I tested it with that :

is_anagram('¡€•¡', '¡¡•€')

And it works. My code uses ord, and any character in unicode has a unique ord value. So I think it really works with the whole unicode.

Hi, although one-line usuallay improves code readability, this isn't the case :)