As GrandFather advised, I would also use HTML::TreeBuilder. Now for the   thing, I take a slightly different approach for removing non-breaking spaces, which I explained in Re: How do I create non-breaking space (  ) with HTML::Element? :

my $NBSP = HTML::Entities::decode_entities(' ');

$foo =~ s/$NBSP//g ;
[download]

Hope this helps.

Gu