-224
REGR_AVGX, 5-226
REGR_AVGY, 5-226
REGR_COUNT, 5-226
REGR_INTERCEPT, 5-226
REGR_R2, 5-226
REGR_SLOPE, 5-226
REGR_SXX, 5-226
REGR_SXY, 5-226
REGR_SYY, 5-226
REMAINDER, 5-231
REPLACE, 5-232
ROUND (date), 5-233
ROUND (number), 5-234
ROW_NUMBER, 5-236
ROWIDTOCHAR, 5-238
ROWIDTONCHAR, 5-239

Index-38

RPAD, 5-240
RTRIM, 5-241
SCN_TO_TIMESTAMP, 5-242
SESSIONTIMEZONE, 5-244
SET, 5-245
SIGN, 5-246
SIN, 5-247
single-row functions, 5-3
SINH, 5-248
SOUNDEX, 5-249
SQRT, 5-250
STATS_BINOMIAL_TEST, 5-251
STATS_CROSSTAB, 5-252
STATS_F_TEST, 5-253
STATS_KS_TEST, 5-255
STATS_MODE, 5-256
STATS_MW_TEST, 5-257
STATS_ONE_WAY_ANOVA, 5-259
STATS_T_TEST_INDEP, 5-261, 5-265
STATS_T_TEST_INDEPU, 5-261, 5-265
STATS_T_TEST_ONE, 5-261, 5-263
STATS_T_TEST_PAIRED, 5-261, 5-264
STATS_WSR_TEST, 5-267
STDDEV, 5-268
STDDEV_POP, 5-270
STDDEV_SAMP, 5-272
SUBSTR, 5-274
SUBSTR2, 5-274
SUBSTR4, 5-274
SUBSTRB, 5-274
SUBSTRC, 5-274
SUM, 5-276
SYS_CONNECT_BY_PATH, 5-278
SYS_CONTEXT, 5-279
SYS_DBURIGEN, 5-286
SYS_EXTRACT_UTC, 5-287
SYS_GUID, 5-288
SYS_TYPEID, 5-289
SYS_XMLAGG, 5-290
SYS_XMLGEN, 5-291
SYSDATE, 5-292
SYSTIMESTAMP, 5-293
TAN, 5-294
TANH, 5-295
TIMESTAMP_TO_SCN, 5-296
TO_BINARY_DOUBLE, 5-297
TO_BINARY_FLOAT, 5-299
TO_BLOB, 5-300
TO_CHAR (character), 5-301
TO_CHAR (datetime), 5-302
TO_CHAR (number), 5-305
TO_CLOB, 5-307
TO_DATE, 5-308
TO_DSINTERVAL, 5-310
TO_LOB, 5-312
TO_MULTI_BYTE, 5-313
TO_NCHAR (character), 5-314
TO_NCHAR (datetime), 5-315
TO_NCHAR (number), 5-316
TO_NCLOB, 5-317

TO_NUMBER, 5-318
TO_SINGLE_BYTE, 5-319
TO_TIMESTAMP, 5-320
TO_TIMESTAMP_TZ, 5-321
TO_YMINTERVAL, 5-323
TRANSLATE, 5-324
TRANSLATE ... USING, 5-325
TREAT, 5-327
TRIM, 5-328
TRUNC (date), 5-329
TRUNC (number), 5-330
t-test, 5-261
TZ_OFFSET, 5-331
UID, 5-332
UNISTR, 5-333
UPDATEXML, 5-334
UPPER, 5-336
USER, 5-337
USERENV, 5-338
VALUE, 5-340
VAR_POP, 5-341
VAR_SAMP, 5-343
VARIANCE, 5-344
VSIZE, 5-346
WIDTH_BUCKET, 5-347
XML functions, 5-8
XMLAGG, 5-349
XMLCAST, 5-351
XMLCDATA, 5-352
XMLCOLATTVAL, 5-353
XMLCOMMENT, 5-354
XMLCONCAT, 5-355
XMLDIFF, 5-356
XMLELEMENT, 5-358
XMLEXISTS, 5-361
XMLFOREST, 5-362
XMLISVALID, 5-363
XMLPARSE, 5-364
XMLPATCH, 5-365
XMLPI, 5-367
XMLQUERY, 5-368
XMLROOT, 5-370
XMLSEQUENCE, 5-371
XMLSERIALIZE, 5-373
XMLTABLE, 5-375
XMLTRANSFORM, 5-377
SQL. See Structured Query Language (SQL)
SQL statements
ALTER FLASHBACK ARCHIVE, 10-74
auditing
by access, 13-34
by session, 13-34
stopping, 18-79
successful, 13-34
CREATE FLASHBACK ARCHIVE, 14-55
DDL, 10-2
determining the execution plan for, 18-20
DML, 10-2
DROP FLASHBACK ARCHIVE, 17-47
organization of, 10-4

rolling back, 18-96
session control, 10-3
space allocation, resumable, 11-48
storage in the result cache, 16-63
suspending and completing, 11-48
system control, 10-3
tracking the occurrence in a session, 13-29
transaction control, 10-3
type of, 10-1
undoing, 18-96
SQL:99 standards, 1-1
SQL*Loader inserts, logging, 10-85
SQL/DS data types, 3-28
restrictions on, 3-29
SQRT function, 5-250
standalone procedures
dropping, 17-64
standard SQL, C-1
Oracle extensions to, C-26
standby database
synchronizing with primary database, 11-48
standby databases
activating, 10-34
and Data Guard, 10-38
committing to primary status, 10-36
controlling use, 10-42
converting to physical standby, 10-38
designing media recovery, 10-19
mounting, 10-18
recovering, 10-21
STAR_TRANSFORMATION hint, 3-106
STAR_TRANSFORMATION_ENABLED initialization
parameter, 3-106
START LOGICAL STANDBY APPLY clause
of ALTER DATABASE, 10-37
START WITH clause
of ALTER MATERIALIZED VIEW ...
REFRESH, 11-13
of queries and subqueries, 19-26
of SELECT and subqueries, 19-10
START WITH parameter
of CREATE SEQUENCE, 15-69
startup_clauses
of ALTER DATABASE, 10-9
STATEMENT_QUEUING hint, 3-107
statistics
collection during index rebuild, 10-86
deleting from the data dictionary, 13-23
forcing disassociation, 17-35
on index usage, 10-91
on scalar object attributes
collecting, 13-17
on schema objects
collecting, 13-17
deleting, 13-17
user-defined
dropping, 17-50, 17-52, 17-62, 18-6, 18-13
statistics types
associating
with columns, 13-25

Index-39

with domain indexes, 13-25
with functions, 13-25
with indextypes, 13-25
with object types, 13-25
with packages, 13-25
disassociating
from columns, 17-34
from domain indexes, 17-34
from functions, 17-34
from indextypes, 17-34
from object types, 17-34
from packages, 17-34
STATS_BINOMIAL_TEST function, 5-251
STATS_CROSSTAB function, 5-252
STATS_F_TEST function, 5-253
STATS_KS_TEST function, 5-255
STATS_MODE function, 5-256
STATS_MW_TEST function, 5-257
STATS_ONE_WAY_ANOVA function, 5-259
STATS_T_TEST_INDEP function, 5-261, 5-265
STATS_T_TEST_INDEPU function, 5-261, 5-265
STATS_T_TEST_ONE function, 5-261, 5-263
STATS_T_TEST_PAIRED function, 5-261, 5-264
STATS_WSR_TEST function, 5-267
STDDEV function, 5-268
STDDEV_POP function, 5-270
STDDEV_SAMP function, 5-272
STOP LOGICAL STANDBY clause
of ALTER DATABASE, 10-37, 10-38
STORAGE clause
of ALTER CLUSTER, 10-6
of ALTER INDEX, 10-85
of ALTER MATERIALIZED VIEW LOG, 11-19
of CREATE MATERIALIZED VIEW LOG, 15-29
of CREATE MATERIALIZED VIEW LOG. See
CREATE TABLE
of CREATE TABLE, 8-46
storage parameters
default, changing, 12-93
resetting, 19-67, 19-69
STORE IN clause
of ALTER TABLE, 12-40, 16-56
stored functions, 14-58
string literals. See text literals.
strings
converting to ASCII values, 5-23
converting to unicode, 5-51
strings. See text literals.
Structured Query Language (SQL)
description, 1-1
functions, 5-2
keywords, A-2
Oracle Tools support of, 1-3
parameters, A-2
standards, 1-1, C-1
statements
determining the cost of, 18-20
syntax, 10-4, A-1
structures
locking, B-4

Index-40

subexpressions
of regular expressions, 5-219, 5-224
SUBMULTISET condition, 7-13
SUBPARTITION BY HASH clause
of CREATE TABLE, 16-22, 16-59
SUBPARTITION BY LIST clause
of CREATE TABLE, 16-59
SUBPARTITION clause
of ANALYZE, 13-19
of DELETE, 17-29
of INSERT, 18-59
of LOCK TABLE, 18-72
of UPDATE, 19-76
subpartition template
creating, 12-58
replacing, 12-58
subpartition-extended table names
in DML statements, 3-120
restrictions on, 3-120
syntax, 3-120
subpartitions
adding, 12-59, 12-60
adding rows to, 18-54
allocating extents for, 12-36
coalescing, 12-60
converting into nonpartitioned tables, 12-72
creating, 16-22
creating a template for, 12-58, 16-58
deallocating unused space from, 12-36
exchanging with tables, 12-27
hash, 16-59
inserting rows into, 18-59
list, 16-59
list, adding, 12-60
locking, 18-71
logging insert operations, 12-35
moving to a different segment, 12-63
physical attributes
changing, 12-34
removing rows from, 12-67, 17-29
renaming, 12-67
revising values in, 19-76
specifying, 16-56
template, creating, 16-58
template, dropping, 12-58
template, replacing, 12-58
subqueries, 9-1, 9-14, 19-4, 19-5
containing subqueries, 9-14
correlated, 9-14
defined, 9-1
extended subquery unnesting, 9-15
inline views, 9-14
nested, 9-14
of past data, 19-17
scalar, 6-14
to insert table data, 16-67
unnesting, 9-15
using in place of expressions, 6-14
SUBSTR function, 5-274
SUBSTR2 function, 5-274

SUBSTR4 function, 5-274
SUBSTRB function, 5-274
SUBSTRC function, 5-274
subtotal values
deriving, 19-27
subtypes
dropping safely, 18-14
SUM function, 5-276
supplemental logging
identification key (full), 10-32
minimal, 10-32
SUSPEND clause
of ALTER SYSTEM, 11-66
sustained standby recovery mode, 10-22
SWITCH LOGFILE clause
of ALTER SYSTEM, 11-66
SYNC WITH PRIMARY
clause of ALTER SESSION, 11-48
synonyms
changing the definition of, 18-3
creating, 16-2
granting system privileges for, 18-45
local, 16-4
private, dropping, 18-3
public, 16-3
dropping, 18-3
remote, 16-4
removing from the database, 18-3
renaming, 18-85
synonyms for, 16-2
syntax diagrams, A-1
loops, A-3
multipart diagrams, A-4
SYS user
assigning password for, 14-22
SYS_CONNECT_BY_PATH function, 5-278
SYS_CONTEXT function, 5-279
SYS_DBURIGEN function, 5-286
SYS_EXTRACT_UTC function, 5-287
SYS_GUID function, 5-288
SYS_NC_ROWINFO$ column, 16-69, 17-21
SYS_SESSION_ROLES namespace, 5-279
SYS_TYPEID function, 5-289
SYS_XMLAGG function, 5-290
SYS_XMLGEN function, 5-291
SYSAUX clause
of CREATE DATABASE, 14-27
SYSAUX tablespace
creating, 14-27, 16-87
SYSDATE function, 5-292
SYSDBA system privilege, 18-48
SYSOPER system privilege, 18-48
system change numbers
obtaining, 2-8
system control statements, 10-3
PL/SQL support of, 10-3
system global area
flushing, 11-64, 11-65
updating, 11-63
system parameters

GLOBAL_TOPIC_ENABLED, 11-73
system partitioning, 16-61
system privileges
ADMINISTER ANY SQL TUNING SET, 18-40
ADMINISTER DATABASE TRIGGER, 18-46
ADMINISTER SQL MANAGEMENT
OBJECT, 18-40, 18-44
ADMINISTER SQL TUNING SET, 18-40
ADVISOR, 18-40
ALTER ANY CLUSTER, 18-40
ALTER ANY CUBE, 18-43
ALTER ANY CUBE DIMENSION, 18-43
ALTER ANY DIMENSION, 18-41
ALTER ANY INDEX, 18-41
ALTER ANY INDEXTYPE, 18-41
ALTER ANY LIBRARY, 18-42
ALTER ANY MATERIALIZED VIEW, 18-42
ALTER ANY MINING MODEL, 18-43
ALTER ANY OPERATOR, 18-44
ALTER ANY OUTLINE, 18-44
ALTER ANY PROCEDURE, 18-44
ALTER ANY ROLE, 18-44
ALTER ANY SEQUENCE, 18-45
ALTER ANY SQL PROFILE, 18-40
ALTER ANY TABLE, 18-45
ALTER ANY TRIGGER, 18-46
ALTER ANY TYPE, 18-46
ALTER DATABASE, 18-40
ALTER DATABASE LINK, 18-41
ALTER PROFILE, 18-44
ALTER PUBLIC DATABASE LINK, 18-41
ALTER RESOURCE COST, 18-45
ALTER ROLLBACK SEGMENT, 18-44
ALTER SESSION, 18-45
ALTER SYSTEM, 18-40
ALTER TABLESPACE, 18-46
ALTER USER, 18-46
ANALYZE ANY, 18-47
AUDIT ANY, 18-47
AUDIT SYSTEM, 18-40
BACKUP ANY TABLE, 18-45
BECOME USER, 18-47
CHANGE NOTIFICATION, 18-47
COMMENT ANY MINING MODEL, 18-43
COMMENT ANY TABLE, 18-47
CREATE ANY CLUSTER, 18-40
CREATE ANY CONTEXT, 18-40
CREATE ANY CUBE, 18-43
CREATE ANY CUBE BUILD PROCESS, 18-44
CREATE ANY CUBE DIMENSION, 18-43
CREATE ANY DIMENSION, 18-41
CREATE ANY DIRECTORY, 18-41
CREATE ANY INDEX, 18-41
CREATE ANY INDEXTYPE, 18-41
CREATE ANY JOB, 18-42
CREATE ANY LIBRARY, 18-42
CREATE ANY MATERIALIZED VIEW, 18-42
CREATE ANY MEASURE FOLDER, 18-43
CREATE ANY MINING MODEL, 18-43
CREATE ANY OPERATOR, 18-44

Index-41

CREATE ANY OUTLINE, 18-44
CREATE ANY PROCEDURE, 18-44
CREATE ANY SEQUENCE, 18-45
CREATE ANY SQL PROFILE, 18-40
CREATE ANY SYNONYM, 18-45
CREATE ANY TABLE, 18-45
CREATE ANY TRIGGER, 18-46
CREATE ANY TYPE, 18-46
CREATE ANY VIEW, 18-47
CREATE CLUSTER, 18-40
CREATE CUBE, 18-43
CREATE CUBE BUILD PROCESS, 18-43
CREATE CUBE DIMENSION, 18-43
CREATE DATABASE LINK, 18-41
CREATE DIMENSION, 18-41
CREATE EXTERNAL JOB, 18-42
CREATE INDEXTYPE, 18-41
CREATE JOB, 18-42
CREATE LIBRARY, 18-42
CREATE MATERIALIZED VIEW, 18-42
CREATE MEASURE FOLDER, 18-43
CREATE MINING MODEL, 18-43
CREATE OPERATOR, 18-44
CREATE PROCEDURE, 18-44
CREATE PROFILE, 18-44
CREATE PUBLIC DATABASE LINK, 18-41
CREATE PUBLIC SYNONYM, 18-45
CREATE ROLE, 18-44
CREATE ROLLBACK SEGMENT, 18-44
CREATE SEQUENCE, 18-45
CREATE SESSION, 18-45
CREATE SYNONYM, 18-45
CREATE TABLE, 18-45
CREATE TABLESPACE, 18-46
CREATE TRIGGER, 18-46
CREATE TYPE, 18-46
CREATE USER, 18-46
CREATE VIEW, 18-47
DEBUG ANY PROCEDURE, 18-41
DELETE ANY CUBE DIMENSION, 18-43
DELETE ANY MEASURE FOLDER, 18-43
DELETE ANY TABLE, 18-45
DROP ANY CLUSTER, 18-40
DROP ANY CONTEXT, 18-40
DROP ANY CUBE, 18-43
DROP ANY CUBE BUILD PROCESS, 18-44
DROP ANY CUBE DIMENSION, 18-43
DROP ANY DIMENSION, 18-41
DROP ANY DIRECTORY, 18-41
DROP ANY INDEX, 18-41
DROP ANY INDEXTYPE, 18-41
DROP ANY LIBRARY, 18-42
DROP ANY MATERIALIZED VIEW, 18-42
DROP ANY MEASURE FOLDER, 18-43
DROP ANY MINING MODEL, 18-43
DROP ANY OPERATOR, 18-44
DROP ANY OUTLINE, 18-44
DROP ANY PROCEDURE, 18-44
DROP ANY ROLE, 18-44
DROP ANY SEQUENCE, 18-45

Index-42

DROP ANY SQL PROFILE, 18-40
DROP ANY SYNONYM, 18-45
DROP ANY TABLE, 18-45
DROP ANY TRIGGER, 18-46
DROP ANY TYPE, 18-46
DROP ANY VIEW, 18-47
DROP PROFILE, 18-44
DROP PUBLIC DATABASE LINK, 18-41
DROP PUBLIC SYNONYM, 18-45
DROP ROLLBACK SEGMENT, 18-44
DROP TABLESPACE, 18-46
DROP USER, 18-46
EXECUTE ANY CLASS, 18-42
EXECUTE ANY INDEXTYPE, 18-42
EXECUTE ANY LIBRARY, 18-42
EXECUTE ANY OPERATOR, 18-44
EXECUTE ANY PROCEDURE, 18-44
EXECUTE ANY PROGRAM, 18-42
EXECUTE ANY TYPE, 18-46
EXEMPT ACCESS POLICY, 18-47
EXEMPT REDACTION POLICY, 18-40
FLASHBACK ANY TABLE, 18-42, 18-45, 18-47
FLASHBACK ARCHIVE ADMINISTER, 18-41
for job scheduler tasks, 18-42
for the Advisor framework, 18-40
FORCE ANY TRANSACTION, 18-47
FORCE TRANSACTION, 18-47
GLOBAL QUERY REWRITE, 18-42
GRANT ANY OBJECT PRIVILEGE, 18-48
GRANT ANY PRIVILEGE, 18-48
GRANT ANY ROLE, 18-44
granting, 15-59, 18-33
to a role, 18-35
to a user, 18-35
to PUBLIC, 18-35
INSERT ANY CUBE DIMENSION, 18-43
INSERT ANY MEASURE FOLDER, 18-43
INSERT ANY TABLE, 18-45
list of, 13-29, 18-39
LOCK ANY TABLE, 18-45
MANAGE SCHEDULER, 18-42
MANAGE TABLESPACE, 18-46
MERGE ANY VIEW, 18-47
ON COMMIT REFRESH, 18-42
QUERY REWRITE, 18-42
RESTRICTED SESSION, 18-41, 18-45
RESUMABLE, 18-48
revoking, 18-87
from a role, 18-89
from a user, 18-89
from PUBLIC, 18-89
SELECT ANY CUBE, 18-43
SELECT ANY CUBE DIMENSION, 18-43
SELECT ANY DICTIONARY, 18-48
SELECT ANY MINING MODEL, 18-43
SELECT ANY SEQUENCE, 18-45
SELECT ANY TABLE, 18-45
SYSDBA, 18-48
SYSOPER, 18-48
UNDER ANY TYPE, 18-46

UNDER ANY VIEW, 18-47
UNLIMITED TABLESPACE, 18-46
UPDATE ANY CUBE, 18-43
UPDATE ANY CUBE BUILD PROCESS, 18-44
UPDATE ANY CUBE DIMENSION, 18-43
UPDATE ANY TABLE, 18-45
SYSTEM tablespace
locally managed, 14-26
SYSTEM user
assigning password for, 14-22
SYSTIMESTAMP function, 5-293

T
TABLE clause
of ANALYZE, 13-19
of INSERT, 18-60
of SELECT, 19-20
of TRUNCATE, 19-70
of UPDATE, 19-75, 19-77
TABLE collection expression, 19-20
table compression, 11-10, 12-35, 15-14, 16-34
basic, 16-35
during bulk load operations, 16-35
for archiving data, 16-35
Hybrid Columnar, 16-35
OLTP, 16-35
table locks
disabling, 12-79
duration of, 18-71
enabling, 12-79
EXCLUSIVE, 18-72, 18-73
modes of, 18-72
on partitions, 18-72
on remote database, 18-72
on subpartitions, 18-72
and queries, 18-71
ROW EXCLUSIVE, 18-72, 18-73
ROW SHARE, 18-72
SHARE, 18-72
SHARE ROW EXCLUSIVE, 18-73
SHARE UPDATE, 18-73
table partitions
compression of, 12-35, 16-34
table REF constraints, 8-12
of CREATE TABLE, 16-30
tables
adding a constraint to, 12-55
adding rows to, 18-54
aliases, 3-121
in CREATE INDEX, 14-69
in DELETE, 17-31
allocating extents for, 12-36
assigning to a cluster, 16-41
changing degree of parallelism on, 12-76
changing existing values in, 19-73
collecting statistics on, 13-19
comments on, 13-47
compression of, 12-35, 16-34
creating, 16-6

multiple, 15-65
creating comments about, 13-46
data stored outside database, 16-40
deallocating unused space from, 12-36
default physical attributes
changing, 12-34
degree of parallelism
specifying, 16-6
disassociating statistics types from, 18-6
dropping
along with cluster, 17-36
along with owner, 18-16
indexes of, 18-6
partitions of, 18-6
enabling tracking, 16-66
external, 16-37
creating, 16-40
restrictions on, 16-40
externally organized, 16-37
flashing back to an earlier version, 18-27
granting system privileges for, 18-45
heap organized, 16-37
index-organized, 16-37
overflow segment for, 16-39
space in index block, 16-38
inserting rows with a subquery, 16-67
inserting using the direct-path method, 18-54
joining in a query, 19-23
LOB storage of, 8-46
locking, 18-71
logging
insert operations, 12-35
table creation, 16-34
migrated and chained rows in, 13-22
moving, 12-31
moving to a new segment, 12-76
moving, index-organized, 12-77
nested
storage characteristics, 16-48
object
creating, 16-7
querying, 16-68
of XMLType, creating, 16-69
organization, defining, 16-37
parallel creation of, 16-63
parallelism
setting default degree, 16-63
partition attributes of, 12-57
partitioning, 3-119, 16-6, 16-51
allowing rows to move between
partitions, 12-38
default attributes of, 12-57
physical attributes
changing, 12-34
purging from the recycle bin, 18-83
read-only mode, 12-39
read/write mode, 12-39
reference-partitioned, 12-57, 12-74, 16-60
relational
creating, 16-7

Index-43

remote, accessing, 14-31
removing from the database, 18-5
removing rows from, 17-26
renaming, 12-39, 18-85
restricting
records in a block, 12-38
retrieving data from, 19-4
saving blocks in a cache, 16-62
SQL examples, 16-70
storage attributes
defining, 16-6
storage characteristics
defining, 8-46
storage properties of, 16-33, 16-42
subpartition attributes of, 12-57
synonyms for, 16-2
tablespace for
defining, 16-6, 16-34
temporary
duration of data, 16-32
session-specific, 16-25
transaction specific, 16-25
unclustering, 17-36
updating through views, 17-20
validating structure, 13-21
XMLType, querying, 16-69
TABLESPACE clause
of ALTER INDEX ... REBUILD, 10-88
of CREATE CLUSTER, 14-5
of CREATE INDEX, 14-72
of CREATE MATERIALIZED VIEW, 15-14
of CREATE MATERIALIZED VIEW LOG, 15-30
of CREATE TABLE, 16-34
tablespaces, 12-93
allocating space for users, 17-11
allowing write operations on, 12-99
automatic segment-space management, 16-91
backing up data files, 12-95
bigfile, 16-86
database default, 14-23
default temporary, 14-28
resizing, 12-94
undo, 14-29
bringing online, 12-98, 16-90
coalescing free extents, 12-94
converting
from permanent to temporary, 12-100
from temporary to permanent, 12-100
creating, 16-83
data files
adding, 12-96
renaming, 12-96
default, 10-39
specifying for a user, 13-9
default permanent, 14-27
default temporary, 10-39
learning name of, 10-39
designing media recovery, 10-19
dropping contents, 18-10
encrypting, 8-56, 16-89

Index-44

ending online backup, 12-96
extent size, 16-88
granting system privileges for, 18-46
in FLASHBACK mode, 12-100, 16-92
in FORCE LOGGING mode, 12-98, 16-89
locally managed, 8-51
altering, 12-93
logging attribute, 12-98, 16-89
managing extents of, 16-90
read only, 12-99
reconstructing lost or damaged, 10-19, 10-26
recovering, 10-19, 10-21
removing from the database, 18-9
renaming, 12-94
size of free extents in, 12-94
smallfile, 16-86
database default, 14-23
default temporary, 14-28
undo, 14-29
specifying
data files for, 16-87
for a table, 16-33
for a user, 17-10
for index rebuild, 12-78
taking offline, 12-98, 16-90
temp files
adding, 12-96
temporary
creating, 16-93
defining for the database, 14-21
shrinking, 12-94
specifying for a user, 13-9, 17-11
undo
altering, 12-93
creating, 14-29, 16-94
dropping, 18-10
TAN function, 5-294
TANH function, 5-295
TDE. See Transparent Data Encryption
temp files
bringing online, 10-27
defining for a tablespace, 16-84, 16-85, 16-86
defining for the database, 14-21
disabling autoextend, 10-27
dropping, 10-27, 12-96
enabling autoextend, 8-35, 10-27
extending automatically, 8-35
renaming, 10-25
resizing, 10-27
reusing, 8-35
shrinking, 12-97
size of, 8-34
specifying, 8-29
taking offline, 10-27
TEMPFILE clause
of ALTER DATABASE, 10-11, 10-27
TEMPORARY clause
of ALTER TABLESPACE, 12-100
of CREATE TABLESPACE, 16-93
temporary tables

creating, 16-6, 16-25
session-specific, 16-25
transaction-specific, 16-25
TEMPORARY TABLESPACE clause
of ALTER USER, 13-9
of ALTER USER. See CREATE USER
of CREATE USER, 17-11
temporary tablespace groups
reassigning for a user, 13-9
specifying for a user, 17-11
temporary tablespaces
creating, 16-93
default, 10-39
specifying extent management during database
creation, 14-21
specifying for a user, 13-9, 17-11
TEST clause
of ALTER DATABASE ... RECOVER, 10-22
testing for a set, 7-12
text
date and number formats, 3-56
literals
in SQL syntax, 3-45
properties of CHAR and VARCHAR2 data
types, 3-46
syntax of, 3-46
text literals
conversion to database character set, 3-46
TH datetime format element suffix, 3-66
throughput
optimizing, 3-79
THSP datetime format element suffix, 3-66
TIME data type
DB2, 3-29
SQL/DS, 3-29
time format models
short, 3-59, 3-63
time zone
changing time zone data file, 5-175
converting data to particular, 6-8
determining for session, 5-244
formatting, 3-63
setting for the database, 14-30
TIME_ZONE session parameter, 11-53
timestamp
converting to local time zone, 6-8
TIMESTAMP data type, 3-18
DB2, 3-29
SQL/DS, 3-29
TIMESTAMP WITH LOCAL TIME ZONE data
type, 3-19
TIMESTAMP WITH TIME ZONE data type, 3-18
TIMESTAMP_TO_SCN function, 5-296
TO SAVEPOINT clause
of ROLLBACK, 18-96
TO_BINARY_DOUBLE function, 5-297
TO_BINARY_FLOAT function, 5-299
TO_BLOB function, 5-300
TO_CHAR (character) function, 5-301
TO_CHAR (datetime) function, 5-302

format models, 3-60, 3-67
TO_CHAR (number) function, 5-305
format models, 3-57, 3-67
TO_CLOB function, 5-307
TO_DATE function, 5-308
format models, 3-60, 3-65, 3-67
TO_DSINTERVAL function, 5-310
TO_LOB function, 5-312
TO_MULTI_BYTE function, 5-313
TO_NCHAR (character) function, 5-314
TO_NCHAR (datetime) function, 5-315
TO_NCHAR (number) function, 5-316
TO_NCLOB function, 5-317
TO_NUMBER function, 5-318
format models, 3-57
TO_SINGLE_BYTE function, 5-319
TO_TIMESTAMP function, 5-320
TO_TIMESTAMP_TZ function, 5-321
TO_YMINTERVAL function, 5-323
top-level SQL statements, 13-32
top-N reporting, 2-10, 5-82, 5-209, 5-236
tracking
enabling for a table, 12-39, 16-66
transaction control statements, 10-3
PL/SQL support of, 10-3
transactions
allowing to complete, 11-63
assigning
rollback segment to, 19-64
automatically committing, 13-49
changes, making permanent, 13-49
commenting on, 13-50
distributed, forcing, 11-46
ending, 13-49
implicit commit of, 10-2, 10-3
in-doubt
committing, 13-49
forcing, 13-51
resolving, 19-65
isolation level, 19-64
locks, releasing, 13-49
naming, 19-65
read-only, 19-64
read/write, 19-64
rolling back, 11-64, 18-96
to a savepoint, 18-96
savepoints for, 19-2
TRANSLATE ... USING function, 5-325
TRANSLATE function, 5-324
Transparent Data Encryption, 16-27
master key, 11-69
TREAT function, 5-327
triggers
compiling, 13-2
creating, 16-98
database
altering, 13-3
dropping, 18-12, 18-16
disabling, 12-79, 13-2
enabling, 12-79, 13-2, 13-3, 16-98

Index-45

granting system privileges for, 18-46
INSTEAD OF
dropping, 17-16
re-creating, 16-98
removing from the database, 18-12
renaming, 13-3
TRIM function, 5-328
TRUNC (date) function, 5-329
format models, 5-379
TRUNC (number) function, 5-330
TRUNCATE PARTITION clause
of ALTER TABLE, 12-67
TRUNCATE SUBPARTITION clause
of ALTER TABLE, 12-67
TRUNCATE_CLUSTER statement, 19-67
TRUNCATE_TABLE statement, 19-69
type constructor expressions, 6-14
TYPES clause
of ASSOCIATE STATISTICS, 13-25, 13-26
of DISASSOCIATE STATISTICS, 17-34
types. See object types or data types
TZ_OFFSET function, 5-331

U
UID function, 5-332
unary operators, 4-2
UNDER ANY TYPE system privilege, 18-46
UNDER ANY VIEW system privilege, 18-47
UNDER clause
of CREATE VIEW, 17-19
UNDER object privilege
on a type, 18-49
on a view, 18-51
UNDER_PATH condition, 7-21
undo
rollback, 11-40, 14-29
system managed, 11-40, 14-29
UNDO tablespace clause
of CREATE DATABASE, 14-29
of CREATE TABLESPACE, 16-94
undo tablespaces
creating, 14-29, 16-94
dropping, 18-10
modifying, 12-93
preserving unexpired data, 12-100, 16-94
UNDO_RETENTION initialization parameter
setting with ALTER SYSTEM, 18-27
UNIFORM clause
of CREATE TABLESPACE, 16-90
UNION ALL set operator, 4-5, 19-33
UNION set operator, 4-5, 19-33
UNIQUE clause
of CREATE INDEX, 14-67
of CREATE TABLE, 16-30
of SELECT, 19-15
unique constraints
conditional, 14-82
enabling, 16-65
index on, 16-65

Index-46

unique elements of, 5-245
unique indexes, 14-67
unique queries, 19-15
UNISTR function, 5-333
universal rowids. See urowids
UNLIMITED TABLESPACE system privilege, 18-46
UNNEST hint, 3-107
unnesting collections, 19-20
examples, 19-54
unnesting subqueries, 9-15
unpivot operations, 19-22
examples, 19-48
syntax, 19-8
UNQUIESCE clause
of ALTER SYSTEM, 11-67
UNRECOVERABLE, 10-86, 16-37
See also NOLOGGING clause
unsorted indexes, 14-72
UNUSABLE clause
of ALTER INDEX, 10-90
UNUSABLE LOCAL INDEXES clause
of ALTER MATERIALIZED VIEW, 11-10
of ALTER TABLE, 12-74
UPDATE ANY CUBE BUILD PROCESS system
privilege, 18-44
UPDATE ANY CUBE DIMENSION system
privilege, 18-43
UPDATE ANY CUBE system privilege, 18-43
UPDATE ANY TABLE system privilege, 18-45
UPDATE BLOCK REFERENCES clause
of ALTER INDEX, 10-92, 10-93
UPDATE GLOBAL INDEXES clause
of ALTER TABLE, 12-76
UPDATE object privilege
on a table, 18-51
on a view, 18-51
on an OLAP cube, 18-50
on an OLAP cube dimension, 18-50
update operations
collecting supplemental log data for, 10-32
UPDATE SET clause
of MERGE, 18-75
UPDATE statement, 19-73
updates
and simultaneous insert, 18-74
using MERGE, 18-75, 18-76
UPDATEXML function, 5-334
UPGRADE clause
of ALTER DATABASE, 10-19
of ALTER TABLE, 12-37
UPPER function, 5-336
URLs
generating, 5-286
UROWID data type, 3-28
urowids
and foreign tables, 3-28
and index-organized tables, 3-28
description of, 3-28
USE_CONCAT hint, 3-107
USE_HASH hint, 3-108

USE_MERGE hint, 3-108
USE_NL hint, 3-108
USE_NL_WITH_INDEX hint, 3-109
USE_PRIVATE_OUTLINES session
parameter, 11-54
USE_STORED_OUTLINES session
parameter, 11-54, 11-72
USER function, 5-337
user groups
adding or dropping a member, 10-69
adding to a disk group, 10-69
dropping from a disk group, 10-69
USER SYS clause
of CREATE DATABASE, 14-22
USER SYSTEM clause
of CREATE DATABASE, 14-22
USER_COL_COMMENTS data dictionary
view, 13-47
USER_INDEXTYPE_COMMENTS data dictionary
view, 13-47
USER_MVIEW_COMMENTS data dictionary
view, 13-47
USER_OPERATOR_COMMENTS data dictionary
view, 13-47
USER_TAB_COMMENTS data dictionary
view, 13-47
user-defined functions, 5-380
name precedence of, 5-382
naming conventions, 5-382
user-defined operators, 4-9
user-defined statistics
dropping, 17-50, 17-52, 17-62, 18-6, 18-13
user-defined types, 3-29
USERENV function, 5-338
USERENV namespace, 5-279
users
allocating space for, 17-11
and database links, 14-33
assigning
default roles, 13-9
profiles, 17-11
authenticating, 13-11
authenticating to a remote server, 14-34
changing authentication, 13-11
creating, 17-7
default tablespaces for, 13-9, 17-10
denying access to tables and views, 18-71
external, 15-60, 17-9
global, 15-60, 17-10
granting system privileges for, 18-46
local, 15-60, 17-9
locking accounts, 17-12
operating system
adding to a disk group, 10-69
dropping from a disk group, 10-69
password expiration of, 17-12
removing from the database, 18-16
SQL examples, 17-12
temporary tablespaces for, 13-9, 17-11
USING BFILE clause

of CREATE JAVA, 14-94
USING BLOB clause
of CREATE JAVA, 14-94
USING clause
of ALTER INDEXTYPE, 10-98
of ASSOCIATE STATISTICS, 13-26, 13-27
of CREATE DATABASE LINK, 14-34
of CREATE INDEXTYPE, 14-88
USING CLOB clause
of CREATE JAVA, 14-94
USING INDEX clause
of ALTER MATERIALIZED VIEW, 11-11
of ALTER TABLE, 12-32
of constraints, 8-17
of CREATE MATERIALIZED VIEW, 15-16
of CREATE TABLE, 16-65
USING NO INDEX clause
of CREATE MATERIALIZED VIEW, 15-16
USING ROLLBACK SEGMENT clause
of ALTER MATERIALIZED VIEW ...
REFRESH, 11-14
of CREATE MATERIALIZED VIEW, 15-19
UTC
extracting from a datetime value, 5-287
UTC offset
replacing with time zone region name, 3-52
UTLCHN.SQL script, 13-22
UTLEXPT1.SQL script, 12-73
UTLXPLAN.SQL script, 18-20

V
VALIDATE clause
of DROP TYPE, 18-14
VALIDATE REF UPDATE clause
of ANALYZE, 13-21
VALIDATE STRUCTURE clause
of ANALYZE, 13-21
validation
of clusters, 13-21
of database objects
offline, 13-22
of database objects, online, 13-22
of indexes, 13-21
of tables, 13-21
VALUE function, 5-340
VALUES clause
of CREATE INDEX, 14-75
of INSERT, 18-61
VALUES LESS THAN clause
of CREATE TABLE, 16-53
VAR_POP function, 5-341
VAR_SAMP function, 5-343
VARCHAR data type, 3-10
VARCHAR2 data type, 3-10
converting to NUMBER, 3-57
VARGRAPHIC data type
DB2, 3-29
SQL/DS, 3-29
VARIANCE function, 5-344

Index-47

VARRAY clause
of ALTER TABLE, 12-13, 12-14
VARRAY column properties
of ALTER TABLE, 12-13, 12-14, 12-44
of CREATE MATERIALIZED VIEW, 15-10
of CREATE TABLE, 16-12, 16-47
varrays, 3-30
changing returned value, 12-52
compared with nested tables, 3-39
comparison rules, 3-39
creating, 17-3
dropping the body of, 18-15
dropping the specification of, 18-13
modifying column properties, 12-17
storage characteristics, 12-44, 12-54, 16-47
storing out of line, 3-30
varying arrays. See varrays
version queries
pseudocolumns for, 2-5
view constraints, 8-19, 17-18
and materialized views, 8-17
dropping, 18-18
modifying, 13-15
views
base tables
adding rows, 18-54
changing
definition, 18-18
values in base tables, 19-73
creating
before base tables, 17-16
comments about, 13-46
multiple, 15-65
creating object subviews, 17-19
defining, 17-14
dropping constraints on, 13-15
editioning, 17-17
granting system privileges for, 18-47
modifying constraints on, 13-15
object, creating, 17-19
recompiling, 13-14
re-creating, 17-16
remote, accessing, 14-31
removing
from the database, 18-18
rows from the base table of, 17-26
renaming, 18-85
retrieving data from, 19-4
subquery of, 17-20
restricting, 17-22
synonyms for, 16-2
updatable, 17-20
with joins
and key-preserved tables, 17-21
making updatable, 17-21
XMLType, 17-21
XMLType, creating, 17-25
XMLType, querying, 17-21
virtual columns
adding to a table, 12-43

Index-48

creating, 16-29
modifying, 12-43
VSIZE function, 5-346

W
WHENEVER NOT SUCCESSFUL clause
of NOAUDIT, 18-81
WHENEVER SUCCESSFUL clause
of AUDIT sql_statements, 13-34
of NOAUDIT, 18-81
WHERE clause
of DELETE, 17-30
of queries and subqueries, 19-25
of SELECT, 9-4
of UPDATE, 19-79
WIDTH_BUCKET function, 5-347
WITH ... AS clause
of SELECT, 19-13
WITH ADMIN OPTION clause
of GRANT, 18-36
WITH CHECK OPTION clause
of CREATE VIEW, 17-16, 17-22
of DELETE, 17-29
of INSERT, 18-61
of SELECT, 19-8
of UPDATE, 19-76
WITH GRANT OPTION clause
of GRANT, 18-38
WITH HIERARCHY OPTION
of GRANT, 18-38
WITH INDEX CONTEXT clause
of CREATE OPERATOR, 15-37
WITH OBJECT ID clause
of CREATE MATERIALIZED VIEW LOG, 15-31
WITH OBJECT IDENTIFIER clause
of CREATE VIEW, 17-19
WITH OBJECT OID. See WITH OBJECT IDENTIFIER.
WITH PRIMARY KEY clause
of ALTER MATERIALIZED VIEW, 11-13
of CREATE MATERIALIZED VIEW ...
REFRESH, 15-18
of CREATE MATERIALIZED VIEW LOG, 15-31
WITH READ ONLY clause
of CREATE VIEW, 17-16, 17-22
of DELETE, 17-29
of INSERT, 18-61
of SELECT, 19-8
of UPDATE, 19-76
WITH ROWID clause
of column ref constraints, 8-13
of CREATE MATERIALIZED VIEW ...
REFRESH, 15-19
of CREATE MATERIALIZED VIEW LOG, 15-31
WITH SEQUENCE clause
of CREATE MATERIALIZED VIEW LOG, 15-31
WRITE clause
of COMMIT, 13-50
WRITE object privilege
on a directory, 18-48

X
XML
conditions, 7-20
data
storage of, 16-49
database repository
SQL access to, 7-20, 7-21
documents
producing from XML fragments, 5-290
retrieving from the database, 5-286
examples, F-8
format models, 3-70
fragments, 5-94
functions, 5-8
XMLAGG function, 5-349
XMLCAST function, 5-351
XMLCDATA function, 5-352
XMLCOLATTVAL function, 5-353
XMLCOMMENT function, 5-354
XMLCONCAT function, 5-355
XMLDATA pseudocolumn, 2-11
XMLDIFF function, 5-356
XMLELEMENT function, 5-358
XMLEXISTS function, 5-361
XMLFOREST function, 5-362
XMLGenFormatType object, 3-70
XMLIndex
creating, 14-78
modifying, 10-89
XMLISVALID function, 5-363
XMLPARSE function, 5-364
XMLPATCH function, 5-365
XMLPI function, 5-367
XMLQUERY function, 5-368
XMLROOT function, 5-370
XMLSchemas
adding to a table, 16-50
single and multiple, 16-50
XMLSEQUENCE function, 5-371
XMLSERIALIZE function, 5-373
XMLTABLE function, 5-375
XMLTRANSFORM function, 5-377
XMLType columns
properties of, 12-45, 16-49
storage of, 16-49
storing in binary XML format, 16-49
XMLType storage clause
of CREATE TABLE, 16-49
XMLType tables
creating, 16-69, 16-75
creating index on, 14-81
XMLType views, 17-21
querying, 17-21

Index-49

Index-50


Source Exif Data: 
File Type                       : PDF
File Type Extension             : pdf
MIME Type                       : application/pdf
PDF Version                     : 1.6
Linearized                      : Yes
Language                        : en
XMP Toolkit                     : Adobe XMP Core 5.2-c001 63.139439, 2010/10/03-12:08:50
Format                          : application/pdf
Description                     : Oracle Database
Title                           : Oracle Database SQL Language Reference
Creator                         : Oracle Corporation
Marked                          : True
Producer                        : Acrobat Elements 9.0.0 (Windows)
Copyright                       : 1996, 2015
Creator Tool                    : FrameMaker 10.0.2
Modify Date                     : 2016:01:05 06:07:10-08:00
Create Date                     : 2016:01:05 05:47:03Z
Metadata Date                   : 2016:01:05 06:07:10-08:00
Document ID                     : uuid:fa8c3a5e-b467-470b-8f65-07416cf2c65c
Instance ID                     : uuid:8c35d217-0b53-4ebe-a592-1ba515e60ac8
Page Mode                       : UseOutlines
Page Count                      : 1634
Author                          : Oracle Corporation
Subject                         : Oracle Database


EXIF Metadata provided by EXIF.tools









 






Navigation menu