Thread:User talk:CodeCat/Characters to strip/reply

What I made now works, and it's clearer than octal because hexadecimal is much more widely used. But we can't use it if we move this into Module:languages because function calls are not allowed there.

Just a list of characters to strip will not work, because of things like SC ć. That happens to other letters as well, like Cyrillic ё and й. Decomposing might have been a nice idea at first, but it seems that may cause more problems than it solves because characters that we take to be fundamentally different turn out to be composed.