Closed balmas closed 4 years ago
I believe the changes necessary for this include the following:
In LatinLanguageDataset, everywhere we compare a word to an inflection value we need to use modelCompareWord.
E.g.
const lemma = this.irregularLemmas.get(pofs).find(item => item.word ===inflection[Feature.types.word].value)
this.irregularLemmas.get(pofs).some(item => item.word === inflection[Feature.types.word].value)
return this.irregularLemmas.get(pofs).some(item => item.word === inflection[Feature.types.word].value)
each need to be changed
In addition, we need to implement an override to morpheme.matchingValues
in the Form
derived class, to have it do modelCompareLanguage on features of type word and form.
confirm that it's still not working
Going through the closed issues - I checked this - and both desum and defuturus returns unplemented Verb irregular
Irina, the fix for desum should be included in the next build. Hopefully it will work.
When lookup up desum (or even dēsum) we don't get the inflection table for dēsum . I think this is because we aren't doing a model-specific compare of the word form everywhere we need to be.
Originally posted by @balmas in https://github.com/alpheios-project/inflection-tables/issues/250#issuecomment-534598699