You would think the native DOM parsing engine would be quicker than trying to add another layer of abstraction and parsing which would incur additional overhead