Change MediaKeys.setServerCertificate() to return Promise<boolean>

Change MediaKeys.setServerCertificate() to return Promise<boolean>

https://github.com/w3c/encrypted-media/issues/145 changed the EME spec
so that setServerCertificate() returns Promise<boolean>. Update the
implementation to match.

This change updates the IDL and returns true on success.
NOT_SUPPORTED_ERRORs will resolve with false.

BUG=618800
TEST=EME tests pass, www.netflix.com plays

Committed: https://crrev.com/e3d85c9614d5864bc3a9506b529036f2d703e0ec
Cr-Commit-Position: refs/heads/master@{#399595}

[jrummell, 2016-06-09 23:13:50]

jrummell@chromium.org changed reviewers:
+ ddorwin@chromium.org, xhwang@chromium.org

[jrummell, 2016-06-09 23:13:51]

PTAL. This is just updating the IDL to return Promise<boolean>. Actual plumbing
through Chromium will come later.

[xhwang, 2016-06-10 21:40:11]

https://chromiumcodereview.appspot.com/2056053003/diff/1/third_party/WebKit/S...
File third_party/WebKit/Source/modules/encryptedmedia/MediaKeys.cpp (right):

https://chromiumcodereview.appspot.com/2056053003/diff/1/third_party/WebKit/S...
third_party/WebKit/Source/modules/encryptedmedia/MediaKeys.cpp:106:
resolve(true);
Can we translate completeWithError(NOT_SUPPORTED_ERROR) to resolve(false)? Then
we don't need to really change chromium other than making sure the right
exception is used.

NOT_SUPPORTED_ERROR means "The operation is not supported", which seems
appropriate for this use:

https://www.w3.org/TR/dom/#notsupportederror

[jrummell, 2016-06-10 23:00:47]

Description was changed from

==========
Change MediaKeys.setServerCertificate() to return Promise<boolean>

https://github.com/w3c/encrypted-media/issues/145 changed the EME spec
so that setServerCertificate() returns Promise<boolean>. Update the
implementation to match.

This change only updates the IDL and returns true on success.
Returning false if the CDM does not support a server certificate
will be implmented in a future CL.

BUG=618800
TEST=EME tests pass, www.netflix.com plays
==========

to

==========
Change MediaKeys.setServerCertificate() to return Promise<boolean>

https://github.com/w3c/encrypted-media/issues/145 changed the EME spec
so that setServerCertificate() returns Promise<boolean>. Update the
implementation to match.

This change updates the IDL and returns true on success.
NOT_SUPPORTED_ERRORs will resolve with false.

BUG=618800
TEST=EME tests pass, www.netflix.com plays
==========

[jrummell, 2016-06-10 23:01:15]

Updated so that NOT_SUPPORTED_ERRORs will resolve with false.

https://codereview.chromium.org/2056053003/diff/1/third_party/WebKit/Source/m...
File third_party/WebKit/Source/modules/encryptedmedia/MediaKeys.cpp (right):

https://codereview.chromium.org/2056053003/diff/1/third_party/WebKit/Source/m...
third_party/WebKit/Source/modules/encryptedmedia/MediaKeys.cpp:106:
resolve(true);
On 2016/06/10 21:40:11, xhwang wrote:
> Can we translate completeWithError(NOT_SUPPORTED_ERROR) to resolve(false)?
Then
> we don't need to really change chromium other than making sure the right
> exception is used.
> 
> NOT_SUPPORTED_ERROR means "The operation is not supported", which seems
> appropriate for this use:
> 
> https://www.w3.org/TR/dom/#notsupportederror

Done. I checked all the calls for SetServerCertificate(), and the only one
rejecting with NOT_SUPPORTED_ERROR is in aes_decryptor, so it provides the
desired behavior without lots of changes on the Chromium side.

[ddorwin, 2016-06-13 17:28:09]

Thanks.

https://codereview.chromium.org/2056053003/diff/20001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/media/encrypted-media/encrypted-media-async-setcert-with-gc.html
(right):

https://codereview.chromium.org/2056053003/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/media/encrypted-media/encrypted-media-async-setcert-with-gc.html:24:
});
Do we need something like this here to catch an incorrect rejection? This
appears to pass without a done() call in the expected path.
.catch(function(error) {
  forceTestFailureFromPromise

https://codereview.chromium.org/2056053003/diff/20001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/media/encrypted-media/encrypted-media-syntax.html
(right):

https://codereview.chromium.org/2056053003/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/media/encrypted-media/encrypted-media-syntax.html:974:
assert_false(ArrayBuffer.isView(cert.buffer));
What does isView() of the buffer tell us? Aren't we trying to pass an
ArrayBuffer?

https://codereview.chromium.org/2056053003/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/media/encrypted-media/encrypted-media-syntax.html:991:
navigator.requestMediaKeySystemAccess('org.w3.clearkey',
[{}]).then(function(access) {
FYI, an empty configuration is no longer allowed by the spec. We're going to
need to change this to getDummyConfiguration() or something like that
throughout.

https://codereview.chromium.org/2056053003/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/media/encrypted-media/encrypted-media-syntax.html:1001:
test.done();
We don't check the result. What is the result for all()?

[xhwang, 2016-06-13 17:45:49]

looking good % nit and ddorwin's comments

https://chromiumcodereview.appspot.com/2056053003/diff/20001/third_party/WebK...
File third_party/WebKit/Source/modules/encryptedmedia/MediaKeys.cpp (right):

https://chromiumcodereview.appspot.com/2056053003/diff/20001/third_party/WebK...
third_party/WebKit/Source/modules/encryptedmedia/MediaKeys.cpp:109: // false."
So convert any NOTSUPPORTEDERRORS into resolving with false.
s/NOTSUPPORTEDERRORS/NOTSUPPORTEDERROR/

[jrummell, 2016-06-13 18:54:13]

Updated.

https://codereview.chromium.org/2056053003/diff/20001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/media/encrypted-media/encrypted-media-async-setcert-with-gc.html
(right):

https://codereview.chromium.org/2056053003/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/media/encrypted-media/encrypted-media-async-setcert-with-gc.html:24:
});
On 2016/06/13 17:28:09, ddorwin wrote:
> Do we need something like this here to catch an incorrect rejection? This
> appears to pass without a done() call in the expected path.
> .catch(function(error) {
>   forceTestFailureFromPromise

This is a promise_test(), so it fails if the promise fails. The older tests use
async_test(), where we need to fail the test if the promise is rejected.

https://codereview.chromium.org/2056053003/diff/20001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/media/encrypted-media/encrypted-media-syntax.html
(right):

https://codereview.chromium.org/2056053003/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/media/encrypted-media/encrypted-media-syntax.html:974:
assert_false(ArrayBuffer.isView(cert.buffer));
On 2016/06/13 17:28:09, ddorwin wrote:
> What does isView() of the buffer tell us? Aren't we trying to pass an
> ArrayBuffer?

isView() tells us that cert.buffer is not an ArrayBufferView (and thus must be
an ArrayBuffer). However, I can create an ArrayBuffer directly, so changed this
to make it more obvious.

https://codereview.chromium.org/2056053003/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/media/encrypted-media/encrypted-media-syntax.html:991:
navigator.requestMediaKeySystemAccess('org.w3.clearkey',
[{}]).then(function(access) {
On 2016/06/13 17:28:09, ddorwin wrote:
> FYI, an empty configuration is no longer allowed by the spec. We're going to
> need to change this to getDummyConfiguration() or something like that
> throughout.

Acknowledged. I'll do a subsequent CL to handle this, since it will affect all
the EME layout tests.

https://codereview.chromium.org/2056053003/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/media/encrypted-media/encrypted-media-syntax.html:1001:
test.done();
On 2016/06/13 17:28:09, ddorwin wrote:
> We don't check the result. What is the result for all()?

It returns an array of the results from each promise. The promise created in
test_setServerCertificate() already checks the result, but I've changed it to
use the promise results.

https://codereview.chromium.org/2056053003/diff/20001/third_party/WebKit/Sour...
File third_party/WebKit/Source/modules/encryptedmedia/MediaKeys.cpp (right):

https://codereview.chromium.org/2056053003/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/modules/encryptedmedia/MediaKeys.cpp:109: // false."
So convert any NOTSUPPORTEDERRORS into resolving with false.
On 2016/06/13 17:45:49, xhwang wrote:
> s/NOTSUPPORTEDERRORS/NOTSUPPORTEDERROR/

Done.

[ddorwin, 2016-06-13 19:13:21]

lgtm

[jrummell, 2016-06-13 23:21:02]

The CQ bit was checked by jrummell@chromium.org

[commit-bot: I haz the power, 2016-06-13 23:21:42]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2056053003/40001

[commit-bot: I haz the power, 2016-06-13 23:27:02]

Description was changed from

==========
Change MediaKeys.setServerCertificate() to return Promise<boolean>

https://github.com/w3c/encrypted-media/issues/145 changed the EME spec
so that setServerCertificate() returns Promise<boolean>. Update the
implementation to match.

This change updates the IDL and returns true on success.
NOT_SUPPORTED_ERRORs will resolve with false.

BUG=618800
TEST=EME tests pass, www.netflix.com plays
==========

to

==========
Change MediaKeys.setServerCertificate() to return Promise<boolean>

https://github.com/w3c/encrypted-media/issues/145 changed the EME spec
so that setServerCertificate() returns Promise<boolean>. Update the
implementation to match.

This change updates the IDL and returns true on success.
NOT_SUPPORTED_ERRORs will resolve with false.

BUG=618800
TEST=EME tests pass, www.netflix.com plays
==========

[commit-bot: I haz the power, 2016-06-13 23:27:03]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2016-06-13 23:27:05]

CQ bit was unchecked

[commit-bot: I haz the power, 2016-06-13 23:28:22]

Description was changed from

==========
Change MediaKeys.setServerCertificate() to return Promise<boolean>

https://github.com/w3c/encrypted-media/issues/145 changed the EME spec
so that setServerCertificate() returns Promise<boolean>. Update the
implementation to match.

This change updates the IDL and returns true on success.
NOT_SUPPORTED_ERRORs will resolve with false.

BUG=618800
TEST=EME tests pass, www.netflix.com plays
==========

to

==========
Change MediaKeys.setServerCertificate() to return Promise<boolean>

https://github.com/w3c/encrypted-media/issues/145 changed the EME spec
so that setServerCertificate() returns Promise<boolean>. Update the
implementation to match.

This change updates the IDL and returns true on success.
NOT_SUPPORTED_ERRORs will resolve with false.

BUG=618800
TEST=EME tests pass, www.netflix.com plays

Committed: https://crrev.com/e3d85c9614d5864bc3a9506b529036f2d703e0ec
Cr-Commit-Position: refs/heads/master@{#399595}
==========

[commit-bot: I haz the power, 2016-06-13 23:28:23]

Patchset 3 (id:??) landed as
https://crrev.com/e3d85c9614d5864bc3a9506b529036f2d703e0ec
Cr-Commit-Position: refs/heads/master@{#399595}