use HTML::TokeParser::Simple;

  my $p = HTML::TokeParser::Simple->new(shift || die);

   while ( my $token = $p->get_token ) {

if ($token->is_text) {

print $token->as_is;
    }
else {
chomp();

  }
}