[webkit-reviews] review cancelled: [Bug 18608] [Gtk] WebKitNetworkRequest needs to be finished : [Attachment 20895] almost complete implementation of a gobject exposing ResourceRequest

[bugzilla-daemon at webkit.org]

Gustavo Noronha Silva <gns at gnome.org> has cancelled Gustavo Noronha Silva
<gns at gnome.org>'s request for review:
Bug 18608: [Gtk] WebKitNetworkRequest needs to be finished
http://bugs.webkit.org/show_bug.cgi?id=18608

Attachment 20895: almost complete implementation of a gobject exposing
ResourceRequest
http://bugs.webkit.org/attachment.cgi?id=20895&action=edit

------- Additional Comments from Gustavo Noronha Silva <gns at gnome.org>
Messing around with the code answered some of my doubts. By looking at
FrameLoaderClientGtk's needs, and webkitwebframe.cpp I understand that the
original ResourceRequest which I'm mapping needs to be carried with
WebKitNetworkRequest, and updated with the values. Currently I'm carrying a
copy, though, and I update the original where it needs to be updated, such as
in the dispatchWillSendRequest delegate.

Some rough edges still need thought and work; a more memory efficient way of
handling the internal hash table which contains the headers, only creating it
when really needed, for instance.

Also, we still have no implementation for ResourceResponse as
WebKitNetworkResponse, so the new signal I created for webkitwebview,
request-started is missing the redirectResponse argument it should have. I will
write a placeholder to fix this, or perhaps even implemente
WebKitNetworkResponse completely for this patch to land, if you prefer it that
way.

Sounds about right?