I've used
HTML::TokeParser to parse regular HTML pages but I think you might want to look at
HTML::TableExtract for what you're after.
Revolution. Today, 3 O'Clock. Meet behind the monkey bars.
I would love to change the world, but they won't give me the source code