These are the files for running tamnon, a dialect analyzer for Ancient Greek. For now, it does binary differentiation between Attic and Doric forms.
Save the input data as a [textname].txt file in the "texts" folder (the full text of Pindar's work, from the Perseus Digital Library, has been included).
Then, run
python2 tamnon_preprocess.py [textname] (this may take a few minutes)
This will create a set of intermediate files in the "intermediate files" folder.
Next, run python2 tamnon_process.py [textname] (this may also take a minute or two, but less than the preprocess)
which will create a variety of result files, stored in results/[textname]/
The result files are as follows:
- overall.txt contains an overview of the results, without individual tokens.
- dialect.txt contains the number of forms and tokens for each dialect.
- rules.txt contains the number of forms and matching tokens for each rule.
- tokens.txt contains the list of every token (in text order) with its parses.
- evaluation.txt contains comparisons between Morpheus' dialect analyses and tamnon's dialect analyses.
- all.txt contains all information combined in a single file.
- the graphs folder contains the rules results in graph form.
Alternatively, you can run python2 tamnon_gui.py
to open a small graphical user interface that allows instant feedback on text
placed into the upper textarea.
If you only want to examine portions from a text you have already preprocessed,
you can run python2 tamnon_gui.py [textname]
to avoid rerunning the morphological parses each time.
To test the tamnon rules, run "python2 tamnon_get_test_forms.py" every time you change the forms that are to be tested on, then run "python2 tamnon_test_rules.py" to test all of the rules and ensure they are working. Any errors will appear under the header of the appropriate rule.
MIT