Veoh: Fix grabber
[videosite.git] / videosite / CollegeHumorGrabber.pm
index 192ece5..7fb3cae 100644 (file)
@@ -9,7 +9,6 @@ use videosite::GrabberBase;
 @ISA = qw(videosite::GrabberBase);
 
 use videosite::HTMLHelper;
-use LWP::Simple qw(!get);
 use XML::Simple;
 use Data::Dumper;
 
@@ -20,7 +19,10 @@ sub new {
     my $self = $class->SUPER::new();
 
     $self->{'NAME'} = 'collegehumor';
-    $self->{'PATTERNS'} = ['(http://www.collegehumor.com/video:(\d+))'];
+    $self->{_SELFTESTURL} = 'http://www.collegehumor.com/video/5635400/pixar-intro-parody';
+    $self->{_SELFTESTTITLE} = 'Pixar Intro Parody';
+    $self->{'PATTERNS'} = ['(http://www.collegehumor.com/video:(\d+))',
+                           '(http://www.collegehumor.com/video/(\d+))'];
 
     bless($self, $class);
     $self->_prepare_parameters();
@@ -49,7 +51,7 @@ sub _parse {
     $metadata->{'DLURL'} = undef;
 
     # Get the XML file containing the video metadata
-    unless(defined($content = LWP::Simple::get(sprintf('http://www.collegehumor.com/moogaloop/video:%s', $2)))) {
+    unless(defined($content = $self->simple_get(sprintf('http://www.collegehumor.com/moogaloop/video/%s', $2)))) {
         $self->error('Could not download XML metadata');
         return undef;
     }
@@ -60,20 +62,7 @@ sub _parse {
     }
 
     $metadata->{'DLURL'} = $t->{'video'}->{'file'};
-
-    # The XML does not contain the full title of the video, for
-    # reasons possibly known to some jerk at CollegeHumor.
-    # So we'll have to parse the actual HTML, too.
-    $p = HTMLHelper->new();
-    unless(defined($content = $p->load(sprintf('http://www.collegehumor.com/video:%s', $2)))) {
-        $self->error('Could not download HTML');
-        return undef;
-    }
-
-    $t = $p->findnodes('meta[@name="title"]');
-    if (defined($t)) {
-        $metadata->{'TITLE'} = $t->{'content'};
-    }
+    $metadata->{'TITLE'} = $t->{'video'}->{'caption'};
 
     unless(defined($metadata->{'DLURL'}) && defined($metadata->{'TITLE'})) {
         $self->error('Could not extract download URL and title');