From: Ralf Ertzinger Date: Fri, 18 Jan 2008 19:42:19 +0000 (+0100) Subject: - Add zeropuctuation grabber X-Git-Url: https://git.camperquake.de/gitweb.cgi?a=commitdiff_plain;h=a018c395f1f0d469b199512ded838d4f6f6061b7;p=videosite.git - Add zeropuctuation grabber --- diff --git a/videosite/ZeroPunctuationGrabber.pm b/videosite/ZeroPunctuationGrabber.pm new file mode 100644 index 0000000..2a93840 --- /dev/null +++ b/videosite/ZeroPunctuationGrabber.pm @@ -0,0 +1,77 @@ +# (c) 2007 by Ralf Ertzinger +# licensed under GNU GPL v2 +# +# Grabber for www.escapistmagazine.com/articles/view/editorials/zeropunctuation + +package ZeroPunctuationGrabber; + +use GrabberBase; +@ISA = qw(GrabberBase); + +use LWP::Simple qw(!get); +use HTML::TokeParser; +use Data::Dumper; + +use strict; + +sub new { + my $class = shift; + my $self = $class->SUPER::new(); + + $self->{'NAME'} = 'zeropunctuation'; + $self->{'PATTERNS'} = ['(http://www.escapistmagazine.com/articles/view/editorials/zeropunctuation/([-A-Za-z0-9]+))']; + + bless($self, $class); + $self->_prepare_parameters(); + + return $self; +} + +sub _parse { + my $self = shift; + my $url = shift; + my $pattern = shift; + my $content; + my $metadata = {}; + my $p; + my $e; + + $url =~ m|$pattern|; + $url = $1; + + $metadata->{'URL'} = $url; + $metadata->{'ID'} = $2; + $metadata->{'TYPE'} = 'video'; + $metadata->{'SOURCE'} = $self->{'NAME'}; + $metadata->{'TITLE'} = undef; + $metadata->{'DLURL'} = undef; + + unless(defined($content = LWP::Simple::get($url))) { + $self->error('Could not download %s', $url); + return undef; + } + + $p = HTML::TokeParser->new(\$content); + + # Look for the title + if ($p->get_tag('title')) { + $metadata->{'TITLE'} = $p->get_text(); + $metadata->{'TITLE'} =~ s/^The Escapist : Zero Punctuation: (.*)$/$1/im; + } + + while ($e = $p->get_tag('script')) { + my $c = $p->get_text(); + if ($c =~ m|var vars = \{file:\x27([^\x27]+)\x27|) { + $metadata->{'DLURL'} = 'http://' . $1 . '_high.flv'; + } + } + + unless(defined($metadata->{'DLURL'}) && defined($metadata->{'TITLE'})) { + $self->error('Could not determine download URL'); + return undef; + } + + return $metadata; +} + +1;