C:\>perl perlbot_randomnode_request.pl CGI cookie help
Out of 431 possibilities
I think this url : "<A HREF="/index.pl?node_id=20485&lastnode_id=864">cookie usi
ng cgi.pm</a>" may best help you understand CGI cookie help
. You may also try this random hit : "<A HREF="/index.pl?node_id=18572&lastnode_
id=864">Answer:  How can I use a CGI script to return an image?</a>"


##</code><code>##

use LWP::UserAgent;
use HTTP::Request;
use HTTP::Request::Common;

my @junk=qw(what's what i need want how to can do does has the had should will anyone help me did for please hey);
my $request="@ARGV";
my $original_request= "@ARGV";
my @monkurls;
my %responses;
my $responsecount;

# strip out crap words.
foreach (@junk) { $request=~s/\b$_\b//ig};

#
#   arrayify the request.
#   if there's only one word, bail.
#

my @requestwords = split/ /,$request;
if (scalar(@requestwords)==1){print "You expect me to search on that? No way. Maybe with some more useful words."; exit}

#
#   start the request.
#

my $ua = LWP::UserAgent->new;
my $url =  "http://www.perlmonks.com/index.pl?node_id=864;node=$request";
my $response = $ua->request(GET $url);

#
#   if the title is 'not found' then there's no matches.
#   if the title matches $request exactly, there's a node with the exact title.
#   return that.
#

if ($response->content!~m/<title>(search results)<\/title>/i) {
        if ($response->content=~m/<title>(not found)<\/title>/i) {
                print "I couldn't find any nodes about $request. Sorry, eh?.";
                exit;
        }
        if ($response->content=~m/<title>($original_request)<\/title>/i) {
                print "I found but one node that matched your request : \"http://www.perlmonks.com/index.pl?node=$request\" may help you understand $request\n";
                exit;
        }
}
my @text=split /\n/, $response->content;


while ($text[0] !~ /here's the stuff:/i) {shift @text}
shift @text;
my $thisline = shift @text;
while ($thisline !~ /<FORM METHOD=GET ACTION="http:\/\/search.cpan.org\/search">/i){
    if ($thisline=~m/(<title>(.*)<\/title>)/i) {
        print "$1\n";
    }
    $thisline =~ s/<LI>//;
    if ($thisline !~ />re:/i) {
        $countgrep=grep{$thisline =~/$_/i }@requestwords;
        if ($countgrep) {
            $responsecount++;
            $responses{$thisline}=$countgrep;#urls, %{line=>$thisline, confidence=>$countgrep};
        }
    }
    $thisline = shift @text;
}
;
@monkurls = sort {$responses{$a} <=> $responses{$b}} keys %responses;

my $index   = rand @monkurls;
my $quip = $monkurls[rand @monkurls];

print "Out of ". scalar@monkurls . " possibilities\n";
print "I think this url : \"$monkurls[-1]\" may best help you understand @ARGV\n. ";
scalar@monkurls > 1 && print "You may also try this random hit : \"$quip\"";