Connelly Barnes offers his attempt at an HTML-like parser-thingy: htmldata.

Try it. You might like it.