could make it build opml files by crawling the links for titles and feeds.