I've got to recommend PyICU here. unidecode is good, but has some holes (I noticed the Azeri letter schwa, for example). PyICU is a binding for IBMs Internation

I've got to recommend PyICU here. unidecode is good, but has some holes (I noticed the Azeri letter schwa, for example). PyICU is a binding for IBMs International Components for Unicode, and basically has a coding language for unicode transforms. Here's an example that is equivalent to unidecode: https://github.com/pudo/normality/blob/master/normality/transliteration.py#L31 https://github.com/pudo/normality/blob/master/normality/tran...