I have problems running Java Readline on UTF-8 terminals. From the source code, it seems as if the implementation assumes every string entered to be in Latin-1. This matches the experience, where Java will retrieve the Latin-1 characters that makes the multibyte characters.
