Merge branch 'master' of ssh://sun@ryoko:22003/home/sun/GIT/quotesite

[quotesite.git] / quotesite / QdbGrabber.pm

diff --git a/quotesite/QdbGrabber.pm b/quotesite/QdbGrabber.pm
new file mode 100644 (file)
index 0000000..0a65255
--- /dev/null
+++ b/quotesite/QdbGrabber.pm
@@ -0,0 +1,74 @@
+# (c) 2007 by Ralf Ertzinger <ralf@camperquake.de>
+# licensed under GNU GPL v2
+#
+# Grabber for qdb.us
+
+package QdbGrabber;
+
+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'} = 'qdb.us';
+    $self->{'PATTERNS'} = ['(http://(?:[-a-zA-Z0-9_.]+\.)*qdb\.us/(\d+))'];
+
+    bless($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 $t;
+    my $t2;
+
+    $url =~ m|$pattern|;
+    $url = $1;
+
+    $metadata->{'URL'} = $url;
+    $metadata->{'ID'} = $2;
+    $metadata->{'TYPE'} = 'quote';
+    $metadata->{'SOURCE'} = $self->{'NAME'};
+    $metadata->{'CONTENT'} = undef;
+
+    # Get the HTML file containing the quote
+    unless(defined($content = LWP::Simple::get(sprintf('http://qdb.us/?%s', $2)))) {
+        $self->error('Could not download quote');
+        return undef;
+    }
+
+    $p = HTML::TokeParser->new(\$content);
+
+    OUTER: while ($t = $p->get_tag('table')) {
+        if (exists($t->[1]->{'class'}) && ($t->[1]->{'class'} eq 'quote')) {
+            while ($t2 = $p->get_tag('p')) {
+                $metadata->{'CONTENT'} = $p->get_text('/p');
+                last OUTER;
+            }
+        }
+    }
+
+    unless(defined($metadata->{'CONTENT'})) {
+        $self->error('Could not extract quote content');
+        return undef;
+    }
+
+    return $metadata;
+}
+
+1;