I'm sure you could do what you asked with that HTMLParser project. I doubt that there's a specific method to do it, though. Although now that I look through the API documentation, there's
a Parser class and it has a method named "extractAllNodesThatMatch". Possibly -- actually quite likely now that I look at the docs more -- you could use that.
But yeah, the project really doesn't have much in the way of useful examples. You're pretty much left to trawl through the docs and figure it out for yourself. Although looking at their examples wouldn't do you any harm.