Open tiff opened 4 years ago
@MikeUnwalla I would like to remove the plural form "aboves" of "above" by adding it to the removed.txt? Would this be okay for you?
@tiff , yes. As best I know, 'aboves' is not a standard English word.
I'm also noticing that this rule is missing a lot incorrect cases, perhaps because they involve words listed as NNP in the dictionary. For example, "I have 2 kid" is acceptable, . Is the right course of action to move "kid" as NNP to removed as well? (Also as NN:UP for that matter)
@donovanrichardson , thanks for the report.
'kid' (lowercase k) is not NNP. Thus, the false negative is not because of Kid/NNP.
kid/NN:UN is correct. NN:UN means that the word can be both a count noun and a non-count noun. 'Kid' is both: https://www.lexico.com/en/definition/kid
@MikeUnwalla Aha, I've never heard of that definition for kid!
Nevertheless, I am concerned about the exception that CD_NN provides for NN:UN, because sentences including "three kid" should not be acceptable, whether the noun is uncountable or countable singular. For reference the line with the problematic NN:UN exception is below.
<exception postag="NN:UN?|NNP?S|JJ.*|IN|CD" postag_regexp="yes"/>
Testing grammar.xml without the NN:UN exception, I immediately found a sentence that should be acceptable ("I'll be in my office from ten tomorrow.") but failed the test. If those failures aren't too numerous, perhaps I can begin adding antipatterns for sentences that break when the NN:UN exception is removed.
@donovanrichardson,
perhaps I can begin adding antipatterns for sentences that break when the NN:UN exception is removed.
Yes, that is fine by me. Given that you will add antipatterns, I will leave you to deal with the removal of the NN:UN exception.
I think there are two issues here: