If you want to "scrub" or "sanitize" html input in a reliable and flexible fashion, then html::scrubber is for you.

the author was not satisfied with html::sanitizer because it is based on html::treebuilder, so he wrote something similar that works directly with html::parser.