HTML::TokeParser is just a wrapper on top of HTML::Parser. The example I gave is in API 2 callback style. You can also use the API 3 style new constructor but I find it is not really as clear. That bit of logic - increment a counter on open tag, decrement on close, print if counter on, don't if not, and fix the comments gets you a lot of mileage..... Toke::Parser is far easier to use, and much more pleasing to the eye but you loose some of the full power of Parser.