This class can parse HTML documents using regular expressions to split the document character data, the begin and end tag elements.
For each type of elements, the class call custom callback that can do any type of processing. The callback functions may bu global or be functions of a class of a given object.
Yesterday, I was working on PRAjax. The UpdatePanel did not work completely as I wanted it to work: in the background, the whole page was still fetched and updated. A cleaner way would be to just fetch updated content and not the whole page...