Have you checked the various HTML modules on CPAN? I, for one, would think that modifying the regex engine to recognize HTML entities, such as   as a member of the \s character-class is far too specific to HTML-processing for a general purpose language, such as Perl.