It doesn't need to recognize language. Just an option to change the character set, globaly and per note. Because there are similar characters in cyrillic and roman languages for example. That way when you search, it will not confuse cyrillic 'а' and roman 'a', for example.

I have a headache right now, because i've a bad hangover, so sorry if this doesn't make sense. :/

Anyway, here's an other idea - make it syncable with an online repository of some sort. An user defined FTP or something. And lastly, port the app to the desktop, so that the information can be available there also.