Update hotword private tests to not need auto whitelisting.

Update hotword private tests to not need auto whitelisting.

Component apps are currently auto whitelisted for all whitelisted APIs.
To allow this to be removed the hotword private tests are being updated
to manually whitelist the test extensions used.

See https://codereview.chromium.org/447673002/ for more context.

BUG=370375
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=290200

[rpetterson, 2014-08-14 19:57:52]

drive by:

Does this still work? My experience was that in cases where we run two test
extensions in one test (setEnabled, setAudioLoggingEnabled) the two extensions
had to have two separate extension IDs.

[benwells, 2014-08-15 03:12:25]

On 2014/08/14 19:57:52, rpetterson wrote:
> drive by:
> 
> Does this still work? My experience was that in cases where we run two test
> extensions in one test (setEnabled, setAudioLoggingEnabled) the two extensions
> had to have two separate extension IDs.

Yes the tests still pass locally and on the bots. The redness is unrelated.

I believe when you have one extension loaded with an id, and another extension
with the same id is loaded, the second extension replaces the first. If you ever
needed both loaded (e.g. to communicate) it wouldn't work, but in this case it
is fine as the two extensions don't interact at all.

I can change it if you think it is better to have different ids. In that case
we'd probably want a different ID for each extension, I suppose, or it would get
a bit confusing.

[Matt Giuca, 2014-08-15 04:31:48]

https://codereview.chromium.org/470893002/diff/1/chrome/test/data/extensions/...
File
chrome/test/data/extensions/api_test/hotword_private/experimentalHotwordEnabled/manifest.json
(right):

https://codereview.chromium.org/470893002/diff/1/chrome/test/data/extensions/...
chrome/test/data/extensions/api_test/hotword_private/experimentalHotwordEnabled/manifest.json:2:
"key":
"MIGfMA0GCSqGSIb3DQEBAQUAA4GNADCBiQKBgQDCXAxIB5iu+XGtMYYJgSwMrqO+zNa3FlWeMJLOV+U1T2VL6wDU3WD9YNlioI6a6wG49AFquEbRxQwwxlvAZC1c95LBvRlnQAkEVum0KbrJ8WHTxxDEPOfITE0J1AP5j8V0WQ9jbYvUxgefIPhDPXHpdPRAxDotygTrPa33x1075wIDAQAB",
It looks like experimentalHotwordEnabled and isAvailable were sharing a key, and
setAudioLoggingEnableFalse and setEnabledFalse were sharing a key, and you've
changed them all to share the one key. (Is there a reason for this? Just so you
can simplify the apitest.cc file?)

Was anything else depending on these old IDs? In my experience we don't have
keys in test manifests unless we explicitly refer to them elsewhere, which makes
me skeptical.

Can you put a comment above the key line on each manifest like this:
// chrome-extension://cpfhkdbjfdgdebcjlifoldbijinjfifp
(see
chrome/test/data/extensions/platform_apps/custom_launcher_page/manifest.json)
this means I can grep around for the key and discover both the manifest and any
references to it.

[Matt Giuca, 2014-08-15 05:05:56]

https://codereview.chromium.org/470893002/diff/1/chrome/test/data/extensions/...
File
chrome/test/data/extensions/api_test/hotword_private/experimentalHotwordEnabled/manifest.json
(right):

https://codereview.chromium.org/470893002/diff/1/chrome/test/data/extensions/...
chrome/test/data/extensions/api_test/hotword_private/experimentalHotwordEnabled/manifest.json:2:
"key":
"MIGfMA0GCSqGSIb3DQEBAQUAA4GNADCBiQKBgQDCXAxIB5iu+XGtMYYJgSwMrqO+zNa3FlWeMJLOV+U1T2VL6wDU3WD9YNlioI6a6wG49AFquEbRxQwwxlvAZC1c95LBvRlnQAkEVum0KbrJ8WHTxxDEPOfITE0J1AP5j8V0WQ9jbYvUxgefIPhDPXHpdPRAxDotygTrPa33x1075wIDAQAB",
FYI, the old extension keys are:

pnelnpgljhcpjnllfncjcbcfijbdnihl (experimentalHotwordEnabled and isAvailable)
and
ifkaogagfghjdpfcibghheaghgodpjla (setAudioLoggingEnableFalse and
setEnabledFalse)

Neither of these terms can be found in the codebase, so you're good to go!

[Matt Giuca, 2014-08-15 05:07:56]

lgtm if you put in // chrome-extension://blah

[benwells, 2014-08-15 07:15:46]

On 2014/08/15 05:07:56, Matt Giuca wrote:
> lgtm if you put in // chrome-extension://blah

Done.

rlp: is this change ok with you?

[rpetterson, 2014-08-15 21:36:49]

On 2014/08/15 07:15:46, benwells wrote:
> On 2014/08/15 05:07:56, Matt Giuca wrote:
> > lgtm if you put in // chrome-extension://blah
> 
> Done.
> 
> rlp: is this change ok with you?

Totally. The cleanup is great; I just wasn't sure it would work. Glad to hear it
does!

[benwells, 2014-08-17 22:20:46]

The CQ bit was checked by benwells@chromium.org

[commit-bot: I haz the power, 2014-08-17 22:21:21]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/benwells@chromium.org/470893002/20001

[commit-bot: I haz the power, 2014-08-18 01:53:47]

Committed patchset #2 (20001) as 290200