Add CookieStore::DeleteSessionCookiesAsync method.

Add CookieStore::DeleteSessionCookiesAsync method.

This is needed by the Android port of Chromium, which has a different
startup and shutdown control flow from other platforms. We also need
to support the Android framework's android.webkit package, which has
a public CookieManager.removeSessionCookie() API.

BUG=None
TEST=CookieMonsterTest


Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=133912

[Iain Merrick, 2012-03-30 12:57:49]

I don't understand the full details of the tests so there's a bit of cargo-cult
copy-and-pasting here. Happy to clean it up if you can give me a few pointers.
The new tests pass and I believe they're testing the right thing, however.

I was able to remove some duplicated code from the SPDY tests, however. Two
identical MockCookieStore classes replaced with a new SimpleCookieStore in
cookie_store_test_helpers.

[erikwright (departed), 2012-04-02 14:12:24]

On 2012/03/30 12:57:49, Iain Merrick wrote:
> I don't understand the full details of the tests so there's a bit of
cargo-cult
> copy-and-pasting here. Happy to clean it up if you can give me a few pointers.
> The new tests pass and I believe they're testing the right thing, however.
> 
> I was able to remove some duplicated code from the SPDY tests, however. Two
> identical MockCookieStore classes replaced with a new SimpleCookieStore in
> cookie_store_test_helpers.

The changes in net/cookies look fine to me.

As for the movement of the mock cookie store from websocket_job_spdy{2,3} to
cookie_store_test_helpers, I'm not that keen on having it there. I would suggest
that you make this a separate CL for the owner of net/websockets and create a
utility file in that directory.

If you want to submit the current CL beforehand, go ahead and add the method to
both of the existing implementations.

(I AM happy to see code duplication removed, but I don't want to get into a
discussion about how to do it in this otherwise non-controversial CL).

[Iain Merrick, 2012-04-02 14:15:52]

On 2012/04/02 14:12:24, erikwright wrote:
> If you want to submit the current CL beforehand, go ahead and add the method
to
> both of the existing implementations.
> 
> (I AM happy to see code duplication removed, but I don't want to get into a
> discussion about how to do it in this otherwise non-controversial CL).

Makes sense. I'll update this patch to keep the duplication, and make a separate
one to remove it.

[Iain Merrick, 2012-04-04 11:14:29]

On 2012/04/02 14:15:52, Iain Merrick wrote:

> I'll update this patch to keep the duplication

Done

[erikwright (departed), 2012-04-04 15:16:22]

LGTM. Note that you will require an owner from net/ as well (I'm only
net/cookies).

[Iain Merrick, 2012-04-16 13:28:36]

+yutak for net/websockets/OWNERS review.

[Yuta Kitamura, 2012-04-16 20:41:29]

lgtm

[commit-bot: I haz the power, 2012-04-25 13:23:38]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/husky@chromium.org/9959011/15001

[commit-bot: I haz the power, 2012-04-25 14:49:42]

Change committed as 133912