[Boost-users] Question about Boost string tokenizer