Sort_Merge_Ver_5_Ref_Man_60484800D_Sep86 Sort Merge Ver 5 Ref Man 60484800D Sep86

Sort_Merge_Ver_5_Ref_Man_60484800D_Sep86 Sort_Merge_Ver_5_Ref_Man_60484800D_Sep86

User Manual: Pdf Sort_Merge_Ver_5_Ref_Man_60484800D_Sep86

Open the PDF directly: View PDF .
Page Count: 108 [warning: Documents this large are best viewed by clicking the View PDF Link!]

60484800

'-"" CONTRPL DATA

SORT/MERGE

VERSION 5

REFERENCE MANUAL

CDC®OPERATING SYSTEMS:

NOS 2

NOS/BE 1

REVISION RECORD

Revision

A (11/20/81)

B (02/26/82)

C (02/20/84)

D (09/05/86)

Description

Original release under NOS/BE 1 at PSR level 552.

This manual is revised to reﬂect the support of NOS 2 at PSR level 552. Various

technical and editorial changes have been made. This is a complete reprint.

This manual is revised to reﬂect the support of NOS/BE 1.5 at PSR level 601. Various

technical and editorial changes have been made.

This manual is revised at PSR level 664 to document support of the CYBER 170 800 series

models and the CYBER 180 Computer Systems and support of overlapping keys. Various

technical and editorial changes have been made.

REVISION LETTERS I, 0, Q, AND X ARE NOT USED

©COPYRIGHT CONTROL DATA CORPORATION 1981, 1982, 1984, 1986

Printed in the United States of America

Address comments concerning this manual to:

CONTROL DATA CORPORATION

Technology and Publications Division

P. 0. Box 3492

SUNNYVALE, CALIFORNIA 94088-3492

or use Comment Sheet in the back of this manual

ii 60484800 D

LIST OF EFFECTIVE PAGES

New features, as well as changes, deletions, and additions to information in this manual are indicated by bars

in the margins or by a dot near the page number if the entire page is affected. A bar by the page number

indicates pagination rather than content has changed.

Page Revision

Front Cover

Inside Front Cover

Title Page

iii/iv

v/vi

v i i

v i i i

1-1

1-2

1-3

1-4

2-1

2-2

2-3

2-4

2-5

3-1

3-2 thru 3-5

3-6

3-7

3-8 thru 3-11

4-1

4-2 thru 4-4

4-4.1/4-4.2

4-5

5-1

5-2 thru 5-7

6-1

7-1

7-2

7-3

7-4

8-1

8-2

8-3 thru 8-9

A-l thru A-13

B-l thru B-10

C-l thru C-4

D-l

D-2

E-l

E-2 thru E-5

F-l

Index-1 thru -4

Comment Sheet/Mailer

Back Cover

60484800 D i i i / i v •

PREFACE

This manual describes CONTROL DATA® Sort/Merge

Version 5.0, a generalized sorting and merging

utility. As described in this publication, Sort/

Merge Version 5.0 operates under control of the

NOS 2 and the NOS/BE 1 operating systems for the

CDC® CYBER 170 and CYBER 180 Computer Systems.

This manual is written for the application program

mer. You are assumed to be familiar with the CDC

operating system installed at your site and, if you

are using procedure calls to Sort/Merge, with the

calling language.

You can ﬁnd related information in the publica

tions listed below; the publications are listed

within groupings that indicate relative importance

to you.

The NOS manual abstracts and the NOS/BE manual

abstracts are pocket-sized manuals containing brief

descriptions of the contents and intended audience

of all NOS and NOS product set manuals and NOS/BE

and NOS/BE product set manuals, respectively. The

manual abstracts can be useful in determining which

manuals are of greatest interest to you. The

Software P u b l i c a t i o n s Release History s e r v e s as a

guide in determining which revision level of

software documentation corresponds to the

Programming System Report (PSR) level of software

installed at your site.

0F*^

The following manuals are of primary interest:

Publication

COBOL Version 5 Reference Manual

CYBER Record Manager

Basic Access Methods Version 1.5

Reference Manual

FORTRAN Version 5 Reference Manual

SORT/MERGE Version 5 Instant

Publication

Number NOS NOS/BE 1

60497100

60495700

60481300

60484900

The following manuals are of secondary interest:

J^**\

Publication

INTERCOM Version 5 Reference Manual

Network Products

Remote Batch Facility Version 1

Reference Manual

NOS Version 2 Manual Abstracts

NOS Version 2 Reference Set, Volume 1

Introduction to Interactive Usage

NOS Version 2 Reference Set, Volume 3

System Commands

NOS/BE Version 1 Manual Abstracts

NOS/BE Version 1 Reference Manual

Software Publications Release History

Publication

Number NOS 2 NOS/BE 1

60455010

60499600

60485500

60459660

60459680

84000470

60493800

60481000

CDC manuals can be ordered from Control Data Corporation, Literature and

Distribution Services, 308 North Dale Street, St. Paul, Minnesota 55103.

This product is intended for use only as described in this

document. Control Data cannot be responsible for the proper

functioning of undescribed features or parameters.

60484800 D v / v i

CONTENTS

NOTATIONS

1. GENERAL DESCRIPTION

yims.

Functions and Features

Conﬁguration and Software Relationships

Job Flow

2. SORT KEYS

Deﬁnition

Describing Sort Keys

Key Length and Position

Key Type

Collating Sequences

Numeric Data Formats

Sort Order

3. CONTROL STATEMENT SORTS OR MERGES

S0RT5 Control Statement

MERGE Control Statement

Directive Files

Sort/Merge Parameters

FROM

KEY

DIR

DIALOG

ENR

OWNF

OWNFL

OWNMRL

OWNn

RETAIN

SEQx

STATUS

SUM

VERIFY

FASTIO

Specifying Parameters by Position

FILE Control Statement

Specifying Maximum Record Length

Field Length Requirements

Listings

1-1

1-2

2-1

2-2

2-5

3-1

3-2

3-3

3-4

3-5

3-7

3-8

3-9

3-10

3-11

5. PROCEDURE CALLS

SM5S0RT

SM5MERG

SM5FR0M

SM5TO

SM5KEY

SM5E

SM5EL

SM5ENR

SM5FAST

SM5N0DA

SM50FL

SM50MRL

SM50WNn

SM5RETA

SMSSEQx

SM5ST

SM5SUM

SM5END

b. COBOL INTERFACE

COBOL5 Control Statement Option

Direct Calls

7. OWNCODE ROUTINES

Specifying Owncode Routines

Owncode Routine Parameters

Owncode 1: Processing Input Records

Owncode 2: Processing Input Files

Owncode 3: Processing Output Records

Owncode 4: Processing Output Files

Owncode 5: Processing Records With

Equal Keys

8. EXAMPLES

Control Statement Sort on One Key

Sort Speciﬁcation Using Interactive Dialog

Control Statement Sort on Multiple Keys

Control Statement Merge

Creating and Using a Directive File

SUM Parameter Use

A Sort Using Procedure Calls

5-1

5-2

5-3

5-4

5-5

5-6

5-7

6-1

b-1

7-1

7-2

7-3

7-4

8-1

8-3

8-4

8-7

4. INTERACTIVE USE

Terminal Access to Sort/Merge

Logging in Under NOS

Logging in Under NOS/BE

In t era c ti v e D ial o g

Interactive Commands

HELP

RESTART

QUIT

Specifying Parameters Interactively

Logging Out

4-1

4-2

4-3

4-5

APPENDIXES

A Character Data Input, Output, and

Central Memory Representation A-l

B Diagnostics B-l

C Glossary C-l

D Sort/Merge Parameter Summary D-l

E Interactive Dialog E-l

F Sort/Merge 5, Sort/Merge 4 Comparison F-l

INDEX

60484800 D v i i

FIGURES

1-1 Batch Job Flow 1-3

1-2 Interactive Job Flow 1-4

2-1 Internal Representation of Character

and Noncharacter Data 2-2

3-1 FROM Parameter Format 3-2

3-2 TO Parameter Format 3-2

3-3 KEY Parameter Format 3-3

3-4 DIR Parameter Format 3-3

3-5 L Parameter Format 3-3

3-6 E Parameter Format 3-4

3-7 EL Parameter Format 3-4

3-8 DIALOG Parameter Format 3-4

3-9 ENR Parameter Format 3-4

3-10 OWNF Parameter Format 3-4

3-11 OWNFL Parameter Format 3-5

3-12 OWNMRL Parameter Format 3-5

3-13 OWNn Parameter Format 3-5

3-14 RETAIN Parameter Format 3-5

3-15 SEQx Parameter Formats 3-5

3-16 STATUS Parameter Format 3-7

3-17 SUM Parameter Format 3-8

3-18 SUM Parameter Example 3-9

3-19 VERIFY Parameter Format 3-9

3-20 FASTIO Parameter Format 3-9

3-21 FILE Control Statement Format 3-10

4-1 NOS Login Example 4-1

4-2 NOS/BE Login Example 4-2

4-3 Incorrect Response Example 4-3

4-4 Output File Disposition Example 4-4

4-5 Interactive Command Example 4-4.1

4-6 Interactive Parameter Speciﬁcation

Example 4-5

5-1 SM5S0RT Format 5-1

5-2 SM5MERG Format 5-2

5-3 SM5FR0M Format 5-2

5-4 SM5T0 Format 5-2

5-5 SM5KEY Format 5-2

5-6 SM5E Format 5-3

5-7 SM5EL Format 5-3

5-8 SM5ENR Format 5-3

5-9 SM5FAST Format 5-4

5-10 SM5N0DA 5-4

5-11 SM50FL Format 5-4

5-12 SM50MRL Format 5-4

5-13 SM50WNn Format 5-4

5-14 SM5RETA Format 5-5

5-15 SM5SEQx Formats 5-5

5-16 SM5ST Format 5-5

5-17 SM5SUM Format 5-6

5-18 SM5SUM Procedure Call Example 5-7

5-19 SM5END Format

6-1 C0B0L5 Control Statement Example

7-1 Owncode Routine Call and Subroutine

Statement Formats

7-2 Owncode Routine Record Flow

8-1 File UNIVER Record Layout

8-2 File UNIVER

8-3 Job Structure for Sorting File UNIVER

on Last Name

8-4 File UNIVER Sorted on Last Name

8-5 I nterac t i v e D i a log E x ample

8-6 Job Structure for Control Statement

Sort on Multiple Keys

8-7 File UNIVER Sorted on Multiple Keys

8-8 Input Files for Control Statement Merge

8-9 Job Structure for Control Statement

Merge

8-10 Merged File NEWU

8-11 Creating a Directive File

8-12 Using a Directive File (NOS)

8-13 File NUMFIL

8-14 Output From Sort Using a Directive File

8-15 File GRADES Record Layout

8-16 File GRADES

8-17 Interactive Parameter Speciﬁcation

for SUM Example

8-18 Output From SUM Example

8-19 Job Structure for Sort Called From a

FORTRAN Program

8-20 FORTRAN Program With Sort/Merge

Procedure Calls

8-21 Owncode Routine

8-22 Output From FORTRAN Program

TABLES

2-1 Data in Sort Key Fields

2-2 Predeﬁned Collating Sequences

2-3 Numeric Data Formats

2-4 Sign Overpunch Representation

3-1 $CHAR(n) Character Numbers

3-2 SEQS Parameter Summary

3-3 Error Level Codes, STATUS Parameter

3-4 Parameter Positional Order

3-5 File Characteristics

5-1 Result Array Format

5-2 Error Level Speciﬁcation

5-3 Error Level Codes, SM5ST Call

7-1 Owncode Routine Summary

5-7

6-1

7-1

7-3

8-1

8-2

8-3

8-4

8-5

8-6

8-b

8-6

8-7

8-8

8-9

8-y

2-2

2-3

2-4

2-5

3-b

3-7

3-8

3-10

3-11

5-1

5-3

5-b

7-2

/^vPfLrS

""^^IS

'*^%.

viii 60484800 D

Jp^V NOTATIONS

0^S

Unless otherwise speciﬁed, the conventions de

scribed in the following paragraphs are used

throughout this manual in parameter and procedure

call formats, examples, and diagnostic messages.

UPPERCASE Uppercase letters indicate words,

acronyms, or mnemonics required as

input by Sort/Merge or produced as

output by Sort/Merge. Parameter

keywords and interactive commands

are included in this category.

lowercase Lowercase letters indicate words

or symbols that you must supply.

... Ellipsis in a parameter or pro

cedure call format indicates that

omitted portions of the format can

be repeated.

[ 1

Brackets indicate an optional por

tion of a parameter or procedure

call format.

Two periods in a parameter format

indicate a range of letters or

digits.

Delta indicates a space (blank).

Boxes indicate character positions

in storage.

Unless otherwise indicated, all numbers in text are

decimal values.

60484800 A ix

GENERAL DESCRIPTION

S o r t / M e r g e V e r s i o n 5 i s a s e t o f p o w e r f u l a n d

efﬁcient routines that provides high-speed sorting

or merging of re c o rds wi t h a single c o ntrol s t ate

ment* This section introduces the features of

Sort/Merge and provides you with an overview of the

Sort/Merge relationship with system software.

FUNCTIONS AND FEATURES

The purpose of sorting is to arrange items in

order; the purpose of merging is to combine two or

more sets of preordered items. Using Sort/Merge,

records are rearranged in the order you specify, or

t w o o r m o r e s o r t e d ﬁ l e s a r e c o m b i n e d i n t o o n e

ﬁle. Sort/Merge offers the following features:

• A sort or merge specification with a single

control statement

• An optional directive file for sort or merge

speciﬁcation

• The capability of sorting or merging records

from as many as 100 ﬁles with one call to

Sort /Merge

• Character and noncharacter sort key types

• Five predefined collating sequences

• User-defined collating sequences

• Summing of numeric fields

• Sorting of any CYBER Record Manager (CRM)

sequential ﬁle and any record type (except U)

• A tutorial dialog for interactive sort or merge

speciﬁcation

• A set of procedures for sort or merge proc

essing from within a program written in FORTRAN

or some other language

• Interface with COBOL programs through a C0B0L5

control statement option or direct calls

• Owncode routines to insert, substitute, modify,

or delete records during Sort/Merge processing

These features are described brieﬂy in the fol

lowing paragraphs.

Sorts are initiated with the S0RT5 control state

ment; merges are initiated with the MERGE control

statement. You specify processing requirements for

the sort or merge with parameters in the respective

control statement. Merge processing requires ﬁles

that have already been sorted. Sort/Merge gener

ates code based on the parameters and parameter

values that you include in the S0RT5 or MERGE

control statement.

The S0RT5 or the MERGE control statement can be up

to 240 characters in length. You can specify more

complex sorts or merges by using a directive ﬁle

for additional parameters. Sort/Merge reads the

parameters in a directive ﬁle after all of the

parameters in the control statement have been

read. Using a single control statement to specify

your sort or merge provides simplicity and elimi

nates the necessity of creating, saving, and

subsequently retrieving directive ﬁles. Directive

ﬁles, on the other hand, provide standardization

of installation parameters and collating sequences.

Records are sorted or merged on ﬁelds of data

within each input record; the ﬁelds are called

sort keys. Sort keys can be 6-bit character data,

signed or unsigned binary integers, or ﬂoating

point numbers. Data can be sorted or merged

according to a predeﬁned or user-deﬁned collating

sequence for character data, or according to

numeric value for numeric data. Numeric data can

be signed or unsigned.

Records with equal sort key ﬁelds can be combined

into one new record. You can specify that numeric

ﬁelds be added to the corresponding ﬁelds in the

other records with equal keys. The one resultant

r e c o r d c o n t a i n s t h e k e y ﬁ e l d s a n d t h e n u m e r i c

ﬁelds that are the sums of the speciﬁed numeric

ﬁelds in all the records with equal keys; data

ﬁ e l d s t h a t a r e n o t k e y o r s u m ﬁ e l d s a r e s e t t o

the corresponding ﬁelds in one of the old records.

Sort/Merge uses CRM ﬁle processing and input/

output capabilities. Any of the CRM sequential

ﬁles or record types (except U) can be sorted or

m e r g e d ; t h i s i n c l u d e s b o t h ﬁ x e d a n d v a r i a b l e

length records. Because Sort/Merge uses CRM, the

characteristics of ﬁles input to Sort/Merge and

ﬁles output from Sort/Merge must be described to

CRM with FILE control statements unless default

characteristics apply.

You can specify any sort or merge from your termi

nal by entering the S0RT5 or MERGE control state

ment and processing parameters. An interactive

d i a l o g i s a l s o a v a i l a b l e t o g u i d e y o u , s t e p b y

step, through your sort or merge speciﬁcation.

The processing options are presented, and you

choose the option you want as well as identify your

ﬁles and sort key ﬁelds to Sort/Merge. Several

interactive commands are available to control the

use of the interactive dialog. During the dialog,

you can obtain an explanation of Sort /Merge usage

by entering HELP at your terminal.

A set of procedures exists that allows you to

initiate a sort or merge from within your program

writte n in F O RTR A N or some o t her l a n quage. T h e

same processing options are available using calls

to the procedures as are available using the S0RT5

or the MERGE control statement; the procedure calls

closely correspond to the Sort/Merge parameters.

60484800 B 1-1

A COBOL5 control statement option exists that

a l l o w s y o u t o u s e S o r t / M e r g e 5 . I f y o u s p e c i f y

S0RT5 in your C0B0L5 control statement, Sort/

Merge 5 is used when the COBOL SORT or MERGE state

ment in your program executes.

You can write routines, called owncode routines, to

insert, substitute, modify, or delete records

during Sort/Merge processing. Owncode routines are

s u b r o u t i n e s w r i t t e n i n F O R T R A N o r s o m e o t h e r

language. The routines must be compiled; the ﬁle

containing the compiled routines is loaded during

your sort or merge. Depending on the exit number

you specify, owncode routines are executed to

process input records, output records, or records

with equal keys. You can use owncode routines with

contro l statement sorts o r merges, or w i t h proce

dure calls to Sort/Merge.

CONFIGURATION AND SOFTWARE

RELATIONSHIPS

The hardware conﬁguration needed to support Sort/

Merge Version 5 is the same as the minimum conﬁg

uration required for the NOS or NOS/BE operating

| s y s tem. O ne C D C CYB E R 170 or 1 80 s e r ies h o st

computer, together with peripheral processing units

and disk subsystem is necessary. For installation,

one tape unit is needed. Additionally, the inter

active feature of Sort/Merge requires one 255x

series communication subsystem.

Sort/Merge Version 5 resides in a library named

SRT5LIB on the NOS or NOS/BE operating system.

Sort/Merge requires only the basic NOS or NOS/BE

software. Sort/Merge interfaces with CYBER Record

Manager, which is one of the standard products

released as part of both operating systems.

Interactive use of Sort/Merge requires the Inter

a c t i v e F a c i l i t y ( I A F ) V e r s i o n 1 u n d e r N O S o r

INTERCOM Version 5 under NOS/BE. IAF in turn

interfaces with the network software Network Access

Method (NAM). The NOS Version 2 Reference Set,

Volume 1 (for NOS users) or the INTERCOM reference

manual (for NOS/BE users) contains information

concerning terminal requirements.

JOB FLOW

You can use Sort/Merge for batch processing or

i n t e r a c t i v e l y. F o r b a t c h s o r t j o b s , r e c o r d s t o b e

sorted can be in permanent ﬁles, input in the same

job with your SORT5 control statement and param

eters, or supplied by an owncode routine* For

m e r g e p r o c e s s i n g , r e c o r d s m u s t b e i n p r e s o r t e d

ﬁles assigned to the job and named in the MERGE

control statement. Sort/Merge uses CYBER Record

Manager for input and output. Sorted or merged

records can be written to a ﬁle and saved, listed

by a printer, or processed by an owncode routine.

Figure 1-1 illustrates batch job ﬂow.

Interactively, records to be sorted or merged can

be in local ﬁles or in permanent ﬁles you assign

to your job; records to be sorted can also be

supplied by an owncode routine. You can enter the

SORT 5 or MERGE control statement and processing

parameters from your terminal, or you can use the

i n t e r a c t i v e d i a l o g t o sp e c i f y y o u r so r t o r m e r g e .

At the end of the sort or merge, records are in a

local ﬁle that can be displayed at your terminal,

saved as a permanent ﬁle, or listed by a printer;

output records can also be processed by an owncode

routine. Figure 1-2 illustrates interactive

j o b ﬂ o w.

1-2 60484800 D

jSﬁSfF^^S,

0ms.

A. Sort

Records supplied

by an owncode

routine

Permanent

ﬁle to be

sorted

Sort

Parameters

CYBER

I R e c o r d |

• Manager .

Sort/Merge

INPUT

ﬁle to be

sorted

Sorted ﬁle

saved as a

permanent ﬁle

Sorted ﬁle

printed out

Sorted records

processed by an

owncode routine

B. Merge

Sorted ﬁles

to be

merged

1 CYBER

| R e c o r d |

Manager .

Sort/Merge

Merge

Parameters

Merged ﬁle

saved as a

permanent ﬁle

Merged ﬁle

printed out

Merged records

processed by an

owncode routine

Figure 1-1. Batch Job Flow

^ m s

60484800 A 1-3

A. Sort

Permanent

ﬁle to be

sorted

Records supplied

by an owncode

routine

Interactive

sort speciﬁcation

1 C Y B E R '

| R e c o r d |

. Manager

.s: INTER-

I C O M I

I I

Sort/Merge

Sorted ﬁle

saved as a

permanent ﬁle

Sorted ﬁle

printed out

Sorted records

processed by an

owncode routine

B. Merge

Interactive

I merge speciﬁcation

I C Y B E R I

| R e c o r d |

Manager .

I I

Sort/Merge

Merged ﬁle

saved as a

permanent ﬁle

Merged ﬁle

printed out

Merged records

processed by an

owncode routine

^S^v

Figure 1-2. Interactive Job Flow

1-4 60484800 B

SORT KEYS

j^ss-

0m*s

ym^s

All sort or merge operations are based on the

ordering of values assigned to the data to be

sorted or merged, which are called sort keys. This

section includes an explanation of sort keys and a

discussion of the information you must supply to

describe sort keys to Sort/Merge.

DEFINITION

A sort key is a ﬁeld of data within each record in

an input ﬁle to be sorted or in a presorted ﬁle

t o b e m e r g e d w i t h o n e o r m o r e o t h e r ﬁ l e s . T h e

ﬁeld is used by Sort/Merge to determine the order

in which records are output. A ﬁle can be sorted

or merged on more than one ﬁeld of data; key

ﬁelds must occur in the same position and be the

same length in each record, and the total number of

key characters must be less than 256 (if you also

specify sum ﬁelds, the total number of characters

in the key and sum ﬁelds together must be less

than 256). Sort keys in ﬁles with variable length

r e c o r d s c a n n o t e x t e n d b e y o n d t h e l e n g t h o f t h e

shortest record in the ﬁle; the sort key ﬁeld

mu s t be pr e se nt in ev er y r e co rd . F o r ex a mp le , i f

the records range from a minimum of 25 characters

to a maximum of 80 characters, all sort keys must

be in the ﬁrst 25 characters.

The ﬁrst key you specify is the most important key

and is called a major sort key. The keys you

specify after the ﬁrst key are of lesser

importance and are called minor sort keys. Each

sort key describes the length of the ﬁeld, the

po s it i on of the ﬁe ld w it h in ea c h r ec o rd , t h e t ype

o f d a t a e x p e c t e d i n t h a t ﬁ e l d , a n d t h e e f f e c t o f

that data on the sort order. When you specify

multiple sort keys, the keys can differ as to type

of data, collating sequence, and sort order. The

alignment of data in sort keys is important.

Character data must be left-justiﬁed in the ﬁeld;

numbers must be right-justiﬁed in the ﬁeld.

An example of a ﬁle with multiple sort keys is a

university student ﬁle with a record for each

student. Each record includes the last name and

ﬁrst and middle initials, the student number, the

date of birth, the ﬁeld of study, the grade point

average, and a code representing class (freshman,

sophomore, junior, senior); all the ﬁelds are

written with character data. The ﬁle could be

maintained with the student number as the major

key - records are normally retrieved by specifying

the student number. The ﬁle can be sorted by the

name in standard alphabetic order when a list of

student names is needed*

When a university department needs to know which

students are majoring in ﬁelds within the

depart m e n t, t h e ﬁle can b e sort e d on the ﬁ e l d of

st u dy. T he s am e s o rt can sp e ci f y t he n am e a s a

minor key so that records with the same ﬁeld of

study are also sorted in alphabetic order by the

name. The ﬁle can be sorted by the class code as

the major key, a n d by the g r ade poi n t averag e in

descending numeric order as a minor key to give a

lis t by clas s wi t h t h e st u dent s ha v ing the h igh e st

grade point average at the beginning of the list.

DESCRIBING SORT KEYS

You must describe to Sort/Merge every ﬁeld of data

that you want used as a sort key. Sort key

descriptions include the following information:

• Starting location of the key within the record

• K e y l e n g t h

• Ty p e o f da t a i n th e k e y f i e l d

• Sort order

You describe sort keys with either the KEY

parameter in a control statement or directive ﬁle,

or with the SM5KEY procedure call. The options and

defaults for describing sort keys are discussed in

the following paragraphs.

KEY LENGTH AND POSITION

Y o u d e ﬁ n e k e y ﬁ e l d l e n g t h a n d p o s i t i o n b y

specifying the ﬁrst byte or bit of the ﬁeld and

either the number of bytes or bits in the ﬁeld

( l e n g t h o f t h e ﬁ e l d ) , o r t h e l a s t b y t e o r b i t o f

the ﬁeld. The leftmost byte and the leftmost bit

in a record are counted as number 1. For character

data, each character is six bits and occupies one

byte. For example, if you want to specify the name

of the university student ﬁle as a sort key, and

the name ﬁeld is the leftmost ﬁeld in the record,

you specify the ﬁrst byte as 1. If the name ﬁeld

is 20 characters long, you specify length as 20.

Sort/Merge interprets the integers you specify for

key length and position as bit numbers when the key

type (discussed later in this section) speciﬁes

bits; otherwise, byte numbers are assumed. Unless

the SEOA parameter or SM5SEQA procedure is used,

k e y ﬁ e l d s c a n o v e r l a p o n e a n o t h e r . N o d e f a u l t

e x i s t s f o r t h e s t a r t i n g l o c a t i o n o f a k e y w i t h i n a

record; you must specify the starting location. If

you do not specify the length of a key, length

defaults to 1.

KEY TYPE

Yo u s p e c i f y t h e t y p e o f d a t a i n a k e y ﬁ e l d w i t h

the name of a collating sequence or with the name

of a nu m e ric da t a forma t . T he data in a key ﬁeld

ca n b e ch ar ac te r o r no nc ha ra ct er. C h a r a c t e r da ta

is represented in the computer as display code

values. To indicate the key type for character

data, you specify the name of a collating sequence

or, for numeric character data, the name of a

numeric data format. Noncharacter data is

represented in the computer as binary values. To

indicate the key type for noncharacter data, you

specify the name of a numeric data format.

60484800 D 2-1

Figure 2-1 illustrates the difference between the

internal representation of character and nonchar

acter data. Table 2-1 summarizes character and

noncharacter data types and the associated sort

key type.

A. Character data

—1

Octal equivalent of display code character

2 9 2 4 1 8 1 2 6 0

t46 34 35 36 37

B. N oncharacter data

-1234

Octal equivalent of binary value

5 9 0

77...775455

Figure 2-1. Internal Representation of Character

and Noncharacter Data

If a sort key ﬁeld contains any characters that

are not meaningful for the key type you specify (an

alphabetic character in a ﬁeld deﬁned as a

numeric key for example), the sort order for that

k e y ﬁ e l d i n t h a t r e c o r d i s u n d e ﬁ n e d . I n t h e

o u t p u t ﬁ l e , t h e d a t a f o r t h a t k e y ﬁ e l d i n t h a t

record is also undeﬁned. The record is still

sorted according to any other sort keys you have

speciﬁed.

The collating sequences and numeric data formats

you can specify are discussed in the following

paragraphs.

Collating Sequences

A collating sequence determines the precedence

given to each character in relation to the other

characters. You specify a collating sequence for

c h a r a c t e r d a t a t o d e t e r m i n e t h e s o r t o r d e r. T h e

data must be display code characters.

Five predeﬁned collating sequences are available

to you as a Sort/Merge user: ASCII6, C0B0L6,

DISP L AY, EBCDIC6, and INTBCD. The default col

lating sequence is ASCII6. Table 2-2 shows the

predeﬁned collating sequences.

If none of the predeﬁned collating sequences

orders data as you would like, you can deﬁne a

nonstandard collating sequence. If you are using

c o n t r o l s t a t e m e n t c al l s t o S o r t / M e r g e , y o u d e ﬁ n e

your own collating sequence with SEQx parameters in

the S0RT5 or MERGE control statement. If you are

using the procedure calls, you deﬁne your own

collating sequence with SM5SEQx procedure calls.

Numeric Data Formats

Numeric data can appear in a key ﬁeld in one of

the formats listed in table 2-3. Numeric data can

be signed or unsigned. For character numeric data

that is signed, the sign can be a ﬂoating sign, an

overpunch representation over the leading (left

most) digit, a leading separate character, an over-

punch representation over the trailing (rightmost)

di g it , o r a t r ai l in g s ep a ra te ch a ra c te r.

A ﬂoating sign is a negative sign embedded between

leading blanks and the numeric characters; leading

zeros must be converted to blanks. Positive values

in this format are not signed. The following

examples are valid ﬂoating sign formats:

AA - 1

A A A 1

A A A 0

12 3 4

0*™*W$$\

TABLE 2-1. DATA IN SORT KEY FIELDS

Type Internal

Representation

Data

in Field Type

Speciﬁed b y

Data Ordered

According to

Alphabetic Name of a collating

sequence

Speciﬁed

collating

sequence

Character Display code

Numeric

Name of a collating

sequence

Name of a numeric

data format

S p e c i ﬁ e d

collating

sequence

Numeric

value

Noncharacter Binary value Numeric Name of a numeric

data format

Numeric

value

'"^y

2-2 60484800 A

TABLE 2-2. PREDEFINED COLLATING SEQUENCES

ym^s

Collating

Sequence ASCI16 C0B0L6 DISPLAY EBCDIC6 INTBCD

Decimal Octal Graphics Display

Code Graphics Display

Code Graphics

CDC

INTBCO

Code

00 00 blank 55 blank 55 :t 00 blank 55 00

01 01 66 no. 01 57 01

02 02 ii 64 xt 63t 02 72 02

03 03 60 61 03 51 03

04 04 It 53. -* 65 04 45 04

05 05 63? 60 05 66 05

06 06 67 A\ 67 06 67 06

07 07 70 70 07 53 07

08 10 51 71 10 47 10

09 11 52 73 11 52 11

10 12 47 75 12 11 12

11 13 45 —1 76 13 -i 76 13

12 14 56 57 14 46 14

13 15 46 52 15 50 15

14 16 57 77 16 56. 16

15 17 50 45 17 xt 63t 17

16 20 33 53 20 65 20

17 21 34 47 21 73 21

18 22 35 46 22 71 22

19 23 36 50 23 00 23

20 24 37 56 24 60 24

21 25 40 51 25 <a 74 25

22 26 41 54 26 70 26

23 27 42 64 27 54 27

24 30 43 72 30 it 64 30

25 31 44 01 31 61 31

26 32 :t 00 02 32 01 32

27 33 11 03 33 02 33

28 34 11 04 34 03 34

29 35 54 05 35 04 35

30 36 73 06 36 05 -i 36

31 37 71 07 37 06 37

32 40 74 10 40 07 40

33 41 01 11 41 10 41

34 42 02 66 42 11 42

35 43 03 12 43 62 43

36 44 04 13 44 12 44

37 45 05 14 45 13 45

38 46 06 15 46 14 46

39 47 07 16 47 15 47

40 50 10 17 50 16 50

41 51 11 20 51 17 51

42 52 12 21 52 20 52

43 53 13 22 53 21 53

44 54 14 62 54 22 54

45 55 15 23 blank 55 none 75 55

46 56 16 24 56 23 56

47 57 17 25 57 24 57

48 60 20 26 60 25 blank 60

49 61 21 27 61 26 61

50 62 22 30 ] + 62* 27 62

51 63 23 31 xt 63t 30 63

52 64 24 32 64 31 64

53 65 25 00 -» 65 32 65

54 66 26 33 66 33 66

55 67 27 34 A\ 67 34 67

56 70 30 35 70 35 70

57 71 31 36 71 36 71

58 72 32 37 72 37 72

59 73 61 40 73 40 73

60 74 75 41 74 41 74

61 75 62 42 75 42 75

62 76 SS 76 43 —1 76 43 76

63 11 65 44 11 44 A\ 11

'I n in st a l l a t ions using a 63-gra phic set, the X grap hie does not exist. The : graphic is display code 63.

60484800 A 2-3

TABLE 2-3. NUMERIC DATA FORMATS

Name Data Type Sign Comments

NUMERIC_LO Numeric characters Leading overpunch All characters are decimal digits except the

leading character, which indicates a sign by

an overpunch. Data is ordered according to

numeric value with negative zero ordered

before positive zero.

NUMERIC_LS Numeric characters Leading separate All characters are decimal digits except the

leading character, which is a negative or

p o s i t i v e s i g n . S p e c i f y i n g a ﬁ e l d t h a t i s

not at least two characters in length causes

a fatal error. Data is ordered according to

numeric value with negative zero ordered

before positive zero.

NUMERIC_TO Numeric characters Trailing overpunch All characters are decimal digits except the

trailing character, which indicates a sign by

an overpunch. Data is ordered according to

numeric value with negative zero ordered

before positive zero.

NUMERICJTS Numeric characters Trailing separate All characters are decimal digits except the

trailing character, which is a negative or

p o s i t i v e s i g n . S p e c i f y i n g a ﬁ e l d t h a t i s

not at least two characters in length causes

a fatal error. Data is ordered according to

numeric value with negative zero ordered

before positive zero.

NUMERIC_NS Numeric characters None All characters are decimal digits. Data is

ordered according to numeric value.

NUMERICJFS Leading blanks,

numeric characters

- sign for negative

values; a + character

is not allowed

The ﬁeld contains leading blanks (leading

zeros must be converted to blanks); if the

value is negative, the right-most leading

blank must be converted to a minus sign. If

the ﬁeld contains no leading blanks, the

value must be positive. This format is

equivalent to the FORTRAN I format. Data is

ordered according to numeric value with

negative zero ordered before positive zero.

BINARY Binary integer None The ﬁeld starts and ends on character bound

aries. Data is ordered according to numeric

value.

BINARYJBITS Binary integer None The ﬁeld does not start or end on character

boundaries. Data is ordered according to

numeric value.

INTEGER One's complement

binary integer

Positive if leftmost

bit is 0; negative if

leftmost bit is 1

The ﬁeld starts and ends on character bound

aries. Data is ordered according to numeric

value with negative zero ordered before posi

tive zero.

INTEGERJBITS One's complement

binary integer

Positive if leftmost

bit is 0; negative if

leftmost bit is 1

The ﬁeld does not start or end on character

boundaries. Data is ordered according to

numeric value with negative zero ordered

before positive zero.

REAL Normalized binary

real or ﬂoating

point number

Signed The ﬁeld occupies a full computer word and

is aligned on word boundaries. Data is

ordered according to numeric value with all

forms of zero ordered equally. The order of

indeﬁnite values is undeﬁned.

2-4 60484800 C

^^*^-

The following examples are not valid ﬂoating sign

formats:

A A 0 1 Leading zero not allowed

A - 0 1 Leading zero not allowed

+12 3 Positive sign not allowed

AAAA All blank ﬁeld not allowed

the collating sequence you specify. For numeric

keys sorted in descending order, the record having

the key with the lowest value is written last on

the output ﬁle. Descending character keys are

sorted in reverse order of the collating sequence

you specify. The default sort order is ascending.

TABLE 2-4. SIGN OVERPUNCH REPRESENTATION

A negative sign overpunch is equivalent to

overstriking a digit with a - , which is a punch in

ro w 11. A positive si gn overpunch is equivalent to

overstriking a digit with a + , which is a punch in

row 12. When a signed overpunch digit is received

as input, the digit is punched as indicated in the

second column of table 2-4. When a signed

overpunch digit is displayed as output, the digit

appears as indicated in the third column of

table 2-4. Noncharacter numeric data can be signed

or unsigned binary integers or normalized ﬂoating

point numbers.

You deﬁne numeric key ﬁelds by specifying the

ﬁrst byte of the ﬁeld and either the length of

the ﬁeld or the last byte of the ﬁeld. For types

BINARYJBITS and INTEGER_BITS, you specify the ﬁrst

bit position of the ﬁeld and either the length of

the ﬁeld in bits or the last bit of the ﬁeld.

SORT ORDER

The order for sorting keys is ascending or

descending. For numeric keys sorted in ascending

order, the record having the key with the highest

value is written last on the output ﬁle.

Ascending character keys are sorted according to

Sign and Input Output

Digit Punch Representation

+9 12-9

+8 12-8

+7 12-7

+6 12-6

+5 12-5

+4 12-4

+3 12-3

+2 12-2

+1 12-1 A.

+0 12-0 [t

-0 11-0 ]t

-1 11-1

-2 11-2

-3 11-3

-4 11-4

-5 11-5

-6 11-6

-7 11-7

-8 11-8

-9 11-9

tvalid on NOS with 029 keypunches. See

appendix A, table A-2, for NOS/BE repre

sentations of +0 and -0.

60484800 B 2-5

CONTROL STATEMENT SORTS OR MERGES

y^$S

0^s

This section

through control

that can appear

statement are

d i r e c t i v e ﬁ l e s ,

requirements for

are explained.

describes Sort/Merge processing

statement calls. The parameters

in a S0RT5 or in a MERGE control

discussed, along with optional

The CYBER Record Manager (CRM)

co ntrol sta temen t so rts or merges

SORT5 CONTROL STATEMENT

Sort/Merge sort capabilities are available through

the S0RT5 control statement. Sort/Merge generates

code based on the parameters you specify in the

S0RT5 control statement (and directive ﬁles, if

used) or on default values. Sort/Merge generates

the code to make efﬁcient use of memory, mass

storage, and the central processor.

The control statement begins with the word S0RT5,

which must be followed by a period. The control

statement can appear in one of the following forms:

S0 RT5 .p l , p 2 , . .. ,p n

S0RT5.plAp2A ...Apn

In the preceding control statement examples, p

represents a processing parameter. Parameters must

be separated by commas or spaces; commas and spaces

can be used interchangeably.

The S0RT5 control statement can be processed in a

batch job or entered interactively from a ter

minal. The control statement must be CRM block

type C (BT=C) and record type Z (RT=Z) format (see

the CRM Basic Access Methods reference manual for

information on block types and record types). Each

l i n e o f t h e c o n t r o l s t a t e m e n t c a n b e u p t o 1 0 0

characters in length; however, characters in excess

of 80 are ignored.

In interactive mode, control statements cannot be

continued on more than one line. Instead, you can

either create a directive ﬁle or use a procedure

ﬁle so you can use a text editor to verify or

change complex sort speciﬁcations.

In batch mode, you can continue a control statement

on more than one line by terminating the line with

two or more periods and beginning the new line with

one period.

NOTE

Continuation is not allowed for ranges

(indicated by two periods); ranges must be

started and completed on the same line.

You can continue a control statement on any number

of lines up to a maximum of 240 characters; periods

used as continuation symbols are not included in

the character count.

MERGE CONTROL STATEMENT

Sort/Merge merge capabilities are available through

the MERGE control statement. The input ﬁles for a

merge must be presorted. Input ﬁles for a merge

with summing must be presorted and presummed. When

called with the MERGE control statement, Sort/Merge

generates code based on the parameters you specify

i n t h e c o n t r o l s t a t e m e n t ( a n d d i r e c t i v e ﬁ l e s , i f

used) or on default values. Sort/Merge generates

code to make efﬁcient use of memory, mass storage,

and the central processor.

The control statement begins with the word MERGE,

which must be followed by a period. The control

statement can appear in one of the following forms:

MERGE.pl,p2,...,pn

MERGE.pl A p2 A ...Apn

In the preceding control statement examples, p

represents a processing parameter. Parameters must

be separated by commas or spaces; commas and spaces

can be used interchangeably.

The MERGE control statement can be processed in a

batch job or entered interactively from a ter

minal. The control statement must be CRM BT-C and

RT=Z format. Each line of the control statement

can be up to 100 characters in length; however,

characters in excess of 80 are ignored.

In interactive mode, control statements cannot be

continued on more than one line. Instead, you can

either create a directive ﬁle or use a procedure

ﬁle so you can use a text editor to verify or

change complex sort speciﬁcations.

In batch mode, you can continue a control statement

on more than one line by terminating the line with

two or more periods and beginning the new line with

one period. You can continue a control statement

on any number of lines up to a maximum of 240 char

acters; periods used as continuation symbols are

not included in the character count.

DIRECTIVE FILES

A control statement sort or merge can use optional

directive ﬁles that contain processing parameters

in addition to the parameters in the S0RT5 or the

M E R G E c o n t r o l s t a t e m e n t . U s e o f d i r e c t i v e ﬁ l e s

allows speciﬁcation of more complex sorts or

merges (speciﬁcations that could be longer than

allowed in a control statement) and permits stan

dardization of installation parameters and collat

ing sequences.

You can specify multiple directive ﬁles for a

single control statement sort or merge. All

directive ﬁles must be CRM sequential ﬁles of

block type C (BT=C) and record type Z (RT=Z) with

record length (FL) of 100 characters or less;

however, characters in excess of 80 are ignored.

60484800 C 3-1

In a directive ﬁle, each statement can be up to

240 characters in length. Each statement must

begin with the word SORT or with the word MERGE

followed immediately by a comma or a space, and

then followed by processing parameters, which are

separated by commas or spaces. The following

ex a m p le s i ll u st ra t e d i r ec t iv e ﬁ le fo rm a ts :

S0RT,pl,p2,...,pn

MERGE plAp2A...Apn

You can continue a directive ﬁle statement on more

than one line by terminating the line with two or

more periods. The new line must not begin with a

period but can begin with one or more spaces. You

specify directive ﬁles in the S0RT5 or MERGE

control statement with the DIR parameter, discussed

in the Sort/Merge Parameters subsection.

SORT/MERGE PARAMETERS

You specify various parameters in the S0RT5 or the

MERGE control statement, or in a directive ﬁle, to

direct Sort/Merge processing. Most parameters can

appear either in a control statement or in a direc

tive ﬁle. The DIALOG parameter cannot appear in a

directive ﬁle. Unless stated otherwise, a param

eter can appe a r o n l y o n ce i n yo u r s o r t o r me r g e

speciﬁcation; a fatal error results if you specify

duplicate parameters. You can specify parameters

with keywords or by position in the control state

m e n t o r d i r e c t i v e ﬁ l e . W h e n s p e c i ﬁ e d w i t h

keywords, parameters take one of the following

forms, depending on the parameter:

keyword=((value-set)[,(value-set)]...)

keyword=value

Spaces can appear anywhere in a parameter except

within a name, a number, or a range indicator. The

order for specifying parameters by position follows

t h e d e s c r i p t i o n s o f t h e i n d i v i d u a l p a r a m e t e r s .

Refer to section 7 for information on owncode

routines, which are mentioned in the descriptions

of several of the parameters.

FROM

The FROM parameter (ﬁgure 3-1) speciﬁes the input

ﬁles from which the records to be sorted or merged

are read* As many as 100 ﬁles can be sorted or

merged in one job step. The ﬁles are read in the

order that, you specify them; a ﬁle cannot be

speciﬁed more than once for a merge. The ﬁles

are read past end-of-record marks; reading termi

nates when end-of-ﬁle or end-of-information is

encountered (see the CRM Basic Access Methods

reference manual). The ﬁles are rewound before

and after use unless the name of the ﬁle is INPUT

or a FILE control statement parameter speciﬁes

otherwise. If a ﬁle is open when Sort/Merge is

called, it is not rewound after use because the CF

value in the FIT is undetermined.

If you specify the ﬁle $NULL with the FROM param

eter, records to be sorted are not read from a ﬁle;

records must be supplied by an owncode routine.

You can specify the FROM parameter more than once

in a directive ﬁle. If you omit the FROM param

et er, re co r ds t o b e s or t ed a r e r e a d f r o m t he ﬁ l e

OLD (unless an owncode routine supplies records).

An empty FROM ﬁle results in a null sort. The i

FROM ﬁle cannot be the same as the TO ﬁle.

FROM=lfn

FROM=(lfn1,lfn2,...)

FROM=$NULL

lfn Logical ﬁle name

beginning with a

; 1 to

letter.

letters or digits.

Figure 3-1. FROM Parameter Format

The TO parameter (ﬁgure 3-2) speciﬁes the ﬁle to

which sorted or merged records are written (unless

an owncode routine processes all records). The ﬁle

is not rewound before or after use unless a FILE

control statement parameter speciﬁes otherwise.

TO=lfn

TO=$NULL

lfn Logical ﬁle name; 1 to 7 letters or digits,

beginning with a letter.

Figure 3-2. TO Parameter Format

If you specify the ﬁle $NULL with the TO param

eter, sorted or merged records are not written to a

ﬁle; you must supply an owncode routine to process

all records. If you omit the TO parameter, records

are written to the ﬁle NEW (unless an owncode

routine processes all records). The TO ﬁle cannot

be the same as the FROM ﬁle.

KEY

The KEY parameter (ﬁgure 3-3) speciﬁes the key

ﬁelds that determine the sorted or merged order of

output records. You can specify the KEY parameter

more than once in a directive ﬁle on separate com

mand lines. A maximum of 100 key ﬁelds can be

speciﬁed.

Each key_def deﬁnes a separate key. They are

deﬁned in the order of precedence, from major key

to minor key, read left to right. A key_def may

consist of a range. A range consists of the ﬁrst

and last byte of the key, separated by two periods.

A s in gl e i n t eg er m a y be co ns id er ed a s a sp ec ia l

case of a range in which the ﬁrst and last byte

a r e t h e s a m e . I f a k e y _ d e f c o n s i s t s o f a n y t h i n g

more than a range, it must be enclosed within

p a r e n t h e s e s , o r e l s e i t w i l l b e p r o c e s s e d a s a

multiple key sort.

The key position and length are the ﬁrst byte of

the ﬁeld and either the number of bytes or bits in

the ﬁeld (length of the ﬁeld) or the last byte or

bit of the ﬁeld. You specify ﬁrst and length (or

last) as integers; the leftmost byte and the

leftmost bit in a record are counted as number 1.

For character data, each character is six bits and

o c c u p i e s o n e b y t e . F o r a l l t y p e s o f d a t a , ﬁ r s t

3-2 60484800 D

KEY = "(key_def [,key_def] ...)1

range J

key_def [range "1

(range [.type [,ad] ])

L{ﬁrst,length[,type[,ad] ]) J

range [ﬁrst ]

[ﬁrst.. last J

ﬁrst First byte or bit of the key ﬁeld.

last Last byte or bit of the key ﬁeld.

length Number of bytes or bits in the key ﬁeld;

default is 1.

type Name of a numeric data format or

collating sequence; default is ASCI 16.

ad Order: A for ascending, D for descend

ing; default is ascending.

Figure 3-3. KEY Parameter Format

and length (or last) refer to bytes except for the

formats BINARYJBITS and INTEGER_BITS. You must

specify the ﬁrst byte or bit of a key ﬁeld. If

you do not specify otherwise in the KEY parameter,

the key length is assumed to be 1.

The key type is the name of a predeﬁned or

user-deﬁned collating sequence or the name of a

numeric data format. If you do not specify a key

type, the type is assumed to be the ASCII6

collating sequence. See section 2 for a list of

predeﬁned collating sequences and numeric data

formats.

Sort order is either ascending or descending, and

you specify A or D, respectively. If you specify

neither, the sort order is assumed to be

ascending. For numeric keys sorted in ascending

order, t h e rec o r d hav i n g the key w i th th e high e s t

value is written last on the output ﬁle.

Ascending character keys are sorted according to

the collating sequence you specify. For numeric

keys sorted in descending order, the record having

the key with the lowest value is written last on

the output ﬁle. Descending character keys are

sorted in reverse order of the collating sequence

you specify.

Records are ﬁrst sorted according to the key ﬁeld

described by the leftmost key_def. Records with

equal values for the major key ﬁeld are sorted

according to the ﬁeld described by the next

key_def, and so on. A fatal error occurs if key

ﬁelds overlap one another and the SEQA parameter

is speciﬁed. A fatal error also occurs if the

total number of key characters is more than 255.

(I f y o u a ls o s p e c i f y s um ﬁ e l d s , t h e to ta l n u m b e r

of characters in the key and sum ﬁelds together

must be less than 256.)

You must be careful when deﬁning a sort with a

single key. The following command deﬁnes a single

key that is 20 bytes long, starting in byte 6:

S0RT5. KEY=6..25

The following command deﬁnes the same sort key:

S0RT5. KEY=((6,20))

The following command deﬁnes a sort on two single

byte keys where byte 6 is the major key and byte 20

is the minor key:

S0RT5. KEY=(6,20)

If you omit the KEY parameter, the following

assumptions are made: KEY=1.. (minimum record

length), the record length is the smallest MNR in

the F I L E con t r o l state m e n ts for input ﬁles ( o r the

smallest FL for F and Z records), the key type is

the ASCII6 collating sequence, and the sort order

is ascending.

DIR

The DIR parameter (ﬁgure 3-4) speciﬁes a direc

tive ﬁle or ﬁles from which sort or merge param

eters are read. Parameters in a directive ﬁle are

read only after all of the control statement param

eters have been read. The ﬁles are read in the

order that you specify them. The ﬁles are rewound

before and after use unless the name of the ﬁle is

INPUT or a FILE control statement parameter speci

ﬁes otherwise.

DIR=lfn

DIR=(lfn1,lfn2 )

lfn Logical ﬁle name; 1 to 7 letters or digits,

beginning with a letter.

Figure 3-4. DIR Parameter Format

You can specify the DIR parameter more than once in

a directive ﬁle. If you omit the DIR parameter,

no parameters are read from a directive ﬁle; the

sort or merge is completely speciﬁed in the SORTS

or the MERGE control statement.

The L parameter (ﬁgure 3-5) speciﬁes the ﬁle to

w h i c h l i s t i n g i n f o r m a t i o n i s w r i t t e n . T h e l i s t i n g

ﬁl e i s wr it te n i n C R M B T= C, RT = Z f o r m a t wi th a

default full length of 132. Listing information

in cl ud es t h e S o r t / M e rg e v e r s i o n n u m b e r, d ir ec ti v e

ﬁle stat e ment s , t h e ti m e a n d d a te, and d iag n osti c

messages, if any.

L=lfn

L=$NULL

lfn Logical ﬁle name; 1 to 7 letters or digits,

beginning with a letter.

Figure 3-5. L Parameter Format

0ms

60484800 D 3-3

The listing ﬁle is not rewound before or after use

unless a FILE control statement parameter speciﬁes

rewind. If you specify the ﬁle $NULL with the L

parameter, listing information is not written. If

you omit the L parameter, listing information is

written to the ﬁle OUTPUT.

o t h e r pa r a m e t er s ( e x c ep t t h e S TAT U S p a r a m et er )

speciﬁed in a control statement along with the

DIALOG parameter do not affect sort or merge proc

essing; the sort or merge is performed solely on

the basis of your dialog responses. Interactive

u s e o f t h e d i a l o g i s d e s c r i b e d i n s e c t i o n 4 , a n d

the text of the dialog is included in appendix E.

/■^"^ ■fek

The E parameter (ﬁgure 3-6) speciﬁes the ﬁle to

which diagnostic messages are written. The error

ﬁle is written in CRM BT=C, RT=Z, and FL=72

format* The ﬁle is not rewound before or after

u s e u n l e s s a F I L E c o n t r o l s t a t e m e n t p a r a m e t e r

speciﬁes to rewind.

E=lfn

E=$NULL

lfn Logical ﬁle name; 1 to 7 letters or digits,

beginning with a letter.

Figure 3-6. E Parameter Format

If you specify a listing ﬁle with a ﬁle name

different from the error ﬁle name, diagnostic

messages are written to the listing ﬁle and to the

error ﬁle. If you specify the ﬁle $NULL with the

E parameter, diagnostic messages are not written.

If you omit the E parameter, diagnostic messages

are written to the listing ﬁle.

The EL parameter (ﬁgure 3-7) speciﬁes the error

level to be reported. .If EL=T, all trivial errors,

plus all errors of levels W, F, and C, are reported.

If EL=W, all warning errors, plus all errors of

levels F and C, are reported. If EL=F, all fatal

errors, plus all errors of level C, are reported.

If EL^C, all catastrophic errors are reported.

Omitting the EL parameter is equivalent to EL=W.

EL=T

EL=W

EL=F

EL=C

Figure 3-7. EL Parameter Format

Errors are written to the error ﬁle or, if no

e r r o r ﬁ l e i s s p e c i ﬁ e d , t o t h e l i s t i n g ﬁ l e . S e e

appendix B for an explanation of the four levels of

errors.

DIALOG

The DIALOG parameter (ﬁgure 3-8) invokes an inter

active dialog between you and Sort/Merge. The

DIALOG parameter can appear only in a control

statement. If DIALOGUES, the dialog is invoked;

if DIALOG=NO, the dialog is not invoked. Omitting

the DIALOG parameter is equivalent to DIALOG=NO.

The DIALOG parameter can be abbreviated as DIA, and

Y or N can be speciﬁed instead of YES or NO. Any

3-4

DIALOG=YES (or DIA=Y)

DIALOG=NO (or DIA=N)

Figure 3-8. DIALOG Parameter Format

NOTE

Before entering a S0RT5 or MERGE command

w i t h t h e D I A L O G p a r a m e t e r, e n s u r e t h a t

your terminal is in normal mode by enter

ing the NORMAL terminal command. If your

terminal is in ASCII mode during the

interactive dialogue, Sort/Merge cannot

recognize your responses unless you enter

them using uppercase letters only.

If you are already in the dialog, and need

to change the terminal mode, enter a

control-T to return to the system.

ENR

The ENR parameter (ﬁgure 3-9) speciﬁes the esti

mated number of records to be sorted or merged.

You specify either a single value or a range of

values, and decimal integers (from 0 through

1000000000) or the CYBER Control Language (CCL)

variables Rl, R2, R3, RIG, EF, or EFG. If ENR is

speciﬁed as less than 1500, Sort/Merge uses less

memory for the initial sort phase. This causes

noticeably less memory required for the entire sort

if the actual number of records is a low number.

ENR=expr

ENR=expr . . expr

expr Expression that is a decimal integer,

variable R1,R2,R3,R1G,EF, or EFG.

or CCL

Figure 3-9. ENR Parameter Format

OWNF

The OWNF parameter (ﬁgure 3-10) speciﬁes the ﬁle

that is the source of owncode routines; the ﬁle,

which must contain relocatable object modules, is

loaded by Sort/Merge by means of the user call

loader. The load map from loading the routines is

written to ﬁle ZZZZZLM.

OWNF=lfn

lfn Logical ﬁle name; 1 to 7 letters or digits,

beginning with a letter.

Figure 3-10. OWNF Parameter Format

60484800 D

■^^\

0jm*s

J^S

OWNFL

The OWNFL parameter (ﬁgure 3-11) speciﬁes the

exact number of characters in any record from an

owncode routine. The integer you specify can be

from 1 to 5000. The OWNFL parameter can be

a b b r e v i a t e d a s O F L . S e e t h e d i s c u s s i o n o f t h e

OWNMRL parameter for default values.

OWNFL=integer

OFL=integer

Figure 3-11. OWNFL Parameter Format

OWNMRL

The OWNMRL parameter (ﬁgure 3-12) speciﬁes the

maximum length in characters of any record from an

owncode routine. The integer you specify can be

from 1 to 5000. The OWNMRL parameter can be

abbreviated as OMRL. You do not need to specify

the OWNMRL parameter if the sort or merge has an

input or output ﬁle with a maximum record length

at least as long as the longest record supplied by

an owncode routine.

OWNMRL=integer

OMRL=integer

Figure 3-12. OWNMRL Parameter Format

If you omit both the OWNMRL and the OWNFL param

eters, the record length speciﬁcation of records

I from an owncode routine depends on the length

speciﬁcations of the input and output ﬁles. If

all input and output ﬁles have ﬁxed length

records of the same length, this length is used as

a default OWNFL value; otherwise, the largest MRL

or FL from any input or output ﬁle is used as a

default OWNMRL value.

I f t h e s o r t h a s n o i n p u t o r o u t p u t ﬁ l e s ( r e c o r d s

to be sorted are supplied by an owncode routine and

sorted records are processed by an owncode rou

tine), you must specify either OWNFL or OWNMRL;

otherwise, a fatal error results. You cannot

specify both the OWNFL and OWNMRL parameters for

the same sort. Use of OWNFL rather than OWNMRL can

increase performance.

OWNn

The OWNn parameter (ﬁgure 3-13) speciﬁes the name

of an o w n code ro u t ine tha t is executed each t i m e

point n is reached during the sort or merge; n must

be 1, 2, 3, 4, or 5* Owncode routines named by the

OWNn parameter must be in the OWNF ﬁle. All

o w n c o d e r o u t i n e s a r e l o a d e d b e f o r e t h e s o r t o r

merge is started. If you omit the OWNn parameter,

no owncode routine is executed.

OWNn=proc

n 1 , 2 , 3 , 4 , o r 5

proc Procedure-name; 1 to 31 letters, digits or

the special characters ? # @ _. Only 7

characters are signiﬁcant and the ﬁrst must

be a letter.

Figure 3-13. OWNn Parameter Format

RETAIN

The RETAIN parameter (ﬁgure 3-14) directs Sort/

Merge to output records with equal sort keys in the

same order as the records are input (RETAIN=YES).

If RETAIN=N0, or if you omit the RETAIN parameter,

records with equal sort keys might not be output in

the same order as they are input. If RETAIN=YES

and you specify more than one sort or merge input

ﬁl e w i th the FR O M p ar a met e r, t he ord e r i n w hi c h

you specify the ﬁles is the order in which records

with equal keys are output. The RETAIN parameter

can be abbreviated as RET, and Y or N can be speci

ﬁed instead of YES or NO.

RETAIN=YES {or RET=Y)

RETAIN=NO (or RET=N)

Figure 3-14. RETAIN Parameter Format

You cannot specify both the RETAIN and the SUM

parameters in the same sort or merge. If you do

specify both, a catastrophic error occurs.

SEQx

The SEQx parameters (ﬁgure 3-15) deﬁne your own

collating sequence. The SEQN parameter signals the

start of your collating sequence deﬁnition and

names the sequence; name is used as the key type in

the KEY parameter when records are sorted according

to your collating sequence. Specifying a name that

is the same as the name of one of the predeﬁned

collating sequences results in a fatal error.

SEQN=name

SEQS=('char', , 'char')

[SEQR=YES or SEQR=Y]

[SEQA=YES or SEQA=Y]

char

Collating sequence name; 1 to 31 letters,

digits, or the special characters ? # @ _.

Only 10 characters are signiﬁcant and the

ﬁrst must be a letter.

Character in the collating sequence.

Figure 3-15. SEQx Parameter Formats

60484800 D 3-5

Collating positions are assigned to the characters

in your collating sequence by SEQS parameters; each

SEQS parameter speciﬁes either a single value step

or a range of value steps in the sequence. A value

s t e p c o n s i s t s o f o n e o r m o r e c h a r a c t e r s , a n d

characters speciﬁed in the same value step collate

equally. The ﬁrst SEQS parameter speciﬁes the

ﬁrst value step or range of steps, the second SEQS

parameter speciﬁes the second value step or range

of steps, and so on until your collating sequence

is completely deﬁned. The following paragraphs

describe how to specify steps in your collating

sequence with SEQS parameters.

You can specify a single value step consisting of a

single character by a graphic character (see

appendix A) enclosed in apostrophes as in the

following example:

SEQS='A'

You can also specify a single value step consisting

of a single character by $CHAR(n) where n is the

number of the character in the character set at

your site (see table 3-1). The following example

speciﬁes a step consisting of the character number

08 in the character set, which is the letter H:

SEQS=$CHAR(08)

You can specify a single value step consisting of

several characters by a single character enclosed

in apostrophes, followed by one or more characters

and/or ranges of characters. The characters are

enclosed in apostrophes. All characters speciﬁed

or implied in such a step collate equally. The

following example speciﬁes a step consisting of

t h e b l a n k a n d t h e d i g i t s 0 , 1 , 2 , a n d 3 , a l l o f

which collate equally:

SEQS=(' ','0','1','2','3')

The following example speciﬁes the same value step

a s a r a n g e o f c h a r a c t e r s w h e r e ' 0 ' . . ' 3 ' s p e c i ﬁ e s

the range of digits from 0 through 3:

SEQS=(' ','0'..'3')

You can specify several value steps, each consist

ing of one character, with one value that is a

range of characters. The characters are enclosed

in apostrophes. The following example speciﬁes a

four-step collating sequence consisting of the

digits 0, 1, 2, and 3, with 0 assigned the lowest

collating value:

SEQS=('0'..'3')

-/r^l\

TABLE 3-1. $CHAR(n) CHARACTER NUMBERS

Character CDC Graphic ASCII Graphic Octal

6-Bit Character CDC Graphic ASCII Graphic Octal

6 - B i t

Number (64-Charac (64 Charac Number (64-Charac (64 Charac

(decimal) t e r Se t ) t t e r S e t ) t Display

Code (decimal) ter Set)t ter Set)t Display

Code

00 00 32 40

01 01 33 41

02 02 34 42

03 03 35 43

04 04 36 44

05 05 37 45

06 06 38 46

07 07 39 47

08 10 40 50

09 11 41 51

10 12 42 52

11 13 43 53

12 14 44 £3 54

13 15 45 blank blank 55

14 16 46 56

15 17 47 57

16 20 48 // 60

17 21 49 61

18 22 50 62

19 23 51 63

20 24 52 64

21 25 53 65

22 26 54 66

23 27 55 67

24 30 56 70

25 31 57 71

26 32 58 72

27 33 59 73

28 34 60 74

29 35 61 75

30 36 62 —, 76

31 37 63 77

'For 63-chaicacter set differences, see the subsection entitled Charac.ter Set Anomal Les in appendix A

3-6 60484800 C

The preceding example is the same as specifying the

following:

SEQS='0'

SEQS='l'

SEQS='2'

SEQS='3'

You can specify several value steps, each of which

consists of more than one character, as ranges of

characters where each range has the same number of

characters. One value step in the collating

sequence results for each character implied by the

range. The characters are enclosed in apostro

phes. A diagnostic message is issued if each range

does not consist of the same number of characters.

T h e f o l l o w i n g e x a m p l e s p e c i ﬁ e s a t h r e e - s t e p

collating sequence where the ﬁrst step consists of

1 and A (which are assigned the lowest collating

value); the second step is 2 and B; the third step

is 3 and C:

SEQS=('1'..'3','A'..'C)

The preceding example is the same as specifying the

following:

SEQS=('1','A')

SEQS=('2','B')

SEQS=('3','C)

Any SEQS parameter that speciﬁes a range of

characters followed by a single character results

in a fatal error. An apostrophe is speciﬁed in a

SEQS parameter by four apostrophes (""). You can

specify the SEQS parameter more than once in a

directive ﬁle. SEQS parameter use is summarized

in table 3-2.

You can use one SEQR parameter to deﬁne a special

value step that consists of all characters not

explicitly or implicitly speciﬁed with SEQS param

eters; either YES or Y is allowed. The following

example deﬁnes a sequence in which the digits

c o l l a t e i n n u m e r i c o r d e r, t h e n a l l s p e c i a l c h a r

acters (such as periods, commas, slashes) collate

equally, and then letters collate in alphabetic

order:

SEQS=('0'..'9')

SEQR=YES

SEQS=('A'..'Z')

You can specify one SEQA parameter to alter in

output records characters in the same value step to

the ﬁrst character in the step; either YES or Y is

allowed. The following example alters all aster

isks and ampersands to slashes:

SEQS=(7\ '*','&')

SEQA=YES

Your collating sequence speciﬁcation is terminated

by any parameter other than SEQS, SEQR, or SEQA.

You can deﬁne as many as 100 collating sequences

by specifying a separate series of SEQx parameters

for each collating sequence. The name of each

user-deﬁned collating sequence in a sort or merge

must be unique. If you omit the SEQx parameters, a

nonstandard collating sequence is not deﬁned.

STATUS

The STATUS parameter (ﬁgure 3-16) speciﬁes that a

CYBER Control Language (CCL) variable be set to a

value representing the highest level of error that

occurred during the sort or merge. The CCL vari

ables Rl, R2, R3, RIG, EF, or EFG can be used for

this purpose. Table 3-3 shows values that can be

r e t u r n e d i n t h e s p e c i ﬁ e d v a r i a b l e . F o r e x a m p l e ,

i f y o u s p e c i f y S TAT U S = R 1 a n d R l s u b s e q u e n t l y

contains 20, a trivial error occurred.

STATUS=variable (or ST=variable)

variable CCL variable R1, R2, R3, RIG, EF, or EFG.

Figure 3-16. STATUS Parameter Format

If you include the STATUS parameter in your sort or

merge speciﬁcation, Sort/Merge does not abort if a

catastrophic error occurs before any data records

a r e i n p u t . H o w e v e r , i f S o r t / M e r g e c a l l s a n o t h e r

j4f$P*,\

00^\

TABLE 3-2. SEQS PARAMETER SUMMARY

Number of Value

Steps Speciﬁed by

One SEQS Parameter

1 step

More than one step

Number of Characters

in Each Value Step

1 character

More than one

character

1 character

More than one

character

How Speciﬁed

Graphic character or $CHAR(n)

where n is the number of the char

acter in your character set.

Each character in the step is

enclosed in apostrophes and sepa

rated by commas, or a single char

acter is followed by one or more

ranges of characters.

One range of characters.

Ranges of characters where each

range has the same number of char

acters.

Example

SEQS='H' or

SEQS=$CHAR(08)

SEQS=('I','0','X') or

SEQS=(':','0'..'9')

SEQS=('0'..'9')

SEQS=('A'..'E',/V'..'Z')

60484800 C 3-7

TABLE 3-3. ERROR LEVEL CODES, STATUS PARAMETER

Error Level Code

No errors

Trivial 20

Warning 30

Fata l 40

Catastrophic 50

p r o d u c t a n d a n u n r e c o v e r a b l e e r r o r r e s u l t s , a n

abno r mal j ob t ermin a tion will occu r. For exam p l e,

if Sort/Merge called a product, such as Common

Memory Manager (CMM), and CMM processing results in

an error, a termination will occur. If you omit

the STATUS parameter, no error status variable is

set, and Sort/Merge aborts if a catastrophic error

occurs. The STATUS parameter can be abbreviated as

ST.

SUM

The SUM parameter (ﬁgure 3-17) speciﬁes the

ﬁelds that are to be summed in records with equal

key values. Input ﬁles for a merge with summing

must be presorted and pre summed. As many as 255

bytes can be summed. The records with all key

ﬁelds equal are combined into one new record, and

the other records with equal keys are deleted. The

one resultant record contains the key ﬁelds and

the ﬁelds that are the sums of the speciﬁed

ﬁelds from all the records with equal keys; a data

ﬁeld t h a t is n o t a k e y ﬁeld o r a s u m ﬁeld is set

to the corresponding ﬁeld from one of the old

records.

SUM = ((sum_def)[,(sum_def)] ...)

sum_def j (ﬁrst.length [,type[,rep] ]) 1

[(range[,type[,rep]]) J

range ["ﬁrst 1

[ﬁrst., last J

ﬁrst First byte or bit of the sum ﬁeld.

last Last byte or bit of the sum ﬁeld.

length Number of bytes or bits in the sum ﬁeld;

default is 1.

type Name of a numeric data format (except

REAL); default is INTEGER.

rep Number of ﬁelds to be summed; default

is 1.

Figure 3-17. SUM Parameter Format

All the ﬁelds to be summed must contain numeric

data; otherwise, the contents of the new ﬁelds are

undeﬁned. The ﬁelds described as sum ﬁelds

cannot also be key ﬁelds.

In ﬁgure 3-17, a sum_def describes a ﬁeld to be I

summed by specifying the ﬁeld length and position,

a numeric data format, and the number of ﬁelds to

be summed. The ﬁeld length and position are the

ﬁrst byte or bit of the ﬁeld, and either the

number of bytes or bits in the ﬁeld (length of the

ﬁ e l d ) , o r t h e l a s t b y t e o r b i t o f t h e ﬁ e l d . Yo u

specify ﬁrst and length (or last) as integers; the

leftmost byte and the leftmost bit in a record are

counted as number 1. First and length (or last)

refer to bytes except for formats BINARY_BITS and

INTEGER_BITS. You must specify the ﬁrst byte or

bi t of th e su m ﬁ e ld. If you do no t sp e ci f y o t her

w i s e i n t h e S U M p a r a m e t e r , t h e s u m l e n g t h i s

assumed to be I; sum length can be a maximum of 17 I

bytes. I

Sum ﬁeld type is the name of a numeric data format

except REAL; ﬁelds containing data in REAL format

cannot be summed. You must specify the sum ﬁeld

type; no default exists. See section 2 for a list

of numeric data formats. The number of ﬁelds to

be summed is sp e c iﬁed as an integer; the def a u l t

is 1. If you specify multiple ﬁelds with one SUM

parameter, the ﬁelds must be consecutive, must be

the same length, and must contain the same type of

numeric data* Sum ﬁelds cannot overlap one

another. If a sum ﬁeld contains no data because a

record is short, the sum for that ﬁeld is unde

ﬁned. If you do not specify a sum type, the type

is assumed to be INTEGER.

Figure 3-18 shows an example of a SUM parameter.

Part A of ﬁgure 3-18 shows the layout of the

record containing the ﬁelds to be summed. Part B

shows three SUM parameters, each describing one

ﬁeld to be summed. Part C shows one equivalent

SUM parameter that describes the three consecutive

ﬁelds to be summed. Note that when you specify a

single value-set, two sets of parentheses are still

required.

If the summing of the speciﬁed ﬁelds results in

new numeric values that do not ﬁt into the sum

ﬁelds, a fatal error results. One of the records

that caused the overﬂow is deleted as usual, but

the sum ﬁeld contents in the remaining record are

undeﬁned. A diagnostic message indicating which

ﬁeld caused the overﬂow is issued, and the job

step is terminated at the end of the sort (unless

you include the STATUS parameter).

You can specify the SUM parameter more than once in

a directive ﬁle. If you omit the SUM parameter,

records with equal key values are not automatically

combined into single records. You cannot specify

both the SUM parameter and an owncode 5 routine for

the same sort or merge. You cannot specify both

the RETAIN parameter and the SUM parameter in the

same sort or merge. If you do specify both, a

catastrophic error occurs.

A*ms

A&$S

■*mts

3-8 60484800 U

j0ms

A. Record Layout

1 1 0 1 5 2 0 2 5

Student number

(sort key)

Number of

units registered

Number of

units completed Grade points

B. SUM Parameters

SUM=((11,5.INTEGER))

SUM=((16,5,INTEGER))

SUM=((21,5,INTEGER))

C. SUM Parameter

SUM=((11,5,INTEGER,3))

Figure 3-18. SUM Parameter Example

0^^s VERIFY

The VERIFY parameter (ﬁgure 3-19) directs Sort/

Merge to check merge input records for correct

order. This parameter applies only if you are

performing a merge. If a merge input record is out

of order when you have speciﬁed VERIFY=YES, the

m e r g e i s t e r m i n a t e d ( r e g a r d l e s s o f t h e S TAT U S

parameter) and a diagnostic message is issued. If

VERIFY=N0, merge input records are not checked for

correct order. Omitting the VERIFY parameter is

equivalent to VERIFY=N0. You can specify VER

instead of VERIFY, and Y or N instead of YES or

NO. The VERIFY parameter is ignored if you are

performing a sort.

FASTIO=YES

FASTIO=NO

(or FASTIO=Y)

(or FASTIO=N)

Figure 3-20. FASTI O Parameter Format

If FASTI0=YES, all sort input and output ﬁles with |

block type C and record type F, or block type 1 and

record type W, are read directly by Sort/Merge. If

FASTI0=N0, all sort input and output ﬁles are |

processed using CRM. Omitting the FASTIO parameter

is equivalent to FASTI0=N0. You can specify Y or N

instead of YES or NO.

VERIFY=YES

VERIFY=NO

(or VER=Y)

(or VER=N)

Figure 3-19. VERIFY Parameter Format

FASTIO

The FASTIO parameter (ﬁgure 3-20) speciﬁes that

I certain sort input and output records are to be

read and written directly by Sort/Merge rather than

by means of CYBER Record Manager (CRM). The FASTIO

parameter is ignored during a merge. Files

processed by the FASTIO parameter must reside on

mass storage. The records in these ﬁles must be

block type C and record type F, or block type I and

record type W (see the CRM Basic Access Methods

reference manual for detailed information). All

input ﬁles must have identical record types and

record lengths. If the ﬁle resides on tape, or if

the records are not ail the same type and length, a

warning diagnostic is issued and the FASTIO option

is canceled. The ﬁles are then processed using

CRM. Sort/Merge obtains block and record type

information from the ﬁle information table (FIT);

a FILE control statement is not ignored.

Specifying the FASTIO parameter results in

increased speed, but not all errors are diagnosed;

use this parameter when you are sure your ﬁles are

correctly written and described.

SPECIFYING PARAMETERS

BY POSITION

You can specify any of the Sort/Merge parameters in

a control statement or in a directive ﬁle without

a keyword by supplying a value in the position

assigned to that particular keyword. Table 3-4

lists the parameter keywords and the position

assigned to each. Omitted parameters are replaced

by commas. For example, the following control

statements are equivalent:

S0RT5.INFL,0UTFL,5..9,,,,F

SORTS.FR0M=INFL,T0=0UTFL,KEY=5..9,EL=F

The two preceding control statements specify that

records to be sorted are in the ﬁle 1NFL, and

sorted records are to be written to ﬁle OUTFL.

T h e s o r t k e y ﬁ e l d b e g i n s w i t h t h e ﬁ f t h c h a ra c t e r

of each record, and the last character of the key

ﬁeld is the ninth character of the record. Fatal

and catastrophic errors are to be reported.

You can mix keyword speciﬁcation and positional

speciﬁcation of parameters in the same control

statement or directive ﬁle. whenever you specify

a parameter with a keyword, the parameter retains

its position number (table 3-4). Any positional

parameters you specify after a keyword parameter

are numbered according to the position number of

60484800 D 3-9

TABLE 3-4. PARAMETER POSITIONAL ORDER

Parameter Position

Keyword Number

FROM

KEY

DIR

Reserved for

future use

DIALOG

ENR 10

Reserved for 11

future use

Reserved for 12

future use

OWNF 13

OWNFL 14

OWNMRL 15

Reserved for 16

future use

0WN1 17

0WN2 18

0WN3 19

0WN4 20

0WN5 21

RETAIN 22

SEQA 23

SEQN 24

SEQR 25

SEQS 26

STATUS 27

SUM 28

Reserved for 29

future use

VERIFY 30

FASTIO 31

the last keyword parameter you speciﬁed. The

following example illustrates mixing keyword and

positional parameters:

S0RT5.DIR=DIRFIL,,,,,,6000,RET=Y,,,,,EFG

The DIR parameter is position number 4. Commas

replace parameters 5 through 9. The estimated

number of records (6000) is position number 10.

Numbering begins again with a keyword parameter

(RETAIN number 22). Commas replace parameters 23

through 26. The positional value EFG is the status

variable (STATUS number 27).

FILE CONTROL STATEMENT

Unless default characteristics apply for a ﬁle,

you must specify the characteristics of all input

and output data ﬁles named in a S0RT5 or a MERGE

control statement, or in a directive ﬁle, with a

CYBER Record Manager (CRM) FILE control statement.

FILE control statements must precede the S0RT5 or

th e M E R GE c o n t r o l st a t e m e n t . T h e fo rm at o f t h e

FILE control statement is shown in ﬁgure 3-21.

/^^%V

FILE{rm[,keyword=option] . . . )

l f n L o g i c a l ﬁ l e n a m e ; 1 t o 7 l e t t e r s o r

digits, beginning with a letter.

keyword-option Symbolic name of the ﬁle information

table (FIT) ﬁeld and the option

selected.

Figure 3-21. FILE Control Statement Format

The ﬁle information table (FIT) is a table through

which ﬁle characteristics are communicated to CRM

Basic Access Methods; the FIT contains information

that describes the ﬁle and speciﬁes how the ﬁle

is accessed. You set FIT ﬁelds by specifying the

s y m b o l i c n a m e o f t h e ﬁ e l d i n t h e F I L E c o n t r o l

statement and selecting one of the options associ

ated with that ﬁeld. Several FIT ﬁelds important

to Sort/Merge processing are discussed in the

following paragraphs. The CRM Basic Access Methods

reference manual contains a complete list of all

FIT ﬁelds and options that you can select. Note

that speciﬁcation of certain FIT ﬁelds can

require setting of additional FIT ﬁelds.

I n p u t ﬁ l e s t o b e s o r t e d o r m e r g e d , a n d o u t p u t

ﬁles to which sorted or merged records are writ

ten, must be sequential ﬁles (records are read and

written in order one after the other). The ﬁle

organization (FO) FIT ﬁeld can be included in the

FILE control statement and sequential organization

speciﬁed (F0=SQ); you can omit this parameter and

sequential organization is assumed.

Sequential ﬁles are organized into groups of

records called blocks. Block type is speciﬁed

with the block type (BT) FIT ﬁeld. The default is

BT= C . C R M a l so o ff ers sev e ral reco r d t y pes , an d

any of these record types except U can be sorted or

merged. Record type is speciﬁed with the record

type (RT) FIT ﬁeld. The default is RT=Z. The

full length (FL) or maximum record length (MRL) FIT

ﬁeld must be set to establish the number of char

a c t e r s i n a r e c o r d . T h e d e f a u l t i s F L = 1 5 0 o r

/^^%.

3-10 60484800 D

MRL=150. The default is also a satisfactory

description for records of less than 150 char

acters. If you have any ﬁles of considerably less

than 150 characters (25 characters for example),

Sort/Merge can perform more efﬁciently if you

include a FILE control statement.

Default CRM characteristics apply for the ﬁles

INPUT, OUTPUT, and PUNCH. You do not need to

include FILE control statements for these ﬁles.

Table 3-5 summarizes the default ﬁle characteris

t i c s f o r S o r t / M e r g e i n p u t a n d o u t p u t d a t a ﬁ l e s ,

directive ﬁles, the error and listing ﬁles, and

the ﬁles INPUT, OUTPUT, and PUNCH.

SPECIFYING MAXIMUM

RECORD LENGTH

Unless default characteristics apply for a ﬁle,

Sort/Merge requires that you specify maximum record

length for input and output data ﬁles. You

specify maximum record length by setting the FL or

MRL FIT ﬁeld in the FILE control statement. Set

the FL FIT ﬁeld for F or Z type records to specify

the length o f each record (i f RT = F ) , or the up p e r

li mi t o f ch ar ac te rs i n a r e c o r d ( if RT = Z ) . F o r a ll

other record types, set the MRL FIT ﬁeld to

specify the maximum length of a record.

taneously open. The default ﬁle buffer size (BFS)

is 2001 g words. You can change this value with

FILE control statement to save memory.

TABLE 3-5. FILE CHARACTERISTICS

> w F I T

^sField BT RT FL

File ^"s.

FROM ﬁles FL<150

TO ﬁ l e FL<150

DIR ﬁles FLO00

L ﬁ l e FL=132

E ﬁ l e FL=72*

INPUT FL=80

OUTPUT FL=»140

PUNCH FL=80

'If E=0UTPUT, FL=140 is used.

FIELD LENGTH REQUIREMENTS

Large sort or merge operations can require

additional central memory. The memory needed for a

merge is directly proportional to the number of the

input ﬁles you have, because all ﬁles are simul-

LISTINGS

If an error occurs during the sort or merge, the

message S0RT5 - USER ERROR, JOB ABORTED is sent to

your job dayﬁle. Otherwise, when the sort or

merge is complete, a message indicating the number

of records sorted or merged is sent to the dayﬁle.

r60484800 D 3-11

INTERACTIVE USE

Sort/Merge capabilities are available interac

tively; you can specify any of the Sort/Merge

parameters from your terminal. Sort/Merge also

p r o v i d e s a t u t o r i a l d i a l o g t o a i d i n s p e c i f y i n g

parameters. This section describes the use of this

dia l og, S ort/Merg e inte ractive comma n ds, a n d log in

and logout processes. The complete text of the

dialog is included in appendix E.

In interactive mode, control statements cannot be

continued on more than one line. Instead, you can

either create a directive ﬁle or use a procedure

ﬁle so you can use a text editor to verify or

change complex sort speciﬁcations.

PASSWORD:

E n t e r t h e 4 - t o 7 - c h a r a c t e r p a s s w o r d

assigned to you to provide access security;

some terminals overtype to preserve privacy.

termnam-APPLICATION:

E n t e r I A F ( f o r I n t e r a c t i v e F a c i l i t y , t h e

network software that provides you with the

time- s h a ring capabi l i ties o f NO S ) ; te r m n am

represents the terminal identiﬁcation.

JSN: zzzz, NAMIAF

CHARGE NUMBER:

yms

TERMINAL ACCESS TO

SORT/MERGE

To access Sort/Merge from your terminal, you ﬁrst

e s t a b l i s h t h e p h y s i c a l c o n n e c t i o n b e t w e e n t h e

terminal and the computer. The method of estab

lishing the connection varies depending on the type

of terminal and the coupling between the terminal

and computer. Consult the NOS Version 2 Reference

Set, Volume 1 (for NOS users) or the INTERCOM

reference manual (for NOS/BE users) for information

on connection methods.

Connecting your terminal to the computer initiates

the login process. The operating system, either

NOS or NOS/BE, identiﬁes itself to you, and you

must supply information that indicates you are a

valid system user.

LOGGING IN UNDER NOS

The login procedure to NOS begins with initial

information from the system (including date, time,

terminal name, installation name, and operating

system version). The second line of this informa

tion is dependent on the installation. You are

then prompted to enter information to indicate you

are a valid system user. The prompts and required

responses are as follows:

FAMILY:

JSN: zzzz, NAMIAF

READY.

If a charge number is requested, enter your

assigned charge number and project number.

The system will then respond with READY.

The login procedure is now complete. You

can then enter the batch subsystem. Do not

change to ASCII mode since Sort/Merge does

not process data in this mode.

Figure 4-1 shows an example of a NOS login. You

are now ready to perform sort or merge operations

as de s c r i b e d later i n this s e c t i on.

82/01/08. 10.42.16. T143A

CDC NOS 2

FAMILY:

USER NAME: xxxxxxx

PASSWORD: xxxx

T143A - APPLICATION: iaf

JSN: AADI, NAMIAF

CHARGE NUMBER:

? $X*&£@ftK

PROJECT NUMBER:

READY,

/batch

RFL,0.

Figure 4-1. NOS Login Example

00"*S

Enter the family name of the mass storage

device that contains your permanent ﬁles.

Some installations do not request a family

name; the family name is the default family

for the system.

USER NAME:

Enter the user name assigned to you by your

system administrator.

LOGGING IN UNDER NOS/BE

As soon as your terminal is connected to the

system, the system responds with a message indi

cating date, time, and the INTERCOM version number

(INTERCOM is the software that directs the ﬂow of

data between your terminal and the central site

computer). The system requests the following:

PLEASE LOGIN

60484800 B 4-1

Enter LOGIN. You are then prompted for validation

information. The prompts and required responses

are as follows:

ENTER USER NAME-

Enter the user name assigned to you by your

system administrator.

J3GSmQQQQCKX ENTER PASSWORD-

Enter the password (up to 10 letters and

digits) assigned to you to provide access

security; some terminals overtype to pre

serve privacy.

When the user number and password are accepted, a

2-character user code and the login time, followed

by the equipment number and port number, are dis

played or printed at your terminal. The message

COMMAND- appears, indicating the login process is

complete. Figure 4-2 shows an example of NOS/BE

operations as described in the following subsection.

CONTROL DATA INTERCOM 5.0

D AT E 0 1 / 2 3 / 8 1

TIME 09.35.50.

PLEASE LOGIN

ENTER USER NAME-sort

HtiMftttfcnr ENTER PASSWORD-

01/23/81 LOGGED IN AT 09.36.23.

WITH USER-ID OK

EQUIP/PORT 63/076

C0MMAND-

Figure 4-2. NOS/BE Login Example

INTERACTIVE DIALOG

Sort/Merge provides an interactive dialog to aid In

specifying parameters. The tutorial dialog is

designed to guide you, step by step, through your

sort or merge speciﬁcation. You can specify any

control statement sort or merge, including a sort

or merge with owncode routines and a nonstandard

collating sequence, using the interactive dialog.

After login is complete, and before beginning the

dialog, you must assign any ﬁles used in your sort

or merge (ﬁles containing input records to be

sorted, owncode routines, or presorted ﬁles to be

merged) to your terminal with an ATTACH command or

the equivalent. You must enter CYBER Record Man

ager (CRM) FILE control statements for any of your

ﬁles that are not written in standard unit record

format. Standard unit record format ﬁles include

ﬁles created by punched cards in a batch job or

li ne s e n t e r e d a t a t er mi na l, a n d th e s p e c i a l ﬁ le s

INPUT, OUTPUT, and PUNCH. CRM deﬁnes these ﬁles

as block type C (BT=C) and record type Z (RT=Z).

Sort/Merge assumes each record contains 150 char

acters or less. If you have any ﬁles of consider

ably less than 150 characters (25 characters for

example), Sort/Merge can perform more efﬁciently

i f y o u e n t e r a F I L E c o n t r o l s t a t e m e n t . S e e y o u r

operating system reference manual for information

about retrieving permanent ﬁles and the CRM Basic

Access Methods reference manual for information

about FILE control statements.

For NOS/BE users, Sort/Merge automatically connects

and disconnects ﬁles INPUT and OUTPUT; Sort/Merge

writes the dialog to OUTPUT, which is displayed or

printed at your terminal. Your dialog responses

are read from INPUT.

NOTE

Before entering a S0RT5 or MERGE command

with the DIALOG parameter, ensure that

your terminal is in normal mode by enter

ing the NORMAL terminal command. If your

terminal is in ASCII mode during the

interactive dialogue, Sort/Merge cannot

recognize your responses unless you enter

them using uppercase letters only.

If you are already in the dialog, and need

to change the terminal mode, enter a

control-T to return to the system.

The interactive dialog begins when you type

S0RT5.DIAL0G=YES or MERGE. DIALOG=YES. Any other

parameters (except STATUS) that you enter along

with DIALOG are ignored. During the dialog, Sort/

Merge asks various questions, one at a time, and

your response to each question determines the next

question. For example, Sort/Merge asks if you have

just one input ﬁle and no 0WN1 or 0WN2 subrou

tines. If you enter YES, the next question requests

the name of your input ﬁle. If you enter NO, the

next question asks if you have an 0WN1 subroutine.

The expected response to each question is evident

from the question. For example, one question asks

how many key ﬁelds you have; you enter a number.

For questions requiring yes or no answers, you can

type YES or Y, or NO or N. If you make a mistake

(misspell the word yes, for example), Sort/Merge

gives you an explanation of the expected answer and

asks the question again. If you enter an incorrect

answer a second time, Sort/Merge asks you if you

want help. If you type YES, an explanation of

Sort/Merge usage is displayed or printed at your

terminal. If you do not want help and type NO, the

original question is repeated. Figure 4-3 shows an

example of this sequence.

The last question Sort/Merge asks is how many

r e c o r d s a r e t o b e s o r t e d o r m e r g e d . A f t e r y o u

en te r t h e a p p r o p r i a t e nu m b e r, So rt /M e r g e in fo rm s

you that the sort or merge you have speciﬁed is

beginning. When the sort or merge is complete,

control of your terminal is returned to the oper

ating system; the operating system prompt for input

appears (/ under NOS, COMMAND- under NOS/BE).

/<^K

4-2 60484800 D

>^*v

/sort5.dialog=yes-*

IS YOUR TERMINAL ALREADY SET TO NORMAL MODE OR UPPERCASE

ASCIII MODE. ALSO, ARE YOU SURE THAT ALL YOUR FILES (IF ANY)

ARE EITHER STANDARD UNIT RECORD FORMAT OR HAVE HAD FILE

CONTROL STATEMENTS SPECIFIED FOR THEM

? yes

DO YOU HAVE JUST ONE INPUT FILE AND NO OWN1 OR OWN2 SUBROUTINES

? no

DO YOU HAVE AN OUN1 SUBROUTINE

? y g « » t i .

DO YOU HAVE AN OWN1 SUBROUTINE ** .

? y s e - * .

Begin interactive dialog.

DO YOU WANT HELP

? y

SINCE YOU DID NOT SAY "NO" YOU ARE GOING TO BE HELPED. **

■Y» IS THE SAME AS 'YES' AND 'N' IS THE SAME AS 'NO'.

TYPE 'RESTART' TO GO BACK TO THE BEGINNING.

TYPE 'QUIT' TO STOP RUN.

THE OWN1 ROUTINE CAN BE USED TO PROCESS INPUT RECORDS. A DETAILED

OF THIS OWNCODE ROUTINE IS IN THE SORT/MERGE REFERENCE MANUAL.

DO YOU HAVE AN OWN1 SUBROUTINE^ .

? yse

DO YOU HAVE AN OWN1 SUBROUTINE

? yse

DO YOU WANT HELP

? no

DO YOU HAVE AN OWN1 SUBROUTINE-*

? yes

Incorrect response.

Question asked again.

Second incorrect response.

Help after second incorrect response.

DESCRIPTION

- Question asked again.

No help after second incorrect

response; question asked again.

Figure 4-3. Incorrect Response Example

If sorted or merged records are written to a ﬁle

(not processed by an owncode routine), Sort/Merge

creates a local ﬁle and gives this ﬁle the name

you entered in response to the question "What is

the name of your output ﬁle?" Upon completion of

the sort or merge, you can display or print this

ﬁle at your terminal, or route the ﬁle to a

p r i n t e r t o b e l i s t e d . I f y o u w a n t t o u s e t h e ﬁ l e

at a later time, you must save it as a permanent

ﬁle before logging out; otherwise, the ﬁle is

lost. Figure 4-4 shows an example of output ﬁle

disposition.

After the explanation, the question Sort/Merge

asked you immediately before you entered the HELP

command is asked again, resuming the dialog.

RESTART

If you type RESTART, the interactive dialog is

started over at the beginning. All. previous

answers to questions are discarded. You can use

this command if you give Sort/Merge incorrect

information such as the wrong number of key ﬁelds.

INTERACTIVE COMMANDS QUIT

Several commands exist to control the use of the

interactive dialog. These commands are HELP,

RESTART, and QUIT. A command can be entered at any

time during the dialog, but HELP, RESTART, and QUIT

cannot be the names of user-supplied entities such

as ﬁles and collating sequences. Use of the

commands is described in the following paragraphs.

HELP

If you type QUIT, the interactive dialog is ter

minated. The sort or merge is not performed, and

control of your terminal is returned to the oper

ating system. You can then perform other terminal

operations or logout. If permanent ﬁles con

t a i n i n g i n p u t r e c o r d s t o b e s o r t e d o r m er g e d , o r

owncode routines have been assigned to your job,

you should release these ﬁles by using the RETURN

c o m m a n d ( s e e y o u r o p e r a t i n g s y s t e m r e f e r e n c e

manual)*

If you type HELP, an explanation of Sort/Merge

u s a g e i s d i s p l a y e d o r p r i n t e d a t y o u r t e r m i n a l .

Figure 4-5 shows an example of interactive command

use*

60484800 D 4-3

/sort5.dialog=yes

WHAT IS THE NAME OF YOUR OUTPUT FILE

? s r t o u t «

THANK YOU SORT/MERGE NOW BEGINS.

SORT5.diaLog=YES—

/copysbf,srtout,output -*-

1REYES

MAYER

CHARLES

MARTIN

NEECE

S L 100246031558ANTHRO

M I 100991122359ANTHR0

S H 101418032459ANTHR0

R C 100955082157ART

M L 99911121358ART

Begin interactive dialog.

Local ﬁle created for sorted output.

Dialog completion; sort begins.

NOS repeats control statement (up to 48 characters)

Display or print sorted ﬁle.

L A S E U R P T

SUGARMAN B T

S M I T H F R

DOUGLAS M L

O K A D A N A

EOI ENCOUNTERED,

/rewind,srtout

REWIND,SRTOUT.

/copysbf,srtout,myﬁ I

EOI ENCOUNTERED,

/route,myf i L,dc=pr -*-

ROUTE COMPLETE,

/save,srtout —

100678042256PSYCH

100528070457SOC

101062120758SOC

101325071558UNDEC

100103111750UNDEC

3341

2382

2453

2891

2291

2233

3501

2913

2585

2225

File SRTOUT.

Copy sorted ﬁle to be listed by printer.

Route copy of sorted ﬁle to printer.

Make sorted ﬁle a permanent ﬁle.

Figure 4-4. Output File Disposition Example

I *-* 60484800 D

jm&s

/sort5.dia Log=yes Begin interactive dialog.

Dialog started over

at beginning.

IS YOUR TERMINAL ALREADY SET TO NORMAL MODE OR UPPERCASE

ASCII MODE. ALSO, ARE YOU SURE THAT ALL YOUR FILES (IF ANY)

ARE EITHER STANDARD UNIT RECORD FORMAT OR HAVE HAD FILE

CONTROL STATEMENTS SPECIFIED FOR THEM

? y

YOU HAVE JUST ONE INPUT FILE AND NO OWN1 OR 0WN2 SUBROUTINES

? n

DO YOU HAVE AN OWN1 SUBROUTINE

? y

WHAT IS THE NAME OF THE FILE CONTAINING YOUR OWNCODE SUBROUTINE(S)

? myrthns

WHAT IS THE NAME OF YOUR OWN1 SUBROUTINE

? restart-* Command entered.

IN RESPONSE TO YOUR REQUEST RESTART HAS BEEN INITIATED.

IS YOUR TERMINAL ALREADY SET TO NORMAL MODE OR UPPERCASE )

ASCII MODE. ALSO, ARE YOU SURE THAT ALL YOUR FILES (IF ANY)

ARE EITHER STANDARD UNIT RECORD FORMAT OR HAVE HAD FILE

CONTROL STATEMENTS SPECIFIED FOR THEM

? y

DO YOU HAVE JUST ONE INPUT FILE AND NO OWN1 OR OWN2 SUBROUTINES

? n

DO YOU HAVE AN OWN1 SUBROUTINE

? y

WHAT IS THE NAME OF THE FILE CONTAINING YOUR OWNCODE SUBROUTINE(S)

? ownsrtn

WHAT IS THE NAME OF YOUR OWN1 SUBROUTINE

? myrtn

DO YOU HAVE AN 0WN2 SUBROUTINE

? h e l p — ■ Command entered.

*Y' IS THE SAME AS 'YES' AND 'N' IS THE SAME AS 'NO'. ^

TYPE 'RESTART' TO GO BACK TO THE BEGINNING. J „

TYPE 'QUIT' TO STOP RUN. > USa9e ..

THE OWN2 ROUTINE CAN BE USED TO PROCESS INPUT FILES. A DETAILED DESCRIPTION 1 exPlanar,on

OF THIS OWNCODE ROUTINE IS IN THE SORT/MERGE REFERENCE MANUAL. /

DO YOU HAVE AN 0WN2 SUBROUTINE

? quit-* Conmand entered.

sort/merge is quitting.-* Sort/Merge terminated.

Figure 4-5. Interactive Command Example

r60484800 D 4-4.1/4-4.2

SPECIFYING PARAMETERS

INTERACTIVELY

You can specify any of the Sort/Merge parameters

from your terminal in a S0RT5 or MERGE control

statement. After you login, assign all necessary

ﬁles (including directive ﬁles) to your job, and

type in FILE control statements for ﬁles with

other than default characteristics. You enter

S0RT5 or MERGE followed by parameters In one of the

following forms:

S0 RT5 .p l , p 2 , . .. ,p n

MERGE.plA p2A ...A pn

In the preceding example, p represents a processing

parameter. Parameters must be separated by commas

or spaces; commas and spaces can be used inter

changeably. You can enter the control statement

and parameters in input line character positions 1

through 80. If your sort or merge speciﬁcation Is

longer than 80 characters, you must use a directive

ﬁle for the additional parameters.

Figure 4-6 shows an example of specifying param

eters interactively.

LOGGING OUT

To end your terminal session, you type GOODBYE,

BYE, or LOGOUT if you are using the Interactive

Facility of NOS. If you are using INTERCOM under

NOS/BE, you type LOGOUT. Under either system, a

message indicating that you are logged out is dis

played or printed at your terminal.

0^S

/attach,univer -*

/ﬁle(univer,bt=c,rt=z,ﬂ=38)

FILE(UNIVER,BT=C,RT=Z,FL=38)

/ﬁle(oout,bt=c,rt=z,f1=38) \

FILE(00UT,BT=C,RT=Z,FL=38) ;

/sort5.from=univer,to=oout,key=((35,3,asci i6,d)) -*■

S0RT5.FR0M=UNIVER,T0=00UT,KEY=((35,3,ASCI16,D)) —

Input ﬁle assigned to the job.

FILE control statements for input

and output ﬁles because records

are only 35 characters in length.

-- S0RT5 control statement entered.

— NOS repeats control statement

(up to 48 characters).

— Control returns to NOS operating system.

Figure 4-6. Interactive Parameter Speciﬁcation Example

i0&s.

60484800 B 4-5

PROCEDURE CALLS

0?*^.

So r t/ M er g e a ls o e x is t s a s a s e t o f p ro c ed ure s; a

sort or merge can be invoked from within a program

by a sequence of procedure calls. The procedures

can be called from FORTRAN or by a program in any

language that uses the standard FORTRAN calling

sequence. The same type of Sort/Merge processing

is available through the procedures as is available

through the S0RT5 or the MERGE control statement;

t h e p r o c e d u r e c a l l s c l o s e l y c o r r e s p o n d t o t h e

Sort/Merge parameters. This section describes the

procedure calls; COBOL programmers should see

section 6 for information on calling the procedures.

All conventions for coding FORTRAN statements must

be observed in procedure calls (see the FORTRAN 5

reference manual). You must assign any permanent

ﬁles named in procedure calls to your job with an

ATTACH command (or the equivalent) in the control

statement section of the job. The procedures

cannot be used in a static memory management

environment. A merge could require twice as much

mass storage for intermediate merge ﬁles as is

required for the merge input ﬁles.

All ﬁles to be sorted, or presorted ﬁles to be

me r g e d , mu st b e s e q u e n t i a l ﬁ l e s . T he d e f au lt ﬁ l e

characteristics are block type C, record type Z,

and full length of 150 characters or less. For

ﬁles with other block types and record types, you

must include CYBER Record Manager (CRM) FILE

control statements in the control statement section

of your job. See the CRM Basic Access Methods

reference manual for information on block types and

record types.

The procedures can be called in any order with two

exceptions: SM5S0RT or SM5MERG must be the ﬁrst

procedure called, and SM5END must be the last

procedure called. Sort/Merge collects processing

information until SM5END is called; the sort or

merge is then performed. Unless stated otherwise,

a procedure can be called only once during a sort

or merge. Refer to section 7 for information on

owncode routines, which are mentioned in the

descriptions of several of the procedures.

SM5SORT

Execution of the SM5S0RT procedure signals the

b e g i n n i n g o f a s o r t s p e c i ﬁ c a t i o n . T h e S M 5 S 0 R T

procedure must be the ﬁrst procedure called for a

so rt. The format of the SM5S0RT ca ll i s sh own in

ﬁgure 5-1.

CALL SM5SORT(option)

option Result array name (1 to 7 letters or digits,

beginning with a letter) or the integer 0.

s t a t i s t i c s a n d r e s u l t s t o y o u r p r o g r a m w h e n t h e

sort is completed. You set the ﬁrst element of

the result array to the number of elements (as many

as 15) in which you want to receive information.

If you set the ﬁrst element to 0 or specify CALL

SM5SORT(0), no statistics or results are returned.

The type of result that is returned in each element

is shown in table 5-1.

TABLE 5-1. RESULT ARRAY FORMAT

Figure 5-1. SM5SORT Format

The array associated with the SM5S0RT procedure is

a 16-element integer array that returns sort

Array

Element

Number

Contents

1Number of elements of results

you want returned (0 through

15)

2Number of records read from

sort or merge input ﬁles

3Number of records deleted by

an owncode 1 routine

4Number of records inserted by

an owncode 1 routine

5Number of records inserted by

an owncode 2 routine

6Number of records sorted or

merged

7Number of records deleted by

an owncode 3 routine

8Number of records inserted by

an owncode 3 routine

9Number of records inserted by

an owncode 4 routine

10 Reserved

11 Number of records deleted by

an owncode 5 routine

12 Number of records combined by

summing

13 Number of records written to

o u t p u t ﬁ l e

14 Minimum record length in

characters

15 Average record length in

characters

16 Maximum record length in

characters

60484800 A 5-1

SM5MERG

Execution of the SM5MERG procedure signals the

beginning of a merge speciﬁcation. The SM5MERG

procedure must be the ﬁrst procedure called for a

merge. The format of the SM5MERG call is shown in

ﬁgure 5-2.

CALL SM5MERG(option)

option Result array name (1 to 7 letters or digits,

beginning with a letter) or the integer 0.

Figure 5-2. SM5MERG Format

The input ﬁles for a merge must be presorted.

Input ﬁles for a merge with summing must be

presorted and presummed.

The array asociated with the SM5MERG procedure is a

16-element integer array that returns merge statis

tics and results to your program when the merge is

complet e d . You set the ﬁrst e l e ment of the r e s u l t

array to the number of elements (as many as 15) in

which you want to receive information. If you set

the ﬁrst element to 0 or specify CALL SM5MERG(0),

no statistics or results are returned. The type of

result that is returned in each element is shown in

table 5-1. Note that the array is not set by Sort/

Merge until the end of the merge.

SM5FROM

Execution of the SM5FR0M procedure speciﬁes the

input ﬁles from which the records to be sorted or

merged are read. The format of the SM5FR0M call is

shown in ﬁgure 5-3.

CALL SMSFROMCﬁlel'I/ﬁle?:

'ﬁle' File name; 1 to 7 letters or digits, beginning

with a letter.

Figure 5-3. SM5FROM Format

Yo u c a n s p e c i f y a s m a n y a s 1 0 0 ﬁ l e s w i t h o n e

SM5FR0M procedure, or you can call the SM5FR0M

procedure as many as 100 times. The ﬁles are read

in the ord e r t h at you spe c ify the m ; a ﬁl e ca n not

be speciﬁed more than once for a merge. The ﬁles

are rewound before and after use unless the name of

the ﬁle is INPUT or a FILE control statement

parameter speciﬁes otherwise. If a ﬁle is open

when Sort/Merge is called, it is not rewound after

use because the CF value in the FIT is undetermined.

Sort/Merge assumes that the ﬁles you specify with

the SM5FR0M procedure are CRM block type C and

record type Z with FL of 150 characters or less.

If you specify ﬁles that have different character

istics, you must describe these ﬁles to Sort/Merge

by including FILE control statements in the control

statement section of your job. You should open the

FROM ﬁles before doing a sort or merge to be sure

that Sort/Merge gets passed the correct ﬁle

attributes. The ﬁles are not read past embedded

e n d - o f - r e c o r d o r e n d - o f - ﬁ l e m a r k s ( s e e t h e C R M

Basic Access Methods reference manual).

If the SM5FR0M procedure is not called, records to

be sorted are read from the ﬁle OLD (unless an

owncode routine supplies records). An empty

SM5FR0M ﬁle results in a null sort. The SM5FR0M

ﬁle cannot be the same as the SM5T0 ﬁle.

SM5TO

Execution of the SM5T0 procedure speciﬁes the ﬁle

to which sorted or merged records are written

(unless an owncode routine processes all records).

The format of the SM5T0 call is shown in ﬁgure 5-4.

CALL SM5TOCﬁle')

'ﬁle' File name; 1 to 7 letters or digits, beginning

with a letter.

Figure 5-4. SM5TO Format

The ﬁle is not rewound before or after use unless

the name of the ﬁle is OUTPUT or a FILE control

statement parameter speciﬁes otherwise. Sort/

Merge assumes that the ﬁle you specify with the

SM5T0 procedure is CRM block type C and record type

Z with FL of 150 characters or less. If you

specify a ﬁle with different characteristics, you

must describe the ﬁle to Sort/Merge by including a

FILE control statement in the control statement

section of your job. You should open the TO ﬁle

before doing a sort or merge to be sure that

Sort/Merge gets passed the correct ﬁle attributes.

S o r t / M e r g e w r i t e s t h e ﬁ l e d i r e c t l y, a n d t h e ﬁ l e

i s c l o s e d w h e n t h e s o r t o r m e r g e i s c o m p l e t e d

unless the name of the ﬁle is INPUT. If your

program also reads or writes the ﬁle, you must

include an OPEN or REWIND statement in your program

after calling the procedures. If the SM5TO proce

dure is not called, records are written to the ﬁle

NEW (unless an owncode routine processes all

records). The SM5T0 ﬁle cannot be the same as the

SM5FR0M ﬁle.

SM5KEY

E x e c u t i o n o f t h e S M 5 K E Y p r o c e d u r e s p e c i ﬁ e s a

single key ﬁeld for the sort or merge. The format

of the SM5KEY call is shown in ﬁgure 5-5.

CALL SMSKEYfﬁrst.length.'typeVad')

ﬁrst First byte or bit of the key ﬁeld.

length Number of bytes or bits in the key ﬁeld.

'type' Name of a numeric data format or collating

sequence.

'ad' Order, either 'A' for ascending or 'D' for

descending.

Figure 5-5. SM5KEY Format

5-2 60484800 D

J0&S

In the SM5KEY procedure call, you specify the ﬁrst

byte or bit of the key ﬁeld, the length of the

ﬁeld in bytes or bits, the key type (the name of a

collating sequence or a numeric data format), and

the sort order. You must specify the key charac

teristics in the order shown.

You specify ﬁrst and length as integers; the

leftmost byte or bit in a record is counted as

number 1. For character data, each character is

six bits and occupies one byte. For numeric data,

ﬁrst and length refer to bytes except for formats

BINARY_BITS and INTEGER_BITS. You must specify the

ﬁrst byte or bit of a key ﬁeld. If you do not

specify the length, length defaults to 1.

The key type is the name of a predeﬁned collating

sequence or your own collating sequence deﬁned by

SM5SEQx procedures, or the name of a numeric data

format. The name must be enclosed in apostrophes.

The default type is the ASCII6 collating sequence.

See section 2 for a list of predeﬁned collating

sequences and numeric data formats.

The sort order is either ascending or descending,

and you specify 'A' or 'D', respectively. The

default order is ascending. For numeric keys

sorted in ascending order, the record having the

key with the highest value is written last on the

output ﬁle. Ascending character keys are sorted

according to the collating sequence you specify.

For numeric keys sorted in descending order, the

record having the key with the lowest value is

written last on the output ﬁle. Descending

character keys are sorted in reverse order of the

collating sequence you specify.

You can call the SM5KEY procedure as many as 100

times during a sort or merge to specify multiple

sort keys. Output records are sorted or merged

according to the key ﬁeld described by the ﬁrst

SM5KEY procedure called, then according to the key

ﬁeld described by the second SM5KEY procedure

called, and so on. The total number of key char

acters must be less than 256 (if you also specify

sum ﬁelds, the total number of characters in the

key and sum ﬁelds together must be less than

256). Unless the SM5SEQA procedure is used, key

ﬁelds can overlap one another. If the SM5KEY

procedure is not called, the entire record is used

as a character key ﬁeld, and the default key type

and sort order are used.

SM5E

Execution of the SM5E procedure speciﬁes the ﬁle

to which diagnostic messages are written. The

format of the SM5E call is shown in ﬁgure 5-6.

CALL SM5E('ﬁle')

'ﬁle' File name; 1 to 7 letters or digits, beginning

with a letter.

Figure 5-6. SM5E Format

The ﬁle is written in CRM BT=C, RT=Z, and FL=72

f o r m a t . T h e ﬁ l e i s n o t r e w o u n d b e f o r e o r a f t e r

use unless a FILE control statement parameter

s p e c i ﬁ e s r e w i n d . I f t h e S M 5 E p r o c e d u r e i s n o t

call e d, d i agno s tic m essa g e s a r e w r i tten to t he ﬁ l e

OUTPUT. If you specify the ﬁle $NULL with the

SM5E procedure, diagnostic messages are not written.

SM5EL

Execution of the SM5EL procedure speciﬁes the

error level to be reported. The format of the

SM5EL call is shown in ﬁgure 5-7.

CALL SM5EL(error-level)

error-level Alphabetic character enclosed in apostrophes

or integer (see table 5-2).

Figure 5-7. SM5EL Format

The error levels that you can select are shown in

tab l e 5 - 2. Yo u c an s peci f y t h e a l pha b eti c ch a rac

ter enclosed in apostrophes or the integer. For

example if you specify 'W' or 2, any warning,

fatal, and catastrophic error messages are re

ported. Errors are written to the ﬁle speciﬁed

by the SM5E procedure or, if the SM5E procedure is

n o t c a l l e d , t o t h e ﬁ l e O U T P U T. I f t h e S M 5 E L

procedure is not called, errors of levels T, W, F,

and C are reported. See appendix B for an explana

tion of the four levels of errors.

TABLE 5-2. ERROR LEVEL SPECIFICATION

Error Level Errors Reported

'T' or 1

'W or 2

'F' or 3

'C or 4

T, W, F, C

W, F, C

F, C

SM5ENR

Execution of the SM5ENR procedure speciﬁes the

estimated number of records to be sorted or

merged. The format of the SM5ENR call is shown in

ﬁgure 5-8.

CALL SM5ENR(integer)

integer Number of records to be sorted or merged;

0 to 1000000000.

Figure 5-8. SM5ENR Format

The integer you specify is the approximate number

of records to be sorted or merged. If SM5ENR is

speciﬁed as less than 1500, Sort/Merge uses less

memory for the initial sort phase. This causes

no ticeably less memory required for the entire sort

if the actual number of records is a low number.

60484800 D 5-3

SM5FAST

Execution of the SM5FAST procedure can specify that

certain sort input and output records are to be

read and written directly by Sort/Merge rather than

by means of CYBER Record Manager (CRM). The

SM5FAST procedure is ignored during a merge. The

format of the SM5FAST call is shown in ﬁgure 5-9.

CALL SM5FAST(option)

option 'YES' or 'Y'

IN©' or 'N'

Figure 5-9. SM5FAST Format

T h e op t i o n c a n b e 'Y E S ' o r ' N O ' (o r 'Y ' or ' N ' ).

| I f t h e o p t i o n i s ' Y E S ' , c e r t a i n s o r t i n p u t a n d

output records are read and written directly by

Sort/Merge. If the option is 'NO', CRM is used.

Fi l es pr o ce s se d b y t h e S M5 FA ST p r oc edu re mu s t

reside on mass storage. The records in these ﬁles

must be block type C and record type F, or block

type I and record type W (see the CRM Basic Access

Methods reference manual for detailed infor

m a t i o n ) . A l l i n p u t ﬁ l e s m u s t h a v e a n i d e n t i c a l

re c or d t yp e a n d r ec or d l e ng t h. If t h e ﬁ le re s id es

on tape, or if the records are not all the same

type and length, a warning diagnostic is issued and

the SM5FAST option is canceled. The ﬁles are then

processed using CRM. Sort/Merge obtains block and

r e c o r d t y p e i n f o r m a t i o n f r o m t h e ﬁ l e i n f o r m a t i o n

table (FIT); a FILE control statement is not

ignored. Calling the SM5FAST procedure with 'YES'

results in increased speed, but not all errors are

diag n o sed; use t his p roced u r e wh e n yo u are sure

your ﬁles are correctly written and described.

SM5NODA

Execution of the SM5N0DA procedure speciﬁes that

Sort/Merge dayﬁle messages are to be suppressed.

The format of the SM5N0DA call is shown in ﬁgure

5-10.

CALL SM5NODA

Figure 5-10. SM5NODA Format

The integer you specify is the exact number of

characters in each record; a catastrophic error

results if a record from an owncode routine does

n o t h a v e t h e e x a c t n u m b e r o f c h a r a c t e r s . I f t h e

SM50FL procedure is not called, records from an

owncode routine can be no longer than the longest

allowed input or output record. You cannot call

both the SM50FL procedure and the SM50MRL procedure

in the same sort or merge.

SM50MRL

Execution of the SM50MRL procedure speciﬁes the

maximum length of any record from an owncode

routine. The format of the SM50MRL call is shown

in ﬁgure 5-12.

CALL SM50MRLIinteger)

Figure 5-12. SM50MRL Format

The integer you specify is the maximum record

length in characters. (The SM50FL procedure is

recommended if all records from an owncode routine |

are the same length.) The SM50MRL procedure need

not be called if the sort or merge has an input or

outp u t ﬁl e with a m a ximum reco r d len g th a t lea s t

a s l o n g a s t h e r e c o r d l e n g t h s p e c i ﬁ e d b y t h i s

procedure. If the SM50MRL procedure is not called,

records from an owncode routine can be no longer |

than the longest allowed input or output record.

You cannot call both the SM50MRL procedure and the

S M 5 0 F L p r o c e d u r e i n t h e s a m e s o r t o r m e r g e . |

SMSOWNn

Execution of the SMSOWNn procedure speciﬁes an

owncode routine that is executed each time a

certain point is reached during the sort or merge.

The format of the SM50WNn call is shown in ﬁgure

5-13. |

CALL SM50WNn(proc)

n1, 2, 3, 4, or 5

proc Owncode routine

beginning with a

name;

letter.

1 to 7 letters or digits.

SM50FL Figure 5-13. SM50WNn Format

Execution of the SM50FL procedure speciﬁes the

number of characters In ﬁxed length records from

an owncode routine. The format of the SM50FL call

is shown in ﬁgure 5-11.

CALL SM50FL(integer)

Figure 5-11. SM50FL Format

The procedure you specify is the name (address) of

the owncode routine; in a FORTRAN program, the name

must be declared in an EXTERNAL statement. The n

associated with the SM50WNn call must be 1, 2, 3,

4, or 5. If the SM50WNn procedure is not called,

no owncode routine is executed.

SM5RETA

Execution of the SM5RETA procedure can specify that

records with equal sort keys are output in the same

5-4 60484800 D

order as they are input. The format of the SM5RETA

call is shown in ﬁgure 5-14.

CALL SM5RETA(option

option 'YES' or 'Y'

' N O ' o r ' N '

| Figure 5-14. SM5RETA Format

T h e o p t i o n c a n b e ' YE S ' o r ' N O ' ( o r ' Y ' o r ' N ' ) .

If the option is 'YES', records with equal keys

retain their original order. If the option is

' N O ' , r e c o r d s w i t h e q u a l k e y s m i g h t n o t r e t a i n

their original order.

If RETAIN=YES and you specify more than one sort or

merge input ﬁle with the SM5FROM procedure, the

order in which you specify the input ﬁles is the

order in which records with equal keys are output.

If the SM5RETA procedure is not called, records

with equal keys might not be output in the same

order as they are input.

I You cannot specify both the SM5RETA procedure and

the SM5SUM procedure in the same sort or merge. If

you do specify both, a catastrophic error occurs.

SM5SEQx

Execution of a series of SM5SEQx procedures

speciﬁes a user-deﬁned collating sequence that is

referenced by the SM5KEY procedure. The formats of

the SM5SEQx calls are shown in ﬁgure 5-15.

CALL SM5S EQN ('name')

CALL SM5SEQS('char* 'char')

[CALL SM5SEQRCYES') or CALL SM5SEQR{'Y')

[CALL SM5SEQA('YES') or CALL SM5SEQA('Y')

'name' Collating sequence name; 1 to 31 letters,

digits, or the special characters ? # @ —.

Only 10 characters are signiﬁcant and the

ﬁrst must be a letter.

'char' Character in the collating sequence.

Figure 5-15. SM5SEQx Formats

The SM5SEQN procedure names your collating sequence.

The name is a literal enclosed in apostrophes; the

name cannot be the same as the name of one of the

predeﬁned collating sequences. The SM5SEQN proce

dure must be the ﬁrst called in the series of

SM5SEQx procedures.

Co l la t in g p osi ti o ns ar e a s si g ne d t o t h e c ha r ac t er s

in your collating sequence by the SM5SEQS proce

dure, which can be called more than once. Each

SM5SEQS procedure speciﬁes a value step in the

sequence. A value step consists of one or more

characters, and characters speciﬁed in the same

step collate equally. Characters speciﬁed with

the ﬁrst SM5SEQS procedure called collate ﬁrst,

characters speciﬁed with the second SM5SEQS

procedure called collate second, and so on. The

following paragraphs describe how to specify steps

in your collating sequence with SM5SEQS procedures.

You can specify a value step consisting of one

character by a graphic character (see appendix A)

enclosed in apostrophes as in the following example:

CALL SM5SEQS('A')

You can specify a value step consisting of several

characters by single characters enclosed in

apostrophes and separated by commas. Tne following

example speciﬁes a step consisting of the blank

and the digits 0, 1, 2, and 3, all of which collate

equally:

CALL SM5SEQSC ' ,'0','1','2','3')

If the SM5SEQR procedure is called, equal collating

position is assigned to any characters not

sp eciﬁe d by SM5 SEQS procedure s. F or e xampl e, if

you deﬁne a collating sequence that includes

digits and letters but not any of the special

characters (such as commas, periods, slashes), a

call to SM5SEQR includes the special characters in

your collating sequence. The special characters

a l l c o l l a t e e q u a l l y. T h e S M 5 S E Q R p r o c e d u r e c a n

only be called once for each call to SM5SEQN; the

SM5SEQR procedure can be called anywhere in the

SM5SEQx series.

If the SM5SEQA procedure is called, characters in

the same value step are altered on output to the

ﬁrst character in the step. The following example

alters all asterisks and ampersands to slashes:

CALL SM5SEQS('/','*','&')

CALL SM5SEQA('YES')

The SM5SEQA procedure can only be called once for

each call to SM5SEQN; the SM5SEQA procedure can be

called anywhere in the SM5SEQx series.

Your collating sequence speciﬁcation is terminated

by any procedure call other than SM5SEQS, SM5SEQR,

or SM5SEQA. You can deﬁne more than one collating

sequence by calling a separate series of SM5SEQx

procedure s for e ach co l l a ting s e q u ence. The name

of ea c h c ol l ati ng seq uen ce m us t b e u n iq u e. If no

SM5SEQx procedures are called, a nonstandard

collating sequence is not deﬁned.

SM5ST

E x e c u t i o n o f t h e S M 5 S T p r o c e d u r e s p e c i ﬁ e s a n

integer variable that is set to a value represent

ing the highest level of error that occurred during

the sort or merge. The format of the SM5ST call is

shown in ﬁgure 5-16. |

CALL SM5ST(variable)

variable Symbolic name of an integer variable; 1 to

7 letters or digits, beginning with a letter.

Figure 5-16. SM5ST Format

0ms,

60484800 D 5-5

The error level that is represented by the value

returned to the SM5ST procedure is shown in table

5-3. For example if a 3 is returned to the SM5ST

procedure, a fatal error occurred during the sort

or merge. If you call the SM5ST procedure, Sort/

Merge does not abort if a catastrophic error occurs

before any data records are input. However, if

Sort/Merge calls another product and an unrecover

able error results, an abnormal job termination

w i l l o c c u r . F o r e x a m p l e , i f S o r t / M e r g e c a l l s a

product, such as Common Memory Manager (CMM), and

CMM processing results in an error, a termination

wi ll o c c u r.

TABLE 5-3. ERROR LEVEL CODES, SM5ST CALL

Error Level Code

No errors

Trivial 10

Warning 20

Fatal 30

Catastrophic 40

SM5SUM

Execution of the SM5SUM procedure speciﬁes ﬁelds

that are to be summed in records with equal key

values. Input ﬁles for a merge with summing must

be presorted and pre summed. As many as 255 bytes

can be summed. The records with all key ﬁelds

equal are combined into one new record, and the

other records with equal keys are deleted. The one

resultant record contains the key ﬁelds and the

ﬁ e l d s t h a t a r e t h e s u m s o f t h e s p e c i ﬁ e d ﬁ e l d s

from all the records with equal keys; a data ﬁeld

t h a t i s n o t a k e y ﬁ e l d o r a s u m ﬁ e l d i s s e t t o

the corresponding ﬁeld from one of the old records.

The format of the SM5SUM call is shown in ﬁgure

5-17.

CALL SM5SUM(ﬁrst,length,'type',rep)

ﬁrst First byte or bit of the sum ﬁeld.

length Number of bytes or bits in the sum ﬁeld.

'type* Name of a numeric data format.

rep Number of ﬁelds to be summed.

Figure 5-17. SM5SUM Format

All the ﬁelds to be summed must contain numeric

data; otherwise, the contents of the new ﬁeld are

undeﬁned. The ﬁelds described as sum ﬁelds

cannot also be key ﬁelds.

The SM5SUM procedure call speciﬁes the ﬁrst byte

or bit of the sum ﬁeld, the length of the sum

ﬁeld in bytes or bits (length can be a maximum of

17 bytes), the type of data in the sum ﬁeld (the

name of a numeric data format except REAL), and the

number of ﬁelds to be summed. See section 2 for a

list of numeric data formats.

If you specify multiple ﬁelds with one SM5SUM

proc e d ure, the ﬁ e lds m ust b e c o n secut i v e, b e the

same length, and contain the same type of numeric

data. Sum ﬁelds cannot overlap one another. If a

s u m ﬁ e l d c o n t a i n s n o d a t a b e c a u s e a r e c o r d i s

short, the sum for that ﬁeld is undeﬁned.

Figure 5-18 shows an example of an SM5SUM procedure I

c a l l . P a r t A o f ﬁ g u r e 5 - 1 8 s h o w s t h e l a y o u t o f |

the record containing ﬁelds to be summed. Part B

shows three SM5SUM procedure calls, each describing

one ﬁeld to be summed. Part C shows one equiva

lent SM5SUM procedure call that describes the three

consecutive ﬁelds to be summed.

If the summing of the speciﬁed ﬁelds results in

new numeric values that do not ﬁt into the sum

ﬁelds, a fatal error results. One of the records

that caused the overﬂow is deleted as usual, but

the sum ﬁeld contents in the remaining record are

undeﬁned. A diagnostic message indicating which

ﬁeld caused the overﬂow is issued, and the job is

terminated at the end of the sort (unless you call

the SM5ST procedure).

You can call the SM5SUM procedure as many as 100

ti me s d u r in g a s or t o r m e r g e t o s pe ci fy d i ff e r e nt

sum ﬁelds. If the SM5SUM procedure is not called,

records with equal key values are not automatically

combined into single records. You cannot call both

the SM5SUM procedure and the SM50WN5 procedure in

the same sort or merge.

You cannot specify both the SM5RETA and the SM5SUM

procedure in the same sort or merge. If you do

specify both, a catastrophic error occurs.

A&^^S

5-6 60484800 D

A. Record Layout

10 15 20 25

Student number

(sort key)

Number of

units registered

Number of

units completed Grade points

B. SM5SUM Procedure Calls

CALL SM5SUM(11,5,'INTEGER',1)

CALL SM5SUM(16.5,'INTEGER', D

CALL SM5SUM(21,5,'INTEGER', 1)

C. SM5SUM Procedure Call

CALL SM5SUM(11,5,'INTEGER', 3)

Figure 5-18. SM5SUM Procedure Call Example

/$*S SM5END

Execution of the SM5END procedure terminates your

sort or merge speciﬁcation and initiates Sort/

Merge processing. The format of the SM5END call Is

shown in ﬁgure 5-19. The SM5END procedure must be

the last called for any one sort or merge.

CALL SM5END

Figure 5-19. SM5END Format

yims.

60484800 D 5-7

COBOL INTERFACE

/ f ^ v

Two options are available for using Sort/Merge 5

from COBOL 5: a control statement option when you

use the COBOL SORT or MERGE statement and direct

calls. This section describes these two options.

COBOL5 CONTROL STATEMENT

OPTION

Sort/Merge is used by COBOL when the COBOL SORT or

MERGE statement is included in your program. You

can select Sort/Merge 5 by means of a C0B0L5

control statement option. If you specify S0RT5 in

your C0B0L5 control statement, Sort/Merge 5 is used

when the SORT or MERGE statement in your program

executes. Figure 6-1 shows a sample C0B0L5 control

statement that includes the SORTS option. See the

COBOL 5 reference manual for information on coding

the SORT or MERGE statement in a program.

COBOL5,l=PROGA,SORT5,L=REPTO.

Figure 6-1. COBOL5 Control Statement

Example

DIRECT CALLS

Direct calls from a COBOL 5 program to the Sort/

Merge procedures described in section 5 can be made

with ENTER FTN5 SM5xxx statements. Calls to the

procedures cannot be mixed with the SORT or MERGE

s t a t e m e n t ; t h e p r o c e d u r e s c a n n o t b e c a l l e d i n a

COBOL input or output procedure, and the SORT or

MERGE statement cannot be executed in a series of

direct calls (before SM5END is called). See the

COBOL 5 reference manual for information on coding

ENTER statements.

All ﬁles to be sorted or merged must be sequential

ﬁles. The default ﬁle characteristics are block

type C, record type Z, and full length of 150

characters or less. For ﬁles with other block

types and record types, you must include CYBER

Record Manager (CRM) FILE control statements in the

control statement section of your job (see the CRM

Basic Access Methods reference manual).

r60484800 A 6-1

0ms OWNCODE ROUTINES

Yo u c a n w r i t e a r o u t i n e t o i n s e r t , s u b s t i t u t e ,

modify, or delete input and output records during

Sort/Merge processing. Your routine, called an

owncode routine, is executed each time the sort or

merge reaches a certain point. Sort/Merge passes a

record to the owncode routine, which processes the

record. When the record is returned to Sort/Merge

from the owncode routine, Sort/Merge processes the

record according to an option speciﬁed by the

owncode routine. Owncode routines can also supply

the records to be sorted; when Sort/Merge is ready

for a record, the owncode routine passes a record

to Sort/Merge.

point in processing at which your owncode routine

is executed. Figure 7-1 shows the format for

naming an owncode routine in a control statement

and in a procedure call, and the subrou tine state

ment format, which is a standard FORTRAN subroutine

statement.

OWNn=proc CALL SM50WNn(proc)

SUBROUTINE proc(retum_code,reca,rla[,recb,rlb])

Owncode routines are routines written in FORTRAN

(subroutine subprograms) or some other language

that have been compiled and saved in relocatable

binary form. The ﬁle containing your owncode

routines is automatically loaded during Sort/Merge

processing. This section describes specifying

owncode routines and the available processing

options.

SPECIFYING OWNCODE ROUTINES

To specify owncode routines when you are using

control statement calls to Sort/Merge, you use the

OWNF parameter. The OWNF parameter gives Sort/

Merge the name of the ﬁle containing your owncode

routines; the ﬁle is loaded by Sort/Merge by means

of the user call loader so that the routines are

av ailable during Sort/Merge processing. The names

of your owncode routines are speciﬁed with the

OWNn parameter; n is an integer from 1 through 5

that tells Sort/Merge at which point in processing

the routine is executed. You must supply a sepa

rate subroutine for each OWNn speciﬁed.

When you are using procedure calls to Sort/Merge,

you call SM50WNn to specify the name of an owncode

routine and to indicate at which point in process

ing the routine is executed. Owncode routines must

be loaded along with the Sort/Merge procedures.

You cannot use AAM to process ﬁles within an

owncode routine when executing a control card

Sort/Merge. In a FORTRAN program, you must name

the owncode routines in an EXTERNAL statement. You

cannot call SMSOWNn from a COBOL program because

there is no way (other than CALL or ENTER) to tell

COBOL the name of an external subroutine. You can

write a subroutine in FORTRAN or some other

language to call the owncode routine.

Ownc o d e exits 1 a n d 2 can be taken from a sort

only; exits 3, 4, and 5 can be taken from a sort or

merge.

OWNCODE ROUTINE PARAMETERS

Owncode routines are called with parameters. Up to

ﬁve parameters are passed between Sort/Merge and

your owncode routine. Parameters are passed at the

Figure 7-1. Owncode Routine Call and Subroutine

Statement Formats

If you specify an owncode 1, owncode 2, owncode 3,

or ow n co d e 4 r o ut i ne , t he ret ur n _c o de , r e ca , a n d

rla parameters are passed by Sort/Merge. If you

specify an owncode 5 routine, the recb and rib

parameters are passed in addition to the other

three parameters. The parameters are described in

the following paragraphs.

The return_code parameter is passed by Sort/Merge

to an owncode routine as an integer with value 0.

The return_code parameter can be altered by the

owncode routine to the integer value 1, 2, or 3.

The value returned to Sort/Merge by this parameter

indicates a speciﬁc action to be taken by Sort/

Merge. A return_code value that is not deﬁned

ca u se s a cat ast r op h ic er r or. T h e m ea n in g s o f t he

various return_codes are discussed later in this

section.

The reca parameter is an otherwise empty array or

string variable that is used to pass the current

record; except f o r the c u r r e n t record, the conte n t s

of reca are undeﬁned. The rla parameter passes a

value that is the number of characters in the

current record being sorted or merged.

The recb and rib parameters are used when proc

essing two records with equal keys. The recb

parameter passes the second record; the rib

p a r a m e t e r p a s s e s a v a l u e t h a t i s t h e n u m b e r o f

characters in the second record.

The allowed length of records passed to an owncode

routine depends on how you have speciﬁed record

length. If you have speciﬁed the OWNFL parameter

or have called the SM50FL procedure, the number of

characters in the current record must be the same

a s t h e O W N F L o r S M 5 0 F L v a l u e . O t h e r w i s e , t h e

maximum record length is determined as the largest

MRL or FL FIT ﬁeld in a CYBER Record Manager FILE

control statement, or the largest value from the

OWNMRL parameter or the SM50MRL procedure; all

records must contain between one and this maximum

record length number of characters, inclusive. An

rla or rib parameter value that does not correspond

to a record speciﬁcation causes a catastrophic

error.

60484800 D 7-1

The reca, rla, recb, and rib parameters can be

altered by an owncode routine; the routine can pass

a different record back to Sort /Merge in reca or

recb, and t h e num b e r of characte r s in the r e c o rd

can b e d i ffe r e nt. The parame t e rs m u s t be passe d

from an owncode routine in order: return_code,

reca, rla, recb, and rib. The arrays reca and recb

can be any type of array.

p r o c e s s i n g o f t h e r e c o r d , c o n t r o l r e t u r n s t o

Sort/Merge, which processes the record passed back

in reca according to the return_code value set by

the owncode 1 routine (the record passed back to

Sort/Merge in reca can be different from the record

passed to the routine). The return_code value and

the associated processing performed by Sort/Merge

can be as follows:

Table 7-1 summarizes the owncode routines and the

parameters passed. Figure 7-2 illustrates record

movement from Sort/Merge to an owncode routine and

back to Sort/Merge.

OWNCODE 1: PROCESSING

INPUT RECORDS

The 0WN1 parameter or the SM50WN1 procedure call

names an owncode 1 routine that is executed in one

of two instances.

I f y o u h a v e s p e c i ﬁ e d i n p u t ﬁ l e s w i t h t h e F R O M

parameter or the SM5FR0M procedure call, the

owncode 1 routine is executed after reading each

record. The return_code, reca, and rla parameters

are passed to the routine by Sort/Merge. The

return_code is 0, reca contains the record, and rla

is the record length in characters. After owncode

0 The record passed back to Sort/Merge in

reca is sorted.

1 The record passed back to Sort/Merge in

reca is deleted.

2 A n a d d i t i o n a l r e c o r d i s i n s e r t e d i n t o t h e

sort. The record in reca is entered into

the sort, and the owncode 1 routine is

executed again with reca and rla set to the

record that just entered the sort.

3 Input from the current ﬁle is terminated.

The record in reca and remaining records in

the ﬁle are not sorted. If more input

ﬁles are speciﬁed, records are read from

the next input ﬁle. The owncode 1 routine

is executed after reading e a ch rec o r d from

this next ﬁle.

TABLE 7-1. OWNCODE ROUTINE SUMMARY

Routine

Type Processing

Parameters Passed

Return__code Value

Returnjcode reca rla recb rib

Owncode 1 Input records 0 Sort current record.

1 Delete current record.

2 Insert new record.

3 Terminate input from

current ﬁle.

Owncode 2 Input ﬁles 0 Begin processing next

i n p u t ﬁ l e .

1 Insert new record.

Owncode 3 Output records 0 M o d i f y c u r r e n t r e c o r d .

1 D e l e t e c u r r e n t r e c o r d .

2 Create new record.

3 Terminate output.

Owncode 4 Output ﬁles 0 Sort or merge is complete.

1 Insert new record.

Owncode 5 Equal keys 0 Retain both records.

1 Replace both records with

new record.

7-2 60484800 A

Record

y^s

Sort/Merge Owncode

Routine

"s^

Record "*"~~

Figure 7-2. Owncode Routine Record Flow

If you have not speciﬁed any input ﬁles, the

owncode 1 routine is executed when Sort/Merge

is ready for another record to process. The

return_code, reca, and rla parameters are passed to

the routine by Sort/Merge. The return_code is 0,

reca is an empty array with room for the largest

record, and rla is 0. When control is returned

t o S o r t / M e r g e f r o m t h e o w n c o d e 1 r o u t i n e , t h e

return_code value and the associated processing

performed by Sort/Merge can be as follows:

0 The record passed back to Sort/Merge in

reca is sorted.

2 A n a d d i t i o n a l r e c o r d i s i n s e r t e d i n t o t h e

sort. The record in reca is entered into

t h e s o r t , a n d t h e o w n c o d e 1 r o u t i n e i s

executed again with reca and rla set to the

record that just entered the sort.

3 Input is terminated.

OWNCODE 2: PROCESSING

INPUT FILES

The 0WN2 parameter or the SM50WN2 procedure call

names an owncode 2 routine that is executed in one

of two instances.

If you have speciﬁed input ﬁles with the FROM

p a r a m e t e r o r t h e S M 5 F R 0 M p r o c e d u r e c a l l , t h e

owncode 2 routine is executed after input from a

ﬁle has terminated (input terminates when end-

of-section is found on ﬁle INPUT, when end-of-ﬁle

is found on any other ﬁle, or when an owncode 1

ro ut i n e p a ss es a r e t ur n_ c o d e v a l ue o f 3 to So rt /

Merge). The return_code, reca, and rla parameters

a r e p a s s e d t o t h e r o u t i n e b y S o r t / M e r g e . T h e

return__code is 0, and reca and rla are passed as a

null record; reca is an empty array, and record

length is 0. When control is returned to Sort/

Merge from the owncode 2 routine, the return_code

value and the associated processing performed by

Sort/Merge can be as follows:

0 Processing of the next input ﬁle, if any,

begins.

1 A n a d d i t i o n a l r e c o r d i s i n s e r t e d i n t o t h e

sort after the last record. The record

i n s e r t e d i s t h e ﬁ r s t r l a c h a r a c t e r s i n

reca, which have been provided by the

routine. The owncode 2 routine is executed

again.

If you have not speciﬁed any input ﬁles, the

owncode 2 routine is executed after an owncode 1

routine has terminated input. The return_code,

reca, and rla parameters are passed to the routine

by Sort/Merge. The return_code is 0, and reca and

rla are passed as a null record. When control is

returned to Sort/Merge from the owncode 2 routine,

the return_code value and the associated processing

performed by Sort/Merge can be as follows:

0 Signals the end of input.

1 A n a d d i t i o n a l r e c o r d i s i n s e r t e d i n t o t h e

sort after the last record. The record

i n s e r t e d i s t h e ﬁ r s t r l a c h a r a c t e r s i n

reca, which have been provided by the

routine. The owncode 2 routine is executed

again.

OWNCODE 3: PROCESSING

OUTPUT RECORDS

The 0WN3 parameter or the SM50WN3 procedure call

names an owncode 3 routine that is executed in one

of two instances.

I f y o u h a v e s p e c i ﬁ e d a n o u t p u t ﬁ l e w i t h t h e T O

parameter or the SM5T0 procedure call, the own-

code 3 routine is executed before writing a record

to the output ﬁle. The return_code, reca, and rla

parameters are passed to the routine by Sort/

Merge. The return_code is 0, reca is the record,

and rla is the record length in characters. After

owncode processing of the record, control returns

to Sort/Merge, which processes the record passed

back in reca according to the return_code value set

by the owncode 3 routine. The return_code value

and the associated processing performed by Sort/

Merge can be as follows:

0 The record passed back to Sort/Merge in

reca is written to the output ﬁle.

1 The record passed back to Sort/Merge in

reca is deleted.

2 An additional record is written to the

output ﬁle. The record in reca is written

out, and the owncode 3 routine is executed

again with reca and rla set to the original

record.

3 O u t p u t f r o m t h e c u r r e n t ﬁ l e i s t e r m i

nated. The record in reca is not written

ou t . I f a n o wnc ode 4 r ou t in e i s s p ec i ﬁe d ,

i t i s e x e c u t e d ; o t h e r w i s e , t h e s o r t o r

merge is terminated.

60484800 C 7-3

If you have not speciﬁed an output ﬁle, the

owncode 3 routine is executed when a record is

ready for output; sorted or merged records are

processed by the routine. The return_code, reca,

and rla parameters are passed to the routine by

S o r t / M e r g e . T h e r e t u r n _ c o d e i s 0 , r e c a i s t h e

record, and rla is the record length in char

acters. When control is returned to Sort/Merge

from the owncode 3 routine, the return_code value

and th e associated p r o c e s s i n g performed by Sort/

Merge can be as follows:

1 T h e r e c o r d p a s s e d b a c k t o S o r t / M e r g e i n

reca is deleted.

3 Output is terminated. If an owncode 4

routine is speciﬁed, it is executed;

otherwise, the sort or merge is terminated.

OWNCODE 4: PROCESSING

OUTPUT FILES

The OWN4 parameter or the SM50WN4 procedure call

names an owncode 4 routine that is executed in one

of two instances.

If you have speciﬁed an output ﬁle with the TO

parameter or the SMSTO procedure call, the owncode

4 r o u t i n e i s e x e c u t e d a f t e r t h e l a s t r e c o r d h a s

been written to the output ﬁle. The return_code,

reca, and rla parameters are passed to the routine

by Sort/Merge. The return_code is 0, and reca and

rla are passed as a null record. When control is

returned to Sort/Merge from the owncode 4 routine,

the return_code value and the associated processing

performed by Sort/Merge can be as follows:

0 The sort or merge is completed.

1 This value is not allowed since no output

ﬁle has been speciﬁed.

OWNCODE 5: PROCESSING

RECORDS WITH EQUAL KEYS

The 0WN5 parameter or the SM50WN5 procedure call

names an owncode 5 routine that is executed when

two records with equal key values are encountered

during a sort or merge (when you have deﬁned your

own collating sequence, key values can be equal

even if key characters are not identical). The key

ﬁ e l d s m u s t n o t b e m o d i ﬁ e d o r e l s e t h e s o r t i n g

order is undeﬁned. The return__code, reca, rla,

recb, and rib parameters are passed to the routine

by Sort/Merge. The return_code is 0; reca and recb

contain the ﬁrst and second records, respectively,

and rla and rib contain the record lengths in

characters of the ﬁrst and second records,

respectively. After owncode 5 routine processing

of the two records, control is returned to

Sort/Merge, which processes the records according

to the return_code value set by the owncode 5

routine. The return_code value and the associated

processing performed by Sort/ Merge can be as

follows:

0 The ﬁrst rla characters of reca are

a c c e p t e d a s t h e ﬁ r s t r e c o r d ; t h e ﬁ r s t r i b

characters of recb are accepted as the

second record (the records and record

lengths passed back to Sort/Merge can be

different from the records and record

lengths passed to the owncode routine).

X^ifﬁjJV

An additional record is inserted after the

l a s t r e c o r d . T h e r e c o r d i n s e r t e d i s t h e

ﬁrst rla characters in reca. The owncode

4 routine is executed again.

One duplicate record is deleted. The other

record is replaced with the ﬁrst rla char

acters of reca.

If you have not speciﬁed an output ﬁle, an

owncode 4 routine is executed after the last call

t o a n o w nc o d e 3 r o ut i n e i n d i c at i n g t h a t t h e s o r t

has been completed. If output is terminated by an

owncode 3 routine, an owncode 4 routine is called

immediately. The return_code, reca, and rla

parameters are passed to the routine by Sort/Merge.

The return_code is 0, and reca and rla are passed

as a null record. When control is returned to

Sort/Merge from the owncode 4 routine, the

return_code value and the associated processing

performed by Sort/Merge can be as follows:

If you specify RETAIN=YES or call the SM5RETA

procedure in a sort with an owncode 5 routine, the

record that entered the sort ﬁrst is passed to the

owncode routine as reca; otherwise, either of the

two records with equal keys could be passed to the

routine as reca. The owncode 5 routine can control

the order in which the two records are returned to

Sort/Merge. The record returned to Sort/Merge as

reca is entered into the sort before the record

returned as recb. An owncode 5 routine cannot be

speciﬁed in the same sort as a SUM parameter or

SM5SUM procedure.

7-4 60484800 D

EXAMPLES

T h i s s e c t i o n c o n t a i n s e x a m p l e s o f s o r t s a n d

merges. The ﬁrst example is a simple control

statement sort on one key. The same sort is illus

t r a t e d u s i n g t h e i n t e r a c t i v e d i a l o g . A c o n t r o l

st a t e m e n t so rt o n m u l t i p l e ke y s , a c o n t r o l s ta te

me n t m er ge , d i r e ct iv e ﬁ l e c re a t i o n an d u s e , S U M

parameter use, and a FORTRAN program with Sort/

Merge procedure calls and an owncode routine are

also included.

CONTROL STATEMENT SORT ON

ONE KEY

Figure 8-1 shows the record layout in a university

student ﬁle named UNIVER. Each record includes

the last name and ﬁrst and middle initials, the

s t u d e n t n u m b e r, t h e d a t e o f b i r t h , t h e ﬁ e l d o f

study, the grade point average, and a code repre

senting class (freshman, sophomore, junior,

s e n i o r ) ; a l l ﬁ e l d s a r e w r i t t e n w i t h c h a r a c t e r

data. The ﬁle is maintained with the student

number as the major key. Records are normally

ordered in ascending order according to the student

number as shown in ﬁgure 8-2 with ﬁle UNIVER.

Figure 8-3 shows the job structures for sorting

ﬁle UNIVER to give an alphabetic list of

students. Under both NOS and NOS/BE operating

systems, the job statements and accounting infor

mation are followed by ATTACH commands to assign

the ﬁle to the job. CYBER Record Manager (CRM)

FILE control statements are included, because the

input and output records contain only 38 characters

each, considerably less than the default of 150.

The S0RT5 control statement calls for records from

UNIVER to be sorted on a key that occupies charac

ter positions 1 through 10 in each record according

to the default ASCII6 collating sequence in ascend

i n g o r d e r . S o r t e d r e c o r d s a r e w r i t t e n t o ﬁ l e

SRTOUT, which is created as a local ﬁle during the

sort. File SRTOUT is then copied to ﬁle OUTPUT

with a right shift, because the ﬁrst character in

the print line is reserved for carriage control.

Figure 8-4 shows the output from the job.

WALLACE 87366110255ENGIR 2861

JOHNSON 90248063051HATH 2253

SANDERS 99855022858BUS 3011

NEECE 99911121358ART 2291

TERRELL 99998040356ENG 3861

OKADA 100103111750UNDEC 2225

REYES 100246031558ANTHR0 3341

SUGARMAN 100528070457S0C 3501

PHILLIPS 100531121158EDU 2112

KRUTZ

•100532010353P0LISCI 1981

WARNES 102116060861P0LISCI 2814

CARLSON 102126022355ENGIR 3454

FUHRMAN 102212111859CHEM 3204

MCMAHON 102223061260ENG 2784

JUNG 102301052561PHYSED 2214

P0P0VICH 102311100961BUS 2434

JONES 102318081555EDU 2844

PEAKE 102326111960J0URN 3024

BRISCOE 102343121157ENVIR0 2544

HORNE 102377070861COMPSCI 3894

Figure 8-2. File UNIVER

SORT SPECIFICATION USING

INTERACTIVE DIALOG

Figure 8-5 shows the interactive dialog used to

specify the same sort of ﬁle UNIVER on the last

name. The input ﬁle is assigned to the job with

an ATTACH command under NOS or a FETCH command

under NOS/BE. CRM FILE control statements are

en t er e d f or the in put an d o u tp u t ﬁ le s . Th e d i al o g

is invoked by typing DIAL0G=»YES. Sorted records

are written to ﬁle SORTOUT, which Sort/Merge

creates as a local ﬁle. File SORTOUT can be

printed or displayed at the terminal; SORTOUT can

also be routed to a printer or saved as a permanent

ﬁle.

1 1 1 1 3 1 5 2 1 2 7 3 5 3 8

LAST NAME STUDENT

NO. DOB STUDY GPA

FIRST INITIAL MIDDLE INITIAL CQDE

Figure 8-1. File UNIVER Record Layout

60484800 B 8-1

NOS Operating System

Job statement

USER statement

CHARGE statement

ATTACH,UNIVER.

FILE(UNIVER,BT=C,RT=Z,FL=38)

FILE(SRTOUT,BT=C,RT=Z,FL=38)

SORT5.FROM=UNIVER,TO=SRTOUT,KEY=1. .10

COPYSBF,SRTOUT,OUTPUT.

EOI

NOS/BE Operating System

Job statement

ACCOUNT statement

ATTACH,UNIVER,ID=PUBS.

FILE(UNIVER,BT=C,RT=Z,FL=38)

FILE(SRTOUT,BT=C,RT=Z,FL=38)

SORT5.FROM=UNIVER,TO=SRTOUT,KEY=1. .10

COPYSBF,SRTOUT,OUTPUT.

EOI

Figure 8-3. Job Structure for Sorting File

UNIVER on Last Name

BARTLETT 100800100957ART 2735

BILLINGS 101579111855MUS 2965

BRISCOE 102343121157ENVIR0 2544

CARLSON 102126022355ENGIR 3454

CHARLES 101418032459ANTHRO 2453

CLARK 101400102954EC0N 3782

CLARK 101023101956ENG 2083

COCHRAN 100725111857BI0 3011

DAVIES 100812080656JOURN 2031

DAVIS

•100972071650ENG 3541

•

WALLIN 101056041659P0LISCI 3151

WARNES 102116060861P0LISCI 2814

WILSON 101967010261MATH 3454

WONG 101001012755PSYCH 2152

WOO 101315100159BUS 3223

WOODSTOCK 101497030160CHEM 3483

YEH 102005120645ART 2764

YOST 100880111158ENG 2582

ZEITZ 100963111858MATH 2612

ZIMMERS 101075063059MATH 2992

/"^S

Figure 8-4. File UNIVER Sorted on Last Name

/attach,univer

tiile(univer,bt=c,rt=z,f1=38)

FILE<UNIVER,BT=C,RT=Z,FL=38)

/ﬁle(sortout,bt=c,rt=z,f1=38)

FILE(SORTOUT,BT=C,RT=Z,FL=38)

/sort5.dialog=yes

IS YOUR TERMINAL ALREADY SET TO NORMAL MODE OR UPPERCASE

ASCII MODE. ALSO, ARE YOU SURE THAT ALL YOUR FILES (IF ANY)

ARE EITHER STANDARD UNIT RECORD FORMAT OR HAVE HAD FILE

CONTROL STATEMENTS SPECIFIED FOR THEM

? yes

DO YOU HAVE JUST ONE INPUT FILE AND NO 0WN1 OR 0WN2 SUBROUTINES

? yes

WHAT IS THE NAME OF YOUR INPUT FILE

? univer

DO YOU HAVE ANY 0UN3 OR 0WN4 SUBROUTINES

? no

WHAT IS THE NAME OF YOUR OUTPUT FILE

? sortout

THE STANDARD COLLATING SEQUENCE IS ASCII6.

HERE IS THE WHOLE ASCII6 COLLATING SEQUENCE LISTED

!,,#$%&,()*+,-./0123456789:;<=>?aABCDEFGHIJKLMNOPQRSTUVWXYZC\3',

DO ALL OF YOUR KEYS CONSIST OF CHARACTER DATA TO BE SORTED

ACCORDING TO THE ASCII6 COLLATING SEQUENCE

? yes

ASCENDING ORDER OF THE ASCII6 COLLATING SEQUENCE MEANS THE

NORMAL ORDER.

!«•#$%&•()*+,-./0123456789:;<=>?aABCDEFGHIJKLMNOPQRSTUVWXYZC\D*

DESCENDING ORDER MEANS THE REVERSE ORDER.

_*3\CZYXWVUTSRQP0NMLKJIHGFEDCBA3?>=<;:9876543210/.-,+*)C8%$#"!

DO YOU WANT ALL YOUR KEYS SORTED IN ASCENDING ORDER

? yes

HOW MANY KEY FIELDS DO YOU HAVE

? 1

THE LEFT-MOST CHARACTER OF A RECORD IS NUMBER 1.

FOR THE MOST IMPORTANT KEY FIELD, WHAT IS THE NUMBER

OF THE LEFT MOST CHARACTER

? 1

HOW MANY CHARACTERS ARE IN THE MOST IMPORTANT KEY FIELD

? 10

DO YOU WANT RECORDS WITH EQUIVALENT KEY VALUES TO BE WRITTEN

IN THE SAME ORDER THEY ARE READ

? no

GIVE AN ESTIMATE OF THE MAXIMUM NUMBER OF RECORDS TO BE SORTED

? 75

THANK YOU. SORT/MERGE NOW BEGINS.

S0RT5.DIAL0G=YES

/*s%

s*^S

8-2 Figure 8-5. Interactive Dialog Example 60484800 D

CONTROL STATEMENT SORT ON

MULTIPLE KEYS

Figure 8-6 shows the job structures for sorting

ﬁle UNIVER on three keys. Under both NOS and

NOS/BE, the job statements and accounting informa

tion are followed by ATTACH commands to assign the

ﬁle to the job. CRM FILE control statements for

the input and output ﬁles are included. The S0RT5

control statement calls for records to be ﬁrst

sorted on the ﬁeld of study, which occupies char

acter positions 27 through 34 in each record.

Records with equal keys for the major key are then

sorted on the class code, which is a 1-character

ﬁeld in position 38. The third key sorts students

with the same ﬁeld of study and class by the last

name.

Aj^^S

NOS Operating System

Job statement

USER statement

CHARGE statement

ATTACH,UNIVER.

Fl LE(UNIVER,BT=C,RT=Z,FL=38)

FILE(SRTOUT,BT=C,RT=Z,FL=38)

SORT5.FROM=UNIVER,TO=SRTOUT,

.KEY=((27. .34),(38,1),(1. .10))

COPYSBF,SRTOUT,OUTPUT.

EOI

NOS/BE Operating System

Job statement

ACCOUNT statement

ATTACH,UNIVER,ID=PUBS.

Fl LE(UNI VER,BT=C,RT=Z,FL=38)

FILE(SRTOUT,BT=C,RT=Z,FL=38)

SORT5.FROM=UNIVER,TO=SRTOUT,

.KEY=((27. .34),(38,1),(1. .10))

COPYSBF,SRTOUT,OUTPUT.

EOI

0jm*B\,

Figure 8-6. Job Structure for Control Statement

Sort on Multiple Keys

Figure 8-6 i l l ustrates c o n t i nuing the S0RT 5 control

statement beyond one line. The ﬁrst line of the

control statement is terminated with three periods;

the second line begins with one period. Sorted

r e c o r d s a r e w r i t t e n t o t h e ﬁ l e S R TO U T, w h i c h i s

then copied to ﬁle OUTPUT with a right shift,

because the ﬁrst character in the print line is

reserved for carriage control. Figure 8-7 shows

the output from the job.

CONTROL STATEMENT MERGE

Figure 8-8 shows two ﬁles (UNIVER, ADDSTU) ordered

according to the student number. The two ﬁles can

be merged because they are sorted on the same key.

Fig u r e 8 - 9 s h ows the j ob s truc t ures to m erge the

two ﬁles. Under both NOS and NOS/BE, the two

ﬁ l e s a r e at t ac h e d . A n e w ﬁ l e (N E W U ) i s c r e a t e d

as a permanent ﬁle to which merged records are

written; under NOS, the DEFINE command creates an

empty direct access permanent ﬁle. Under NOS/BE,

the REQUEST command assigns NEWU to a permanent

ﬁle device; the CATALOG command then makes the

ﬁle permanent. CRM FILE control statements for

the input ﬁles and the merged ﬁle are included.

REYES 100246031558ANTHRO 3341

MAYER 100991122359ANTHR0 2882

CHARLES 101418032459ANTHRO 2453

MARTIN 100955082157ART 2891

NEECE 99911121358ART 2291

NAKAMURA 101529051260ART 2594

YEH 102005120645ART 2764

BARTLETT 100800100957ART 2735

COCHRAN 100725111857BI0 3011

H0Y0

•101925103060BI0 3014

•

KRUTZ 100532010353P0LISCI 1981

WALLIN 101056041659P0LISCI 3151

WARNES 102116060861P0LISCI 2814

WONG 101001012755PSYCH 2152

LANGDON 101754080549PSYCH 2013

LASEUR 100678042256PSYCH 2233

SUGARMAN 100528070457S0C 3501

SMITH 101062120758SOC 2913

DOUGLAS 101325071558UNDEC 2585

OKADA 100103111750UNDEC 2225

Figure 8-7. File UNIVER Sorted on Multiple Keys

A. File UNIVER

WALLACE 87366110255ENGIR 2861

JOHNSON 90248063051MATH 2253

SANDERS 99855022858BUS 3011

NEECE 99911121358ART 2291

TERRELL 99998040356ENG 3861

OKADA 100103111750UNDEC 2225

REYES 100246031558ANTHR0 3341

SUGARMAN 100528070457SOC 3501

PHILLIPS 100531121158EDU 2112

KRUTZ

•

100532010353P0LISCI 1981

•

WARNES 102116060861P0LISCI 2814

CARLSON 102126022355ENGIR 3454

FUHRMAN 102212111859CHEM 3204

MCMAHON 102223061260ENG 2784

JUNG 102301052561PHYSED 2214

P0P0VICH 102311100961BUS 2434

JONES 102318081555EDU 2844

PEAKE 102326111960J0URN 3024

BRISCOE 102343121157ENVIR0 2544

H0RNE 102377070861COMPSCI 3894

B. File ADDSTU

QUINTERA 90154101253BI0 3451

KING 100012090848BUS 2431

ANDRUS 100478042855J0URN 2121

UNGERMAN 100933120356PHYSED 3012

KLEIN 100987051260ENGIR 2762

GIBSON 101728070160MATH 2012

IRVING 101750111855ENG 3943

ALLEN 102056012561LNGUIS 3854

GREENWOOD 102168101961EDU 2264

ANDERSEN 102308032160P0LISCI 2544

EBERHARD 102320061158BUS 3014

GOMEZ 102379Q22260COMPSCI 2984

Figure 8-8. Input Files for Control Statement Merge

60484800 B 8-3

NOS Operating System

Job statement

USER statement

CHARGE statement

ATTACH,UNIVER,ADDSTU.

DEFINE,NEWU.

Fl LE(UNI VER,BT=C,RT=Z,FL=38)

FILE(ADDSTU,BT=C,RT=Z,FL=38)

FILE(NEWU,BT=C,RT=Z,FL=38)

MERGE.FROM=(UNIVER,ADDSTU),TO=NEWU,KEY=15. .20

COPYSBF,NEWU,OUTPUT.

EOI

NOS/BE Operating System

Job statement

ACCOUNT statement

ATTACH,UNIVER,ID=PUBS.

ATTACH,ADDSTU,ID=PUBS.

REQUEST,NEWU,*PF.

FILE(UNIVER,BT=C,RT=Z,FL=38)

Fl LE(ADDSTU,BT=C,RT=Z,FL=38)

FILE(NEWU,BT=C,RT=Z,FL=38)

MERGE.FROM=(UNIVER,ADDSTU),TO=NEWU,KEY=15. .20

CATALOG,NEWU,ID=PUBS,XR=CHNG.

COPYSBF,NEWU,OUTPUT.

EOI

Figure 8-9. Job Structure for Control Statement Merge

The MERGE control statement names the two input

ﬁles and the new output ﬁle. The key on which

the ﬁles are merged is the student number (the

ﬁeld on which the ﬁles are presorted). The

merged ﬁle is copied to ﬁle OUTPUT with a right

shift, because the ﬁrst character In the print

line is reserved for carriage control. Figure 8-10

shows the output from the job.

WALLACE 87366110255ENGIR 2861

QUINTERA 90154101253BI0 3451

JOHNSON 90248063051MATH 2253

SANDERS 99855022858BUS 3 0 11

NEECE 99911121358ART 2291

TERRELL 99998040356ENG 3861

KING 100012090848BUS 2431

OKADA 100103111750UNDEC 2225

REYES 100246031558ANTHR0 3341

ANDRUS

•100478042855J0URN 2121

•

MCMAHON 102223061260ENG 2784

JUNG 102301052561PHYSED 2214

ANDERSEN 102308032160P0LISCI 2544

P0P0VICH 102311100961BUS 2434

JONES 102318081555EDU 2844

EBERHARD 1023200611589US 3014

PEAKE 102326111960J0URN 3024

BRISCOE 102343121157ENVIR0 2544

HORNE 102377070861COMPSCI 3894

GOMEZ 102379022260C0MPSCI 2984

Figure 8-10. Merged File NEWU

CREATING AND USING A

DIRECTIVE FILE

Figure 8-11 shows a portion of a terminal session

in which a directive ﬁle is created. In the ﬁrst

p a r t o f ﬁ g u r e 8 - 11 , t h e d i r e c t i v e ﬁ l e i s c r e a t e d

using XEDIT under NOS; (cr) denotes carriage

return. In the second part of the ﬁgure, the

directive ﬁle is created using EDITOR under NOS/BE.

Under NOS, XEDIT is entered after login, and the

directive ﬁle to be created is given the name

DIRFIL. The directive ﬁle statements begin with

the word SORT followed by a comma. The statements

specify a user-deﬁned collating sequence named

MYSEQ. The ﬁrst SEQS parameter speciﬁes ten

steps from 0 to 9; the other SEQS parameter speci

ﬁes one step consisting of hyphens, blanks, and

slashes. The SEQA parameter speciﬁes that blanks

and slashes are to be output as hyphens. The

directive ﬁle is then reviewed (still using XEDIT)

to make s ure the ﬁle is acc u r a te. The ﬁ l e DIRFIL

is saved as an indirect access permanent ﬁle.

Under NOS/BE, EDITOR is entered after login. The

statements of the directive ﬁle are identical to

the statements entered under NOS, and the directive

ﬁ l e i s t h e n r e v i e w e d u s i n g t h e L I S T c o m m a n d .

Using the SAVE command, DIRFIL Is made a local ﬁle

without sequence numbers; the CATALOG command saves

the ﬁle as a permanent ﬁle.

Figure 8-12 shows a portion of a NOS terminal ses

sion in which a ﬁle named NUMFIL (ﬁgure 8-13) is

sorted according to the MYSEQ collating sequence.

The directive ﬁle and the input ﬁle are assigned

to the job with GET commands for indirect access

permanent ﬁles. CRM FILE control statements are

entered for the input and output ﬁles.

The S0RT5 control statement is then entered naming

the input and output ﬁles, the key ﬁeld, and the

d i r e c t i v e ﬁ l e ; t h e n a m e o f t h e u s e r - d e ﬁ n e d

collating sequence is the key type. After Sort/

Merge reads the directive ﬁle, an image of this

ﬁle is printed or displayed at the terminal (the L

parameter default is OUTPUT). When the sort is

complete, the S0RT5 control statement is printed or

displayed at the terminal (note that NOS repeats

only 48 characters of the control statement).

Control is then returned to the operating system.

Figure 8-14 shows the output from the job using the

directive ﬁle with the MYSEQ collating sequence.

The records are in numeric order, and the separa

tors in the number have been standardized as

hyphens in all records

SUM PARAMETER USE

Figure 8-15 shows the record layout in a ﬁle named

GRADES. Each record includes three numeric ﬁelds:

number of units attempted, number of units com

p l e t e d , a n d g r a d e p o i n t s . F i g u r e 8 - 1 6 s h o w s ﬁ l e

GRADES with multiple records for each student.

Records are to be sorted according to student

number and, using the SUM parameter, records with

the same student number are to be combined into one

record by adding together the numeric ﬁelds to

give the total number of units attempted, units

completed, and the total number of grade points. ^%

8-4 60484800 B

0^s

NOS Operating System

/xedi t,di rf i L,c -*-

XEDIT 3.1.00

INPUT -*-

sort,seqn=myseq

sort,seqs=('0'..

sort,seqs=('-','

sort,seqa=yes

9')

/')!

? —

EDIT -.

?? top -m

?? print*-*

SORT,SEQN=MYSEQ

S0RT,SEQS=(,0,..'9')

S0RT,SEQS=(,-,,, ','/

S0RT,SEQA=YES

END OF FILE

?? end,dirﬁL,replace

DIRFIL REPLACED

NOS/BE Operating System

COMMAND- editor —

..1=sort,seqn=myseq

2=sort,seqs=('0'..'9')

3=sort,seqs=(,-,,, ■,'/')

4=sort,seqa=yes

l i s t , a , s u p «

SORT,SEQN=MYSEQ

S0RT,SEQS=(,0'..,9') I

S0RT,SEQS=(,-,,, ,,*/,)l

SORT,SEQA=YES

.. save,di rf i I ,n -«

..end -«

COMMAND- catalog,dirﬁl,id=pubs

Enter XEDIT in create mode; DIRFIL is the name

given the ﬁle to be created.

Enter XEDIT input mode.

Directive ﬁle statements.

Enter XEDIT edit mode.

Move pointer to the beginning of the ﬁle.

Print the entire ﬁle.

Directive ﬁle statements.

Leave XEDIT and save DIRFIL as an indirect access

permanent ﬁle.

Enter EDITOR.

Directive ﬁle statements.

List the entire ﬁle.

Directive ﬁle statements.

Make edit ﬁle a local ﬁle without sequence numbers.

Leave EDITOR.

Save ﬁle as a permanent ﬁle named DIRFIL.

Figure 8-11. Creating a Directive File

/ g e t , d i r ﬁ I

/get,numﬁI

/ﬁle(numﬁl,bt=c,rt=z,f1=35)

FILE(NUMFIL,BT=C,RT=Z,FL=35)

/ﬁle(seqﬁ l,bt = c,rt = z,f1=35)

FILE(SEQFIL,BT=C,RT=Z,FL=35)

/sort5.from=numﬁl,to=seqﬁl,key=((25,11,myseq)),dir=dirﬁl

SORT/MERGE 5.0 * LISTING OF DIRECTIVE FILE - DIRFIL

12/04. 16.45.45.

1. SORT,SEQN=MYSEQ

2. S0RT,SEQS=('0,..,9')

3. S0RT,SEQS=(,-,,, •,'/■)

4. SORT,SEQA=YES

SORT/MERGE 5 - 18 RECORDS SORTED

Figure 8-12. Using a Directive File (NOS)

60484800 B 8-5

DOE CHERYL 706-87-5430

THOMPSON LOUISE 819 95 6445

GORDON ROBERT 176/81/0013

WATSON JOSEPH 091-15-1938

ANDERSON PETER 778 21 2847

HALL MARY 029-55-1789

SMITH JOHN 555-70-6351

NORTH PATRICIA 234-58-3190

ROLLINS SARAH 770-25-8147

SHARP DANIEL 447/46/0043

LYNCH DENNIS 549-22-6741

SMITH CATHERINE 387 14 9998

NOE JANE 880-25-3058

BROWN WILLIAM 120 59 4049

HARRIS JEAN 229-80-7776

JONES CHRISTOPHER 667 /26/2514

CARTER BARBARA 114 48 5112

HENDERS GERALD 998-72-6203

Figure 8-13. File NUMFIL

HALL MARY 029-55-1789

WATSON JOSEPH 091-15-1938

CARTER BARBARA 114-48-5112

BROWN WILLIAM 120-59-4049

GORDON ROBERT 176-81-0013

HARRIS JEAN 229-80-7776

NORTH PATRICIA 234-58-3190

SMITH CATHERINE 387-14-9998

SHARP DANIEL 447-46-0043

LYNCH DENNIS 549-22-6741

SMITH JOHN 555-70-6351

JONES CHRISTOPHER 667-26-2514

DOE CHERYL 706-87-5430

ROLLINS SARAH 770-25-8147

ANDERSON PETER 778-21-2847

THOMPSON LOUISE 819-95-6445

NOE JANE 880-25-3058

HENDERS GERALD 998-72-6203

GREENWOOD 102168101961EDU 002002000

IRVING 101750111855ENG 004004016

IRVING 101750111855ENG 098095375

GREENWOOD 102168101961EDU 003003009

ANDRUS 100478042855J0URN 005005010

ANDRUS 100478042855JOURN 004000000

ALLEN 102056012561LNGUIS 003003012

KLEIN 100987051260ENGIR 003003006

QUINTERA 90154101253BI0 003000000

IRVING 101750111855ENG 003003012

ANDRUS 100478042855J0URN 190173371

ANDRUS 100478042855J0URN 003003009

KING 100012090848BUS 005005015

ALLEN 102056012561LNGUIS 005000000

ALLEN 102056012561LNGUIS 025020077

ALLEN 102056012561LNGUIS 004004012

ALLEN 102056012561LNGUIS 005005020

KING 100012090848BUS 004004016

EBERHARD 102320061158BUS 005005015

EBERHARD 102320061158BUS 001001000

EBERHARD 102320061158BUS 004004016

KLEIN 100987051260ENGIR 005005015

EBERHARD 102320061158BUS 003003006

GIBSON 101728070160MATH 003000000

KLEIN 100987051260ENGIR 003003006

KLEIN 100987051260ENGIR 135135373

ANDERSEN 102308032160P0LISCI 005005015

ANDERSEN 102308032160P0LISCI 003003009

QUINTERA 90154101253810 180176607

UNGERMAN 100933120356PHYSED 150148445

ANDERSEN 102308032160P0LISCI 005005010

UNGERMAN 100933120356PHYSED 003003012

UNGERMAN 100933120356PHYSED 003003009

KING 100012090848BUS 002002006

KING 100012090848BUS 178174423

ANDRUS 100478042855J0URN 001001000

GREENWOOD 102168101961EDU 003003006

GREENWOOD 102168101961EDU 030030068

IRVING 101750111855ENG 003003009

GIBSON 101728070160MATH 145143290

GIBSON 101728070160MATH 005005010

QUINTERA 90154101253BI0 005005010

GOMEZ 102379022260C0MPSCI 003003009

GOMEZ 102379022260C0MPSCI 005005010

GOMEZ 102379022260C0MPSCI 002002008

GREENWOOD 102168101961EDU 005005020

QUINTERA 90154101253BIO 004004012

Figure 8-14. Output From Sort Using a Directive File Figure 8-16. File GRADES

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

LAST NAME

FIRST INITIAL

STUDENT

NO. DOB STUDY

MIDDLE INITIAL UNITS

ATTEMPTED

UNITS COMPLETED

GRADE POINTS

Figure 8-15. File GRADES Record Layout

8-6 60484800 B

Figure 8-17 shows the S0RT5 control statement with

the SUM parameter. The input and output ﬁles are

named, and the student number (character positions

15 through 20) is speciﬁed as the sort key. The

SUM parameter speciﬁes that a 3-character numeric

ﬁeld of type NUMERIC_NS begins in position 36 in

each record. The repetition indicator 3 speciﬁes

that three contiguous ﬁelds are to be summed.

F i g u r e 8 - 1 8 s h o w s t h e o u t p u t f r o m t h e j o b . T h e

output ﬁle contains one record for each student.

The numeric ﬁelds are the totals of the units

attempted, units completed, and grade points.

A SORT USING PROCEDURE CALLS

Figure 8-19 shows the job structures for performing

a sort from within a FORTRAN program. Under both

NOS and NOS/BE, the job statements and accounting

information are followed by ATTACH commands to

a s s i g n t h e i n p u t d a t a ﬁ l e ( U N I V E R ) a n d t h e ﬁ l e

containing the FORTRAN source program to the job.

The FORTRAN 5 compiler is called, and the program

is compiled. The program is then executed (LGO).

Figure 8-20 shows the FORTRAN program containing

the Sort/Merge procedure calls. File UNIVER is

read, and student records with grade point average

of 3.50 or better are written to an intermediate

ﬁle (INT1). Sort/Merge is called to sort the ﬁle

on grade point average in descending order (highest

grade point average to lowest grade point aver

age). The SM5S0RT call names a result array that

is dimensioned as a 16-element integer array and

the ﬁrst element set to 15 to receive the maximum

number of s o r t results and s t a tistics.

T h e S M 5 0 W N 1 c a l l s p e c i ﬁ e s t h a t a n o w n c o d e 1

routine named CCODE is to be executed after Sort/

Merge reads each record from INT1. Records are

passed to the routine by Sort/Merge. Figure 8-21

shows the routine, which is a FORTRAN subroutine.

The SUBROUTINE statement names the routine and the

parameters passed by Sort/Merge. Parameter RETCODE

is t he r etur n _cod e pa s sed a s 0 , RE C is an a rra y

containing the record, and RL is the record length

in charact e rs. The routine converts the class c ode

in each record to the class name. The records are

r e t u r n e d t o S o r t / M e r g e i n a r r a y R E C . T h e

return_code value is left as 0 because each record

in this example is to be sorted. Sort/Merge then

s o r t s t h e r e c o r d s t o ﬁ l e I N T 2 . T h e s o r t e d ﬁ l e i s

returned to the FORTRAN program to be written out

in a formatted report. Figure 8-22 shows the

output from the job, which includes the report and

the contents of the result array.

'get,grades

/ﬁle(grades,bt=c,rt=z,f1=44)

FILE(GRADES,BT=C,RT=Z,FL=44)

/ﬁle(sumex,bt=c,rt=z,ﬂ=44)

FILE(SUMEX,BT=C,RT=Z,FL=44)

/sortS.f rom=grades,to=sumex,key=15..20,sum=( (36,3,numeri c__ns,3))

SORT/MERGE 5 - 47 RECORDS SORTED

Figure 8-17. Interactive Parameter Speciﬁcation for SUM Example

Total units completed

Total units attempted

Total grade points

QUINTERA 090154101253BI0 19218Ji.

5629

KING 100012090848BUS 189185460

ANDRUS 100478042855J0URN 203182390

UNGERMAN 100933120356PHYSED 156154466

KLEIN 100987051260ENGIR 146146400

GIBSON 101728070160MATH 153148300

IRVING 101750111855ENG 108105412

ALLEN 102056012561LNGUIS 042032121

GREENWOOD 102168101961EDU 043043103

ANDERSEN 102308032160P0LISCI 013013034

EBERHARD 102320061158BUS 013013037

GOMEZ 102379022260C0MPSCI 010010027

Figure 8-18. Output From SUM Example

60484800 B 8-7

NOS Operating System

Job statement

USER statement

CHARGE statement

ATTACH,UNIVER,SM5EX.

FILE(UNIVER,BT=C,RT=Z,FL=38)

FTN5,l=SM5EX,L=OUTEX.

LGO.

COPY,OUTEX,OUTPUT.

EOI

NOS/BE Operating System

Job statement

ACCOUNT statement

ATTACH,UNIVER,ID=PUBS.

ATTACH,SM5EX,ID=F.

FILE(UNIVER,BT=C,RT=Z,FL=38)

FTN5,l=SM5EX,L=OUTEX.

LGO.

COPY,OUTEX,OUTPUT.

EOI

Figure 8-19. Job Structure for Sort Called From a FORTRAN Program

1PROGRAM DLIST

2 INTEGER GPA

3CHARACTER SNAME*14, MAJ0R*8, C0DE*1, CLASS*12

4DIMENSION IARAY(16)

5COMMON /C0MN/ CODE, CLASS

EXTERNAL CC0DE

1READ(L"UNIVER",100,END=10) SNAME, MAJOR, GPA, CODE

9IF (GPA .GE. 350) WRITE(L"INT1",200) SNAME, MAJOR, GPA, CODE

10 GO TO 1

12 10 IARAY(1)=15

13 CALL SM5S0RTUARAY)

14 CALL SM50FL(80)

15 CALL SM5FR0M('INT1')

16 CALL SMSKEY^^'ASCIU'^D')

17 CALL SM50WNKCC0DE)

18 CALL SM5TO(,INT2,)

19 CALL SM5END

20 REWIND(L"INT2")

22 PRINT *, IARAY

23 WRITECL"0UTEX",400)

24 15 READ(L,,INT2",300,END=20) SNAME, MAJOR, GPA, CLASS

25 WRITE(LMOUTEX",500) SNAME, MAJOR, CLASS, GPA

26 GO TO 15

28 100 FORMAT <A14,12X,A8,I3,A1)

29 200 FORMAT (A14,5X,A8,5X,I3,5X,A1,39X)

30 300 FORMAT <A14,5X,A8,5X,I3,5X,A12,28X)

31 400 FORMAT (36X,"DEAN'S LIST" // 15X,"STUDENT",

32 * 12X,"MAJOR",8X,"CLASS",12X,"GPA",65X /)

33 500 FORMAT (15X,A14,5X,A8,5X,A12,5X,I3,59X)

35 20 STOP

36 END

Figure 8-20. FORTRAN Program With Sort/Merge Procedure Calls

8-8 60484800 B

j4$Wm*\

SUBROUTINE CCODE(RETCODE,REC,RL)

3INTEGER RETCODE, RL

4CHARACTER C1*6, C2*6, C3*9, C4*8, C5*12,

5* C0DE*1, CLASS*12, REC(*)*12

6COMMON /SUBCOM/ CODE, CLASS

7D ATA C 1 , C 2 / ' S E N I 0 R ' , ' J U N I O R ' /

8DATA C3,C4/'SOPHOMORE','FRESHMAN'/

9DATA C5/'UNCLASSIFIED'/

11 CODE = REC(4)(5:5)

12 IF (CODE .EQ. '1') THEN

13 CLASS = C1

14 E L S E I F (C O D E . E Q . ' 2 ') T H E N

15 CLASS = C2

16 E L S E I F (C O D E . E Q . ' 3 ') T H E N

17 CLASS = C3

18 E L S E I F (C O D E . E Q . ' 4 ') T H E N

19 CLASS = C4

20 E L S E I F (C O D E . E Q . ' 5 ') T H E N

21 CLASS = C5

22 ELSE

23 PRINT *, CODE

24 END IF

25 REC(4)(5:12) = CLASS(1:8)

26 REC(5)(1:4) = CLASS(9:12)

27 RETURN

28 END

Figure 8-21. Owncode Routine

A. Report

DEAN'S L I S T

STUDENT MAJOR CLASS GPA

SHIELDS COMPSCI SENIOR 390

HORNE COMPSCI FRESHMAN 389

TERRELL ENG SENIOR 386

SMITH PHIL SOPHOMORE 385

SMITH MATH SENIOR 379

CLARK ECON JUNIOR 378

TIEMON LNGUIS SOPHOMORE 376

FRANKLIN PHIL JUNIOR 370

HANSEN BUS SOPHOMORE 358

DAVIS ENG SENIOR 354

SUGARMAN SOC SENIOR 350

B. Result Array

15 11 0 0

Element 1;

11 0 0 0 0 11 80 80 80

user set. Elements 2 thru 16; set by

Sort/Merge . See table 5-1.

Figure 8-22. Output From FORTRAN Program

0PS

60484800 B 8-9

<*%

CHARACTER DATA INPUT, OUTPUT, AND,

CENTRAL MEMORY REPRESENTATION

0^S

0*S

This appendix describes the code and character sets

used by host computer operating system local batch

device drivers, magnetic tape drivers, and terminal

communication products. Some software products

assume that certain graphic or control characters

are associated with speciﬁc binary code values for

collating or syntax processing purposes. This

appendix does not describe those associations.

All references within this manual to the ASCII

character set or the ASCII code set refer to the

character set and code set deﬁned in the American

Na tional Standard Code for Information Interchange

(ASCII, ANSI Standard X3.4-1977). References in

this manual to the ASCII character set do not

n e c e s s a r i l y a p p l y t o t h e A S C I I c o d e s e t . S o r t /

Merge uses a 63- or 64-character subset of the full

ASCII set. Only characters in 6-bit or internal

BCD display code should be used with Sort/Merge.

CHARACTER SETS AND CODE SETS

A character set differs from a code set. A char

acter set is a set of graphic and/or control char

acters. A code set is a system of symbols used to

r e p r e s e n t e a c h ch a r a c t e r wi t hi n a c h a r a c t e r s e t .

Characters exist outside the computer system and

communication network; codes are received, stored,

retrieved, and transmitted within the computer

system and network.

GRAPHIC AND CONTROL

CHARACTERS

A graphic character can be displayed at a terminal

or printed by a line printer. Examples of graphic

characters are the characters A through Z, a blank,

and the digits 0 through 9. A control character

initiates, modiﬁes, or stops a control operation.

An example of a control character is the backspace

character, which moves the terminal carriage or

cursor back one space. Although a control char

acter is not a graphic character, some terminals

can produce a graphic representation when they

receive a control character.

CODED AND BINARY

CHARACTER DATA

Character codes can be interpreted as coded char

acter data or as binary character data. Coded

c h a r a c t e r d a t a i s c o n v e r t e d f r o m o n e c o d e s e t

representation to another as it enters or leaves

the computer system; for example, data received

from a terminal or sent to a magnetic tape unit is

converted. Binary character data Is not converted

as it enters or leaves the system. Character codes

are not converted when moved within the system; for

example, data transferred to or from mass storage

is not converted.

The distinction between coded character data and

binary character data is important when reading or

60484800 B

punching cards and when reading or writing magnetic

tape. Only coded character data can be properly

reproduced as characters on a line printer. Only

binary character data can properly represent char

acters on a punched card when the data cannot be

stored as display code.

The distinction between binary character data and

characters represented by binary data (such as

peripheral equipment instruction codes) is also

important. Only such binary noncharacter data can

properly reproduce characters on a plotter.

FORMATTED AND UNFORMATTED

CHARACTER DATA

Character codes can be interpreted by a product as

f o r m a tt e d c ha r a c te r d a ta o r a s u n f o r ma t t e d c h a r

acter data. Formatted data can be stored or

retrieved by a product in the form of the codes

described for coded character data in the remainder

of this appendix, or formatted data can be altered

to another form during storage or retrieval; for

example, 1 can be stored as a character code or as

an integer value. Treatment of unformatted data by

a product includes both coded character data and

binary character data as described in this appendix.

NETWORK OPERATING SYSTEMS

The NOS and NOS/BE operating systems support the

following character sets:

• CDC graphic 64-character set

• CDC graphic 63-character set

• ASCII graphic 64-character set

• ASCII graphic 63-character set

• ASCII graphic 95-character set

In addition, NOS supports the ASCII 128-character

graphic and control set.

Each installation must select either a 64-character

set or a 63-character set. The differences between

the codes of a 63-character set and the codes of a

64-character set are described under Character Set

Anomalies. Any reference in this appendix to a 64-

character set implies either a 63- or 64-character

set unless otherwise stated.

To represent its six listed character sets in

ce n t r a l me mo ry, NO S s u p p o r ts t h e fo ll o w i n g co de

sets:

• 6-bit display code

• 12-bit ASCII code

• 6/12-bit display code

A-l

To represent its ﬁve listed character sets in

central memory, NOS/BE supports the following code

sets:

• 6-bit display code

• 12-bit ASCII code

Under both NOS and NOS/BE, the 6-bit display code

is a set of 6-bit codes from OOg to 77s.

Under both NOS and NOS/BE, the 12-bit ASCII code is

the ASCII 7-bit code (as deﬁned by ANSI Standard

X3.4-1977) right-justiﬁed in a 12-bit byte.

Assuming that the bits are numbered from the right

starting with 0, bits 0 through 6 contain the ASCII

code, bits 7 through 10 contain zeros, and bit 11

distinguishes the 12-bit ASCII 00008 code from

the end-of-line byte. The 12-bit codes are OOOlg

through 01778 and 4000g.

Under NOS, the 6/12-bit display code is a combina

tion of 6-bit codes and 12-bit codes. The 6-bit

codes are OOg through 77g, excluding 748 and

76g. (The interpretation of the 008 and 63s

codes is described under Character Set Anomalies in

this appendix.) The 12-bit codes begin with either

74g or 768 and are followed by a 6-bit code.

Thus, 74g and 76g are considered escape codes

a n d a r e n e v e r u s e d a s 6 - b i t c o d e s w i t h i n t h e

6/12-bit display code set. The 12-bit codes are

74018, 74028, 74048, 74078, and 76018

through 76778. All other 12-bit codes (74xx8

and 7600g) are undeﬁned.

CHARACTER SET ANOMALIES

The operating system input/output software and some

products interpret two codes differently when the

installation selects a 63-character set rather than

a 64-character set. If an installation uses a

63-character set, the colon graphic character is

always represented by a 638 display code, display

code 008 is undeﬁned (it has no associated

graphic or punched card code), and the % graphic

does not exist.

However, under NOS, if the installation uses a

64-character set, output of a 74048 6/12-bit

d i s p l a y c o d e o r a O O g d i s p l a y c o d e p r o d u c e s a

c o l o n . A c o l o n c a n b e i n p u t o n l y a s a 7 4 0 4 8

6/12-bit display code. The use of undeﬁned

6/12-bit display codes in output ﬁles produces

unpredictable results and should be avoided.

Under NOS/BE, if the installation uses a 64-

character set, output of a OOg display code

produces a colon. Display code 638 is the colon

when a 63-character set is used. The % graphic and

related card codes do not exist and translations

yield a blank (558).

Under both NOS and NOS/BE, two consecutive OOg

codes can be confused with an end-of-line byte and

should be avoided.

CHARACTER SET TABLES

The character set tables A-l and A-2 are designed

so that you can ﬁnd the character represented by a

code (such as in a dump) or ﬁnd the code that

represents a character. To ﬁnd the character

represented by a code, look up the code in the

column listing the appropriate code set and then

ﬁnd the character on that line in the column

l i s t i n g t h e a p p r op r i a t e c h a ra c te r s e t . To ﬁn d t h e

code that represents a character, look up the

character and then ﬁnd the code on the same line

in the appropriate column.

Conversational Terminal Users

Table A-l shows the character sets and code sets

a v a i l a b l e t o a n I n t e r a c t i v e F a c i l i t y ( I A F ) o r

INTERCOM user at an ASCII code terminal using an

ASCII character set. The Full ASCII Character Set

table (later in this appendix) shows the octal and

hexadecimal 7-bit ASCII code for each ASCII char

acter, and can be used to convert codes from octal

to hexadecimal. (Under NOS using network product

software, certain Terminal Interface Program com

mands require speciﬁcation of an ASCII code.

IAF Usage

When in normal time-sharing mode (speciﬁed by the

IAF NORMAL command), IAF assumes the ASCII graphic

64-character set is used and translates all input

and output to or from display code. When in ASCII

time-sharing mode (speciﬁed by the IAF ASCII

command), IAF assumes the ASCII 128-character set

is us e d a nd tr a ns l at es all in pu t a n d o ut p ut to or

from 6/12-bit display code.

The IAF user can convert a 6/12-bit code ﬁle to a

12-bit ASCII code ﬁle using the NOS FCOPY control

statement. The resulting 12-bit ASCII ﬁle can be

routed to a line printer but cannot be output

through IAF.

IAF supports both character mode and transparent

mode transmissions through the network. These

t r a n s m i s s i o n m o d e s a r e d e s c r i b e d u n d e r N e t w or k

Access Method Terminal Transmission Code Sets in

this appendix. IAF treats character mode trans

missions as coded character data; IAF converts

these transmissions to or from either 6-bit or

6/12-bit display code. IAF treats transparent mode

transmissions as binary character data; transparent

mode communication between IAF and ASCII terminals

using any parity setting occurs in the 12-bit ASCII

code shown in table A-l.

INTERCOM Usage

By default, INTERCOM displays the ASCII graphic

64-character set and interprets all input and

output as display code. Refer to the INTERCOM

reference manual*

COMPASS and FORTRAN users can elect to use 12-bit

ASCII code if the terminal in use supports the code

set selected. BASIC users can elect to send and

receive lowercase and uppercase character codes

using the 12-bit ASCII code if the terminal in use

supports the code set selected; BASIC represents

c o d e d c h a r a c t e r d a t a i n c e n t r a l m e m o r y u s i n g

6/12-bit display code in both the NOS and NOS/BE

systems.

A-2 60484800 B

TABLE A-l. CONVERSATIONAL TERMINAL CHARACTER SETS

ASCII ASCII Octal Octal Octal ASCII ASCII Octal Octal Octal

Graphic Character 6-Bit 6/12-Bit 12-Bit Graphic Character 6-Bit 6/12-Bit 12-Bit

(64-Char (128-Char- Display Display ASCII (64-Char (128-Char- Display Display ASCII

acter Set) acter Set) Code Code* Code acter S e t ) a c t e r S e t ) Code Codet Code

: c o l o n n 00" **" circumﬂex 7402 0136

A A 01 01 0101 : colon" 7404tt 0072

B B 02 02 0102 * grave accent 7407 0140

CC03 03 0103 7601 0141

D D 04 04 0104 7602 0142

E E 05 05 0105 7603 0143

FF06 06 0106 7604 0144

G G 07 07 0107 7605 0145

H H 10 10 0110 7606 0146

I I 11 11 0111 7607 1047

J J 12 12 0112 7610 0150

K K 13 13 0113 7611 0151

L L 14 14 0114 7612 0152

M M 15 15 0115 7613 0153

N N 16 16 0116 7614 0154

0 0 17 17 0117 ra 7615 0155

P P 20 20 0120 7616 0156

QQ21 21 0121 7617 0157

R R 22 22 0122 7620 0160

S S 23 23 0123 7621 0161

T T 24 24 0124 7622 0162

U U 25 25 0125 7623 0163

V V 26 26 0126 7624 0164

Ww27 27 0127 7625 0165

XX30 30 0130 7626 0166

Y Y 31 31 0131 7627 0167

ZZ32 32 0132 7630 0170

0 0 33 33 0060 7631 0171

1134 34 0061 7632 0172

2 2 35 35 0062 ■C left brace 7633 0173

3 3 36 36 0063 | v e r t , l i n e 7634 0174

4 4 37 37 0064 > right brace 7635 0175

5 5 40 40 0065 " tilde 7636 0176

6641 41 0066 NUL 7640 4000

7 7 42 42 0067 SOH 7641 0001

8 8 43 43 0070 STX 7642 0002

9 9 44 44 0071 ETX 7643 0003

+ plus + p lus 45 45 0053 EOT 7644 0004

- minus - minus 46 46 0055 ENQ 7645 0005

* asterisk * asterisk 47 47 0052 ACK 7646 0006

/ slash / slash 50 50 0057 BEL 7647 0007

( I . p a r e n . ( I . p a r e n . 51 51 0050 BS 7650 0010

) r. p a r e n . ) r. p a r e n . 52 52 0051 HT 7651 0011

$ dollar $ dollar 53 53 0044 LF 7652 0012

= equal to = equa l to 54 54 0075 VT 7653 0013

space space 55 55 0040 FF 7654 0014

, comma , comma 56 56 0054 CR 7655 0015

. period . period 57 57 0056 SO 7656 0016

tt number ft number 60 60 0043 SI 7657 0017

H 1. bracket H 1. bracket 61 61 0133 DEL 7637 0177

1 r. b r a c k e t ] r. b r a c k e t 62 62 0135 DLE 7660 0020

X p e r c e n t " % percent1* 63" 63" 0045 DC1 7661 0021

" quote " q u ote 64 64 0042 DC2 7662 0022

_ underline _ underline 65 65 0137 DC3 7663 0023

! exclam. ! exclam. 66 66 0041 DC4 7664 0024

& ampersand & ampersand 67 67 0046 NAK 7665 0025

• apostrophe ' apostro phe 70 70 0047 SYN 7666 0026

? question ? question 71 71 0077 ETB 7667 0027

< less than < less than 72 72 0074 CAN 7670 0030

> g r t r. t h a n > g r t r. t h a n 73 73 0076 EM 7671 0031

a coml. at 74 SUB 7672 0032

\ r e v. s l a n t \ r e v. s l a n t 75 75 0134 ESC 7673 0033

■*■ circumflex 76 FS 7674 0034

; sem ico lon ; sem ico lon 77 77 0073 GS 7675 0035

3 com I. a t 7401 0100 RS

7676

7677

0036

0037

^Generally available only on NOS, or through BASIC on NOS/BE.

" T h e i n t e r p re t a t i o n o f t h i s charac ter or coc e depends on its context. Refer to Character Set Anomalies in

the text.

60484800 B A-3

TABLE A-2. LOCAL BATCH DEVICE CHARACTER SETS

CDC ASCII ASCII Octal Octal Octal Card Keypunch Code

Graphic

(64-Character

Graphic

(64-Character

Graphic

(95-Character

6-Bit 6/12-Bit

Display

12-Bit

ASCII

Display 026 029

Set) Set) Set) Code Codet Code

: c o l o n " : c o l o n " 00" 8-2 8-2

AA A 01 01 0101 12-1 12-1

B B B02 02 0102 12-2 12-2

CCC 03 03 0103 12-3 12-3

D D D 04 04 0104 12-4 12-4

EEE 05 05 0105 12-5 12-5

FFF 06 06 0106 12-6 12-6

G G G 07 07 0107 12-7 12-7

H H H 10 10 0110 12-8 12-8

III 11 11 0111 12-9 12-9

J J J 12 12 0112 11-1 11-1

K K K 13 13 0113 11-2 11-2

L L L 14 14 0114 11-3 11-3

M M M 15 15 0115 11-4 11-4

NN N 16 16 0116 11- 5 11-5

0 0 0 17 17 0117 11 - 6 11-6

P P P 20 20 0120 11-7 11-7

Q Q Q 21 21 0121 11-8 11-8

R R R 22 22 0122 11 - 9 11-9

S S S 23 23 0123 0-2 0-2

T T T 24 24 0124 0-3 0-3

U U U 25 25 0125 0-4 0-4

V V V 26 26 0126 0-5 0-5

u u W27 27 0127 0-6 0-6

X X X 30 30 0130 0-7 0-7

YY Y 31 31 0131 0-8 0-8

zZ Z 32 32 0132 0-9 0-9

0 0 0 33 33 0060

1 1 134 34 0061

2 2 2 35 35 0062

333 36 36 0063

44437 37 0064

555 40 40 0065

666 41 41 0066

77 7 42 42 0067

888 43 43 0070

9 9 9 44 44 0071

+ plus + plus + plus 45 45 0053 12 12-8-6

- minus - minus - minus 46 46 0055 11 11

* asterisk * a st er i sk * as t er is k 47 47 0052 11-8-4 11-8-4

/ slash / s l a sh / s lash 50 50 0057 0-1 0-1

( Left paren. ( left paren. ( left paren. 51 51 0050 0-8-4 12-8-5

> right paren. ) right paren. ) right paren. 52 52 0051 12-8-4 11-8-5

$ do l l a r $ d o l l a r $ d o l lar 53 53 0044 11 -8- 3 11-8-3

= equal to = equal to = equal to 54 54 0075 8-3 8-6

space space space 55 55 0040 no punch no punch

, comma , comma , comma 56 56 0054 0-8-3 0-8-3

. period . peri od . peri od 57 57 0056 12-8-3 12-8-3

= equivalence # number # number 60 60 0043 0-8-6 8-3

Z left bracket C left bracket C I. bracket 61 61 0133 8-7 12-8-2

or 12-Ot"

1 right bracket 1 right b r acket ] r. b r ac ke t 62 62 0135 0-8-2 11-8-2

or 11-0"t

% percent" X p ercent" X pe rcent" 63" 63" 0045 8-6 0-8-4

A-4 60484800 B

TABLE A-2. LOCAL BATCH DEVICE CHARACTER SETS (Contd)

0m*s

CDC

Graphic

(64-Character

Set)

t* not equal

r*concat.

V logical OR

A logical AND

t superscript

I subscript

< less than

> greater than

£ Less/equal

>^ greater/equal

—i logical NOT

; semicolon

ASCII

Graphic

(64-Character

Set)

quote

underline

! exclamation

& ampersand

' apostrophe

? question

< less than

> greater than

3 commercial at

\ reverse slant

circumﬂex

semicolon

ASCII

Graphic

(95-Character

Set)

" quote

underline

! exclamation

& ampersand

' apostrophe

? question

< less than

> greater than

\ r e v. s l a nt

; semicolon

S com I. at

-*• circumflex

: colon

' grave accent

•C left brace

| vert, line

> right brace

" tilde

Octal

6 - B i t

Display

Code

Octal

6/12-Bit

Display

Codet

tGenerally available only on NOS, or through BASIC on NOS/BE.

7401

7402

7404

7407

7601

7602

7603

7604

7605

7606

7607

7610

7611

7612

7613

7614

7615

7616

7617

7620

7621

7622

7623

7624

7625

7626

7627

7630

7631

7632

7633

7634

7635

7636

Octal

12-Bit

ASCII

Code

0042

0137

0041

0046

0047

0077

0074

0076

0134

0073

0100

0136

0072

0140

0141

0142

0143

0144

0145

0146

0147

0150

0151

0152

0153

0154

0155

0156

0157

0160

0161

0162

0163

0164

0165

0166

0167

0170

0171

0172

0173

0174

0175

0176

Card Keypunch Code

026

12-8-7

029

8-4 8_7R

0-8-5 0-8-5§

11-0

11-o§ 12-8-7

11-8-2

0-8-7 12

11-8-5 8-5

11-8-6 0-8-7

12-0 12-8-4

11-8-7 0-8-6

8-5 8-4

12-8-5 0-8-2

12-8-6 11-8-7

11-8-6

tt The interpretation of this character or code depends on its context. Refer to Character Set Anomalies

i n t h e t e x t .

"t Available for input only, on NOS.

§ Available for input only, on NOS/BE.

60484800 B A-5

Local Batch Users

Table A-2 lists the CDC graphic 64-character set,

the ASCII graphic 64-character set, and the ASCII

graphic 95-character set. This table also lists

the code sets and card keypunch codes (026 and 029)

that represent the characters.

The 64-character sets use display code as their

code set; the 95-character set uses 12-bit ASCII

code. The 95-character set is composed of all the

characters in the ASCII 128-character set that can

be printed at a line printer (refer to Line Printer

O u t p u t ) . O n l y 1 2 - b i t A S C I I c o d e ﬁ l e s c a n b e

printed using the ASCII graphic 95-character set.

To p r i n t a 6 / 1 2 - b i t d i s p l a y c o d e ﬁ l e ( u s u a l l y

created in IAF ASCII mode), you must convert the

ﬁle to 12-bit ASCII code. To do this, the NOS

FCOPY control statement must be issued. The

95-character set Is represented by the 12-bit ASCII

codes 00408 through 01768.

Line Printer Output

The batch character set printed depends on the

print train used on the line printer to which the

ﬁle is sent. The following are the print trains

corresponding to each of the batch character sets:

Character Set

CDC graphic 64-character set

ASCII graphic 64-character set

ASCII graphic 95-character set

P r i n t

Train

596-1

596-5

596-6

The characters of the default 596-1 print train are

listed in the table A-2 column labeled CDC Graphic

(64-Character); the 596-5 print train characters

are listed in the table A-2 column labeled ASCII

Graphic (64-Character); and the 596-6 print train

characters are listed in the table A-2 column

labeled ASCII Graphic (95-Character).

If a transmission error occurs during the printing

o f a l i n e , N O S p r i n t s t h e l i n e a g a i n . T h e C D C

g r a p h ic p ri n t t ra i n p r i n t s a co n c a t e na t i o n s y m b ol

( r * ) i n t h e ﬁ r s t p r i n t a b l e c o l u m n o f a l i n e

containing errors. The ASCII print trains print an

underline instead of the concatenation symbol.

If an unprintable character exists in a line (that

is, a 12-bit ASCII code outside of the range

00408 through 01768), the number sign (i?)

a p p e a r s i n t h e ﬁ r s t p r i n t a b l e c o l u m n o f a p r i n t

line and a space replaces the unprintable character.

Punched Card Input and Output

Under NOS, coded character data is exchanged with

local batch card readers or card punches according

to the translations shown in table A-2. As indi

cated in the table, additional card keypunch codes

are available for input of the ASCII and CDC char

acters [ and ]. The 95-character set cannot be

read or punched as coded character data.

Depending on an installation or deadstart option,

NOS assumes an input deck has been punched either

In 026 or 029 keypunch code (regardless of the

character set in use). The alternate keypunch

codes can be speciﬁed by a 26 or 29 punched in

columns 79 and 80 of any job card, 6/7/9 card, or

7/8/9 card. The speciﬁed code translation remains

in effect throughout the job unless it is reset by

speciﬁcation of the alternate code translation on

a subsequent 6/7/9 card or 7/8/9 card.

NOS keypunch code translation can also be changed

by a card containing a 5/7/9 punch in column 1. A

blank (no punch) in column 2 indicates 026 conver

sion mode; a 9 punch in column 2 indicates 029

conversion mode. The conversion change remains in

effect unti l ano ther change card is encountered or

the job ends.

The 5/7/9 card also allows literal input when

4/5/6/7/8/9 is punched in column 2. Literal input

can be used to read 80-column binary character data

within a punched card deck of coded character data.

L i t e r a l c a r d s a r e s t o r e d w i t h e a c h c o l u m n i n a

12-bit byte (a row 12 punch is represented by a 1

in b it 11, r ow 11 b y b i t 1 0 , ro w 0 by b it 9 , a n d

rows 1 through 9 by bits 8 through 0 of the byte),

16 central memory words per card. Literal input

cards are read until a card identical to the pre

vious 5/7/9 card (4/5/6/7/8/9 in column 2) is

read. The next card can specify a new conversion

mode.

Under NOS/BE, coded character data is exchanged

w i t h l o c a l b a t c h c a r d r e a d e r s o r c a r d p u n c h e s

according to the translations shown in table A-2.

As indicated in the table, additional card keypunch

codes are available for input of the CDC characters

V and < or their ASCII equivalents ! and <. The

95-character set cannot be read or punched as coded

character data.

Depending on an installation option, NOS/BE assumes

an input deck has been punched either in 026 or in

029 keypunch code (regardless of the character set

in use). The alternate keypunch codes can be

speciﬁed by a 26 or 29 punched in columns 79 and

80 of the job statement or In columns 79 and 80 of

a n y 7 / 8 / 9 c a r d . T h e s p e c i ﬁ e d c o d e t r a n s l a t i o n

remains in effect throughout the job unless it is

reset by speciﬁcation of the alternate code

translation on a subsequent 7/8/9 card.

Under NOS/BE, a card with all of column 1 punched

(that is, 12 punches in column 1) and all of one

other column punched can be followed by 80-column

cards of free-form binary data. These binary data

cards are read or punched as described for NOS

literal data until another card with 12 punches in

column 1 and in one other column occurs, or until

the job ends. The next card is interpreted as

coded data.

Remote Batch Users

When card decks are read from remote batch devices,

the ability to select alternate keypunch code

translations depends upon the remote terminal

equipment.

■<-/^\

A-6 60484800 B

NOS Usage

Remote batch terminal line printer, punched card,

and plotter character set support is described in

the Remote Batch Facility (RBF) reference manual.

RBF supports only character mode transmission to

and from conso l e s through t h e network. Character

mode Is described under Network Access Method

Terminal Transmission Code Sets in this appendix.

NOS/BE Usage

The remote batch terminal line printer, punched

card, and plotter character set support is de

scribed in the INTERCOM reference manual.

Magnetic Tape Users

Coded character data to be copied from mass storage

to magnetic tape is assumed to be represented in

display code. NOS converts the data to external

BCD code when writing a coded 7-track tape and to

ASCII or EBCDIC code (as speciﬁed on the tape

assignment statement) when writing a coded 9-track

tape.

Because only 63 characters can be represented in

7-track even parity, one of the 64 display codes is

lost in conversion to and from external BCD code.

Figure A-l shows the differences in conversion that

depend on which character set (63 or 64) the system

uses. The ASCII character for the speciﬁed char

acter code is shown in parentheses. The output

arrow shows how the display code changes when it is

written on tape in external BCD. The input arrow

shows how the external BCD code changes when the

tape is read and converted to display code.

63-Character Set

Display Code External BCD

16(%)

Output 12(0)

12(0)

Display Code

33(0)

63(:)

Input 33(0)

33(0)

64-Character Set

Display Code External BCD

12(0)

Output 12(0)

16(%)

Display Code

00(:)

33(0)

63(%)

33(0)

Input 33(0)

63(%)

Figure A-1. Magnetic Tape Code Conversions

Tables A-3 and A-4 show the character set conver

sions for 9-track tapes. Table A-3 lists the

conversions to and from 7-bit ASCII character code

and 6-bit display code. Table A-4 lists the

conversions between 8-bit EBCDIC character code and

6-bit display code. Table A-5 shows the character

set conversions between 6-bit external BCD and

6-bit display code for 7-track tapes.

If a lowercase ASCII or EBCDIC code is read from a

9 - t r a ck c o d ed t ap e , i t i s c o n ve r t e d t o it s u p pe r

case 6-bit display code equivalent. To read and

write lowercase ASCII or EBCDIC characters, the

user must assign the tape in binary mode and then

convert the binary character data.

During binary character data transfers to or from

9-track magnetic tape, the 7-bit ASCII codes shown

in table A-6 are read or written unchanged; the

8-bit hexadecimal EBCDIC codes shown in table A-7

al so can be read or written uncha nged. ASCII an d

EBCDIC codes cannot be read or written to 7-track

magnetic tape as binary character data.

Two CDC utility products, FORM and the 8-Bit Sub

routines, can be used to convert to and from EBCDIC

data. Table A-7 contains the octal values of each

EBCDIC code right-justiﬁed in a 12-bit byte with

zero ﬁll. This 12-bit EBCDIC code can also be

produced using FORM and the 8-Bit Subroutines.

NETWORK ACCESS METHOD TERMINAL

TRANSMISSION CODE SETS

There are two modes in which coded character data

can be exchanged with a network terminal console.

These two modes, character mode and transparent

mode, correspond to the type of character code

editing and translation performed by the network

software during input and output operations. The

transmission mode used by the network software for

input can be selected by the terminal operator,

using a Terminal Interface Program command

(sometimes referred to as a terminal deﬁnition

command). The transmission mode used by the

network software for output can be selected by the

application program providing the terminal facility

service.

Character Mode Transmissions

Character mode is the initial and default mode used

for both input and output transmissions. When the

network software services the terminal in character

mode, it translates input characters from the

transmission code used by the terminal into the

ASCII code shown in table A-6. The translation of

a speciﬁc transmission code to a speciﬁc ASCII

code depends on the terminal class the network

software associates with the terminal. In char

acter mode input, the parity of the terminal

transmission code is not preserved in the corres

ponding ASCII code; the ASCII code received by the

terminal-servicing facility program always has its

eighth bit set to zero.

Character mode output is translated in a similar

manner. The network software provides the parity

bit setting appropriate for the terminal being

serviced, even though translating from ASCII char

acters with zero parity bit settings.

The general case for code translations of character

mode data is summarized in the following para

graphs. This generalized description permits use

of only table A-6 to explain all speciﬁc cases.

You can logically extend this generalized descrip

tion to allow use of tables A-l through A-5 as

descriptions of character set mapping for various

functions initiated from a terminal. Tables A-l

through A-5 are provided for your use while coding

an application program to run under the operating

s y s t e m . T h e y d o n o t d e s c r i b e c h a r a c t e r t r a n s

missions between an application program and the

network.

60484800 B A-7

TABLE A-3. ASCII 9-TRACK CODED TAPE CONVERSION

ASCII ASCII

Display Display

Code Character and Codet" Code Character and Code"t

Conversion* Code Conversion" Conversion' Code Conversion"

Code Char Code Char ASCII Code Code Char Code Char ASCII Code

(Hex) (Hex) Char (Octal) (Hex) (Hex) Char (Octal)

20 space 00 NUL space 55 40 60 74

21 7D 66 41 61 01

22 ii 02 STX ii 64 42 62 02

23 03 ETX tt 60 43 63 03

24 04 EOT 53 44 64 04

25 05 ENQ 63 45 65 05

25 05 ENQ space 55 46 66 06

26 06 ACK 67 47 67 07

27 07 BEL 70 48 68 10

28 08 BS 51 49 69 11

29 09 HT 52 4A 6A 12

2A 0A LF 47 4B 6B 13

2B OB VT 45 4C 6C 14

2C OC FF 56 4D 6D 15

2D 0D CR 46 4E 6E 16

2E 0E SO 57 4F 6F 17

2F OF SI 50 50 70 20

30 10 DLE 33 51 71 21

31 11 DC1 34 52 72 22

32 12 DC2 35 53 73 23

33 13 DC3 36 54 74 24

34 14 DC4 37 55 75 25

35 15 NAK 40 56 76 26

36 16 SYN 41 57 77 27

37 17 ETB 42 58 78 30

38 18 CAN 43 59 79 31

39 19 EM 44 5A 7A 32

3A 1A SUB 00 5B 1C FS r. 61

Display code 00 is undeﬁned at sites using the 5C 7C 75

63-character set. 5D 01 SOH 62

3A 1A SUB 63 5E J< 7E AU 76

3B 1B ESC 77 5F 7F DEL 65

3C 7B 72

3D 1D GS 54

3E 1E RS 73

3F 1F US 71

twhen these characters are copied from or to a tape, the characters remain the same and the code changes

from/to ASCII to/from display code.

ttThese characters do not exist in display code. When the characters are copied from a tape, each ASCII

character is changed to an alternate display code c laracter. The corresponding codes are also changed.

Example: When the system copies a lowercase a, 61<| 5, from tape, it writes an uppercase A, 018.

'"A display code space always translates to an ASCII space.

Table A-6 contains the ASCII 128-character set

supported by the Network Access Method. A 96-

character subset consists of the rightmost six

columns and . includes the 95-character graphic

subset referenced previously in this appendix; the

deletion character (DEL) is not a graphic char

acter. A 64-character subset consists of the

middle four columns. Note that 6-bit display code

equivalents exist for the characters in this 64-

character subset only.

A l t h o u g h t h e n e t w o r k s u p p o r t s t h e 1 2 8 - c h a r a c t e r

set, some terminals restrict output to a smaller

subset. This restriction is supported by replacing

the control characters in columns 0 and 1 of table

A-6 with blanks to produce the 96-character subset,

and, additionally, replacing the characters in

columns 6 and 7 with the corresponding characters

from columns 4 and 5, respectively, to produce the

64-character subset.

A-8 60484800 B

TABLE A-4. EBCDIC 9-TRACK CODED TAPE CONVERSION

0ms

EBCDIC EBCDIC

Code

Conversiont

Character and

Code Conversion"

Display

Codet" Code

Conversiont

Character and

Code Conversion^

Display

Codet"

Code Char Code Char ASCII Code Code Char Code Char ASCII Code

(Hex) (Hex) Char (Octal) (Hex) (Hex) Char (Octal)

40 space 00 NUL space 55 C6 86 06

4A 1C IFS 61 C7 87 07

4B 0E SO 57 C8 88 10

4C CO {. 72 C9 89 11

4D 16 BS 51 D1 91 12

4E 0B VT 45 D2 92 13

4F DO 66 D3 93 14

50 2E ACK 67 D4 94 15

5A 01 SOH 62 D5 95 16

5B 37 EOT 53 D6 96 17

5C 25 LF 47 D7 97 20

5D 05 HT 52 D8 98 21

5E 27 ESC 77 D9 99 22

5F —« A1 ** 76 E0 6A 75

60 0D CR 46 E2 A2 23

61 OF SI 50 E3 A3 24

6B OC FF 56 E4 A4 25

6C 2D ENQ 63 E5 A5 26

6C 2D ENQ space 55 E6 A6 27

6D 07 DEL 65 E7 A7 30

6E 1E IRS >" 73 E8 A8 31

6F 1F IUS 71 E9 A9 32

7A 3F SUB : • 00 FO 10 DLE 33

Display code 00 is undeﬁned at sites using the F1 11 DC1 34

63-character set. F2 12 DC2 35

7A 3F SUB 63 F3 13 TM 36

7B 03 ETX tt 60 F4 3C DC4 37

7C 79 74 F5 3D NAK 40

7D 2F BEL 70 F6 32 SYN 41

7E 1D IGS 54 F7 26 ETB 42

7F ti 02 STX ii 64 F8 18 CAN 43

C1 81 01 F9 19 EM 44

C2 82 02

C3 83 03

C4 84 04

C5 85 05

'All EBCDIC codes not listed translate to display code 558 (space). A display code space always

translates to an EBCDIC space.

"These characters do not exist in display code. When the characters are copied from a tape, each EBCDIC

character is changed to an alternate display code character. The corresponding codes are also changed.

Example: When the system copies a lowercase a, 81<|£, from tape, it writes an uppercase A/ 018.

"•When these characters are copied from or to a tape, the characters remain the same (except EBCDIC codes

*A16r 4F"16/ 5A-|6, and 5F<|0) and the code changes from/to EBCDIC to/f rom display code.

Sim ilarly, inpu t from a device may be li mited to a

smaller subset by the device itself because the

de v ic e c a nn o t p rod uce th e f u ll 12 8 -c h ar a ct e r s et.

A character input from a device using a character

set other than ASCII is converted to an equivalent

ASCII character; characters without ASCII character

equivalents are replaced by the ASCII blank char

acter.

An application can also cause character replacement

(as described previously for output) as well as

character conversion, by requesting display-coded

input from the network.

The 7-bit hexadecimal code value for each character

consists of the character's column number in the

table, followed by its row number. For example, N

is in row E of column 4, so its value is 4Ejg.

60484800 B A-9

TABLE A-5. 7-TRACK CODED TAPE CONVERSIONS

External ASCII Octal Display External ASCII Octal Display

BCD Character Code BCD Character Code

01 34 40 46

02 35 41 12

03 36 42 13

04 37 43 14

05 40 44 15

06 41 45 16

07 42 46 17

10 43 47 20

11 44 50 21

12t 33 51 22

13 54 52 66

14 •i 64 53 53

15 74 54 47

16t 63 55 70

17 61 56 71

20 space 55 57 73

21 50 60 45

22 23 61 01

23 24 62 02

24 25 63 03

25 26 64 04

26 27 65 05

27 30 66 06

30 31 67 07

31 32 70 10

32 62 71 11

33 56 72 72

34 51 73 57

35 65 74 52

36 60 75 75

37 67 76 AU. 76

77 77

tAs explained in the text of this appendix, conversion of these codes depends on whether the tape is

being read or written.

Transparent Mode Transmissions

Transparent mode is selected separately for input

and output transmissions. During transparent mode

input, the parity bit is stripped from each ter

minal transmission code (unless the N parity option

has been selected by a Terminal Interface Program

command), and the transmission code is placed in an

8 - b i t b y t e w i t h o u t t r a n s l a t i o n t o 7 - b i t A S C I I

code. Line transmission protocol characters are

deleted from a mode 4C terminal input stream.

When the 8-bit bytes arrive in the host computer, a

terminal servicing facility program such as the

I n t e r a c t i v e F a c i l i t y c a n r i g h t - j u s t i f y t h e b y t e s

within a 12-bit byte. Upon transmission of 12-bit

by t es f rom th e h o st c om p ute r, th e l e ftm o st 4 b i ts

(bits 11 through 8) are discarded.

During transparent mode output, processing similar

to that performed for input occurs. The code in

each 8-bit byte received by the network software

from the terminal servicing facility program is not

translated. The parity bit appropriate for the

te rm i n a l cl a s s b e i n g us e d is al te re d a s i n d ic at ed

b y t h e p a r i t y o p t i o n i n e f f e c t f o r t h e t e r m i n a l .

The codes are then output in transmission bytes

appropriate for the codes associated with the

terminal class being used. Line transmission

p r o t oc o l c h a r a c t e r s a r e i n s e r t e d i n t o a m o d e 4 C

terminal output stream.

A-10 60484800 B

■*ms

//ms

TABLE A-6. FULL ASCII CHARACTER SET

•128-Character Set

•96-Character Subset

-64-Character Subset-

J ^ N

Bits b 4 b 3 b 2 b 1

0 0 0 0

000 1

0 0 1 0

0011

010 0

0 1 0 1

011 0

0111

0 0 0

001

010

0 1 1

1 0 0

1 0 1

1 1 0

1 1 1

Column

Row

NUL

000

SOH

001

STX

002

ETX

003

EOT

004

ENQ

005

ACK

006

BEL

007

010

011

012

013

014

015

016

017

DLE

020

DC1

021

DC2

022

DC3

023

DC4

024

NAK

025

SYN

026

ETB

027

CAN

030

031

SUB

032

ESC

033

034

035

036

037

040

041

042

043

044

045

046

047

(

050

)

051

052

053

054

055

056

057

060

061

062

063

064

065

066

067

070

071

072

073

074

075

076

077

100

101

102

103

104

105

106

107

110

111

112

113

114

11 5

11 6

11 7

120

121

122

123

124

125

126

127

130

131

132

133

134

135

136

137

140

141

142

143

144

145

146

147

150

151

152

153

154

155

156

157

160

161

162

163

164

165

166

167

170

171

172

■C

173

174

175

176

DEL

177

LEGEND:

Numbers under characters are the octal values for the 7-bit character codes used within the network.

60484800 B A-ll

TABLE A-7. FULL EBCDIC CHARACTER SET

Hexa

decimal

EBCDIC

Code

Octal

12-Bit

EBCDIC

Code

EBCDIC

Graphic

Charactert

EBCDIC

Control

Character

Hexa

decimal

EBCDIC

Code

Octal

12-Bit

EBCDIC

Code

EBCDIC

Graphic

Charactert

EBCDIC

Control

Character

00 OOOO NUL 41 0101 undeﬁned

01 0001 SOH t h r u t h r u

02 0002 STX 49 0111

03 0003 ETX 4A 0112

04 0004 PF 4B 0113

05 0005 HT 4C 0114

06 0006 LC 4D 0115

07 0007 DEL 4E 0116

08 0010 undeﬁned 4F 0117

09 0011 undeﬁned 50 0120

OA 0012 SMM 51 0121 undeﬁned

OB 0013 VT t h r u t h r u

OC 0014 FF 59 0131

OD 0015 CR 5A 0132

OE 0016 SO 5B 0133

OF 0017 SI 5C 0134

10 0020 DLE 5D 0135

11 0021 DC1 5E 0136

12 0022 DC2 5F 0137 —\

13 0023 TM 60 0140

14 0024 RES 61 0141

15 0025 NL 62 0142 undeﬁned

16 0026 BS t h r u t h r u

17 0027 IL 69 0151

18 0030 CAN 6A 0152

19 0031 EM 6B 0153

1A 0032 CC 6C 0154

1B 0033 CU1 6D 0155

1C 0034 IFS 6E 0156

1D 0035 IGS 6F 0157

1E 0036 IRS 70 0160 undeﬁned

1F 0037 IUS t h r u t h r u

20 0040 DS 78 0170

21 0041 SOS 79 0171

22 0042 FS 7A 0172

23 0043 undeﬁned 7B 0173

24 0044 BYP 7C 0174

25 0045 LF 7D 0175

26 0046 ETB or EOB 7E 0176

27 0047 ESC or PRE 7F 0177 ••

28 0050 undeﬁned 80 0200 undeﬁned

29 0051 undeﬁned 81 0201

2A 0052 SM . 82 0202

2B 0053 CU2 83 0203

2C 0054 undeﬁned 84 0204

2D 0055 ENQ 85 0205

2E 0056 ACK 86 0206

2F 0057 BEL 87 0207

30 0060 undeﬁned 88 0210

31 0061 undeﬁned 89 0211

32 0062 SYN 8A 0212 undeﬁned

33 0063 undeﬁned t h r u t h r u

34 0064 PN 90 0220

35 0065 RS 91 0221

36 0066 UC 92 0222

37 0067 EOT 93 0223

38 0070 undeﬁned 94 0224

39 0071 undeﬁned 95 0225

3A 0072 undeﬁned 96 0226

3B 0073 CU3 97 0227

3C 0074 DC4 98 0230

3D 0075 NAK 99 0231

3E 0076 undeﬁned 9A 0232 undeﬁned

3F 0077 SUB thru t h r u

40 0100 space AO 0240

A-12 60484800 B

TABLE A-7. FULL EBCDIC CHARACTER SET (Contd)

Jpfe..

Hexa

decimal

EBCDIC

Code

Octal

12-Bit

EBCDIC

Code

A1 0241

A2 0242

A3 0243

A4 0244

A5 0245

A6 0246

A7 0247

A8 0250

A9 0251

AA 0252

thru thru

BF 0277

CO 0300

C1 0301

C2 0302

C3 0303

C4 0304

C5 0305

C6 0306

C7 0307

C8 0310

C9 0311

CA 0312

CB 0313

CC 0314

CD 0315

CE 0316

CF 0317

DO 0320

D1 0321

D2 0322

D3 0323

D4 0324

D5 0325

D6 0326

EBCDIC

Graphic

Charactert

EBCDIC

Control

Character

undeﬁned

Hexa

decimal

EBCDIC

Code

Octal

12-Bit

EBCDIC

Code

D7 0327

D8 0330

D9 0331

DA 0332

thru thru

DF 0337

EO 0340

E1 0341

E2 0342

E3 0343

E4 0344

E5 0345

E6 0346

E7 0347

E8 0350

E9 0351

EA 0352

EB 0353

EC 0354

ED 0355

thru thru

EF 0357

FO 0360

F1 0361

F2 0362

F3 0363

F4 0364

F5 0365

F6 0366

F7 0367

F8 0370

F9 0371

FA 0372

FB 0373

thru thru

FF 0377

EBCDIC

Graphic

Charactert

EBCDIC

Control

Character

undeﬁned

Graphic characters shown are those used on the IBM System/370 standard (PN) print train. Other devi<

support subsets or variations of this character graphic set.

0ms

60484800 B A-13

DIAGNOSTICS

Diagnostic messages that can be issued by Sort/

Me r ge a re li s ted in ta b le B-l . I f an er r or occ urs ,

the error level and the error number are written on

the error ﬁle or the listing ﬁle. The diagnostic

message is written on the next line. For errors

numbered 1 through 40, the parameter statement that

Is in error is written on another line, and a

pointer indicates the position of the error in the

parameter statement.

An error can be one of the following levels:

T Trivial; a trivial diagnostic results from

a usage that is syntactically correct but

questionable.

W Warning; a warning diagnostic results from

a usage that is syntactically incorrect,

but from which Sort/Merge has been able to

recover by making an assumption about what

was intended.

Fatal; a fatal diagnostic results when

Sort/Merge cannot resolve a syntactic or

semantic error. The remaining parameter

statements are read, but no Sort/Merge

processing occurs.

Catastrophic; a catastrophic error causes

immediate Sort/Merge termination.

S p e c i f y i n g t h e S TAT U S p a r a m e t e r o r c a l l i n g t h e

SM5ST procedure prevents job step termination after

any data records are input; the status variable

indicates the highest level of error that occurred.

TABLE B-l. SORT/MERGE DIAGNOSTICS

J^1

0ms

Error

Number

Error

Level Message

A SPACE OR , EXPECTED

CHARACTERS USED OUTSIDE

A STRING OR NAME MUST

BE A LETTER DIGIT ) =

( ' , OR SPACE

A VALUE SHOULD FOLLOW

THIS KEYWORD

UNDEFINED COLLATING

SEQUENCE

CHARACTER AFTER KEYWORD

MUST BE SPACE OR =

INVALID KEYWORD

DETECTED

UNLESS IN A STRING, =

SHOULD FOLLOW ONLY A

KEYWORD

DUPLICATE COLLATING

SEQUENCE NAME FOUND.

Signiﬁcance

A space or a comma must sep

arate parameters in a control

statement or directive ﬁle. A

space or a comma must also sep

arate values in a set of values

speciﬁed for a parameter.

Unless in a literal string or

in the name of a user—supplied

entity, only alphabetic and

numeric characters and the

special characters ) = ( ' ,

and space are valid in a con

tr o l s ta te m en t o r d ir e ct i ve

ﬁle.

At least one value must be

speciﬁed following a

parameter keyword.

The name of a key type in a KEY

parameter is not a predeﬁned

collating sequence or a se

quence you have deﬁned with

SEQx parameters.

A parameter keyword must be

separated from its value by an

= or space.

The indicated keyword is not a

valid procedure call keyword.

Unless an - is in a literal

string, = is onl y valid i mmedi

ately following a parameter

keyword.

Collating sequence names must

be unique.

Action

Insert a space or a comma

between each parameter and

between values speciﬁed

for a parameter.

Replace invalid characters

with valid characters.

Insert a value following

the parameter keyword.

Change the name of the key

type to one of the prede

ﬁned collating sequences

or a sequence you have

deﬁned.

Insert an = after the

parameter keyword.

Determine valid procedure

call keywords.

Replace the invalid = with

a valid character.

Eliminate the duplicate

name.

60484800 D B-l

TABLE B-l. SORT/MERGE DIAGNOSTICS (Contd)

Error

Number

Error

Level Message Signiﬁcance Action

9 F ( ) MUST ENCLOSE A

VALUE

Parentheses in a control state

ment or directive ﬁle must

enclose a parameter value.

Place a parameter value

between the parentheses*

10 CHARACTER AFTER = MUST

BE LETTER, DIGIT ( $ *

OR SPACE

Only an alphabetic or numeric

character or one of the special

characters ( $ ' or space can

follow an =.

Replace the character fol

lowing the a with a valid

character.

11 THIS VALUE SHOULD BE A

STRING (DELIMITED WITH

' S )

The indicated value must be a

literal string enclosed in

apostrophes.

Enclose the indicated value

in apostrophes.

13 THIS COMMA SHOULD BE

FOLLOWED BY A LETTER,

DIGIT, SPACE , OR (

The indicated comma must be

followed by an alphabetic or

numeric character or one of the

special characters , ( or

space.

Replace the character fol

lowing the indicated comma

with a valid character.

14 THIS CHARACTER MUST BE

A LETTER, DIGIT ( '

OR ,

The indicated character must be

an alphabetic or numeric char

acter or one of the special

characters ( * or ,.

Replace the indicated char

acter with a valid

character.

15 TOO MANY PARAMETER

POSITIONS HAVE BEEN

SPECIFIED

When specifying parameters by

position rather than by key

word, a maximum of 31 positions

is allowed; you have probably

included too many commas.

Check your control state

ment for excess commas.

17 CHARACTER AFTER ) MUST

BE SPACE ) OR ,

A right parenthesis can only be

followed by another right

parenthesis, a comma, or a

space.

Replace the invalid char

acter after the right

parenthesis with a valid

character.

18 CHARACTER AFTER ( MUST

BE A LETTER, DIGIT $ '

( OR SPACE

A left parenthesis can only be

followed by an alphabetic or

numeric character, another left

parenthesis, a $ ' or space.

Replace the invalid char

acter after the left

parenthesis with a valid

character.

19 THE HIGH AND LOW VALUES

OF THIS RANGE MUST BE

AN INTEGER, Rl, R2, R3,

RIG, EF OR EFG

If a range of values is speci

ﬁed for the KEY, SUM, or ENR

parameter, the high and low

values of the range must be

integers or the CCL variables

Rl, R2, R3, RIG, EF, or EFG.

Replace the invalid range

value with an integer, Rl,

R2, R3, RIG, EF, or EFG.

20 THE HIGH AND LOW VALUES

OF THIS RANGE MUST BE A

STRING

If the SEQS parameter speciﬁes

a range of values, the high and

low values of the range must be

literal strings enclosed in

apostrophes.

Replace the invalid range

with a literal string

enclosed in apostrophes.

21 CHARACTER FOLLOWING AN

INTEGER MUST BE ) .. ,

SPACE OR LETTER

Only a ) .. (indicating a

range of values), a comma, a

space, or a letter can follow

an integer.

Replace the invalid char

acter following the integer

with a valid character.

22 POSITIONAL VALUES ARE

NOT ALLOWED IN A VALUE-

SET

In a KEY or SUM parameter

value-set, key or sum ﬁeld

characteristics must be speci

ﬁed in the order shown in the

parameter formats; values

cannot be omitted and replaced

by commas.

In a KEY parameter, specify

ﬁeld position, length and

type if you specify order;

specify ﬁeld position and

length if you specify type.

In a SUM parameter, specify

ﬁeld position, length and

type i f yo u specif y repet i

tions .

/■*S3tv

B-2 60484800 D

TABLE B-l. SORT/MERGE DIAGNOSTICS (Contd)

0^S

Jp>s

Error

Number

Error

Level Message Signiﬁcance Action

23 A PERIOD CANNOT BE USED

IN A PARAMETER-LIST

A period is not a valid sepa

rator in a parameter list. Replace the invalid period

with a valid separator;

refer to the parameter

syntax to determine valid

separators.

24 A RANGE IS ONLY ALLOWED

AS A VALUE FOR ENR, KEY

OR SUM PARAMETERS

A range of values can be speci

ﬁed only for the ENR, KEY, or

SUM parameters; single values

must be speciﬁed for other

parameters.

Replace the invalid range

with a single value.

25 UNLESS $CHAR(N) IS

USED, ONLY ONE SET OF

NESTED PARENTHESES IS

ALLOWED

More than two parentheses can

appear together only when

$CHAR(n) is used in a SEQS

parameter to specify steps in a

collating sequence.

Delete the extra

parentheses.

26 UNLESS WITHIN A STRING,

NEITHER . NOR = IS

ALLOWED WITHIN A VALUE

SET

A set of values speciﬁed for a

parameter can contain a period

or an = only in a literal

string.

Correct the syntax for the

indicated values.

27 THIS VALUE SHOULD BE A

FILE OR PROC NAME

The indicated value must be the

name of a ﬁle or procedure.

Use a ﬁle or procedure

name for the indicated

value.

29 PARENTHESES MUST BE

MATCHED IN THE

PARAMETER-LIST

If a right or left parenthesis

appears in a parameter list, it

must be matched in the list by

its opposite.

Insert the missing

parenthesis.

30 TOO FEW SETS GIVEN FOR

THE KEYWORD ENDED HERE

The indicated keyword must have

a speciﬁc set of values speci

ﬁed for it; a value or values

are missing.

Insert the values that must

be speciﬁed for the

indicated keyword.

31 TOO MANY VALUE-SETS

GIVEN FOR THIS KEYWORD-

VALUE-LIST

The indicated keyword must have

a speciﬁc set of values

speciﬁed for it; too many

values have been speciﬁed.

Determine the values that

can be speciﬁed for the

indicated keyword.

32 THIS VALUE SHOULD BE AN

INTEGER

The indicated value must be an

integer.

Replace the invalid value

with an integer value.

33 TOO MANY VALUES

SPECIFIED IN THIS

VALUE-SET

Too many values have been

speciﬁed for the indicated set

of values i n a l i s t of

parameter values.

Determine the values that

can be speciﬁed in the

indicated set of values.

34 TOO FEW VALUES

SPECIFIED IN THIS

VALUE-SET

Too few values have been speci

ﬁed for the indicated set of

values in a list of parameter

values.

Determine the values that

can be speciﬁed in the

indicated set of values.

35 ONLY A , ) .. OR SPACE

CAN FOLLOW A PARAMETER

VALUE

Parameter values must be

followed by a , ) .. or space.

Replace the invalid

character following the

parameter value with a

v a l i d c h a r a c t e r.

37 THIS PARAMETER VALUE

MUST BE A SET OR A

RANGE

The indicated parameter value

cannot be a single value; the

value must be a set of values

or a range of values.

Replace the indicated

parameter value with a

valid set or range of

values*

60484800 D B-3

TABLE B-l. SORT/MERGE DIAGNOSTICS (Contd)

Error

Number

Error

Level Message Signiﬁcance Action

50 DIALOG PARAMETER CAN

ONLY BE SPECIFIED IN A

CONTROL STATEMENT

The DIALOG parameter cannot be

in a directive ﬁle. Specify the DIALOG parame

ter in a control statement.

51 ONLY 3 VALUES CAN BE

GIVEN IN KEY/SUM VALUE-

SET IF F..L FORMAT IS

USED

Values in a KEY or SUM parame

ter value-set are separated by

commas or spaces. If you use

the ﬁrst..last form to specify

key or sum ﬁeld position and

length, a maximum of two other

values (separated by commas or

spaces) is allowed.

Correct the KEY or SUM

parameter so that no more

than three values are

speciﬁed*

152 FILE, PROC, AND COL-SEQ

NAMES MUST BEGIN WITH A

LETTER

A letter (A-Z) must be the

ﬁrst character in a ﬁle name,

procedure name, or the name of

a collating sequence you have

deﬁned.

Begin each ﬁle name, pro

cedure name, and collating

sequence name with a

letter.

53 VALUES ALLOWED FOR THE

FASTIO PARAMETER ARE

YES AND NO

Valid values for the FASTIO

parameter are YES or NO, which

can be speciﬁed as Y or N.

Specify a valid value for

the FASTIO parameter.

54 MORE THAN 100 INPUT

FILES GIVEN, ONLY THE

FIRST 100 SORTED

No more than 100 ﬁles can be

sorted in any one job step;

ﬁles speciﬁed in excess of

100 are ignored.

The ﬁles in excess of 100

can be sorted in a separate

job step.

56 IF THE FIRST VALUE OF A

SEQS IS A RANGE THEN

ALL OTHER VALUES OF

THAT SEQS MUST BE

RANGES OF THE SAME

LENGTH

Single characters cannot follow

a range of characters in a SEQS

parameter.

Correct the SEQS parameter

so that it consists of

ranges of characters with

the same number of charac

ters in each range.

57 WHEN LOW*.HIGH IS USED

AS A VALUE FOR SEQS LOW

MUST BE LESS THAN HIGH

If a range of characters is

used in a SEQS parameter, the

ﬁrst character in the range

must be of less value than the

last.

Correct the SEQS parameter

so that the ﬁrst character

in the range is less than

the last*

58 ONLY SINGLE CHARACTER

STRINGS OR $CHAR CAN BE

USED AS VALUES FOR SEQS

Each character speciﬁed in a

SEQS parameter must be enclosed

in apostrophes and separated

from other characters by .. or

a comma; characters cannot be

grouped together. $CHAR(n)

must specify a single charac

t e r .

Correct the SEQS parameter.

59 FOR SCHAR(N), N MUST BE

A POSITIVE DECIMAL

INTEGER

If $CHAR(n) is used in a SEQS

parameter to specify a value

step consisting of one charac

ter, n must be a positive deci

mal integer that corresponds to

the number of the intended

character in your character

s e t .

Specify n as a positive

decimal integer.

61 FILE NAMES MUST NOT

EXCEED 7 CHARACTERS IN

LENGTH

A ﬁle name can be no longer

than 7 characters.

Change the name of the ﬁle

to a name that is 7 charac

ters or fewer in length.

/=s^\

r ^ ^ B \

B-4 60484800 D

TABLE B-l. SORT/MERGE DIAGNOSTICS (Contd)

^ms. Error

Number

yam^s

0^^

J^S

Error

Level Message

THE MAXIMUM LENGTH OF A

LINE IS 240 CHARACTERS

VALUES ALLOWED FOR THE

DIALOG PARAMETER ARE

YES AND NO

A SPACE OR , MUST

FOLLOW THE SORT OR

MERGE VERBS IN DIR.

FILE

A PERIOD MUST FOLLOW

SORT5 OR MERGE

DIRECTIVE FILE NAMES

MUST BE UNIQUE

LINES CONTINUING THE

SORT5 OR MERGE

PARAMETER-LIST MUST

BEGIN WITH A PERIOD

FILE NAMES CAN CONTAIN

ONLY LETTERS AND DIGITS

LENGTH MUST BE

SPECIFIED IF TYPE

SPECIFIED

KEY LENGTH FOR KEYTYPE

REAL MUST BE 10

ONLY THE VALUES A OR D

ARE ALLOWED AS VALUES

FOR KEY ORDER

UNEQUAL FL FOR INPUT

AND OUTPUT FILES WITH

RT=F OR RT=Z

VALUES ALLOWED FOR THE

EL PARAMETER ARE T, W,

F OR C

VALUES ALLOWED FOR THE

LO PARAMETER ARE OFF, S

AND A

Signiﬁcance

The SORT5 or MERGE control

statement and its associated

parameter list can be no longer

than 240 characters (periods

used as continuation indicators

are not counted).

Valid values for the DIALOG

parameter are YES or NO, which

can be speciﬁed as Y or N.

In a directive ﬁle, any state

ment that begins with SORT or

with MERGE must have a space or

a comma immediately following

SORT or MERGE.

In a control statement, the

word SORT5 or the word MERGE

must be immediately followed by

a period.

A directive ﬁle cannot have

the same name as any other

directive ﬁle used in the sort

or merge.

Any line of a SORT5 or a MERGE

control statement that is con

tinued beyond the ﬁrst line

must begin with one period.

A ﬁle name cannot contain any

special characters.

If the key type is other than

the default ASCII6, or if

ASCII6 is explicitly speciﬁed,

key length cannot be omitted.

Keys declared to be of type

REAL must occupy a full compu

ter word and must be speciﬁed

as 10 bytes in length.

Sort order can be speciﬁed as

ascending with A or as descend

ing with D.

If you specify different ﬁxed

length for input and output

ﬁles, Sort/Merge compresses or

expands records to ﬁt into the

ﬁxed length.

The only values that can be

speciﬁed for the EL parameter

are T, W, F, or C.

The only values that can be

speciﬁed for the L0 parameter

are OFF, S, or A.

Action

Parameters can be placed in

a directive ﬁle so that

the control statement is

240 characters or less.

Specify a valid value for

the DIALOG parameter.

Insert a space or a comma

immediately following SORT

or MERGE in your directive

ﬁle.

Insert a period immediately

following the word S0RT5 or

the word MERGE.

Give all directive ﬁles

different names.

Begin each continuation

line with one period.

Omit special characters

from the ﬁle name.

Specify key length.

Specify length of REAL key

as 10.

Specify either A or D for

sort order; or omit speci

ﬁcation of sort order and

ascending is assumed.

For the most efﬁcient pro

cessing, specify the same

FL for input and output

ﬁles.

Specify T, W, F, or C as

the value of the EL

parameter.

Specify OFF, S, or A as the

value of the LO parameter.

60484800 D B-5

TABLE B-l. SORT/MERGE DIAGNOSTICS (Contd) jt^s.

Error

Number

Error

Level Message Signiﬁcance Action

76 VALUES ALLOWED AS

EXPRESSIONS ARE

INTEGER, Rl, R2, R3,

RIG, EF OR EFG

The only values that can be

speciﬁed as expressions are

integers or the CCL variables

Rl, R2, R3, RIG, EF, or EFG.

Specify a valid expression.

77 SNULL SHOULD NOT BE

USED AS A VALUE WITHIN

A SET OF VALUES

SNULL is a reserved ﬁle name

signifying a null ﬁle; it must

not be speciﬁed in a list of

multiple ﬁles. $NULL can be

speciﬁed as a single value for

a parameter.

Delete $NULL from the list

of ﬁles, or specify $NULL

as the single parameter

value.

78 VALUES ALLOWED FOR THE

ST PARAMETER ARE Rl,

R2, R3, RIG, EF OR EFG

The only values that can be

speciﬁed for the STATUS param

eter are the CCL variables Rl,

R2, R3, RIG, EF, or EFG.

Specify a valid value for

the STATUS parameter.

79 VALUES ALLOWED FOR THE

RETAIN PARAMETER ARE

YES AND NO

The only values that can be

speciﬁed for the RETAIN param

eter are YES or NO, which can

be abbreviated as Y or N.

Specify a valid value for

the RETAIN parameter.

80 OWNMRL AND OWNFL SHOULD

NOT BOTH BE SPECIFIED

FOR THE SAME SORT

In a sort with owncode rou

tines, record length must be

speciﬁed with either the

OWNMRL parameter or the OWNFL

parameter.

Specify record length with

the OWNFL parameter for F

or Z type records; specify

record length with the

OWNMRL parameter for all

other record types.

81 SUM AND KEY FIELDS MAY

NOT OVERLAP

The same characters in a record

cannot be included in a sum

ﬁeld and a key ﬁeld.

Correct SUM parameter and/

or KEY parameter so that

the same characters are not

included in both ﬁeld

descriptions.

82 LINES IN A DIRECTIVE

FILE MUST BE NO LONGER

THAN 100 CHARACTERS IN

LENGTH

A directive ﬁle statement,

beginning with SORT or MERGE,

can be 240 characters in

length, but each line in the

statement must contain no more

than 100 characters.

Use continuation lines in

your directive ﬁle so that

each line contains no more

than 100 characters.

83 REPETITION VALUE OF A

SUM FIELD MUST BE AN

INTEGER

In a SUM parameter, the number

of ﬁelds to be summed must be

expressed as an integer value.

Specify number of ﬁelds to

be summed as an integer.

84 VALUES ALLOWED FOR OWNT

PARAMETER ARE OLD AND

NEW

The only values that can be

speciﬁed for the OWNT param

eter are OLD or NEW.

Specify OLD or NEW as the

value for the OWNT param

eter.

85 IN A KEY/SUM VALUE-SET

USING FIRST..LAST,

LAST IS LESS THAN FIRST

In a KEY or SUM parameter, if

ﬁeld length and position are

speciﬁed as the ﬁrst and last

bytes or bits of the ﬁeld then

the ﬁrst byte or bit must be

less than the last byte or bit

(the leftmost byte or bit in a

record is counted as number 1).

Correct the KEY or SUM

parameter.

86 NUMERIC KEY TYPES ARE

THE ONLY TYPES VALID

FOR SUM FIELDS

In a SUM parameter, sum ﬁeld

type must be the name of a num

eric data format except REAL;

type cannot be a collating

sequence.

Correct the SUM parameter.

B-6 60484800 D

j0ms

00s,

TABLE B-l. SORT/MERGE DIAGNOSTICS (Contd)

Error

Number

Error

Level

Message

LINES OF A DIR FILE

MUST BEGIN WITH SORT OR

MERGE VERB, OR BE A

CONTINUATION

BYTE/BIT POSITION OR

LENGTH OF A SUM/KEY

FIELD MUST BE AN

INTEGER

SEQN VALUE MUST NOT BE

ONE OF THE PREDEFINED

KEY-TYPES

SEQR AND SEQA MAY BE

SPECIFIED ONLY ONCE FOR

EACH COLLATING SEQUENCE

SEQN MUST BE SPECIFIED

BEFORE ANY OTHER SEQ

PARAMETER CAN BE GIVEN

VALUES ALLOWED FOR SEQR

AND SEQA PARAMETERS ARE

YES AND NO

SEQN SHOULD NOT BE

GIVEN WITHOUT GIVING

SEQS OR SEQR BEFORE THE

NEXT SEQN

VALUES ALLOWED FOR SEQS

PARAMETER ARE STRINGS

AND $CHAR(N)

TWO RANGES OF

CHARACTERS USED IN THE

SAME SEQS MUST BE THE

SAME LENGTH

Signiﬁcance

I n a d i r e c t i v e ﬁ l e, e ac h li n e

must begin with the word SORT

or MERGE or, if the line is a

continuation line, with a char

acter of your parameter list or

one or more spaces.

In a SUM or KEY parameter,

ﬁeld length and position must

be speciﬁed as integers,

counting the leftmost byte or

bit in a record as number 1.

When you are deﬁning your own

collating sequence with SEQx

parameters, the name of your

collating sequence cannot be

the same as the name of one of

the predeﬁned collating

sequences.

When you are deﬁning your own

collating sequence with SEQx

parameters, SEQR and SEQA must

not be speciﬁed more than

once.

When you are deﬁning your own

collating sequence with SEQx

parameters, the SEQN parameter

naming your collating sequence

must be the ﬁrst SEQx

parameter.

The only values that can be

speciﬁed for the SEQR and SEQA

parameters are YES or NO, which

can be abbreviated as Y or N.

When you are deﬁning your own

collating sequence with SEQx

parameters, you must specify

the characters in your col

lating sequence with SEQS and

SEQR parameters before naming

another collating sequence.

When you are deﬁning your own

collating sequence with SEQx

parameters, the only valid

values for SEQS parameters are

characters enclosed in apostro

phes or $CHAR(n) where n is the

number of the character in your

character set.

When you are deﬁning your own

collating sequence with SEQx

parameters, the ranges in any

one SEQS parameter must all be

the same length.

Action

Correct the directive ﬁle;

determine the proper format

for directive ﬁles.

Correct the SUM or KEY

parameter. (Make sure two

sets of parentheses enclose

the value-set.)

Give your collating

sequence a unique name that

is not the same as one of

the predeﬁned collating

sequences.

Specify SEQR and SEQA only

once.

Specify SEQN before you

specify any other SEQx

parameter.

Specify a valid value for

the SEQR and SEQA param

eters.

Correct the order of the

SEQx parameters.

Correct the SEQS param

eters.

Correct the SEQS parameter

so that it speciﬁes ranges

that are all the same

length.

60484800 D B-7

TABLE B-l. SORT/MERGE DIAGNOSTICS (Contd)

Error

Number

Error

Level Message Signiﬁcance Action

96 VALUE GIVEN IN THE

$CHAR FUNCTION MUST BE

AN INTEGER

When you are deﬁning your own

collating sequence with SEQx

parameters and are using

$CHAR(n) to specify the charac

ters in the sequence, n must be

a decimal integer corresponding

to the number of the intended

character in your character set.

Specify n of $CHAR(n) as a

decimal integer.

97 CHARACTERS GIVEN IN A

SINGLE COLLATING

SEQUENCE MUST APPEAR

ONLY ONCE

When you are deﬁning your own

collating sequence with SEQx

parameters, a character cannot

appear more than once in all

the SEQS parameters.

Specify each character only

once in a collating

sequence.

98 COL-SEQUENCE NAMES

LONGER THAN 10

CHARACTERS ARE

TRUNCATED

You can specify a collating

sequence name of up to 31 char

acters in length, but only the

ﬁrst 10 characters are sig

niﬁcant.

None.

99 NUMBER OF KEY AND/OR

SUM CHARACTERS MUST BE

LESS THAN 256

A record cannot contain more

than 255 characters that are

deﬁned as key or sum ﬁelds.

Specify the total number

of key and sum characters

together as 255 or less.

100 SIGNED NUMERIC CHARAC

TER FIELD IS TOO SHORT

A key ﬁeld deﬁned by a KEY

parameter or SM5KEY call as

one of the signed numeric data

formats is not long enough to

contain the data.

Redeﬁne the key ﬁeld as

long enough to contain all

the numeric characters and

the sign.

101 NO PARAMETER CAN BE

SPECIFIED MORE THAN

ONCE ON ONE LINE

Parameters that can be speciﬁed

more than once must be on separ

at e l i ne s .

Inspect parameters and

correct error(s).

102 DUPLICATE PARAMETER

SPECIFIED

Some parameters can only be

speciﬁed once.

Inspect parameters and

correct error(s).

103 EITHER AN INPUT FILE,

OWN1 OR OWN2 MUST BE

GIVEN

No sort input ﬁle given. Specify a sort input ﬁle,

or 0WN1 or 0WN2.

104 EITHER AN OUTPUT FILE,

OWN3 OR OWN4 MUST BE

GIVEN

No sort output ﬁle given. Specify a sort output ﬁle,

or OWN1 or OWN 2.

107 INCORRECT PARAMETER

VALUE

A value speciﬁed as a proce

dure call parameter is not a

valid value*

Check values that can be

speciﬁed for the procedure

call parameter.

108 TOO FEW/MANY PARAMETERS Either too few or too many

parameters have been speciﬁed

in a procedure call.

Determine parameters that

can be speciﬁed in the

procedure call*

UO CALL OUT OF SEQUENCE A call to a Sort/Merge FORTRAN

procedure is not in the proper

o r d e r.

Determine procedure call

o r d e r.

111 INVALID NAME GIVEN FOR

SM5SEQN

The name of a user-deﬁned col

lating sequence speciﬁed with

the SM5SEQN call must begin

with a letter and consist of

letters, digits, and the

special characters ? # @ _.

Correct the name speciﬁed

with the SM5SEQN call;

enclose the name in apos

trophes.

,*"*S^.

B-8 60484800 D

TABLE B-l. SORT/MERGE DIAGNOSTICS (Contd)

Error Error

Number Level Message Signiﬁcance Action

112 ONLY SINGLE CHARACTER Each character in a user- Specify each character in

LITERALS CAN BE USED AS deﬁned collating sequence your collating sequence as

PARAMETERS OF SM5SEQS speciﬁed as parameters of

the SM5SEQS procedure must be

speciﬁed separately, enclosed

in apostrophes, and separated

from other characters by

commas.

a separate literal enclosed

in apostrophes.

113 RECORD LENGTH IS Record length is set by the full If subsequent errors abort

GREATER THAN MAXIMUM length (FL) or maximum record the sort, reduce the

RL - 5000 length (MRL) ﬁle information

table (FIT). Also, the OWNFL

and OWNMRL parameters or SM50FL

and SM50MRL procedures specify

the length of a record. 5000

is the maximum record length

supported; records that exceed

5000 characters may generate

subsequent errors that may

abort the sort.

number of characters per

record to 5000 or less.

114 SUM FIELDS MAY NOT Summing is handled with integer Correct the speciﬁcation

EVALUATE TO MORE THAN a r i t h m e t ic . I n te ge r a n d b i n a ry of the sum ﬁeld.

60 BITS sum ﬁelds can be no more than

60 bits long.

115 RETAIN AND SUMMING MAY You cannot specify both the Remove the RETAIN or SUM

NOT BE SPECIFIED AT THE RETAIN and the SUM parameter in parameter*

SAME TIME the same sort or merge.

116 MRL/FL IS NOT LARGE The record size (MRL or FL) is Change your key or ﬁle

ENOUGH TO ACCOMMODATE not large enough to accommodate description.

THE SPECIFIED KEYS the speciﬁed keys.

117 FASTIO PARAMETER

DISABLED FOR FILES NOT

ON MASS STORAGE

The FASTIO parameter cannot

process tapes.

None necessary.

118 FASTIO PARAMETER WAS FASTIO ﬁles must have the Recheck ﬁle

SPECIFIED BUT COULD NOT same record type and length, speciﬁcations.

BE PROCESSED and must reside on mass

storage.

119 FOR KEY AND SUM FIELDS, The beginning position and Correct the key/sum

THE BEGINNING POSITION length of key and sum ﬁelds speciﬁcation.

AND THE LENGTH MUST BE must be non-zero.

NON-ZERO.

120 SUMMING OVERFLOW Sum ﬁelds in t h e o u t put ﬁ l e Check data and summing

OCCURRED - SEE DAYFILE have invalid data. s p e c i ﬁ c a t i o n s .

FOR DETAILS.

121 UNABLE TO LOAD USER

OWNCODE - SORT DID

NOT COMPLETE.

A fat al l oader error occurre d. Check your owncode ﬁles.

122 OWNCODE WAS SPECIFIED

BUT NO OWNF FILE WAS

GIVEN.

OWNF parameter was omitted. Specify the OWNF parameter.

151 NOT SORT, MERGE OR In the SMFILE call, ﬁle dis None, or change SMFILE

OUTPUT; 'SORT' ASSUMED position must be indicated as

SORT, MERGE, or OUTPUT.

call to indicate MERGE or

OUTPUT.

60484800 D B-9

TABLE B-l. SORT/MERGE DIAGNOSTICS (Contd)

Error

Niraber

Error

Level Message Signiﬁcance Action

152 RETURNED FROM OWNCODE

WITHOUT CALLING SMRTN

Control has returned to Sort/

Merge from an owncode routine

without using a call to SMRTN;

processing terminates.

Determine the proper method

of calling SMRTN.

153 TOO MANY ERRORS FOUND.

PROCESSING TERMINATED.

The maximum number of errors

for one call to Sort/Merge have

occurred; processing terminates.

Correct errors and resubmit

job.

154 SMRTN CALL ILLEGAL WITH

SM5 PROCEDURES - RUN

ABORTED.

The Sort/Merge 4 compatible

routine SMRTN has been called

by a set of owncode procedures

using SM5xxx calls.

Change SMRTN to RETURN or

the equivalent.

155 RETURN CODE MUST BE IN

RANGE 0-3

The return code from all own-

code routines must be 0, 1, 2,

or 3.

Change the value of the

return code.

156 AN INPUT RECORD WAS

READ WHICH WAS TOO

SHORT TO HOLD THE

KEY/SUM FIELDS

SPECIFIED FOR THIS

SORT/MERGE.

The order of the records is

undeﬁned if the key ﬁelds

are not within the record.

Change your key

description

157 NEITHER SORT NOR MERGE

HAS BEEN SPECIFIED

You have omitted a call to

SM5SORT or SM5MERGE.

Call SM5SORT or

SM5MERGE before any

other SMS procedures.

158 FILE NAMES FOR A MERGE

MUST BE UNIQUE.

Duplicate ﬁle names exist for

TO/FROM parameters.

Eliminate the duplicate

ﬁle names.

159 KEY FIELDS OVERLAP

OTHER KEY FIELDS.

The same characters in a record

are included in two or more key

ﬁelds.

None.

160 ALTER OPTION CANNOT

BE USED FOR OVERLAPPING

KEY FIELDS.

When you deﬁne your own col

lating sequence use the SEQA

parameter or the SM5SEQA proc

edure, you cannot have over

lapping key ﬁelds.

Redeﬁne the key ﬁelds

so they do not overtop.

161 SUM FIELDS CANNOT OVER

LAP OTHER SUM FIELDS

The same characters in a record

cannot be included in more than

one sum ﬁeld.

Redeﬁne the sum ﬁelds

so that they do not

overlap.

• B-10 60484800 D

GLOSSARY

0ms

Th is glossary deﬁnes terms unique to the des crip

tion of Sort/Merge; common terms that have special

connotations within the context of this manual are

also included.

Ascending -

The order of sorting keys so that the record

having the numeric key with the highest value

is written last on the output ﬁle, and charac

ter k eys a r e so r t ed a c c ording to t h e sp e c iﬁed

collating sequence. See Sort Order.

ASCII -

American National Standard Code for Information

Interchange. An 8-bit code representing a

pre scribed set of 12 8 cha racters. Contr ol Da t a

operating systems use a 6-bit display code to

represent a subset (called ASCII6) of these

characters.

Basic Access Methods (BAM) -

A ﬁle manager that processes sequential and

word addressable ﬁle organizations. See CYBER

Record Manager.

Beginning-of-information (BOI) -

The start of your ﬁrst record in a ﬁle.

System information, such as tape labels of

sequential ﬁles, can appear before beginning-

o f-informa 11on.

Block -

A logical or physical grouping of records.

CYBER Record Manager deﬁnes block types I, C,

K, and E for sequential ﬁles.

Byte -

A group of bits. In the body of this manual, a

byte is six bits that represents a single

character.

Character -

A letter, digit, punctuation mark, or mathe

matical symbol forming part of one or more of

the standard character sets.

Collating Sequence -

A sequence in which the characters that are

acceptable to a computer are ordered for

purposes of sorting, merging, and comparing.

Control Statement Section -

The ﬁrst section of a job; contains the

sequence of control statements that speciﬁes

all steps for job execution.

Control Statement Sort -

A sort performed on the basis of parameters in

the S0RT5 control statement and in optional

d i r e c t i v e ﬁ l e s .

CYBER Control Language (CCL) -

A language that allows you to insert a set of

statements into the control statement section

of your job; provides a means for you to

determine the status of ﬁles, initiate tests

and transfers within the control statement

section, and display results in your job

dayﬁle.

CYBER Record Manager (CRM) -

A generic term relating to the common products

Advanced Access Methods and Basic Access

Methods that run under the NOS and NOS/BE

operating systems and that allow a variety of

record types, blocking types, and ﬁle

organizations to be created and processed. The

execution time input/output of COBOL, FORTRAN,

Sort/ Merge, ALGOL, and the DMS-170 products is

implemented through CYBER Record Manager. The

input/output of the NOS and NOS/BE operating

system utilities such as COPY or SKIPF are not

implemented through CYBER Record Manager. All

CYBER Record Manager ﬁle processing requests

ultimately pass through the operating system

input/output routines.

Default -

A value supplied by Sort/Merge when you omit a

spe c iﬁc a tio n fr o m a pa r ame t er l ist .

Descending -

The order of sorting keys so that the record

having the numeric key with the lowest value is

written last on the output ﬁle, and character

keys are sorted in reverse order of the speci

ﬁed collating sequence. See Sort Order.

Dialog -

An interactive aid designed to guide an in

experienced user through a sort or merge

speciﬁcation.

Directive File -

A ﬁle that can be used in a control statement

sort or merge; contains processing parameters

in a d di ti o n to th e p ar a m e te rs sp ec i ﬁe d i n t h e

S 0 RT 5 o r M E R G E co n t r o l s t at e me n t. Th i s ﬁ l e

must be a sequential ﬁle of block type C and

record type Z.

Display Code -

A 6-bit code representing a 63-character or

64-character computer character set.

EBCDIC -

The abbreviation for extended binary-coded

decimal interchange code. Control Data oper

a t i n g s y s t e m s u s e a 6 - b i t v e r s i o n ( c a l l e d

EBCDIC6) to represent this code.

End-of-information (EOI) -

CYBER Record Manager deﬁnes end-of-informat Ion

Ln sequential ﬁles In terms of the ﬁle resi

dence, as shown in table C-l.

Entry Point -

A locati o n with i n a program that can b e refer

enced from other programs. Each entry point

ha8 a unique name with which it is associated.

60484800 B C-l

TABLE C-l. END-OF-INFORMATION BOUNDARIES

File

Organization

File

Residence Physical

Position

Sequential

Mass storage

Labeled tape

in S I , I, S,

or L format

Unlabeled

tape in SI

or I format

Unlabeled

tape in S or

L format

After the last

user record*

After the last

user record and

before any ﬁle

trailer labels.

After the last

user record and

before any ﬁle

trailer labels.

Undeﬁned.

F i l e -

A collection of records treated as a unit;

represented by a logical ﬁle name and delim

ited by beginning-of-information and end-of-

information.

FILE Control Statement -

A control statement that supplies ﬁle informa

tion table (FIT) values after a source language

program is compiled or assembled but before the

program is executed. Basic ﬁle characteris

tics such as organization, record type, and

d e s c r i p t i o n c a n b e s p e c i ﬁ e d i n t h e F I L E

control statement.

File Information Table (FIT) -

A table through which your program communicates

wit h B A M . Al l ﬁl e pr o ces s ing ex e cute s o n th e

basis of information in the FIT. You can set

FIT ﬁelds directly or use parameters in a ﬁle

ac c es s c a ll th a t s et s t h e ﬁ el d s i nd ir e ct l y.

Floating-Point Number -

N u m e r i c d a t a s t o r e d i n t e r n a l l y a s b i n a r y

values (corresponds to a FORTRAN REAL or COBOL

COMPUTATIONAL-2 number).

Graphic -

A character that can be printed or displayed.

Integer -

Numeric data stored internally as a binary

value rather than a character value; in a

parameter or procedure call format, a number

that does not include any character positions

to the right of the assumed decimal point and

is not signed.

I n t e r a c t i v e -

Job processing in which you and the system

communciate with each other, rather than your

submitting a job and receiving output, having

no control over the job while processing occurs.

Interactive Facility (IAF) -

The network host software product that allows

you to enter commands and to communicate with

a n e x e c u t i n g p r o g r a m f r o m a t i m e - s h a r i n g

terminal; operates under NOS.

INTERCOM -

The software that directs the ﬂow of informa

tion and data between the central site computer

and a number of remote terminals; operates

under NOS/BE.

Job -

A set of control statements and the data and

directives used by those control statements. A

batch job begins with the job, USER, and CHARGE

statements on NOS and the job and ACCOUNT

statements on NOS/BE. An interactive job

begins with login to a terminal.

Key Type -

The name of a numeric data format or collating

sequence.

Keyword -

A part of a parameter that you enter exactly as

shown in the parameter format. It is not

changed by any information you supply.

Library -

A collection of programs or routines that is

searched by the loader for entry points refer

enced by a program.

Literal -

A constant completely deﬁned by its own

identity.

Local File -

A ﬁle currently assigned to a job, or a

temporary ﬁle other than the primary ﬁle;

often contains a copy of an indirect access

ﬁle or data from a magnetic tape.

Logical File Name (lfn) -

The one to seven coded letters or digits by

which the operating system recognizes a ﬁle.

Every logical ﬁle name in a job must be unique

and begin with a letter.

Logical Record -

Under NOS, a data grouping that consists of one

or more PRUs terminated by a short PRU or

z e r o - l e n g t h P R U . E q u i v a l e n t t o a s y s t e m -

logical-record under NOS/BE.

The process used at a terminal to gain access

to the system.

Logout -

The process used to end a terminal session.

Major Sort Key -

A s o rt key th a t i s c ons i de r ed to be t he mo s t

important and is speciﬁed ﬁrst. See Sort

Key, Minor Sort Key.

Mass Storage -

A disk pack or other rotating mass storage

device; not a magnetic tape.

r^^S

,A*ms

C-2 60484800 B

0^S Merge -

The process of combining two or more presorted

ﬁles.

Minor Sort Key -

A sort key that is considered of lesser impor

tance and is speciﬁed after the most important

sort key. See Sort Key, Major Sort Key.

Owncode Routine -

A subroutine you write that provides the capa

bility of inserting, substituting, modifying,

or deleting input and output records during

Sort/Merge processing.

Parameter -

A variable identiﬁed by a keyword and assigned

speciﬁc values in the S0RT5 or MERGE control

statement or in a directive ﬁle; parameters

and their speciﬁed values direct Sort/Merge

processing.

Partition -

CYBER Record Manager deﬁnes a partition as a

division within a ﬁle with sequential orga

nization. Generally, a partition contains

several records or sections. Implementation

o f a p a r t i t i o n b o u n d a r y i s a f f e c t e d b y ﬁ l e

structure and residence, as shown in table C-2.

N o t i c e t h a t i n a ﬁ l e w i t h W t y p e r e c o r d s , a

short physical record unit (PRU) of level 0

terminates both a section and a partition.

Physical Record Unit (PRU) -

Under NOS and NOS/BE, the amount of information

transmitted by a single physical operation of a

speciﬁed device. The size of a PRU depends on

the device, as shown in table C-3.

A PRU that is not full of user data is called a

short PRU; a PRU that has a level terminator

but no user data is called a zero-length PRU.

Procedure -

A subroutine that returns values through its

parameters; invoked when the name of the

procedure is referenced in a CALL or ENTER

statement.

Record -

CYBE R Record Manager deﬁnes a rec o r d as a

group of related characters. A record or a

portion thereof is the smallest collection of

information that can be passed between CRM and

your program. Eight different record types

exist, as deﬁned by the RT ﬁeld of the ﬁle

information table.

Record Type -

The term record type can have one of several

m e a n i n g s , d e p e n d i n g o n t h e c o n t e x t . C Y B E R

Record Manager deﬁnes eight record types

established by an RT ﬁeld in the ﬁle infor

mation table. Tables output by the loader

are classiﬁed as record types such as text,

relocatable, or absolute, depending on the

ﬁrst few word8 of the tables.

Relocatable -

An object program that can reside in any part

of central memory. The actual starting address

is established at load time.

TABLE C-2. PARTITION BOUNDARIES

Device

PRU

device

Record

Type

(RT)

Block

Type

(BT)

S or L

format

tape

D, F,R ,

T,U,Z

•Any

other

tape

format

D , F, T,

R,U,Z

Physical Boundary

A short PRU of

level 0 containing

a one-word deleted

record pointing

back to the last I

block boundary,

followed by a con

trol word with a

ﬂag indicating a

partition boundary.

A short PRU of

level 0 containing

a control word

w i t h a ﬂ a g i n d i

cating a partition

boundary.

A short PRU of

level 0 followed by

a zero-length PRU

of level 17g.

A zero-length PRU

of level number 17g.

C,K,E

A separate tape

block containing as

many deleted rec

ords of record

length 0 as re

quired to exceed

noise record size,

followed by a de

leted one-word

record pointing

back to the last I

block boundary,

followed by a con

trol word with a

ﬂag indicating a

partition boundary.

A separate tape

block containing as

many deleted rec

ords of record

length 0 as re

quired to exceed

noise record size,

followed by a con

trol word with a

ﬂag indicating a

partition boundary.

A tapemark.

Undeﬁned.

y$mt.

60484800 B C-3

TABLE C-3. PRU SIZES TABLE C-4. SECTION BOUNDARIES

Device Size in Number

of 60-Bit Words

Mass storage.

Tape in SI format with

binary data (NOS/BE only).

Tape in I format (NOS only).

Tape in any other format.

512

Undeﬁned.

Result Array -

A means of passing statistics and results of a

sort or merge back to a calling procedure.

Rewind -

An operation that positions a ﬁle at the

beginning-of-information.

Section -

CYBER Record Manager deﬁnes a section as a

d i v i s i o n w i t h i n a ﬁ l e w i t h s e q u e n t i a l o r g a

nization. Generally, a section contains more

than one record and is a division within a

partition of a ﬁle. A section terminates

w i t h a p h y s i c a l r e p r e s e n t a t i o n o f a s e c t i o n

boundary, as shown in table C-4.

The NOS and NOS/BE operating systems equate a

section with a system-logical-record of level 0

through 16g.

Sequential File -

A ﬁle with records in the physical order in

which, they were written. No logical order

exists other than the relative physical record

position.

Signed Numeric Data -

Integer data stored internally in display code;

sorted according to numeric order and sign of

the integer the display code represents.

Sort Key -

A ﬁeld of information within each record in a

sort or merge input ﬁle that is used to deter

mine the order in which records a re writt en to

t h e o u t p u t ﬁ l e .

Sort Order -

The order for sorting keys, either ascending or

descending.

System-Logical-Record -

Under NOS/BE, a data grouping that consists

of one or more PRUs terminated by a short PRU

or zero-length PRU. These records can be

transferred between devices without loss of

structure. Equivalent to a logical record

under NOS.

Word -

A g r o u p o f b i ts ( o r 6 - b i t ch a r a c t er s) b e t we e n

boundaries imposed by the computer system. A

word is 60 bits in length (10 characters).

Device

Record

Type

(RT)

Block

Type

(BT)

Physical

Representation

PRU A deleted one-word

device record pointing

back to the last I

block boundary

followed by a con

tro l wor d with

ﬂags indicating a

section boundary.

At least the con

trol w o r d is i n

a short PRU of

l e v e l 0 .

WCA control word with

ﬂags indicating a

section boundary.

The control word

is in a short PRU

of level 0.

D,F,R, A short PRU with a

T,U,Z leve l less than

"8.

S-Undeﬁned.

S or L A separate tape

format block containing

tape as many deleted

records of record

length 0 as re

quired to exceed

noise record size,

followed by a de

leted one-word

record pointing

back to the last I

block boundary,

followed by a con

tro l wor d with

ﬂags indicating a

section boundary.

W C A separate tape

block containing

as many deleted

records of record

length 0 as re

quired to exceed

noise record size,

followed by a con

trol word with

ﬂags indicating a

section boundary.

D,F,R, C,K,E Undeﬁned.

T, U , Z

SUndeﬁned.

Any Undeﬁned.

other

tape

format

C-4 60484800 B

0&S.

SORT/MERGE PARAMETER SUMMARY

The format speciﬁcations for all parameters used

in a S0RT5 or a MERGE control statement are sum

marized and listed in this appendix. Detailed

information for each parameter is referenced by

page number and appears in section 3 of this manual.

The following abbreviations are used in this

appendix:

char character

expr expression

lfn logical ﬁle name

proc procedure name

rep repetition

DIALOG Parameter

DIALOG=YES (or DIA=Y)

DIALOG=NO (or DIA=N)

Page

3-4

FROM Parameter

FROM=lfn

F R O M = ( l f n 1 , l f n 2 , . . . )

FROM=$NULL

KEY Parameter

KEY= j(key_def[,key_def]

L range

where:

..,-,

Page

3-2

[range "j

( r a n g e t , t y p e [ , a d ] ] )

( ﬁ r s t , l e n g t h [ , t y p e [ , a d ] ] ) J

range = ["ﬁrst "J

|_ﬁrst..lastj

rDIR Parameter

DIR=lfn

DIR»(lfnlflfn2i...)

3-3 L Parameter

L = l f n

L=$NULL

3-3

E Parameter

E=lfn

E=$NULL

EL Parameter

EL=T

EL=W

EL=F

EL=C

ENR Parameter

ENR=expr

ENR=expr .. expr

3-4

OWNF Parameter

OWNF=lfn

OWNFL Parameter

OWNFL=integer

0FL=integer

OWNMRL Parameter

OWNMRL=integer

OMRL=integer

OWNn Parameter

OWNn=proc

3-4

3-5

FASTIO Parameter

FASTIO=YES (or FASTIO=Y)

FASTIO=NO (or FASTIO=N)

3-9 RETAIN Parameter

RETAIN=YES (or RET=Y)

RETAIN=NO (or RET=N)

3-5

60484800 C D-l

SEQx Parameters

SEQN=name

SEQS=('char',...,'char')

Page

3-5

Page

where:

sum_def = T(ﬁrst,length,type[ ,rep] )"]

L(ﬁrst..last,type[,rep]) J

[SEQR=YES OR SEQR-Y]

' [SEQA=YES OR SEQA=Y]

STATUS Parameter

STATUS=variable (or ST=variable)

SUM Parameter

SUM=((sum_def)[,(sum_def)]...)

3-7

3-8

TO Paraaeter

TO=lfn

TO=$NULL

VERIFY Parameter

VERIFY=YES (or VER=Y)

VERIFY=NO (or VER=N)

3-2

3-9

■^S%s

A^^S

D-2 60484800 C

0^S

INTERACTIVE DIALOG

This appendix contains the text of the interactive

dialog (ﬁgure E-l) that is invoked when you spec

ify DIALOGUES. (Your terminal must be in normal

mode; enter NORMAL before entering the S0RT5 or

MERGE command. If you are already in the dialogue,

e n t e r a c o n t r o l - T t o r e t u r n t o t h e o p e r a t i n g

system). The questions asked by Sort/Merge are

indicated by uppercase letters. The answers you

s u p p l y a r e i n d i c a t e d b y l o w er c a s e l e t t e r s ; " n e x t "

or a number following an answer indicates the next

question that is asked by Sort/Merge. For

questions requiring yes or no answers, you can

abbreviate and use y or n.

IS YOUR TERMINAL ALREADY SET TO NORMAL MODE OR UPPERCASE ASCII MODE. ALSO, ARE YOU SURE THAT ALL YOUR

FILES (IF ANY) ARE EITHER STANDARD UNIT RECORD FORMAT OR HAVE HAD FILE CONTROL STATEMENTS SPECIFIED FOR

THEM

yes (1)

no (next)

UNLESS YOU SPECIFY OTHERWISE, ALL INPUT FILES AND THE OUTPUT FILE ARE ASSUMED TO HAVE STANDARD UNIT

RECORD FORMAT SUCH AS LINES TYPED FROM A TERMINAL, LINES IN A BATCH JOB, OR LINES TO BE PRINTED ON A

PRINTER OR TERMINAL. THESE ARE TECHNICALLY KNOWN AS BT=C, RT=Z, FL=N WITH N NO MORE THAN 150. DO YOU

HAVE ANY FILES THAT ARE NOT OF THIS FORMAT?

yes (next)

n o ( 1 )

FOR EACH FILE THAT IS NOT OF THIS FORMAT YOU MUST HAVE SPECIFIED A FILE CONTROL STATEMENT. SORT/MERGE

WILL QUIT NOW SO YOU MAY DO SO.

1. DO YOU HAVE JUST ONE INPUT FILE AND NO 0WN1 OR 0WN2 SUBROUTINES?

yes (next)

n o ( 6 )

2. WHAT IS THE NAME OF YOUR INPUT FILE?

ﬁle-name

3. DO YOU HAVE ANY OWN3 OR 0WN4 SUBROUTINES?

yes (9)

no (next)

4. WHAT IS THE NAME OF YOUR OUTPUT FILE?

ﬁle-name

5. THE STANDARD COLLATING SEQUENCE IS ASCII6. HERE IS THE WHOLE ASCII6 COLLATING SEQUENCE LISTED

."#$%&' ()*+,-./0123456789:;<=>?aABCDEFGHIJKLMN0PQRSTUVWXYZC\D^_

DO ALL OF YOUR KEYS CONSIST OF CHARACTER DATA TO BE SORTED ACCORDING TO THE ASCII6 COLLATING

SEQUENCE?

yes (next)

no (12)

ASCENDING ORDER OF THE ASCII6 COLLATING SEQUENCE MEANS THE NORMAL ORDER.

!"0$X&' ()*+,-./0123456789:;<=>?SABCDEFGHIJKLMN0PQRSTUVWXYZC\D^_

DESCENDING ORDER MEANS THE REVERSE ORDER.

_^J\CZYXWVUTSRQP0NMLKJIHGFEDCBAa?>=<;:9876543210/.-,+*)('8%$#,,l

DO YOU WANT ALL YOUR KEYS SORTED IN ASCENDING ORDER?

yes (next)

no (next)

33. HOW MANY KEY FIELDS DO YOU HAVE?

number

WHAT IS THE TYPE OF THE MOST IMPORTANT KEY FIELD?

key-type

THE LEFT-MOST CHARACTER OF A RECORD IS NUMBER 1. FOR THE MOST IMPORTANT KEY FIELD, WHAT IS THE

NUMBER OF THE LEFT-MOST CHARACTER?

number

HOW MANY CHARACTERS ARE IN THE MOST IMPORTANT KEY FIELD?

number

Figure E-1. Interactive Dialog (Sheet 1 of 5)

60484800 D E-l

DO YOU WANT THIS KEY FIELD SORTED IN ASCENDING ORDER?1

yes

yes (next)

(next)

FOR THE SECONDttnOST IMPORTANT KEY FIELD, WHAT IS THE TYPE OF THE KEY?

key-type

FOR THE SECOND"^MOST IMPORTANT KEY FIELD, WHAT IS THE NUMBER OF THE LEFT-MOST CHARACTER?

number

HOW MANY CHARACTERS ARE IN THE SECOND+tHOST IMPORTANT KEY FIELD?

number

DO YOU WANT RECORDS WITH EQUIVALENT KEY VALUES TO BE WRITTEN IN THE SAME ORDER THEY ARE READ?

yes (next)

no (next)

A SUM FIELD IS A SPECIAL FIELD IN A RECORD THAT IS CHANGED BY SORT/MERGE. IF SUM FIELDS ARE

SPECIFIED SUM FIELDS FROM RECORDS WITH EQUAL KEY VALUES WILL BE NUMERICALLY SUMMED AND PUT IN THE

SANE SUM FIELD IN A NEW RECORD. THE REMAINING KEY AND DATA FIELDS IN THE NEW RECORD WILL BE SET TO

THE CORRESPONDING FIELDS IN ONE OF THE OLD RECORDS. THE ORIGINAL RECORDS WILL BE DELETED. RECORDS

MAY HAVE MORE THAN ONE SUM FIELD, AND THEY WILL BE SORTED ACCORDING TO KEY VALUES. THEY TYPE OF A

SUM MAY BE ANY NUMERIC KEY TYPE EXCEPT REAL.

28. HOW MANY SUM FIELDS DO YOU HAVE?

number

WHAT IS THE TYPE OF THE FIRST^SUM FIELD?

sura-type

WHAT IS THE FIRST CHARACTER IN THE FIRSfftsUM FIELD?

number

it •tt«

HOW MANY CHARACTERS* LONG IS THE FIRST" SUM FIELD?

number

GIVE AN ESTIMATE OF THE MAXIMUM NUMBER OF RECORDS TO BE SORTED,

number (end of dialog)

6. DO YOU HAVE AN OWN1 SUBROUTINE?

yes (next)

n o ( 7 )

WHAT IS THE NAME OF THE FILE CONTAINING YOUR OWNCODE SUBROUTINE(S)?

f1le-name

WHAT IS THE NAME OF YOUR OWN1 SUBROUTINE?

entry-name

7. DO YOU HAVE AN OWN2 SUBROUTINE?

yes (next)

n o ( 8 )

WHAT IS THE NAME OF THE FILE CONTAINING YOUR OWNCODE SUBROUTINE(S)?^

ﬁle-name

8. HOW MANY INPUT FILES DO YOU HAVE?

1 ( 2 )

number (next)

WHAT IS THE NAME OF YOUR FIRST^t INPUT FILE?

ﬁle-name

9. WHAT IS THE NAME OF THE FILE CONTAINING YOUR OWNCODE SUBROUTINE(S)?**

ﬁle-name

DO YOU HAVE AN OWN3 SUBROUTINE?

yes (next)

no (10)

Figure E-1. Interactive Dialog (Sheet 2 of 5)

E-2 60484800 A

yfﬁ^s

WHAT IS THE NAME OF YOUR 0WN3 SUBROUTINE?

entry-name

DO YOU HAVE AN 0WN4 SUBROUTINE?

yes (next)

n o ( 1 1 )

10. WHAT IS THE NAME OF YOUR 0WN4 SUBROUTINE?

entry-name

11. DO YOU HAVE AN OUTPUT FILE?

yes (4)

n o ( 5 )

12. DO YOU KNOW THE NAMES OF THE KEY TYPES YOU WANT TO USE?

yes (33)

no (next)

DO YOUR RECORDS CONSIST ENTIRELY OF CHARACTERS?

yes (19)

no (next)

NON-CHARACTER DATA MAY BE ORDERED ACCORDING TO ITS BINARY (SIGNED OR UNSIGNED) VALUE, OR REAL (I.E.

FLOATING POINT) VALUE.

A SIGNED BINARY KEY INTERPRETS THE BITS WITHIN THE KEY FIELD AS A ONE'S COMPLEMENT BINARY NUMBER.

IF THE LEFT-MOST BIT IS ZERO THEN THE NUMBER IS POSITIVE. IF THE LEFT-MOST BIT IS ONE THEN THE

NUMBER IS NEGATIVE AND THE MAGNITUDE OF THE NUMBER IS THAT WITH ALL BITS COMPLEMENTED. IF THE KEY

FIELD STARTS AND ENDS ON A CHARACTER BOUNDARY THEN THE KEY TYPE IS "INTEGER", OTHERWISE IT IS

"INTEGER_BITS".

AN UNSIGNED BINARY KEY INTERPRETS THE BITS WITHIN THE KEY FIELD AS A BINARY NUMBER. THE LEFT-MOST

BIT IN THE FIELD CONTRIBUTES TO THE NUMERIC VALUE JUST LIKE ANY OTHER BIT IN THE FIELD. IF THE KEY

FIELD STARTS AND ENDS ON A CHARACTER BOUNDARY THEN THE KEY TYPE IS "BINARY", OTHERWISE IT IS

"BINARY_BITS".

A REAL OR FLOATING POINT KEY INTERPRETS A 60-BIT VALUE AS A STANDARD FLOATING POINT NUMBER. THE KEY

TYPE IS "REAL".

13. IF YOU WANT MORE INFORMATION ABOUT A KEY TYPE, INCLUDING HOW TO SPECIFY IT, TYPE THE NAME OF THE KEY

TYPE. (OTHERWISE, TYPE A SPACE.)

binary (14)

binary_bits (15)

integer (16)

integer_bits (17)

real (18)

A ( 1 9 )

14. A BINARY KEY FIELD IS SPECIFIED BY NAMING THE FIRST CHARACTER POSITION AND THE LENGTH OF THE FIELD.

(THE FIRST CHARACTER POSITION IN THE RECORD IS 1.)

(13)

15. A BINARY BITS KEY FIELD IS SPECIFIED BY NAMING THE FIRST BIT POSITION AND THE LENGTH OF THE FIELD.

(THE FIRST BIT POSITION IN THE RECORD IS 1.)

(13)

16. AN INTEGER KEY FIELD IS SPECIFIED BY NAMING THE FIRST CHARACTER POSITION AND THE LENGTH OF THE

FIELD. (THE FIRST CHARACTER POSITION IN THE RECORD IS 1.)

(13)

17. AN INTEGERJ3ITS KEY FIELD IS SPECIFIED BY NAMING THE FIRST BIT POSITION AND THE BIT LENGTH OF THE

FIELD. (THE FIRST BIT POSITION IN THE RECORD IS 1.)

(13)

18. A REAL KEY FIELD IS SPECIFIED BY NAMING THE FIRST CHARACTER POSITION AND THE CHARACTER LENGTH OF THE

FIELD. SINCE REAL KEY FIELDS MUST BE FULL WORDS AND ALIGNED ON WORD BOUNDARIES, THE FIRST CHARACTER

POSITION MUST SPECIFY THE FIRST CHARACTER OF A WORD AND THE LENGTH MUST BE 10.

(13)

Figure E-1. Interactive Dialog (Sheet 3 of 5)

60484800 A E-3

19. KEY FIELDS CONSISTING ONLY OF CHARACTERS MAY BE ORDERED ACCORDING TO A CHARACTER COLLATING SEQUENCE

OR ACCORDING TO A NUMERIC INTERPRETATION OF THE CHARACTERS.

DO YOU WANT ANY OF YOUR CHARACTER KEY FIELDS TO BE INTERPRETED NUMERICALLY?

yes (next)

no (27)

THE KEY TYPE FOR NUMERIC CHARACTER KEYS IS "NUMERICJCX" WHERE XX INDICATES HOW THE SIGN (IF ANY) IS

TO BE EXTRACTED AND WHETHER LEADING BLANKS ARE ALLOWED. THE LEGAL NUMERIC CHARACTER KEY TYPES ARE:

NUMERICJJS ALL DIGITS, NO SIGN

NUMERIC_FS LEADING BLANKS, THEN OPTIONAL NEGATIVE SIGN, THEN DIGITS

NUMERICJ.S REQUIRED SIGN IS FIRST, THEN DIGITS

NUMERIC_TS DIGITS FOLLOWED BY REQUIRED SIGN

NUMERICJ.0 LEADING OVERPUNCHED DIGIT, THEN DIGITS

NUMERIC_TO DIGITS FOLLOWED BY TRAILING OVERPUNCHED DIGIT

20. IF YOU WANT MORE INFORMATION ABOUT ONE OF THE NUMERIC CHARACTER KEY TYPES, INCLUDING HOW TO SPECIFY

IT, TYPE THE NAME OF THE KEY TYPE. (OTHERWISE, TYPE A SPACE.)

nuraericjns (21)

numeric_fs (22)

numeric_ls (23)

nuraeric_ts (24)

numeric__lo (25)

numeric_to (26)

A ( 2 7 )

21. NUMERICJIS KEY FIELDS MUST CONTAIN ALL DECIMAL DIGIT CHARACTERS WITHOUT A SIGN. THE FIELD IS

SPECIFIED BY NAMING THE FIRST CHARACTER POSITION AND THE CHARACTER LENGTH OF THE FIELD.

(20)

22. NUMERIC_FS KEY FIELDS MUST BE NUMERIC CHARACTERS IN A FLOATING SIGN FORMAT. LEADING BLANKS ARE

REQUIRED FOR LEADING ZEROS, FOLLOWED BY AN OPTIONAL SIGN CONSISTING OF "-", FOLLOWED BY DECIMAL

DIGIT CHARACTERS. THIS IS THE SAME FORMAT AS IS PRODUCED BY WRITING ACCORDING TO THE DEFAULT

FORTRAN I FORMAT WITH NO MODIFIERS. THE FIELD IS SPECIFIED BY NAMING THE FIRST CHARACTER POSITION

AND THE CHARACTER LENGTH OF THE FIELD.

(20)

23. THE FIRST CHARACTER OF A NUMERICJ.S KEY FIELD MUST BE A "+" OR "-" CHARACTER. THE REMAINING

POSITIONS MUST CONSIST OF DECIMAL DIGIT CHARACTERS. THE FIELD IS SPECIFIED BY NAMING THE FIRST

CHARACTER POSITION AND THE CHARACTER LENGTH OF THE FIELD.

(20)

24. THE TRAILING (RIGHT-MOST) CHARACTER OF A NUMERIC_TS KEY FIELD MUST BE A "+" OR "-" CHARACTER. THE

REMAINING POSITIONS MUST CONSIST OF DECIMAL DIGIT CHARACTERS. THE FIELD IS SPECIFIED BY NAMING THE

FIRST CHARACTER POSITION AND THE CHARACTER LENGTH OF THE FIELD.

(20)

25. THE LEADING (LEFT-MOST) CHARACTER OF A NUMERICJ.0 KEY FIELD MUST BE AN OVERPUNCHED DIGIT. THE

OVERPUNCHED DIGIT DETERMINES BOTH THE SIGN OF THE ENTIRE NUMBER AND THE VALUE OF THE LEADING DIGIT.

CHARACTERS FROM "0" TO "9" MEAN THE SIGN IS POSITIVE AND THE LEADING DIGIT IS 0 TO 9. CHARACTERS

FROM "A" TO "I" MEAN THE SIGN IS POSITIVE AND THE LEADING DIGIT IS 1 to 9. CHARACTERS FROM "J" TO

"R" MEAN THE SIGN IS NEGATIVE AND THE LEADING DIGIT IS 1 TO 9. THE CHARACTER "<" MEANS THE SIGN IS

POSITIVE AND THE LEADING DIGIT IS 0. THE CHARACTER "!" MEANS THE SIGN IS NEGATIVE AND THE LEADING

DIGIT IS ZERO. THE REMAINING POSITIONS OF THE FIELD MUST CONTAIN DECIMAL DIGITS. THE FIELD IS

SPECIFIED BY NAMING THE FIRST CHARACTER POSITION AND THE CHARACTER LENGTH OF THE FIELD.

(20)

26. THE TRAILING (RIGHT-MOST) CHARACTER OF A NUMERIC_TO KEY FIELD MUST BE AN OVERPUNCHED DIGIT. THE

OVERPUNCHED DIGIT DETERMINES BOTH THE SIGN OF THE ENTIRE NUMBER AND THE VALUE OF THE TRAILING

DIGIT. CHARACTERS FROM "0" TO "9" MEAN THE SIGN IS POSITIVE AND THE TRAILING DIGIT IS 0 TO 9.

CHARACTERS FROM "A" TO "I" MEAN THE SIGN IS POSITIVE AND THE TRAILING DIGIT IS 1 TO 9. CHARACTERS

FROM "J" TO "R" MEAN THE SIGN IS NEGATIVE AND THE TRAILING DIGIT IS 1 TO 9. THE CHARACTER "<" MEANS

THE SIGN IS POSITIVE AND THE LEADING DIGIT IS 0. THE CHARACTER "!" MEANS THE SIGN IS NEGATIVE AND

THE TRAILING DIGIT IS ZERO. THE REMAINING POSITIONS OF THE FIELD MUST CONTAIN DECIMAL DIGITS. THE

FIELD IS SPECIFIED BY NAMING THE FIRST CHARACTER POSITION AND THE CHARACTER LENGTH OF THE FIELD.

(20)

Figure E-1. Interactive Dialog (Sheet 4 of 5)

E-4 60484800 A

27. DO YOU WANT ANY OF YOUR CHARACTER KEYS TO BE ORDERED ACCORDING TO A COLLATING SEQUENCE?

yes (next)

no (32)

THE STANDARD COLLATING SEQUENCE IS ASCII6. HERE IS THE WHOLE ASCII6 COLLATING SEQUENCE LISTED

!"#$X8»()*+,-./0123456789:;<=>?aABCDEFGHIJKLMNOPQRSTUVWXYZC\D-^_

THE KEY TYPE OF A CHARACTER KEY FIELD ORDERED ACCORDING TO A COLLATING SEQUENCE IS THE NAME OF THE

COLLATING SEQUENCE. THE KEY FIELD IS SPECIFIED BY NAMING THE FIRST CHARACTER POSITION AND THE

LENGTH OF THE FIELD.

THERE ARE FIVE PRE-DEFINED COLLATING SEQUENCES: DISPLAY, C0B0L6, ASCII6, EBCDIC6 AND INTBCD. THERE

CAN BE FROM 0 TO 100 USER-DEFINED COLLATING SEQUENCES.

DO YOU WANT TO DEFINE YOUR OWN COLLATING SEQUENCE?

yes (next)

no (32)

29. WHEN ASKED TO TYPE IN YOUR COLLATING SEQUENCE PLEASE TYPE IT IN AS C,C,C,C,C,C OR C,C,C,C,

c c

WHERE C IS A CHARACTER OR OCTAL VALUE.

C CAN BE A LETTER, A ONE-DIGIT NUMBER OR A SPECIAL CHARACTER SUCH AS ( ) + - *.

SORT/MERGE ASSUMES A TWO-DIGIT NUMBER IS AN OCTAL VALUE. (THE NUMBER SHOULD NOT BE FOLLOWED BY A

LETTER "B".) THE COMMA HAS TO BE SPECIFIED BY ITS OCTAL VALUE 56. HOW MANY COLLATING SEQUENCES DO

YOU WANT TO DEFINE?

number

30. WHAT IS THE NAME OF YOUR COLLATING SEQUENCE?

sequence-name

PLEASE TYPE IN YOUR COLLATING SEQUENCE.

C/C/ • ■ •

(30 if more to deﬁne)

(next if done deﬁning)

DEFINING EQUATE MEANS SPECIFYING TWO OR MORE CHARACTERS ALREADY IN A COLLATING SEQUENCE AS EQUAL FOR

COMPARISON PURPOSES. MORE THAN ONE EQUATE COLLATING SEQUENCE MAY BE SPECIFIED FOR A COLLATING

SEQUENCE.

DO YOU WANT TO DEFINE EQUATES FOR YOUR COLLATING SEQUENCES?

yes (next)

no (32)

WHEN ASKED TO TYPE IN YOUR EQUATE SEQUENCE(S), PLEASE ENTER IN THE FORM C,C,C,C,C,C WHERE C IS A

CHARACTER OR OCTAL VALUE AS EXPLAINED BEFORE.

HOW MANY EQUATE SEQUENCES DO YOU WANT TO DEFINE?

number

31. WHAT IS THE NAME OF THE COLLATING SEQUENCE FOR THIS EQUATE?

sequence-name

PLEASE TYPE IN YOUR EQUATE SEQUENCE.

(31 if more to deﬁne)

(next if done deﬁning)

32. YOU SHOULD KNOW THE NAME(S) OF ALL KEY TYPES YOU WANT TO USE. IF YOU DO NOT KNOW, START OVER BY

TYPING "RESTART" INSTEAD OF THE NUMBER OF KEY FIELDS.

(33)

'Not asked if you answered "yes" to the previous question about sort order.

''Repeated with successive numbers until all have been speciﬁed.

t"BITS" if you speciﬁed sum type INTEGER_BITS or BINARY_BITS.

++Not asked if you gave the ﬁle name in response to question 6.

Figure E-1. Interactive Dialog (Sheet 5 of 5)

60484800 A E-5

f^%

r*%

SORT/MERGE 5, SORT/MERGE 4 COMPARISON

Sort/Merge 5 is functionally equivalent to

Sort/Merge 4, but these versions of the product are

externally incompatible because the control state

ment for Sort/Merge 5 has been changed. Table F-l

shows a comparison of Sort/Merge 5 and Sort/Merge

4 . A s i n d i c a t e d i n t a b l e F - l , t h e S o r t / M e r g e 4

FORTRAN procedures can be called by Sort/ Merge 5

users, and the Sort/Merge 4 owncode exits can be

taken from FORTRAN 5.

Since Sort/Merge 5 procedures are compatible with

Sort/Merge 4 FORTRAN procedures, no changes are

needed to convert programs calling Sort/Merge 4

FORTRAN procedures FORTRAN compatible owncode

routines can be called by Sort/Merge 4 FORTRAN

procedures or by Sort/Merge 5 without any changes

to the owncode routines. All these Sort/Merge 5

relocatable binaries can be run by including LDSET

(LIB=SRT5LIB) in the load sequence before program

execution.

TABLE F-l. SORT/MERGE 5, SORT/MERGE 4 COMPARISON

Feature Sort/Merge 5 Sort/Merge 4

Control

Statement

Sort

Most sorts can be speciﬁed with the S0RT5

control statement and parameters alone.

A control statement sort requires a separate

directives section of the input ﬁle.

Merge Merge-only processing is available. Merge-only processing is available.

Key Types Sort keys can be 6-bit display code charac

ters, signed or unsigned binary integers,

or ﬂoating-point numbers.

Sort keys can be 6-bit characters written in

display code or internal BCD code, signed or

unsigned binary integers, or ﬂoating-point

numbers.

Collating

Sequences

ASCII6 (default), C0B0L6, DISPLAY, EBCDIC6,

and INTBCD are predeﬁned. User-deﬁned is

also available.

ASCII6 (default if installation character

set is ASCII), C0B0L6 (default if installa

tion character set is CDC), DISPLAY, and

INTBCD are predeﬁned. User-deﬁned is

also available.

Record

Types

Any CYBER Record Manager sequential record

type except U can be sorted.

Any CYBER Record Manager sequential record

type except U can be sorted.

Tape Sorts Tape ﬁles can be used as FROM and TO

ﬁles, but balanced and polyphase sorting

is not performed. Tape scratch ﬁles for

intermediate ﬁles is not supported.

A tape variant provides balanced and poly

phase tape sorting.

Checkpoint/

Restart

No t av ail a bl e . A job abnormally terminated can be captured

on tape so that the job can be restarted

from the checkpoint, rather than from the

beginning.

Interactive

Usage

Interactive dialog and commands are avail

able to aid in sort or merge speciﬁcation.

No interactive dialog or commands are

available.

Procedure

Calls

Sort or merge can be initiated from within

a program (FORTRAN or other language) by a

series of procedure calls.

Sort or merge can be initiated from within

a COMPASS or FORTRAN program by a series of

procedure calls. The Sort/Merge 4 FORTRAN

procedures can be called by Sort/Merge 5

users.

Owncode

Routines

Routines can be written to process input

records, output records, or records with

equal keys during Sort/Merge processing.

Routines can be written to process input

records, output records, records with equal

keys, or nonstandard labels during Sort/

Merge processing. The Sort/Merge 4 FORTRAN

owncode exits can be taken by Sort/Merge 5

users.

COBOL C0B0L5 control statement option can specify

Sort/Merge 5 be used when COBOL SORT or

MERGE statement is executed, or direct

calls to the procedures can be made.

The interface is through the COBOL SORT or

MERGE statement; Sort/Merge 4 is the default

for sorts or merges from COBOL 5.

60484800 D F-l

INDEX

A l t e r

SEQA parameter 3-7

SM5SEQA procedure 5-5

Ascending (see Sort order)

ASCII6 (see Collating sequence)

Basic Access Methods (BAM) 3-1, 3-10, C-l

Batch processing 1-2

Beginning-of-information (BOI) C-l

BINARY (see Numeric data formats)

BINARY_BITS (see Numeric data formats)

Bit 2-1

Blanks (see Spaces)

Block

Deﬁnition C-l

Types 3-11

Byte 2-1, C-l

CALL statement 5-1

Character

Data 2-1

Deﬁnition C-l

S e t s A - l

COBOL

Control statement 1-2, 6-1

Sort/Merge procedures 6-1

C0B0L6 (see Collating sequence)

Collating sequence

ASCII6 2-2

C0B0L6 2-2

Deﬁnition 1-1, 2-2, C-l

DISPLAY 2-2

EBCDIC6 2-2

INTBCD 2-2

User-deﬁned

SEQx parameters 3-5

SM5SEQx procedures 5-4

Commas 3-1, 3-9, 4-5

Conﬁguration

Hardware 1-2

Software 1-2

Continuation lines 3-1

Control statement (see FILE control statement, MERGE

control statement, S0RT5 control statement)

CYBER Control Language (CCL)

Deﬁnition C-l

ENR parameter 3-4

STATUS parameter 3-7

CYBER Record Manager (CRM) 1-1, C-l

Data alignment 2-1

Dayﬁle messages 5-4

Default

Deﬁnition C-l

File characteristics 1-1, 3-11

Descending (see Sort order)

DIA parameter (see DIALOG parameter)

D i a g n o s t i c s B - l

Dialog (see Interactive dialog)

DIALOG parameter 3-2, 3-4, 4-2, D-l

DIR parameter 3-3, D-l

Directive ﬁles

Continuation 3-2

D e ﬁ n i t i o n 1 - 1 , C - l

DIR parameter 3-3

Example 8-4

Format 3-1

Length 3-1

Use 1-1, 3 -1, 8 -4

DISPLAY (see Collating sequence)

Display code C-l

Duplicate parameters 3-2

E parameter 3-4, D-l

EBCDIC6 (see Collating sequence)

EL parameter 3-4, D-l

End-of-ﬁle 3-2, 5-2

End-of-information (EOI) C-l

End-of-record 3-2, 5-2

ENR parameter 3-4, D-l

ENTER statement 6-1

Entry point C-l

Error

Codes

SM5ST procedure 5-5

STATUS parameter 3-7

Level 3-4, 5-3, B-l

Messages B-l

Status variable

SM5ST procedure 5-5, B-l

STATUS parameter 3-7, B-l

Equal keys

Owncode 5 7-4

RETAIN parameter 3-5

SM5RETA procedure 5-4

Summing

SM5SUM procedure 5-6

SUM parameter 3-8

Examples

Control statement merge 8-3

Control statement sort

Multiple keys 8-3

One key 8-1

Directive ﬁle

Creation 8-4

Use 8-4

Interactive commands 4-4

Interactive dialog 8-1

I n t e r a c t i ve p a r a m e t e r s p e c i ﬁ c a t i o n 4 - 5

NOS login 4-1

NOS/BE login 4-2

Output ﬁle disposition 4-4

Owncode routine 8-9

Procedure calls 8-7

EXTERNAL statement 5-4, 7-1

FASTIO parameter 3-9, D-l

Features 1-1

Field length require m e n t s 3-11

60484800 D Index-1

File

Deﬁnition C-2

Description 1-1, 3-10

Format

Data ﬁle 3-2, 5-2

Directive ﬁle 3-1

Error ﬁle 3-4, 5-3

Listing ﬁle 3-3

Local C-2

Organization 3-10

FILE control statement

Description 1-1, 3-10, C-2

DIR parameter 3-3

FASTIO parameter 3-9

Format 3-10

FROM parameter 3-2

Interactive 4-3, 4-4

L parameter 3-3

Owncode routine 7-1

SM5E procedure 5-3

SM5FAST procedure 5-4

SM5FR0M procedure 5-2

SM5T0 procedure 5-2

TO parameter 3-2

File information table (FIT) 3-10, C-2

Floating-point numbers 2-5, C-2

FORTRAN programs 1-1, 5-1, 7-1

FROM parameter 3-2, D-l

Graphic A-l, C-2

3-5, 5-4

L parameter 3-3, D-l

Leading blanks 2-2

Library C-2

L i s t i n g ﬁ l e 3 - 3

Literal C-2

Owncode routines

Procedures 5-1

Lo g ic al ﬁl e n a me (l fn ) C - 2

Logical record C-2

Logout 4-5, C-2

Major sort key 2-1, C-2

Mass storage C-2

MERGE control statement

/*^%.

3-1, 4-5

Continuation

Format 3-1, 4-5

Length 3-1, 4-5

Punctuation

Merge order 3-9

Merging

Control statement

Deﬁnition C-3

I n i t i a t i o n 1 - 1 , 5 - 6

Procedure call 5-1

Purpose 1-1

Minor sort key 2-1, C-3

Multiple sort keys 2-1

3-1, 4-5

3-1

HELP command 4-3

Input ﬁles

FROM parameter 3-2

Owncode routine processing 7-3

SM5FR0M procedure 5-2

INTBCD (see Collating sequences)

In t eg er C- 2

INTEGER (see Numeric data formats)

INTEGER_BITS (see Numeric data formats)

Interactive commands 4-3

Interactive dialog

Te x t E - l

Use 1-1, 4-2, 8-1

Int e ractive processin g 1-2, 4-1, C-2

INTERCOM 1-2, C-2

In te r n a l re p r e se nt at i o n 2 -2

Job

Deﬁnition C-2

Flow

Batch 1-3

Interactive 1-4

Structure

Control statement merge

Control statement sort

D i r e c t i v e ﬁ l e u s e 8 - 7

Procedure calls 8-7

KEY parameter 3-2, D-l

Keys (see Sort keys)

Keyword

Deﬁnition C-2

Sort/Merge parameters 3-2

8-3, 8-4

8-3

Network Access Method (NAM) 1-2

Noncharacter data 2-1

NOS 1-2

NOS/BE 1-2

Notations ix

Numeric data formats

BINARY 2-4

BINARYJBITS 2-4

INTEGER 2-4

INTEGER_BITS 2-4

NUMERIC_FS 2-4 '

NUMERICJ.0 2-4

NUMERICJ.S 2-4

NUMERIC_NS 2-4

NUMERICJTO 2-4

NUMERICJTS 2-4

REAL 2-4

NUMERICJFS (see Numeric data formats)

NUMERICJLO (see Numeric data formats)

NUMERICJLS (see Numeric data formats)

NUMERIC_NS (see Numeric data formats)

NUMERICJTO (see Numeric data formats)

NUMERIC TS (see Numeric data formats)

OFL parameter (see OWNFL parameter)

0MRL parameter (see OWNMRL parameter)

Order (see Sort order)

Output ﬁles

Owncode routine processing 7-4

SM5T0 procedure 5-2

TO parameter 3-2

Overﬂow

SM5SUM procedure 5-6

SUM parameter 3-8

Overpunch character representation 2-5

r**E%

Index-2 60484800 D

0^s

Owncode routines

Calling sequence 3-5

Deﬁnition 1-2, 7-1, C-3

File 3-5

Loading 3--5, 7-1

Name 3-5, 5-4

Owncode 1 7-2

Owncode 2 7-3

Owncode 3 7-3

Owncode 4 7-4

Owncode 5 7-4

Parameters 7-1

Record length 3-5, 5-4

Summary 7-2

OWNF parameter 3-4, 7-1, D-l

OWNFL parameter 3-5, 7-1, D-l

OWNMRL parameter 3-5, D-l

OWNn parameter 3-5, 7-1, D-2

Parameter

Deﬁnition C-3

Owncode routine 7-1

Sort/Merge 3-2, D-l

Partition C-3

Periods 3-1

Physical record unit (PRU) C-3

Positional parameters 3-2, 3-10

Procedure calls

Deﬁnition 1-1, C-3

Loading 5-1

Order 5-1

Owncode routines 7-1

Record length 5-1

Result array 5-1

Punctuation (see Commas, Periods, Spaces)

QUIT command 4-3

Range ix, 3-1, 3-7

REAL (see Numeric data formats)

n „ _ _ 7 _ l f 7 _ 2

7 - 1 , 7 - 2 *

Reca parameter

Recb parameter

Record

Deﬁnition C-3

Estimated number

ENR parameter 3-4

SM5ENR procedure 5-3

Length

Owncode routine 3-5, 5-4

Speciﬁcation 3-11, 5-1

Type 1-1, 3-11, C-3

Relocatable C-3

Resource map 3-3

RESTART command 4-3

Result array 5-1, C-4

RET parameter (see RETAIN parameter)

RETAIN parameter 3-5, D-2

Return_code parameter 7-1

Rewind C-4

REWIND statement 5-2

Rla parameter 7-1

Rib parameter 7-1

3-7, D-2

3-5, D-2

3-7, D-2

3-6, D-2

2-2, C-4

5-3

5-4

5-2

5-4

5-4, 7-1

5-4

5-5

Section C-4

S e q u e n ti a l ﬁ l e 3-11, C-4

SEQx parameters

SEQA parameter

SEQN parameter

SEQR parameter

SEQS parameter

Signed numeric data

SM5E procedure 5-3

SM5EL procedure 5-3

SM5END procedure 5-7

SM5ENR procedure

SM5FAST procedure

SM5FR0M procedure

SM5KEY procedure

SM5MERG procedure

SM5N0DA procedure

SM50FL procedure

SM50MRL procedure

SM50WNn procedure

SM5RETA procedure

SMSSEQx procedures

SM5SEQA procedure

SM5SEQN procedure

SM5SEQR procedure

SM5SEQS procedure

SMSSORT procedure 5-1

SM5ST procedure 5-5

SM5SUM procedure 5-6

SM5T0 procedure 5-2

Sort keys

Deﬁnition 1-1, 2-1, C-4

Key length

General 2-1

KEY parameter 3-2

SM5KEY procedure 5-2

Key position

General 2-1

KEY parameter 3-2

SM5KEY procedure 5-2

Key type

General 2-1, C-2

KEY parameter 3-2

SM5KEY procedure 5-2

S i z e 2 - 1

Sort/Merge 4 F-l

Sort order

Ascending

KEY parameter 3-2

SM5KEY procedure 5-2

D e ﬁ n i t i o n 2 - 5 , C - 4

Descending

KEY parameter 3-2

SM5KEY procedure 5-2

Sorting

Control s t a t e m e n t 3-1

Initiation 1-1, 5-6

Procedure call 5-1

Purpose 1-1

S0RT5 control statement

Continuation 3-1, 4-5

Format 3-1, 4-5

Length 3-1, 4-5

Punctuation 3-1, 4-5

Spaces 3-1, 4-5

ST parameter (see STATUS parame

STATUS parameter 3-7, D-2

Sum ﬁelds

Example 3-9, 5-6, 8-7

Number 3-8, 5-6

Overﬂow 3-8, 5-6

Type 3-8, 5-b

ter)

0ps

60484800 D Index-3

I Summing 1-1, 3-8, 5-6

SUM parameter 3-8, D-2

System-logical-record C-4

Terminal

Access (see Login)

Connection 4-1

TO parameter 3-2, D-2

Value-set

Description 3-2

KEY parameter 3-2

SUM parameter 3-8

Value step

SEQS parameter 3-6

SM5SEQS procedure 5-5

VER parameter (see VERIFY parameter)

VERIFY parameter 3-9, D-2

Unit record format 4-2

Unsigned numeric data 2-2

User-deﬁned collating sequences

SEQx parameters 3-5

SM5SEQx procedures 5-5

Word C-4

$CHAR(n) 3-b

Index-4 60484800 D

A^tms

MANUAL TITLE: Sort/Merge Version 5 Reference Manual

Ji^s

0ms

PUBLICATION NO.: 60484800

REVISION: D

This form is not intended to be used as an order blank. Control Data Corporation

welcomes your evaluation of this manual. Please indicate any errors, suggested

additions or deletions, or general comments on the back (please include page number

references).

FOLD

Please reply No reply necessary

FOID

BUSINESS REPLY MAIL

NO POSTAGE

NECESSARY

IF MAILED

IN THE

UNITED STATES!

FIRST CLASS PERMIT NO. 8241 MINNEAPOLIS, MN.

POSTAGE WILL BE PAID BY ADDRESSEE

CONTRPL DATA

Technology and Publications Division

Mail Stop: SVL104

P.O. Box 3492

Sunnyvale, California 94088-3492

! F O L D FOLD

NO POSTAGE STAMP NECESSARY IF MAILED IN U.S.A.

FOLD ON DOTTED LINES AND TAPE

NAME:

COMPANY:

STREET ADDRESS:

CITY/STATE/ZIP;

TAPE TAPE

"^1

Sort_Merge_Ver_5_Ref_Man_60484800D_Sep86 Sort Merge Ver 5 Ref Man 60484800D Sep86

Sort_Merge_Ver_5_Ref_Man_60484800D_Sep86 Sort_Merge_Ver_5_Ref_Man_60484800D_Sep86

Navigation menu

Versions of this User Manual:

Views

Navigation