source: trollforge/troll/proxen.pl @ 386

#!/usr/bin/perl

use strict;
use LWP::UserAgent;
use HTTP::Request::Common;
use HTTP::Cookies;
use IO::Handle;
STDOUT->autoflush(1);

my $ua = LWP::UserAgent->new;
$ua->agent(
    "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; YPC 3.0.0; .NET CLR 1.1.4322; yplus 4.1.00b)"
);

#Hopefully this will mask us from being detected as a script
$ua->cookie_jar(HTTP::Cookies->new);
$ua->timeout(3);

my $response = $ua->request(GET "http://www.cybersyndrome.net/pla.html");
die "Error fetching new proxies" if !$response->is_success;
my $proxies = $response->content;
$proxies =~ s/\n//g;
$proxies =~ /<ol>(.+)<\/ol>/;
$proxies = $1;
$proxies =~
  s/<a title="\w\w" onMouseOver="s\('\w\w'\)" onMouseOut="d\(\)" class="\w">//g;
$proxies =~ s/<li>//g;
$proxies =~ s/<\/li>//g;
$proxies =~ s/<\/a>/\n/g;

my @proxies = split(/\n/, $proxies);
print "fetched ", scalar @proxies, " proxies\n";
foreach (@proxies)
{
    open(GOODPROXIES, ">>goodproxies.txt") or die "cannot open goodproxies.txt";
    chomp;
    next if ($_ =~ /:80$/);
    if (   $_ =~ /\.mil/
        || $_ =~ /\.gov/
        || $_ =~ /198\.2[5-6]\.[0-9][0-9][0-9]\.[0-9][0-9][0-9]/
        || $_ =~ /199\.2[0-1]:[0-9][0-9]\.[0-9][0-9][0-9]\.[0-9][0-9][0-9]/
        || $_ =~ /198.2[0-9]\.[0-9][0-9][0-9]\.[0-9][0-9][0-9]/)
    {
        print "Skipping, .mil/.gov proxy\n";
        next;
    }
    $ua->proxy('http', "http://$_/");
    my $formkey;
    print "Checking proxy: $_";
    my $reply =
      $ua->request(GET
        'http://slashdot.org/comments.pl?sid=20721&op=Reply&threshold=1&commentsort=0&tid=124&mode=thread&pid=7947680'
      );
    if ($reply->is_success)
    {
        print " || SUCCESS\n";
        print GOODPROXIES "$_\n";
        next;
    }
    print " || FAILURE\n";

    close GOODPROXIES;
}