Implement the calls to GAIA for the IDP IFrame protocol.

Implement the calls to GAIA for the IDP IFrame protocol.

This CL implements Fetchers which can call endpoints defined for the
IDP IFrame protocol, specifically listSessions and getToken. This will
be used as part of the Cross-Device Promo so that Chrome can mint
access tokens of a particular scope based on cookies.

BUG=463611
Committed: https://crrev.com/efc601eba6eb6b7efdaa38d2c49b7d3665e7f9c7
Cr-Commit-Position: refs/heads/master@{#319829}

[Mike Lerman, 2015-03-04 20:01:48]

mlerman@chromium.org changed reviewers:
+ guibinkong@google.com, zelidrag@chromium.org

[Mike Lerman, 2015-03-04 20:01:48]

Hi zelidrag@,

roger's OOO this week. This isn't a super-urgent CL, but if you're up for taking
a look at it, it'll be one less back-logged item for Roger to do when he comes
back :)

Guibin,

I don't know if you've looked at Chrome code before, but this calls your
endpoint. Can you look at the request formatting and response parsing in
gaia_auth_fetcher.cc? 

Thanks!

[Mike Lerman, 2015-03-06 20:08:56]

mlerman@chromium.org changed reviewers:
+ rogerta@chromium.org

[Mike Lerman, 2015-03-06 20:08:57]

Roger, want to look this over when you're back? Thanks.

[Roger Tawa OOO till Jul 10th, 2015-03-09 14:41:57]

lgtm, some comments below.

https://codereview.chromium.org/973953002/diff/20001/google_apis/gaia/gaia_au...
File google_apis/gaia/gaia_auth_fetcher.cc (right):

https://codereview.chromium.org/973953002/diff/20001/google_apis/gaia/gaia_au...
google_apis/gaia/gaia_auth_fetcher.cc:75: const char* kGetTokenResponseRequested
= "get_token";
Use const char xxx[] = "..."; syntax.

https://codereview.chromium.org/973953002/diff/20001/google_apis/gaia/gaia_au...
google_apis/gaia/gaia_auth_fetcher.cc:470: "e=3100087&"               //
temporarily enable the experiment.
Nit: only 2 spaces between " and //.   Same for line 448.

https://codereview.chromium.org/973953002/diff/20001/google_apis/gaia/gaia_au...
google_apis/gaia/gaia_auth_fetcher.cc:569: if
(sessionDictionary->GetString("login_hint", login_hint))
Should that be &login_hint ?

https://codereview.chromium.org/973953002/diff/20001/google_apis/gaia/gaia_au...
google_apis/gaia/gaia_auth_fetcher.cc:1064: ParseListIdpSessionsResponse(data,
&login_hint);
If the parse fails, should call OnListIdpSessionsError().

https://codereview.chromium.org/973953002/diff/20001/google_apis/gaia/gaia_au...
File google_apis/gaia/gaia_auth_fetcher_unittest.cc (right):

https://codereview.chromium.org/973953002/diff/20001/google_apis/gaia/gaia_au...
google_apis/gaia/gaia_auth_fetcher_unittest.cc:856: .Times(1);
Nit: does this fit on line 855?

[Mike Lerman, 2015-03-09 18:11:56]

mlerman@chromium.org changed reviewers:
- zelidrag@chromium.org

[guibinkong, 2015-03-09 18:26:51]

https://codereview.chromium.org/973953002/diff/1/google_apis/gaia/gaia_auth_f...
File google_apis/gaia/gaia_auth_fetcher.cc (right):

https://codereview.chromium.org/973953002/diff/1/google_apis/gaia/gaia_auth_f...
google_apis/gaia/gaia_auth_fetcher.cc:449: "ss_domain=%s&"
remove ss_domain, since ss_domain is same as origin, in which case, ss_domain
isn't required.

https://codereview.chromium.org/973953002/diff/1/google_apis/gaia/gaia_auth_f...
google_apis/gaia/gaia_auth_fetcher.cc:471: "response_type=token+id_token&"
remove '+id_token', since you don't need the ID Token .

[Mike Lerman, 2015-03-09 18:59:57]

Thank you both.

https://codereview.chromium.org/973953002/diff/1/google_apis/gaia/gaia_auth_f...
File google_apis/gaia/gaia_auth_fetcher.cc (right):

https://codereview.chromium.org/973953002/diff/1/google_apis/gaia/gaia_auth_f...
google_apis/gaia/gaia_auth_fetcher.cc:449: "ss_domain=%s&"
On 2015/03/09 18:26:50, guibinkong wrote:
> remove ss_domain, since ss_domain is same as origin, in which case, ss_domain
> isn't required.

Done.

https://codereview.chromium.org/973953002/diff/1/google_apis/gaia/gaia_auth_f...
google_apis/gaia/gaia_auth_fetcher.cc:471: "response_type=token+id_token&"
On 2015/03/09 18:26:50, guibinkong wrote:
> remove '+id_token', since you don't need the ID Token .

Done.

https://codereview.chromium.org/973953002/diff/20001/google_apis/gaia/gaia_au...
File google_apis/gaia/gaia_auth_fetcher.cc (right):

https://codereview.chromium.org/973953002/diff/20001/google_apis/gaia/gaia_au...
google_apis/gaia/gaia_auth_fetcher.cc:75: const char* kGetTokenResponseRequested
= "get_token";
On 2015/03/09 14:41:57, Roger Tawa wrote:
> Use const char xxx[] = "..."; syntax.

Done.

https://codereview.chromium.org/973953002/diff/20001/google_apis/gaia/gaia_au...
google_apis/gaia/gaia_auth_fetcher.cc:470: "e=3100087&"               //
temporarily enable the experiment.
On 2015/03/09 14:41:57, Roger Tawa wrote:
> Nit: only 2 spaces between " and //.   Same for line 448.

Done.

https://codereview.chromium.org/973953002/diff/20001/google_apis/gaia/gaia_au...
google_apis/gaia/gaia_auth_fetcher.cc:569: if
(sessionDictionary->GetString("login_hint", login_hint))
On 2015/03/09 14:41:57, Roger Tawa wrote:
> Should that be &login_hint ?

It's already a pointer, so just login_hint should be fine.

https://codereview.chromium.org/973953002/diff/20001/google_apis/gaia/gaia_au...
google_apis/gaia/gaia_auth_fetcher.cc:1064: ParseListIdpSessionsResponse(data,
&login_hint);
On 2015/03/09 14:41:57, Roger Tawa wrote:
> If the parse fails, should call OnListIdpSessionsError().

Done.

https://codereview.chromium.org/973953002/diff/20001/google_apis/gaia/gaia_au...
File google_apis/gaia/gaia_auth_fetcher_unittest.cc (right):

https://codereview.chromium.org/973953002/diff/20001/google_apis/gaia/gaia_au...
google_apis/gaia/gaia_auth_fetcher_unittest.cc:856: .Times(1);
On 2015/03/09 14:41:57, Roger Tawa wrote:
> Nit: does this fit on line 855?

Yes!

[guibinkong, 2015-03-09 20:51:20]

lgtm

[Mike Lerman, 2015-03-10 02:08:58]

The CQ bit was checked by mlerman@chromium.org

[Mike Lerman, 2015-03-10 02:08:58]

The patchset sent to the CQ was uploaded after l-g-t-m from rogerta@chromium.org
Link to the patchset: https://codereview.chromium.org/973953002/#ps40001 (title:
"Roger and Guibin's nits and comments")

[commit-bot: I haz the power, 2015-03-10 02:09:52]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/973953002/40001

[commit-bot: I haz the power, 2015-03-10 03:02:29]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2015-03-10 03:03:00]

Patchset 3 (id:??) landed as
https://crrev.com/efc601eba6eb6b7efdaa38d2c49b7d3665e7f9c7
Cr-Commit-Position: refs/heads/master@{#319829}