Benchmark

This command runs edktool in benchmarking mode. This mode runs multiple concurrent extraction sessions, several times to test the performance of a grammar. Edktool reads the input document once, and feeds it into each session. It produces timing information after all runs are complete.

The following table describes the parameters for this command.

-l <licensefile>

The file containing a valid license key for Eduction.

This file can include the version key, concatenated to the license key with a semicolon (;), or you can copy the versionkey.dat to the same directory as the license key file. See Eduction Licenses in edktool.

If you do not specify a license key, edktool attempts to load the license licensekey.dat and versionkey.dat in its current working directory. You must specify the license parameter if your license is in a different location.

-i <inputfile>

The file to perform entity extraction on. The input file must be plain text.

-c <configfile>

A configuration file controlling the extraction. See Eduction Configuration File.

You can specify one or more grammar files and one or more entities in place of a configuration file. Specifying a configuration file overrides the grammar or entity parameters.

-g <grammarfile>

A grammar file to use. Edktool ignores this option if you set a configuration file with -c.

If you provide a grammar file but do not specify any entities with -e, Eduction extracts all entities in the grammar file.

NOTE: You can use the MinScore parameter only if you use -c. Without a configuration file, you can specify a grammar that supports scoring, but edktool does not filter out matches based on those scores.

You can use wildcard expressions in this parameter. See Wildcard Expressions in edktool.

-e <entity>

The entities to extract. Separate multiple entities with a comma. Edktool ignores this option if you set a configuration file with -c.

You can use wildcard expressions in this parameter. See Wildcard Expressions in edktool.

-d (Optional) Return details of the matching strings and their locations in the input file.
-s <sessions> The number of sessions to run concurrently during each iteration of the benchmarking test.
-n <number> The number of iterations of the benchmarking test to run.
-b Set this parameter to read the input file in binary mode, rather than text mode. If you create a grammar file that matches entities with only Windows (CR LF) line endings and you run edktool on Windows, edktool must read the input file in binary mode for it to find any matches. OpenText recommends that you create grammar files capable of handling both Windows and Unix line endings.

The benchmarking command runs the specified number of concurrent sessions and iterations and then displays the timing for each run, with a summary showing:

  • the total number of observations.

  • maximum and minimum times.

  • the standard deviation.