The module may actually work well but using regexes instead of a parser is a strong indication that it's got bad/surprise edge cases etc, etc. XML::LibXML and HTML::TokeParser::Simple are a bit more verbose to use but they are easy (once you get the hang of them), robust, and fast.