use std::regex_token_iterator instead of boost::tokenizer

[Jochem Snuverink]

in BDSMySQLWrapper strings are split with boost::tokenizer. While this works fine, this creates a dependency on boost. With C++11 this dependency can be avoided by using std::regex_token_iterator instead (see e.g. http://www.cplusplus.com/reference/regex/regex_token_iterator/regex_token_iterator/ )

[Jochem Snuverink]

fix issue #57 : replace boost tokenizer with std::regex

→ <<cset 56ffe05d2539>>