Integrate throttling logic into AffiliationBackend.

Integrate throttling logic into AffiliationBackend.

This CL also changes ownership of the TickClock object used by the AffiliationFetchThrottler, so that it is now owned by the AffiliationBackend.

Furthermore, this CL makes the AffiliationBackend no longer rely on using ThreadTaskRunnerHandle to schedule asynchronous operations, but instead take a SingleThreadTaskRunner to use in its constructor. This allows delayed tasks (which are now present) to be posted to the correct task runner in tests too (namely, in AffiliationServiceTests).

BUG=437865
Committed: https://crrev.com/4865f4b36ecdda55bd88e9ffc1cacb1ac92f3cbc
Cr-Commit-Position: refs/heads/master@{#320887}

[engedy, 2015-03-16 18:27:05]

engedy@chromium.org changed reviewers:
+ mkwst@chromium.org

[engedy, 2015-03-16 18:29:44]

@Mike, could you please take a look?

[Mike West, 2015-03-16 20:58:04]

LGTM % a small nit.

https://codereview.chromium.org/1008373003/diff/40001/components/password_man...
File components/password_manager/core/browser/affiliation_backend_unittest.cc
(right):

https://codereview.chromium.org/1008373003/diff/40001/components/password_man...
components/password_manager/core/browser/affiliation_backend_unittest.cc:459:
AdvanceTime(GetShortTestPeriod());
Nit: Can you either advance this a bit more than 'GetShortTestPeriod()', or note
explicitly in a comment that the time range is supposed to be exclusive? It just
seems unclear what you're really testing if you're explicitly picking the edge
case (for instance: is no fetch required because the Prefetch expired, or is no
fetch required because the data isn't yet stale, etc).

[engedy, 2015-03-17 08:40:44]

Patchset #4 (id:60001) has been deleted

[engedy, 2015-03-17 08:41:03]

https://codereview.chromium.org/1008373003/diff/40001/components/password_man...
File components/password_manager/core/browser/affiliation_backend_unittest.cc
(right):

https://codereview.chromium.org/1008373003/diff/40001/components/password_man...
components/password_manager/core/browser/affiliation_backend_unittest.cc:459:
AdvanceTime(GetShortTestPeriod());
On 2015/03/16 20:58:04, Mike West wrote:
> Nit: Can you either advance this a bit more than 'GetShortTestPeriod()', or
note
> explicitly in a comment that the time range is supposed to be exclusive? It
just
> seems unclear what you're really testing if you're explicitly picking the edge
> case (for instance: is no fetch required because the Prefetch expired, or is
no
> fetch required because the data isn't yet stale, etc).

Done.

[engedy, 2015-03-17 08:44:00]

The CQ bit was checked by engedy@chromium.org

[engedy, 2015-03-17 08:44:00]

The patchset sent to the CQ was uploaded after l-g-t-m from mkwst@chromium.org
Link to the patchset: https://codereview.chromium.org/1008373003/#ps80001
(title: "Addressed comment.")

[engedy, 2015-03-17 08:44:04]

The CQ bit was checked by engedy@chromium.org

[commit-bot: I haz the power, 2015-03-17 08:44:08]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1008373003/80001

[commit-bot: I haz the power, 2015-03-17 10:44:38]

Committed patchset #4 (id:80001)

[commit-bot: I haz the power, 2015-03-17 10:45:14]

Patchset 4 (id:??) landed as
https://crrev.com/4865f4b36ecdda55bd88e9ffc1cacb1ac92f3cbc
Cr-Commit-Position: refs/heads/master@{#320887}