Use a pre-built parser

Published

July 12, 2024

In addition to building parsers from a data set of local source files, you can save and load compiled parsers as a single delimited-text file.

A very small parser to demonstrate inflectional rules with a minimal vocabulary set is available from the Kanones github repository, at https://raw.githubusercontent.com/neelsmith/Kanones.jl/main/test/data/lgr-rulesparser.cex.

Out of date

The source for the parser used here is out of date.

We can download this file and build a parser from it. (To be tidy, we’ll remove the temporary data file we downloaded once we’ve got a parser.)

using Kanones, Downloads
parserurl = "https://raw.githubusercontent.com/neelsmith/Kanones.jl/main/test/data/lgr-rulesparser.cex"
datafile = Downloads.download(parserurl)
#parser = dfParser(datafile)
rm(datafile)

Now we can use it like any other parser.

s = "ἀγαθόν"
parses = parsetoken(s, parser)

Prebuilt parsers you can use

A prebuilt parser for texts in standard literary Greek orthography is regularly available for download from shot.holycross.edu.

http://shot.holycross.edu/morphology/comprehensive-current.csv is a prebuilt parser that includes vocabulary automatically culled from the digital Liddell-Scott lexicon published by the Perseus project. The vocabulary entries have not been manually verified.

In 2024, this parser is not yet automatically rebuilt on a regular time table. Load it and use it to parse texts the same way you would if you compiled your own parser.

Caution

This is a large file (currently, ca. 700 Mb); depending on your internet connection, it can easily take a couple of minutes to download.

There is also http://shot.holycross.edu/morphology/attic_core-current.cex.