my $link = 'somelinek';
my $page = get($link) or die $!;
my $stream = HTML::TokeParser->new(\$page);
my $doparse = 1; ## 0 means don't parse
while (my $token = $stream->get_token)
{
    if ($token->[0] eq 'S')
    {
       if ($token->[1] eq 'table')
       {
         $doparse = 0;
       }
       elsif ($token->[1] eq 'img')
       {
         ;;
       }
    }
    elsif ($token->[0] eq 'E' and $token->[1] eq 'table')
    {
       $doparse = 1;
    }
    elsif ($token->[0] eq 'C')
    {
       ;;
    }
    elsif ($token->[0] eq 'T' and $doparse eq 1)
    { # text process the text in $token->[1]
        # skip: empty lines, " "
        if (defined ($token->[1]))
        {
          $token->[1] =~ s/ / /ig;
          $token->[1] =~ s/’/'/ig;
          $token->[1] =~ s/&#14[7-8];/"/ig;
          $token->[1] =~ s/—//ig;
          $token->[1] =~ s/&/&/ig;
          $token->[1] =~ s/-{2,}//ig;
          print "$token->[1]";
        }
    }
}