note
tobyink
<p>A few problems:</p>
<ul>
<li><p>You mention a URL, and you create a UA object, but you never actually use the UA to retrieve any HTML from the URL.</p>
<p>Anyway, you don't need to use LWP::UserAgent directly because HTML::TreeBuilder::XPath supplies a <c>new_from_url</c> method.</p></li>
<li><p>You use HTML::Selector::XPath - this module is aimed at turning CSS-style selectors into XPath expressions, but you already have an XPath, so you don't need this module!</p></li>
<li><p>You're using <c>@id</c> and <c>@href</c> within <em>double</em> quotes - this will be interpreted as two Perl arrays! You want to <em>single</em> quote the XPath.</p></li>
<li><p>There shouldn't be a slash at the end of the XPath.</p></li>
</ul>
<p>The following works:</p>
<c>
use 5.010;
use strict;
use warnings;
use HTML::TreeBuilder::XPath;
my $url = "http://www.youtube.com/results?search_query=run+flo+rida";
my $tree = HTML::TreeBuilder::XPath->new_from_url($url);
my $xpath = '(//*[@id="search-results"]/li)[1]/div[2]/h3/a/@href';
my @nodes = $tree->findnodes($xpath);
say $_->getValue for @nodes;
</c>
<div class="pmsig"><div class="pmsig-757127">
<small><small><tt>package Cow { use Moo; has name => (is => 'lazy', default => sub { 'Mooington' }) } say Cow->new->name</tt></small></small>
</div></div>
1014379
1014379