I'm going to write one up that follows the new york times api rather than a meta tag scraper...it seems fairly simple to do...