This looks nice. I'm going to try it out. I've used Xpath before and it (mostly) works on well-formed web pages. Also, I'm not completely sure if the advanced parsing mode allows for conditional non-link tests? Something like if title==X, then scrape the page. Good work.
well right now title==X is not supported. However, you can fork the project and contribute. I am also planning to add a css selector to it. So that i would be much easier