KanĂ³nes User's Guide Build corpus-specific parsers for ancient Greek

Building and using a parser

You can build a parser with the sbt task fst:

sbt fst CORPUS [CONFIG_FILE]

This reads either the parser configuration in CONFIG_FILE or if not specified from the default fileconfiguration.properties, and compiles a SFST parser in parsers/CORPUS/greek.a.

You can use regular SFST tools such as fst-mor or fst-infl with this file. E.g., to Greek words you supply interactively, run

fst-mor parsers/CORPUS/greek.a

There is also an sbt task to apply your parser to a list of words read from a file with one word per line. The syntax is:

sbt parse CORPUS [WORDLIST]

If WORDLIST is not given, the task looks for a file named wordlist.txt.