Only apply extension throttling quota to sync, not local.

Only apply extension throttling quota to sync, not local.

BUG=120596
TEST=browser_tests --gtest_filter=*ExtensionSettings* (new test)

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=134506

[not at google - send to devlin, 2012-04-27 05:21:51]

There were 4 ways I looked into to write this patch:

1. Remove the caching from GetQuotaLimitHeuristics.
2. Make application of the ExtensionsQuotaService done from within the extension
functions themselves.
3. Make the BucketMapper able to handle infinite buckets or something.
4. Add a new override-able method to skip applying any functions.

1. Removing the caching is not something I'm comfortable doing.
2. This would be nice but it'd be a change in behaviour that I don't want to
think about (e.g. it would change the way errors work) and would also require
understanding the way that APIs use this.
3. Had a look at this, seems messy.
4. Also messy since adding a function is unfortunate, but it's the least
invasive change.

... so there you go.

[Matt Perry, 2012-04-27 18:48:02]

Yay for simple changes. LGTM

https://chromiumcodereview.appspot.com/10232004/diff/2001/chrome/test/data/ex...
File chrome/test/data/extensions/api_test/settings/simple_test/background.js
(right):

https://chromiumcodereview.appspot.com/10232004/diff/2001/chrome/test/data/ex...
chrome/test/data/extensions/api_test/settings/simple_test/background.js:362:
clearNTimes(sync, 1001, test.callbackFail(quotaError, function() {
could you do sync and local in parallel?

[not at google - send to devlin, 2012-04-29 23:58:03]

http://codereview.chromium.org/10232004/diff/2001/chrome/test/data/extensions...
File chrome/test/data/extensions/api_test/settings/simple_test/background.js
(right):

http://codereview.chromium.org/10232004/diff/2001/chrome/test/data/extensions...
chrome/test/data/extensions/api_test/settings/simple_test/background.js:362:
clearNTimes(sync, 1001, test.callbackFail(quotaError, function() {
On 2012/04/27 18:48:03, Matt Perry wrote:
> could you do sync and local in parallel?

hm, not sure, I definitely wanted to test those 4 conditions: that local doesn't
exceed, that sync does, that after quota is exceeded local still fails, and that
after local succeeds sync still fails.

After playing with doing it in parallel for a bit, it's a tiny bit messy to
express that "after 1000 writes, sync should have failed, and from then on
should always fail". I dunno, not that complicated, but more than this anyway.

[commit-bot: I haz the power, 2012-04-29 23:59:13]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/kalman@chromium.org/10232004/8002

[commit-bot: I haz the power, 2012-04-30 03:21:09]

Change committed as 134506