retrieve? WWW::Mechanize or LWP::UserAgent or threads here that satisfy search-terms like "fetch" and "html" (or you could open your source in a browser and 'save as')
get nice XML? You may want to understand XML before raising this question... See w3.org/TR/rec-xml if you don't have a pretty good handle on what "eXtensible markup language" is... and search out xml parser here, if you do. See also nodes here satisfying a searchterm like "parse."