# Grabber for break.com # # (c) 2007 by Ralf Ertzinger # licensed under GNU GPL v2 package BreakGrabber; use GrabberBase; @ISA = qw(GrabberBase); use LWP::Simple qw(!get); use HTML::Parser; use Data::Dumper; use strict; sub new { my $class = shift; my $self = $class->SUPER::new(); $self->{'NAME'} = 'break'; $self->{'PATTERNS'} = ['(http://(?:[-a-zA-Z0-9_.]+\.)*break.com/index/([-a-zA-Z0-9_]+?)\.html)']; bless($self, $class); $self->_prepare_parameters(); return $self; } sub _parse { my $self = shift; my $url = shift; my $pattern = shift; my $content; my $metadata = {}; my $p = HTML::Parser->new(api_version => 3); my @accum; my @text; 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($1))) { $self->error('Could not download %s', $url); return undef; } $p->handler(start => \@accum, "tagname, attr"); $p->report_tags(qw(meta)); $p->utf8_mode(1); $p->parse($content); # Look for the title in the meta tags foreach $e (@accum) { if ('meta' eq $e->[0]) { if ('embed_video_title' eq $e->[1]->{'name'}) { $metadata->{'TITLE'} = $e->[1]->{'content'}; } if ('embed_video_thumb_url' eq $e->[1]->{'name'}) { $metadata->{'DLURL'} = $e->[1]->{'content'}; $metadata->{'DLURL'} =~ s/\.jpg$/\.flv/; } } } unless(defined($metadata->{'DLURL'}) && defined($metadata->{'TITLE'})) { $self->error('Could not determine download URL'); return undef; } return $metadata; } 1;