If you'd really like to do this robustly, you should look into one of the HTML modules, such as HTML::Parser or HTML::TokeParser. There is also an HTML::TokeParser Tutorial available on PerlMonks.

-- Mike

--
just,my${.02}