Remove the net_log member from HttpAuthController.

Remove |net_log_| as a class member and pass it in where needed and a little cleanup.

BUG=42795
TEST=existing unit tests

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=52214

[vandebo (ex-Chrome), 2010-07-07 21:10:19]

When you get a chance...

[vandebo (ex-Chrome), 2010-07-12 18:20:53]

ping

[cbentzel, 2010-07-13 02:13:24]

LGTM

You may want to change the subject to be about the NetLog changes to
HttpAuthController, rather than the generic "Clean up HttpAuthController".

http://codereview.chromium.org/2806052/diff/1/6
File net/http/http_proxy_client_socket.h (right):

http://codereview.chromium.org/2806052/diff/1/6#newcode117
net/http/http_proxy_client_socket.h:117: HttpRequestInfo request_;
Is the reordering of member variables intentional? Arguably it may cluster
related variables together which is an improvement.

http://codereview.chromium.org/2806052/diff/1/6#newcode130
net/http/http_proxy_client_socket.h:130: const bool tunnel_;
I haven't seen const used for POD member variables before. It makes sense here,
but it is a bit different from the rest of the code (which usually is of the
pointed-to-object-is-constant form).

[vandebo (ex-Chrome), 2010-07-13 16:53:00]

> You may want to change the subject to be about the NetLog changes to
> HttpAuthController, rather than the generic "Clean up HttpAuthController".

Done.

http://codereview.chromium.org/2806052/diff/1/6
File net/http/http_proxy_client_socket.h (right):

http://codereview.chromium.org/2806052/diff/1/6#newcode117
net/http/http_proxy_client_socket.h:117: HttpRequestInfo request_;
On 2010/07/13 02:13:25, cbentzel wrote:
> Is the reordering of member variables intentional? Arguably it may cluster
> related variables together which is an improvement.

Yes, they were pretty randomly ordered before, and I tried to bring some
semblance of order to them.

http://codereview.chromium.org/2806052/diff/1/6#newcode130
net/http/http_proxy_client_socket.h:130: const bool tunnel_;
On 2010/07/13 02:13:25, cbentzel wrote:
> I haven't seen const used for POD member variables before. It makes sense
here,
> but it is a bit different from the rest of the code (which usually is of the
> pointed-to-object-is-constant form).

I guess const POD isn't used much in the codebase, but I don't see any issues
with this kind of inconsistency. It's easy enough to change if it needs to be
changed, and it provides extra information that makes things easier to
understand.

[cbentzel, 2010-07-13 17:52:26]

LGTM