zverok/spylls

With Korean dictionary, I got random results

hyunwoongko opened this issue · 1 comments

you can get dictionary here.
https://github.com/spellcheck-ko/hunspell-dict-ko/releases/download/0.7.94/ko-aff-dic-0.7.94.zip

I changed name of dictionary to 'ko_KR' from 'ko'.

from spylls.hunspell import Dictionary

dic = Dictionary.from_files('ko_KR')
word = "가방"
print(dic.lookup(word))    

This code returns True sometimes, and it also returns False sometimes.
Please run this code 20 times, then you can see same results with me.
I don't understand why this happens. Thanks.

That’s interesting. For now, I have no consistent theory why it could be stochastic, there are very few possibilities of randomness inside the code 🤔
But as I know very little about the Korean writing systems/encoding, might be related to it somehow.

A couple of questions to try to understand that:

  • is this word actually correct or not?..
  • if you’ll try to print([*dic.lookuper.good_forms(word)]), what it returns? Is it stable?