use HTML::TokeParser::Simple;
   my $p = HTML::TokeParser::Simple->new( $somefile );

   while ( my $token = $p->get_token ) {
     next if ! $token->is_text;
     print $token->return_text;
   }