@sharovatov HTML is limited, at least at this point in history, in true parsing. We have to deal with huge amounts of error handling, etc.