Since this is an HTML parser, send `Accept: text/html` when fetching the page to parse to be sure that's what you'll get.