Add browser tests for clearing sync registrations with site data

Add browser tests for clearing sync registrations with site data

BUG=482083
Committed: https://crrev.com/81b3ca96b0b54bba170a709f515eaae47181066b
Cr-Commit-Position: refs/heads/master@{#357167}

[chasej, 2015-10-29 21:25:17]

chasej@chromium.org changed reviewers:
+ jkarlin@chromium.org

[chasej, 2015-10-29 21:25:18]

jkarlin, please take a look.

[jkarlin, 2015-10-30 13:41:14]

Looks awesome.

https://codereview.chromium.org/1423313006/diff/1/content/browser/background_...
File content/browser/background_sync/background_sync_browsertest.cc (right):

https://codereview.chromium.org/1423313006/diff/1/content/browser/background_...
content/browser/background_sync/background_sync_browsertest.cc:113: return
BrowserContext::GetDefaultStoragePartition(
GetDefaultStoragePartition is deprecated, while you're here can you change this
to GetStoragePartition(shell_->web_contents()->GetBrowserContext(), nullptr); ?

https://codereview.chromium.org/1423313006/diff/1/content/browser/background_...
content/browser/background_sync/background_sync_browsertest.cc:737: }
Can you also test what happens if a sync is in the middle of firing when the
storage partition is cleared?

[chasej, 2015-10-30 15:59:23]

jkarlin, please take another look.

https://codereview.chromium.org/1423313006/diff/1/content/browser/background_...
File content/browser/background_sync/background_sync_browsertest.cc (right):

https://codereview.chromium.org/1423313006/diff/1/content/browser/background_...
content/browser/background_sync/background_sync_browsertest.cc:113: return
BrowserContext::GetDefaultStoragePartition(
On 2015/10/30 13:41:13, jkarlin wrote:
> GetDefaultStoragePartition is deprecated, while you're here can you change
this
> to GetStoragePartition(shell_->web_contents()->GetBrowserContext(), nullptr);
?

Done. I changed to pass the actual site instance for the second parameter, as
GetStoragePosition() indicates that will eventually be required.

https://codereview.chromium.org/1423313006/diff/1/content/browser/background_...
content/browser/background_sync/background_sync_browsertest.cc:737: }
On 2015/10/30 13:41:13, jkarlin wrote:
> Can you also test what happens if a sync is in the middle of firing when the
> storage partition is cleared? 

Done.

[jkarlin, 2015-10-30 16:37:38]

https://codereview.chromium.org/1423313006/diff/20001/content/browser/backgro...
File content/browser/background_sync/background_sync_browsertest.cc (right):

https://codereview.chromium.org/1423313006/diff/20001/content/browser/backgro...
content/browser/background_sync/background_sync_browsertest.cc:759: //
TODO(jkarlin): Use registration.finished to verify that the event actually
registration.finished exists now, you can use that to verify that the event
succeeds once you stop delaying it.

[chasej, 2015-10-30 18:13:28]

jkarlin, hopefully last look.

https://codereview.chromium.org/1423313006/diff/20001/content/browser/backgro...
File content/browser/background_sync/background_sync_browsertest.cc (right):

https://codereview.chromium.org/1423313006/diff/20001/content/browser/backgro...
content/browser/background_sync/background_sync_browsertest.cc:759: //
TODO(jkarlin): Use registration.finished to verify that the event actually
On 2015/10/30 16:37:38, jkarlin wrote:
> registration.finished exists now, you can use that to verify that the event
> succeeds once you stop delaying it.

As discussed, this test shouldn't verify the behaviour of .finished when firing
sync events are deleted.  Removed the TODO.

[jkarlin, 2015-10-30 18:13:43]

lgtm

https://codereview.chromium.org/1423313006/diff/20001/content/browser/backgro...
File content/browser/background_sync/background_sync_browsertest.cc (right):

https://codereview.chromium.org/1423313006/diff/20001/content/browser/backgro...
content/browser/background_sync/background_sync_browsertest.cc:759: //
TODO(jkarlin): Use registration.finished to verify that the event actually
On 2015/10/30 16:37:38, jkarlin wrote:
> registration.finished exists now, you can use that to verify that the event
> succeeds once you stop delaying it.

Actually, the result of .finished shouldn't be tested quite yet as it's unclear
what the result of it should be.

[chasej, 2015-10-30 18:17:43]

The CQ bit was checked by chasej@chromium.org

[commit-bot: I haz the power, 2015-10-30 18:18:55]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1423313006/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1423313006/40001

[commit-bot: I haz the power, 2015-10-30 19:20:45]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2015-10-30 19:22:18]

Patchset 3 (id:??) landed as
https://crrev.com/81b3ca96b0b54bba170a709f515eaae47181066b
Cr-Commit-Position: refs/heads/master@{#357167}