To M51: MSE, MS, and any blob URL: Ignore preload 'none' on resource fetching

To M51: MSE, MS, and any blob URL: Ignore preload 'none' on resource fetching

M51 didn't include MediaStream srcObject support, so this cherry-pick involves
merge conflict resolution to retain the intent of the M52 version, absent the
MediaStream srcObject support.

Prevents the usage of the optional resource fetch algorithm steps for
when preload='none' in the MSE, MediaStream, or any blob URL
cases.  Complies with recent update to MSE spec [1] and fixes the
related crbug.com/539707. Includes test coverage for MSE and invalid
(revoked MSE object) blob URLs. MediaStream test coverage is expected to
come from later import of a new web-platform-tests [2].

[1] https://github.com/w3c/media-source/pull/64
[2] https://github.com/w3c/web-platform-tests/pull/2895

R=guidou@chromium.org
TBR=foolip@chromium.or,srirama.m@samsung.com
BUG=539707
TEST=http/tests/media/media-source/mediasource-preload.html

Review URL: https://codereview.chromium.org/1881733004

Cr-Commit-Position: refs/heads/master@{#389649}
(cherry picked from commit b16f202aa67513f74576312f7f8a3ad17f60458a)

[wolenetz, 2016-06-07 21:33:50]

Description was changed from

==========
To M51: MSE, MS, and any blob URL: Ignore preload 'none' on resource fetching

M51 didn't include MediaStream srcObject support, so this cherry-pick involves
merge conflict resolution to retain the intent of the M52 version, absent the
MediaStream srcObject support.

Prevents the usage of the optional resource fetch algorithm steps for
when preload='none' in the MSE, MediaStream, or any blob URL
cases.  Complies with recent update to MSE spec [1] and fixes the
related crbug.com/539707. Includes test coverage for MSE and invalid
(revoked MSE object) blob URLs. MediaStream test coverage is expected to
come from later import of a new web-platform-tests [2].

[1] https://github.com/w3c/media-source/pull/64
[2] https://github.com/w3c/web-platform-tests/pull/2895

R=foolip@chromium.org,guidou@chromium.org
TBR=srirama.m@samsung.com
BUG=539707
TEST=http/tests/media/media-source/mediasource-preload.html

Review URL: https://codereview.chromium.org/1881733004

Cr-Commit-Position: refs/heads/master@{#389649}
(cherry picked from commit b16f202aa67513f74576312f7f8a3ad17f60458a)
==========

to

==========
To M51: MSE, MS, and any blob URL: Ignore preload 'none' on resource fetching

M51 didn't include MediaStream srcObject support, so this cherry-pick involves
merge conflict resolution to retain the intent of the M52 version, absent the
MediaStream srcObject support.

Prevents the usage of the optional resource fetch algorithm steps for
when preload='none' in the MSE, MediaStream, or any blob URL
cases.  Complies with recent update to MSE spec [1] and fixes the
related crbug.com/539707. Includes test coverage for MSE and invalid
(revoked MSE object) blob URLs. MediaStream test coverage is expected to
come from later import of a new web-platform-tests [2].

[1] https://github.com/w3c/media-source/pull/64
[2] https://github.com/w3c/web-platform-tests/pull/2895

R=guidou@chromium.org
TBR=foolip@chromium.or,srirama.m@samsung.com
BUG=539707
TEST=http/tests/media/media-source/mediasource-preload.html

Review URL: https://codereview.chromium.org/1881733004

Cr-Commit-Position: refs/heads/master@{#389649}
(cherry picked from commit b16f202aa67513f74576312f7f8a3ad17f60458a)
==========

[wolenetz, 2016-06-07 21:35:14]

guidou@: Please take a look. I believe I've resolved the merge conflicts
appropriately to retain the intended fix. Your l*tm will help improve confidence
that this can go directly to stable M51.

[Guido Urdaneta, 2016-06-08 09:31:12]

On 2016/06/07 21:35:14, wolenetz_slow_reviews wrote:
> guidou@: Please take a look. I believe I've resolved the merge conflicts
> appropriately to retain the intended fix. Your l*tm will help improve
confidence
> that this can go directly to stable M51.

lgtm

[Srirama, 2016-06-08 12:11:51]

LGTM but what about the cleanup done in
https://codereview.chromium.org/1920383002/ ?

[wolenetz, 2016-06-08 15:15:04]

On 2016/06/08 12:11:51, Srirama wrote:
> LGTM but what about the cleanup done in
> https://codereview.chromium.org/1920383002/ ?

That cleanup was specific to the layout test, not production code. And without
the test cleanup, it was still a working test.

Thanks for review. Landing this of course will depend on approval of merge to
M51 in the bug, which is unlikely at the moment.

[foolip, 2016-06-09 11:22:09]

lgtm in case you need it.