#!/usr/bin/perl

# torrentreactor.pl v0.1
# Josh Jackson <pturing@firehead.org>

use LWP::UserAgent;
use integer;
use HTML::Tidy;

$categoriesfile = 'categories.txt';

if ($ARGV[0] == "" ) {
 open(INFO, $categoriesfile);
 while ($_ = <INFO>) { print "$_"; }
 close(INFO);
 print "\n\n";
 print "usage: torrentreactor.pl <category number>\n";
 exit;
}

$section = $ARGV[0];
$url = "http://www.torrentreactor.net/torrents/section_${section}";

#start with some atrocious variable names so reading the script is more interesting

$string1  = "\<TABLE\ ALIGN\=\"CENTER\"\ WIDTH\=\"600\"\ CELLSPACING\=\"1\"\ ";
$string1 .= "CELLPADDING\=\"2\"\ BORDER\=\"0\"\ BGCOLOR=\"\#000000\"\>";
$string2  = "\<\/TABLE\>";
$splitrows = "\<\/TR\>\n\<TR";
$splitcols = "\<\/TD\>\n\<TD";

@rows = ();


#if you're still reading, stick around; there's some good stuff later on


#find how many pages there are in this section

$ua = new LWP::UserAgent;
$ua->agent("Mozilla/4.0 (compatible; IAmTheWalrus 5.5; OpenBDSM)");
$res = $ua->request(HTTP::Request->new(GET => $url));
$_ = $res->content;
s/\(([0-9]*)\ total\)/\1/;           #my trusty regex will find it!
$total = $1;
$pages = ($total / 50) + 1;


#the plot thickens
#go through each page of results in this section and get the goods
for ( $page = 0; $page <= $pages; ++$page ) {

  $offset = ($page * 50);
  $url = "http://www.torrentreactor.net/torrents/section_${section}&off=${offset}";

  $ua = new LWP::UserAgent;
  $res = $ua->request(HTTP::Request->new(GET => $url));
  $text = $res->content;

  my $tidy = new HTML::Tidy;       #Tidy is a delightful tool
  $tidy->clean($text);             #It gives me consistent, easy to work with output
                                   #so I don't have to parse messy real-world html

  @text = split(/\>\</, $text);   #insert a newline anywhere that one tag
  $text = join("\>\n\<", @text);  #follows another tag with nothing between

  @string3 = split(/$string1/,$text);     #find the table we want
  $string4 = $string3[1];                 #grab everything after that table opens
  @string5 = split(/$string2/, $string4); #find table endings
  $string6 = $string5[0];                 #grab inside the table

  @theserows = split(/$splitrows/,$string6); #split the table rows
  ($header, @theserows) = @theserows;        #first row is the header

  push(@rows, @theserows);
}


foreach $row (@rows) {
  @cols = split(/$splitcols/,$row);   #split into TDs

       #One Regular Expression to rule them all, One Regular Expression to find them,
       #One Regular Expression to bring them all and in the array bind them
       #In the Land of Perl Code where the $_'s lie.

  $_ = $cols[0];
  if ( s/\>(.*\:.*)/\1/ ) { $time = $1; } #these would be great places for 'else' statements and error
                                          #flagging code, in case one of the regexes didn't always
  $_ = $cols[1];                          #work with the real data. if I cared enough.
  if ( s/view_.*title\=\"(.*)\"/\1/ ) { $title = $1; }

  $_ = $cols[1];
  if ( s/HREF\=\"(http\:\/\/www\.torrentreactor\.net\/torrents\/download_[0-9]*)\"/\1/ ) { $dl = $1; }

  $_ = $cols[1];
  if ( s/HREF\=\"(http\:\/\/www\.torrentreactor\.net\/torrents\/view_[0-9]*)\"/\1/ ) { $view = $1; }

  $_ = $cols[2];
  if ( s/\>([0-9][0-9].*)/\1/ ) { $size = $1; }

  $_   = $cols[3];
  if ( s/\>([0-9].*)/\1/ ) { $seed = $1; }

  $_  = $cols[4];
  if ( s/\>([0-9].*)/\1/ ) { $leech = $1; }

  $_ = $cols[5];
  if ( s/section_.*\"\>(.*)\</\1/ ) { $cat = $1 }  #by this time, I'm composing new regexes
                                                   #with considerable ease and speed
  $_ = $cols[6];
  if ( s/kind_.*\"\>(.*)\</\1/ ) { $kind = $1 }

  $_ = $cols[7];
  if ( s/img\/(.*)\..*/\1/ ) { $lang = $1 }


  print " time:  $time  \n"; # quite the anticlimactic ending here
  print "title:  $title \n"; # 
  print "   dl:  $dl    \n"; # *888888888888888888888888888888888*
  print " view:  $view  \n"; # * This script brought to you      *
  print " size:  $size  \n"; # *          by the                 *
  print " seed:  $seed  \n"; # * Pope Turing Perlcoding Project  *
  print "leech:  $leech \n"; # *       in conjunction with       *
  print "  cat:  $cat   \n"; # *  the Leauge of Satanic Discord  *
  print " kind:  $kind  \n"; # *                                 *
  print " lang:  $lang  \n"; # *888888888888888888888888888888888*
  print "\n";                #
}