> I think that this is only matter of implementing recursive copy in
> http slave. The same way as it is done in file and ftp slave.
> 
> Kget supports copying of directories and because of this, it could
> be used also for recursive copying of some http server, as soon as 
> the above is implemented.
> 
> For recursive copying from ftp and http sites, there could be some
> options e.g. for specifying the depth of copying.

ioslaves should not do recursive copying for http. The http protocol
has no way of telling you what files to get. For this to work you 
have to check whether the file you get is html and parse the html.
This funcionality is in khtml and does not belong in a ioslave.

If you want to have a recursive get, you should get the html file,
parse it, ask which links it has, and then ask the kioslave to get
the links.

You can probably use khtml to do your parsing, without actually
displaying the page. If you can't it should be possible to adapt
khtml to make this possible.

HTML parsing is done in khtml, and only there.

Cheers,
Waldo Bastian
bastian@kde.org