Beg pardon, wrong question.

Would you give me an example of a separate string id parser (like ident?) used in a comma-separated-list parser? In short, I don't know how to resolve the discrepancy in attributes.

Do you mean something like this? #include namespace ecgp { namespace qi = boost::spirit::qi; namespace ascii = boost::spirit::ascii; template <typename Iterator> struct ident : qi::grammar { ident(): ident::base_type(start) { using namespace qi; start %= lexeme[('_' | alpha) >> *(digit | alpha | '_' | '-')]; } qi::rule start; }; template <typename Iterator> struct idlist : qi::grammarstd::string(), ascii::space_type> { ecgp::ident<Iterator> id; idlist(): idlist::base_type(start) { using namespace qi; start %= id % ','; // don't use a sequence here } qi::rulestd::string(), ascii::space_type> start; }; } namespace qi = boost::spirit::qi; int main() { ecgp::idliststd::string::const_iterator g; std::string input("abc,def"); std::string::const_iterator begin = input.begin(); std::string::const_iterator end = input.end(); std::vectorstd::string v; qi::phrase_parse(begin, end, g, qi::ascii::space, v); BOOST_ASSERT(v.size() == 2 && v[0] == "abc" && v[1] == "def"); return 0; } Regards Hartmut --------------- Meet me at BoostCon www.boostcon.com