> i download all html file from a web ,and i know all
> file was downloaded are html file .and 'parsing
> step' spend much time ,any one know how to skip this
> step .

You can't - parsing is necessary to detect links!