IMO The probablity of a given page being well formed, or being non-well formed in a manner that will not completely screw the parser is probably very low. Therefore, you should use a dedicated html parser for parsing html.
3 karma · joined November 15, 2012