Format conventions for SNDlib
The following rules apply to all ASCII input and output files of SNDlib:
- header
- Each input file has to contain a header line at the beginning:
?SNDlib native format; type: <file_type>; version: <version>
-
The file type can be either network, model or solution.
Currently there is only one valid version: 1.0.
Example of a header for a network file of version 1.0:
?SNDlib native format; type: network; version: 1.0
- comment
-
The comment symbol for all input files is the hashmark
symbol '#'. Everything from this symbol to the end of the
current line is ignored, as well as empty lines.
- identifier
- Valid identifiers are of the form [a-zA-Z0-9][a-zA-Z0-9-_]*.
In particular, identifiers must not contain whitespace, since
the latter is used as delimiter between input tokens.
- delimiter
- Valid delimiters between input tokens (int, double, string) are the blank and the tabulator.
- sections
- Sections with a variable number of items in the same format item_format
(e.g., the description of the available network nodes), are written in the format
( item_format* )
- where the '*' after item_format is interpreted as zero or more occurrences of such an item. The notation
( item_format+ )
- refers to one or more occurrences of item_format.
© 2023 Poznan University of Technology (PUT)
www.put.poznan.pl