1) A very valueful article, http://dozb.bokee.com/1976537.html
2) regex_match(str, match_result, expression)
* it will try to match the whole 'str'
* "[/d]+" is wrong, should be "[//d]+"
* "/." is wrong ,should be "//.", but /" and /' are right, because " and ' is not special for regex, /" just is
going to make " a ordinary for the string
* every (...) is a sub_expression which will result match[i], match_result.size() == expression.mark_count()
* /w 任意单词字符 [_0-9a-zA-Z]
* regex esform("(<form)(.*?)(</form>)")
here the '?' is for forbiding greed matching.
3) The usage of regex_search is almost the same with regex_match, the tiny difference is reget_search does not need to match the whole string, only if the expression exists,it returns true with the match_results stored into match_result