[Ocs] Review Request 109504: Do basic HTTP authentication using QNetworkAccessManager, not manually

[Frederik Gladhorn]

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
http://git.reviewboard.kde.org/r/109504/#review29357
-----------------------------------------------------------

Ship it!


There are two style nit-picks that you could fix, but please commit it. I guess master is the right spot.
Thanks!


lib/atticabasejob.h
<http://git.reviewboard.kde.org/r/109504/#comment21942>

    deleting this line is unrelated, but I don't care too much



lib/provider.cpp
<http://git.reviewboard.kde.org/r/109504/#comment21943>

    one set of parenthesis would be enough,
    (QNetworkRequest::Attribute) BaseJob::UserAttribute,
    but I don't care much either way.


- Frederik Gladhorn


On March 16, 2013, 6:51 p.m., Sven Brauch wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> http://git.reviewboard.kde.org/r/109504/
> -----------------------------------------------------------
> 
> (Updated March 16, 2013, 6:51 p.m.)
> 
> 
> Review request for Attica and Frederik Gladhorn.
> 
> 
> Description
> -------
> 
> Until now, attica wrote the credentials into the URL and into the header
> on every request as soon as the user had entered them.
> This could lead to warning dialogs from the HTTP kioslave about
> possible address spoofing, since credentials must only be sent if the
> server requests them explicitly. I noticed this when writing my own OCS server;
> please don't ask me why it even works for opendesktop ;) Possibly it has to do with
> the SSL encryption they use.
> This patch changes attica to use the "official" way to supply credentials, which is
> connecting to the signal the QNetworkAccessManager procides, and filling in
> the credentials there.
> To keep this change as local as possible, the credentials are stored in
> the User attribute fields of the QNetworkRequest, because then they
> can be both set and read in one central place.
> 
> 
> Diffs
> -----
> 
>   lib/atticabasejob.h 9259fa3 
>   lib/atticabasejob.cpp feffab8 
>   lib/provider.cpp 309e117 
> 
> Diff: http://git.reviewboard.kde.org/r/109504/diff/
> 
> 
> Testing
> -------
> 
> Manual testing.
> 
> 
> Thanks,
> 
> Sven Brauch
> 
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freedesktop.org/archives/ocs/attachments/20130317/05da22d5/attachment.html>