- Add HTMLHelper.pm

diff --git a/videosite/HTMLHelper.pm b/videosite/HTMLHelper.pm
new file mode 100644 (file)
index 0000000..e86b193
--- /dev/null
+++ b/videosite/HTMLHelper.pm
@@ -0,0 +1,69 @@
+#
+# A helper class for getting values out of a HTML document
+#
+
+package HTMLHelper;
+
+use HTML::TreeBuilder;
+@ISA = qw(HTML::TreeBuilder);
+
+use strict;
+use Data::Dumper;
+use LWP::Simple qw(!get);
+
+sub new {
+    my $class = shift;
+    my $self = $class->SUPER::new(@_);
+
+    return bless($self, $class);
+}
+
+sub load {
+    my $self = shift;
+    my $URL = shift;
+    my $c;
+
+    unless(defined($c = LWP::Simple::get($URL))) {
+        # Error loading URL
+        return undef;
+    }
+
+    $self->parse($c);
+
+    return 1;
+}
+
+sub findnodes {
+    my $self = shift;
+    my $path = shift;
+    my ($tagname, $classifier);
+    my %matchtag;
+
+    #
+    # Try to make sense of the path specifier.
+    # For the moment, we just allow paths of the following two forms:
+    #
+    # a) <tagname>
+    # b) <tagname>[@<attribute>=<value> (, @<attribute>=<value> ...)]
+    #
+
+    unless (($tagname, $classifier) = $path =~
+            m|(\w+)(?:\[((?:\@[\w-]+=\x22[^\x22]+?\x22)(?:\s*,\s*(?:\@\w+=\x22[^\x22]+?\x22))*)\])?|) {
+        # bad path name
+        return wantarray?():undef;
+    }
+
+    $matchtag{'_tag'} = $tagname;
+    if (defined($classifier)) {
+        foreach (split(/\s*,\s*/, $classifier)) {
+            my ($n, $v) = split(/=/, $_, 2);
+            $n =~ s/^\@//;
+            $v =~ s/"//g;
+            $matchtag{$n} = $v;
+        }
+    }
+
+    return $self->look_down(%matchtag);
+}
+
+1;