Bad cyrrillic simbols sometimes

[PasaOpasen]

So I use this code:

from epitran.backoff import Backoff

backoff = Backoff(['fas-Arab', 'rus-Cyrl'])
                  
backoff.transliterate('Привет дорогой друг пидор')

and it gives

'prʲivʲet doroɡoй druɡ pʲidor'

as u see, there is the russian й in result, which should be (maybe) j. Or am I wrong?

[dmort27]

You are not wrong. There is a bug somewhere in Backoff (this is not a problem for epitran.transliterate). Also, you should present input to Epitran token-by-token rather than as a series of whitespace-delimited tokens, as this will sometimes produce the wrong result.

[PasaOpasen]

Yes, this is not a problem for epitran.transliterate, it's okay