You're right---it doesn't work:); however, as I see it, the problem isn't with CAM::PDF but rather with the google pdf. I think that google is great for searching the web, but when it comes to anything else, it's not so good. I've tried the code with my own non-google pdf's and it works. Try this, and let me know if it works or not:

#!/usr/bin/perl

use strict;
use warnings;
use CAM::PDF;
use LWP::UserAgent;

my $pdf_filename = '/root/Desktop/perl.pdf';
my $pdf_link = 'http://www.greenteapress.com/perl/perl.pdf';

my $client = LWP::UserAgent->new();

my $capture = $client->get("$pdf_link", ":content_file" => "$pdf_filen
+ame");

convert_pdf_to_text();

sub convert_pdf_to_text {
    use CAM::PDF::PageText;
    my $pdf_filename = '/root/Desktop/perl.pdf';
    my $pdf          = CAM::PDF->new($pdf_filename);
    my $y            = $pdf->getPageContentTree(1);
    print CAM::PDF::PageText->render($y);
}
[download]