Re: [Bug-wget] Filename to save to

[Mike Frysinger]

On Thursday 05 January 2012 17:29:48 Volker Kuhlmann wrote:
> Many web servers offer files only as a result of some CGI processing or
> some such mechanism, which doesn't have the filename to save to in the
> URL. The filename is instead supplied as part of the http headers, like:
> 
>   Content-Length: 478508
>   Content-Disposition: attachment; filename=ReceiverShield-36ba699.tar.gz
> 
> It would be really useful if wget was using this filename to save to.
> The current situation is that wget saves to the last component of the
> URL, which requires at least a manual rename. Second-worst case is that the
> last URL component is a kilometre long, worst case is that part is not a
> legal filename and wget dies trying to save the file without saving
> anything.

not a bug.  look up "CVE-2010-2252" as to why this is correct default 
behavior.  use --trust-server-names if you want to use the server name.
-mike

signature.asc
Description: This is a digitally signed message part.