Arabic G2P returns ة in IPA results

[tcourat]

model = epitran.Epitran('ara-Arab')  
res = model.transliterate('علقانة')

The given results is

'ʕlqaːnة'

which contains the letter ة which is not an IPA symbol.

[dmort27]

This is due to the fact that Epitran has encountered a character for which it does not have a mapping. Can you say something more about this character and to what it should be mapped?

[tcourat]

Unfortunately, I'm not competent to deal with that matter. I would have thought that G2P supported the entire alphabet.

Note that I do not have the same issue with the espeak-ng G2P.

[martinarisk]

the character ة becomes a. source:

https://www.neurlang.online/