Beautiful Soup parses arbitrarily invalid XML- or HTML-like substance
into a tree representation. It provides methods and Pythonic idioms
that make it easy to search and modify the tree.

A well-formed XML/HTML document will yield a well-formed data
structure. An ill-formed XML/HTML document will yield a
correspondingly ill-formed data structure. If your document is only
locally well-formed, you can use this library to find and process the
well-formed part of it. The BeautifulSoup class has heuristics for
obtaining a sensible parse tree in the face of common HTML errors.

WWW: http://www.crummy.com/software/BeautifulSoup/