There's a module called 'HTML::TokeParser' that will break a html file into seperate tags. Then you can use a simple 'while and if' to do whatever you want with the text elements.
/\/\averick