It's regex search the of the url itself and the page title. It doesn't match against the full document if that's what you mean. That's an interesting idea though.