Parsing an HTML file is actually a fairly difficult problem.

There are many ways to go wrong and only a few ways to go right.

What Marshall said. I myself have had good luck using HTML::TreeBuilder.