Add the URLRequest's priority to the net-log....

Add the URLRequest's priority to the net-log.

BUG=NONE
TEST=open about:net-internals, look at a request dump. there should be a "priority" field.

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=52186

[cbentzel, 2010-07-13 02:23:37]

LGTM

I didn't look at the mac_valgrind compile failure - but I'm guessing it's an ld
memory exhaustion issue. If not, please check before committing this.

http://codereview.chromium.org/2905007/diff/13001/3003
File net/url_request/url_request.cc (right):

http://codereview.chromium.org/2905007/diff/13001/3003#newcode1
net/url_request/url_request.cc:1: // Copyright (c) 2006-2009 The Chromium
Authors. All rights reserved.
Nit: 2010 copyright.

http://codereview.chromium.org/2905007/diff/13001/3004
File net/url_request/url_request_netlog_params.cc (right):

http://codereview.chromium.org/2905007/diff/13001/3004#newcode25
net/url_request/url_request_netlog_params.cc:25: dict->SetInteger(L"priority",
static_cast<int>(priority_));
Descriptive string instead of integer?

[eroman, 2010-07-13 02:34:54]

Thanks for the review!

> I didn't look at the mac_valgrind compile failure - but I'm 
> guessing it's an ld memory exhaustion issue. If not, please
> check before committing this.

Yeah the mac valgrind trybot is hosed, unrelated to this CL.

http://codereview.chromium.org/2905007/diff/13001/3003
File net/url_request/url_request.cc (right):

http://codereview.chromium.org/2905007/diff/13001/3003#newcode1
net/url_request/url_request.cc:1: // Copyright (c) 2006-2009 The Chromium
Authors. All rights reserved.
On 2010/07/13 02:23:37, cbentzel wrote:
> Nit: 2010 copyright.

Done.

http://codereview.chromium.org/2905007/diff/13001/3004
File net/url_request/url_request_netlog_params.cc (right):

http://codereview.chromium.org/2905007/diff/13001/3004#newcode25
net/url_request/url_request_netlog_params.cc:25: dict->SetInteger(L"priority",
static_cast<int>(priority_));
On 2010/07/13 02:23:37, cbentzel wrote:
> Descriptive string instead of integer? 

My thinking is I want to keep the JSON interchange format between browser and
renderer as structured as possible, and leave any formatting decisions up to the
javascript rather than the C++.

For this case, I would like to leave it to a follow-up CL to improve the
pretty-printing of priorities on the javascript end.

[cbentzel, 2010-07-13 02:40:31]

On Mon, Jul 12, 2010 at 10:34 PM, <eroman@chromium.org> wrote:

>
> My thinking is I want to keep the JSON interchange format between
> browser and renderer as structured as possible, and leave any formatting
> decisions up to the javascript rather than the C++.
>
>
I agree with this approach in general, although in the case of enum's I
could imagine passing a string with the symbolic name is reasonable
(otherwise, you'll need to maintain the mapping in both locations).