onemanga_down.pl

   1 #!/usr/bin/perl
   2 # onemanga_down.pl : A script to download manga from onemanga
   3 #
   4 # Released under ISC license.
   5 #
   6 # Copyright (c) 2008, Abel Camarillo <00z@the00z.org>
   7 #
   8 # Permission to use, copy, modify, and/or distribute this software for any
   9 # purpose with or without fee is hereby granted, provided that the above
  10 # copyright notice and this permission notice appear in all copies.
  11 #
  12 # THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
  13 # WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
  14 # MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
  15 # ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
  16 # WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
  17 # ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
  18 # OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
  19 #
  20
  21 use WWW::Mechanize;
  22 use Getopt::Long ;
  23
  24 Getopt::Long::Configure('bundling');
  25
  26 # default options
  27 my ($debug, $url, $base, $dst_dir) = (0, '', '', '');
  28
  29 GetOptions('debug|d'            => \$debug
  30     , 'url|u=s'                 =>\$url
  31     , 'dest|D=s'                =>\$dst_dir
  32     );
  33
  34 # Initial url
  35 #$url='http://www.onemanga.com/BLAME/1/01/';
  36
  37
  38 my $mech = WWW::Mechanize->new(stack_depth => 2);
  39
  40 while (1)
  41 {
  42   print STDERR "#" x 72, "\n" ;
  43   $mech->get($url);
  44
  45   break unless $mech->success();
  46
  47   my ($base_url) = $url =~  m{(http://[^/]+)/+};
  48
  49   my ($next_url) = $mech->content =~
  50     m{<[[:space:]]*input[[:space:]]+
  51     type="button"[[:space:]]+
  52     value="next[[:space:]]+page"[[:space:]]
  53    onclick="javascript:window.location='([^']+)';"[[:space:]]*/*>}sx;
  54
  55   $next_url = "$base_url/$next_url";
  56
  57   $url =~ s/^$base_url//;
  58
  59   my ($name, $chapter, $page) =  (split qq{/}, $url)[-3..-1] ;
  60
  61 # add leading zeros to $chapter and $page
  62
  63   ($chapter)  = ('000' . $chapter)  =~ m/(.{3})$/;
  64
  65 # to enable that odd `double' pages
  66   $page =~ s/^[[:digit:]]+[^[:digit:]]//;
  67
  68   ($page)     = ('000' . $page)     =~ m/(.{3})$/;
  69
  70   my @dst_file = ($dst_dir,$name,$chapter,$page) ;
  71
  72   shift @dst_file unless $dst_dir;
  73
  74 # create directories
  75   for (my $i=0; $i<$#dst_file;$i++)
  76   {
  77     my $dir =join('/',@dst_file[0..$i]);
  78     unless (-d $dir)
  79     {
  80       print STDERR "mkdir $dir\n";
  81       mkdir  $dir or die "$!" ;
  82     }
  83   }
  84
  85   my %data =
  86   (
  87     'URI: '           => $mech->uri
  88     , 'URL: '         => $url
  89     , 'Base URI: '    => $mech->base
  90     , 'Content : '    => $mech->ct
  91     , 'title : '      => $mech->title
  92     , 'Images : '     => join (qq/\n/, map { $_->url() } $mech->images )
  93     , 'Destination: ' => join ('/', @dst_file)
  94     , 'Chapter : '    => $chapter
  95     , 'Page : '       => $page
  96     , 'Name : '       => $name
  97     , 'Next : '       => $next_url
  98   );
  99
 100
 101   print STDERR map { ($_ , $data{$_}, "\n") } sort {$a cmp $b} keys %data
 102     if $debug;
 103
 104   for ($mech->images)
 105   {
 106     next unless ( $_->url =~ m{/mangas/.*(?:jpe?g|png|gif)$} );
 107
 108     my ($extension) = $_->url =~ /(\.[^.]+)$/;
 109     $dst_file[$#dst_file] .= $extension;
 110
 111     my $dst_file =join ('/', @dst_file);
 112
 113     $mech->get($_->url) ;
 114     print STDERR qq/Downloading Image: `/, $_->url, qq/' to `$dst_file' .../;
 115
 116     open DST , ">$dst_file" or die "\n$!";
 117     print DST $mech->content;
 118     close DST;
 119
 120     print STDERR qq/successfull!\n/ if $mech->success;
 121
 122   }
 123
 124   $url = $next_url;
 125 }