Hi, experts, I using boost.regex parse the content-type header's value in MIME. A typical content-type's BNF like following: Content-Type = "Content-Type:" SP media-type media-type = type "/" subtype *( ";" gen-param ) type = token subtype = token gen-param = pname [ "=" pval ] pname = token pval = token / quoted-string An example content-type as following: multipart/mixed; boundary="theboundary" ;id=33 There may be zero or one more parameters in content-type. My parse code as following: #include <boost/regex.hpp> #include <iostream> #include <string> #define COUT(x) std::cout << x << std::endl #define VAR(v) std::cout << #v << ": " << v << std::endl #define TOKEN "[a-zA-Z0-9]+" #define TYPE "(" TOKEN ")" #define SUBTYPE "(" TOKEN ")" #define PNAME "(" TOKEN ")" #define PVALUE "(" TOKEN ")" #define PARAM "(;" PNAME "=" PVALUE ")" #define CONTENT_TYPE_EXPRESS TYPE "/" SUBTYPE PARAM "*" int main() { std::string ct("multipart/mixed;boundary=theboundary;id=33"); COUT(CONTENT_TYPE_EXPRESS); boost::regex regex_content_type(CONTENT_TYPE_EXPRESS); boost::smatch result; if (boost::regex_match(ct, result, regex_content_type)) { VAR(result.size()); for (unsigned int i=0; i<result.size(); ++i) { if (result[i].matched) { COUT("result[" << i << "] = " << result[i]); } } } else { COUT("match failed."); } return 0; } /** output: ([a-zA-Z0-9]+)/([a-zA-Z0-9]+)(;([a-zA-Z0-9]+)=([a-zA-Z0-9]+))* result.size(): 6 result[0] = multipart/mixed;boundary=theboundary;id=33 result[1] = multipart result[2] = mixed result[3] = ;id=33 result[4] = id result[5] = 33 */ I want to get the parameter "boundary" and its value. Any way to do it? Sebastian Redl said: Not really. Regex doesn't support getting each submatch in a repeated expression. Xpressive does, but in this case, I think you'd be better off modelling the content-type grammar in Spirit and using that. Anyway, this question should have gone to boost-users, so I forward this to boost-users. Thanks in advance. -- Regards. Mockey Chen