#!perl
use strict;
use HTML::TreeBuilder::XPath; 
my $tree = HTML::TreeBuilder::XPath->new;  
$tree->parse_file(\*DATA);

my @items = $tree->findnodes( '//div[@class="review-content"]' )
   or die("no items: $!\n");
for my $item (@items) {
  for ( $item->findnodes( '//meta') ){
    print $_->attr('itemprop');
    print ' = ';
    print $_->attr('content')."\n";
  }
}