michalisek

Results 1 comments of michalisek

UTF-8 is a good choice for internal representation, nevertheless, the modules that are responsible for similar words finding should in my opinion use UTF-32 internally: - Lexicon - the current...