Move UbertokenFetcher from //chrome to //google_apis.

Move UbertokenFetcher from //chrome to //google_apis.

This CL eliminates //chrome- and //content-level dependencies from
UbertokenFetcher and moves it to //google_apis/gaia. Notable changes:
- Change UbertokenFetcher to take in the OAuth2TokenService and
  URLRequestContextGetter to use rather than the Profile.
- Add a FakeOAuth2TokenService to enable moving the UbertokenFetcher unittest
  away from using FakeProfileOAuth2TokenService.

BUG=330292
TBR=thakis

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=245359

[blundell, 2014-01-15 12:09:50]

https://codereview.chromium.org/136723009/diff/1/chrome/browser/signin/google...
File chrome/browser/signin/google_auto_login_helper.cc (right):

https://codereview.chromium.org/136723009/diff/1/chrome/browser/signin/google...
chrome/browser/signin/google_auto_login_helper.cc:84:
uber_token_fetcher_.reset(new UbertokenFetcher(
I didn't want to tackle GoogleAutoLoginHelper in this CL; that work will shortly
follow.

https://codereview.chromium.org/136723009/diff/1/google_apis/gaia/ubertoken_f...
File google_apis/gaia/ubertoken_fetcher_unittest.cc (left):

https://codereview.chromium.org/136723009/diff/1/google_apis/gaia/ubertoken_f...
google_apis/gaia/ubertoken_fetcher_unittest.cc:82:
ProfileOAuth2TokenServiceFactory::GetForProfile(profile())->
These calls were no longer necessary once the account ID was being supplied
explicitly to StartFetchingToken(). I didn't realize this in the prior CL.

[Roger Tawa OOO till Jul 10th, 2014-01-15 14:41:19]

lgtm

Great job, thanks Colin!

https://codereview.chromium.org/136723009/diff/1/google_apis/gaia/ubertoken_f...
File google_apis/gaia/ubertoken_fetcher.cc (right):

https://codereview.chromium.org/136723009/diff/1/google_apis/gaia/ubertoken_f...
google_apis/gaia/ubertoken_fetcher.cc:35: DCHECK(&account_id);
Did you mean: DCHECK(!account_id.empty()) ?

[blundell, 2014-01-15 14:45:59]

I ran into an unexpected problem on the trybots: there is a
FakeOAuth2TokenService class in //chrome/browser/sync (it subclasses
ProfileOAuth2TokenService). zea@, do you have a preference for what to rename
the existing one? The best I can think of is
FakeProfileOAuth2TokenProfileForSync (yikes!).

https://codereview.chromium.org/136723009/diff/1/google_apis/gaia/ubertoken_f...
File google_apis/gaia/ubertoken_fetcher.cc (right):

https://codereview.chromium.org/136723009/diff/1/google_apis/gaia/ubertoken_f...
google_apis/gaia/ubertoken_fetcher.cc:35: DCHECK(&account_id);
Oops, this is just crud that can go away.

On 2014/01/15 14:41:19, Roger Tawa wrote:
> Did you mean: DCHECK(!account_id.empty()) ?

[Roger Tawa OOO till Jul 10th, 2014-01-15 15:02:23]

Should change sync to use FakeProfileOAuth2TokenService instead, but that makes
this CL bigger.  Creating a XXXForSync class for now is fine, but longer term we
should not have fakes outside of google_apis\gaia and possibly
chrome\sync\signin.

[Roger Tawa OOO till Jul 10th, 2014-01-15 15:36:18]

Doh!  In my last comment, I said "chrome\sync\signin", but I meant
"chrome\browser\signin".

[blundell, 2014-01-16 16:01:36]

zea@: I renamed //chrome/browser/sync's FakeOAuth2TokenService to
FakeOAuth2TokenServiceForSync (diff between PS1 and PS2). Please take a look at
//chrome/browser/sync. Thanks!

https://codereview.chromium.org/136723009/diff/1/google_apis/gaia/ubertoken_f...
File google_apis/gaia/ubertoken_fetcher.cc (right):

https://codereview.chromium.org/136723009/diff/1/google_apis/gaia/ubertoken_f...
google_apis/gaia/ubertoken_fetcher.cc:35: DCHECK(&account_id);
On 2014/01/15 14:46:00, blundell wrote:
> Oops, this is just crud that can go away.
> 
> On 2014/01/15 14:41:19, Roger Tawa wrote:
> > Did you mean: DCHECK(!account_id.empty()) ?
> 

Done.

[Nicolas Zea, 2014-01-16 18:50:24]

Kind of an odd name, but not sure what's better :-/
sync LGTM

[blundell, 2014-01-16 19:54:41]

TBR=thakis for the //chrome changes outside of //chrome/browser/signin and
//chrome/browser/sync

[commit-bot: I haz the power, 2014-01-16 19:56:24]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/blundell@chromium.org/136723009/300001

[blundell, 2014-01-16 19:56:47]

The functionality of the fake in //chrome/browser/sync should probably just move
into the one in //chrome/browser/signin in the short-to-medium term.

[commit-bot: I haz the power, 2014-01-16 22:53:01]

Change committed as 245359