Until wiser minds than mine chime in...

You may find it useful to look at HTML::TokeParser::Simple, which has a tutorial on PM here