I just managed to get a script working to pull back search terms for this domain, and boy did I get some crackers, my favourites are:

see men pissing in obscure places
"football socks" sex -school
"george formby grill"
victoian windows (high ranking google spelling mistake)
where are sticklebricks
Chuckle Brothers 18 May 2003

I have a mad script, which I have hacked together, in true cut and paste style which appears to work. I nicked the regular expression stuff from the CPAN's http://www.cpan.org/authors/id/R/RD/RDUARTE/searchengineterms-1.0.pl, which I don't understand, combining it with some Damien Conway 'Cheating Under UNIX' examples from the Data Munging course, and produced this monstrosity, for all you geeks to laugh at.

(movable type seems to have mangled the <> operator in the while loop which has LOGFILE between the sideways hats).

use strict;
use URI::Escape;
my %unique;
foreach my $textfile () {
  if (substr($textfile,-3) eq '.gz'){
    $textfile = "gzip -dc $textfile |";
  }
  open LOGFILE, $textfile;
  while (<LOGFILE>){
    if ( m#\Wq(?:ry|kw|uery\?p)?=([^&\s?]+)# ){
      my $string = uri_unescape($1);
      $string =~ tr/\+/ /;
      chop($string) if (
        (scalar($string=~tr/"//) % 2 != 0) &&
        (rindex($string,'"') == (length($string)-1))
      );
      $unique{$string}++;
    }
  }
}
print "$_\n" foreach (keys %unique);
close(LOGFILE);