Re: [Boost-users] Tokenizer design question