Import Instructions
User Manual:
Open the PDF directly: View PDF 
.
Page Count: 6
Overview
There are five types of data that need to be imported/exported at the same time: SKUs,
ingredients, product lines, formulas, and manufacturing lines. At the customer's request,
manufacturing lines do not need to be imported, but the relation between SKUs and
manufacturing lines still does. In addition, for interoperability with existing spreadsheets, the
formulas and their ingredients are specified in the same file.
CSV files do a good job representing table data, and are extremely easy to parse. In addition,
CSVs are among the most space-efficient formats, so long as the tables they are representing
are themselves the minimal. As discussed above, four CSV (comma-separated values) files are
specified, and their headers are defined below:
Filename Prefix
Header
skus
SKU#,Name,Case UPC,Unit UPC,Unit size,Count per case,PL
Name,Formula#,Formula factor,ML Shortnames,Rate,Mfg setup
cost,Mfg run cost,Comment
ingredients
Ingr#,Name,Vendor Info,Size,Cost,Comment
product_lines
Name
formulas
Formula#,Name,Ingr#,Quantity,Comment
File format
1. The names of the files must be prefixed by an identifying string, as defined above. The
filenames must end in “.csv”. More precisely, the filenames must constitute matches for
the following regular expressions conforming to PCRE standard:
○SKUs: ^skus(\S)*\.csv$
○Ingredients: ^ingredients(\S)*\.csv$
○Product Lines: ^product_lines(\S)*\.csv$
○Formulas: ^formulas(\S)*\.csv$
2. The CSV files must be properly delimited and escaped. RFC4180 specifies the format of
a CSV file. Note specifically:
○Spaces (and other space-like characters) must not be ignored if they are part of
a field.
○Null values, if they are used, must be denoted by an empty string (i.e., two
consecutive commas, or a comma followed by a CRLF character).
○The double quote symbol (ASCII 034) must be properly quoted.
3. A valid header, as defined above, must be included as the first line in each file. The
header is case-insensitive.
4. Data types, uniqueness, requiredness, autogenerability of attributes in the files must
conform to specifications in the Evolution requirements. If an attribute is not
specified, it should follow the requirements below:
○All string-typed attributes must be no more than 1000 characters.
○All Integer-typed attributes must be within range of (i.e., a 32-bit− , )[ 231 231
signed integer).
○All floating-point numbers must be in decimal, and have a leading 0 (before the
decimal place) if the value is less than 1.0.
○The column ML Shortnames in skus.csv contains CSV cells. Each of the CSV cell
items must be a string with no whitespace.
5. Each file must contain no more than 1,000,000 lines. The records may be in no
particular order.
6. For mixed-unit expressions, e.g., ones for the "Size" attribute of an ingredient, the
following processing algorithm must be used.
○The numbers (whole or decimal) are separated from non-number (the unit part).
More formally, the expressions must constitute full matches for the following
PCRE-standard regular expression: l(\d*\.?\d+)\s*(\D.*|)$, where the
number part will be taken from group 1 of the match, and the unit part from
group 2.
○The unit part is then sanitized as follows:
i. Trimmed of whitespace on both ends
ii. Dots (".") and whitespaces (" ") are replaced with empty string
iii. Converted to lowercase
iv. Any trailing ‘s’ is removed
○The unit part (after sanitization) is then interpreted by matching it exactly to the
list below.
Unit
Accepted forms
Mass-based units
Ounce
"oz", "ounce"
Pound
"lb", "pound"
Imperial ton
"ton"
Gram
"g", "gram"
Kilogram
"kg", "kilogram"
Volume-based units
Fluid ounce
"floz", "fluidounce"
Pint
"pt", "pint"
Quart
"qt", "quart"
Gallon
"gal", "gallon"
Milliliter
"ml", "milliliter"
Liter
"l", "liter"
Count-based units
Count
"ct", "count"
Upload Specifications
1. Only one file may be submitted in one import session. This file may be of any type.
2. All submitted CSV file(s) may be assumed to have MIME type text/csv, as stipulated in
RFC4180 §3.
3. The following checks will be performed on each file individually. If any of the check fails,
the file will be rejected as a whole, and an appropriate error message will be displayed
for the administrator.
○File validity: the file must be standards-compliant, and contain the required
headers in the order specified. The matching of headers should be
case-insensitive.
○Duplicate records: the file must not contain duplicate records in itself. The
file may contain collision(s) to records existing in the system. Refer to the
Collision Resolution algorithm in the next section when a collision is detected.
○Data validity: where required, inputs must conform to standards.
■UPC numbers must conform to the UPC-A standard
■Sanitized units in mixed-unit expressions must be one in the accepted
list. Further, units used in formulas.csv must be logically convertible to
the original unit used in the ingredients. Similarly, if a record constitutes
a collision in ingredients.csv, the unit, if different than that of the
existing record, must be logically convertible to the existing unit.
○Empty attributes: all required attribute(s) must be supplied, except when both of
the following are true:
■It is marked as autogeneratable in the Evolution requirements;
■The record whose required attribute is empty contains no collision.
4. Referential integrity must be preserved in the file.
Note that the following relations exist among the files / corresponding entries in the
system:
○In skus.csv,
■PL Name must be existing in the system.
■Formula# must be existing in the system.
■All CSV cell items in ML Shortnames must be existing in the system.
○In formulas.csv:
■Ingr# must be existing in the system.
Example formulas.csv:
Formula#,Name,Ingr#,Quantity,Comment
1001,Formula1,2001,0.1250000000,
1001,Formula1,2002,0.2500000000,
1002,Formula2,2001,0.1250000000,
1002,Formula2,2006,0.0005000000,
Example ingredients.csv:
2001,Bread,Butternut,340g,2.29,This bread is used to make all the burger
2002,Beef,Beef Choice,2lb,12.25,
2006,Turkey,,424 pounds,666.66,
Example product_lines.csv:
Hamburger
Salad
Pasta
Example skus.csv:
SKU#,Name,Case UPC,Unit UPC,Unit size,Count per case,PL Name,Formula#,Formula factor,ML
Shortnames,Rate,Mfg setup cost,Mfg run cost,Comment
1001,Burger Beef,042272010264,725272730706,600g,50,Hamburger,1001,1,"line1,,,,",1,$1,$2,
1002,Burger Turkey,987654321098,705632085943,250g,80,Hamburger,1001,1,"line1",1,$1,$2,
