Adding platform specific checks for VP9 codecs and HLS mime types

Adding platform specific variables for MediaCanPlayTypeTest

Vp9 codec is supported after KitKat (API level 19)
Expectation values are modifed in the test case based on the Android sdk version

HLS mime types are supported after Ice-cream sandwich (API level 14)
A check for identifying android version is added which decides whether the mime type should be added in the media_map_ or not.

Theora is not supported on Android. This check is also added in the "IsCodecSupportedOnAndroid".

MediaCanPlayType test are re-enabled on android.

BUG=357665
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=269748

[amogh.bihani, 2014-04-10 08:44:31]

PTAL.
This is what I have planned for fixing the bug, however I have not removed the
#ifdef which disables the tests for android.
Till now webm, ogg and wav have been fixed by other CL[1].
-------
[1] https://codereview.chromium.org/218303002/

[amogh.bihani, 2014-04-10 09:03:25]

On 2014/04/10 08:44:31, amogh.bihani wrote:
> Till now webm, ogg and wav have been fixed by other CL[1].

Oh sorry, ogg and wav have been fixed by that CL. Webm will be fixed by this CL.

[xhwang, 2014-04-10 23:35:55]

https://codereview.chromium.org/230413003/diff/40001/content/browser/media/me...
File content/browser/media/media_canplaytype_browsertest.cc (right):

https://codereview.chromium.org/230413003/diff/40001/content/browser/media/me...
content/browser/media/media_canplaytype_browsertest.cc:16: const char* kProbably
= "probably";
These (l.16-l.60) should all be "using const char kFoo[]". See comments below.

https://codereview.chromium.org/230413003/diff/40001/content/browser/media/me...
content/browser/media/media_canplaytype_browsertest.cc:72: kVP9Probably = "";
kFoo means it's a const value; changing the pointer is a bit hacky. We are not
using "const char* const kFoo" a lot because that looks crazy. Actually in
chromium, we prefer using "const char kFoo[]" to "const char*" to prevent this
:)

Instead of doing this, is it possible to have some class/function local
expectation variables which can be different on different platforms?

[amogh.bihani, 2014-04-11 10:43:17]

https://codereview.chromium.org/230413003/diff/40001/content/browser/media/me...
File content/browser/media/media_canplaytype_browsertest.cc (right):

https://codereview.chromium.org/230413003/diff/40001/content/browser/media/me...
content/browser/media/media_canplaytype_browsertest.cc:72: kVP9Probably = "";
On 2014/04/10 23:35:56, xhwang wrote:
> kFoo means it's a const value; changing the pointer is a bit hacky. We are not
> using "const char* const kFoo" a lot because that looks crazy. Actually in
> chromium, we prefer using "const char kFoo[]" to "const char*" to prevent this
> :)
> 
> Instead of doing this, is it possible to have some class/function local
> expectation variables which can be different on different platforms?


Oh Ok!
I'll change the type and create local function to handle the expectations.

[amogh.bihani, 2014-04-24 10:03:17]

PTAL.

https://codereview.chromium.org/230413003/diff/70001/net/base/mime_util.cc
File net/base/mime_util.cc (right):

https://codereview.chromium.org/230413003/diff/70001/net/base/mime_util.cc#ne...
net/base/mime_util.cc:416: return false;
This check was not needed before as theora was checked only for video/ogg.
Android does not support video/ogg, so the browser never checked for codecs[1].
However this is not the case with application/ogg. Without this check, the
format_strict_map_ will have theora entry[2] and browser will give wrong result
for canPlayType query.

--------------
[1]
https://code.google.com/p/chromium/codesearch#chromium/src/content/renderer/r...

[2]
https://code.google.com/p/chromium/codesearch#chromium/src/net/base/mime_util...

[Ryan Sleevi, 2014-04-24 21:55:59]

lgtm

[amogh.bihani, 2014-04-28 03:01:01]

xhwang@
PTAL at the testcases.

[xhwang, 2014-04-28 17:48:48]

A question about VP9 support on Android.

https://codereview.chromium.org/230413003/diff/70001/content/browser/media/me...
File content/browser/media/media_canplaytype_browsertest.cc (right):

https://codereview.chromium.org/230413003/diff/70001/content/browser/media/me...
content/browser/media/media_canplaytype_browsertest.cc:145: VP9Probably = "";
Is it guaranteed that VP9 is supported on all devices that run Android KK and
later? At least I don't see this in the doc below:
http://developer.android.com/guide/appendix/media-formats.html

[amogh.bihani, 2014-04-29 03:05:07]

https://codereview.chromium.org/230413003/diff/70001/content/browser/media/me...
File content/browser/media/media_canplaytype_browsertest.cc (right):

https://codereview.chromium.org/230413003/diff/70001/content/browser/media/me...
content/browser/media/media_canplaytype_browsertest.cc:145: VP9Probably = "";
On 2014/04/28 17:48:48, xhwang wrote:
> Is it guaranteed that VP9 is supported on all devices that run Android KK and
> later? At least I don't see this in the doc below:
> http://developer.android.com/guide/appendix/media-formats.html

VP9 has to be supported in all the KK devices. This is mentioned in the
compatibility definition of Android[1]. (section 5.1 - Media Codecs, last entry
in that table)

https://codereview.chromium.org/66863003 this CL had specifically added check
for VP9 on kitkat.
----------
[1]
http://static.googleusercontent.com/media/source.android.com/en/us/compatibil...

[xhwang, 2014-04-29 17:13:41]

Thanks for the link. The CL looks good except for one last question.

https://codereview.chromium.org/230413003/diff/70001/content/browser/media/me...
File content/browser/media/media_canplaytype_browsertest.cc (right):

https://codereview.chromium.org/230413003/diff/70001/content/browser/media/me...
content/browser/media/media_canplaytype_browsertest.cc:145: VP9Probably = "";
On 2014/04/29 03:05:08, amogh.bihani wrote:
> On 2014/04/28 17:48:48, xhwang wrote:
> > Is it guaranteed that VP9 is supported on all devices that run Android KK
and
> > later? At least I don't see this in the doc below:
> > http://developer.android.com/guide/appendix/media-formats.html
> 
> VP9 has to be supported in all the KK devices. This is mentioned in the
> compatibility definition of Android[1]. (section 5.1 - Media Codecs, last
entry
> in that table)
> 
> https://codereview.chromium.org/66863003 this CL had specifically added check
> for VP9 on kitkat.
> ----------
> [1]
>
http://static.googleusercontent.com/media/source.android.com/en/us/compatibil...

Thanks for the link. Can you add a comment about this?

https://codereview.chromium.org/230413003/diff/70001/content/browser/media/me...
content/browser/media/media_canplaytype_browsertest.cc:506: PropAndVP9Probably =
"maybe";
I think I am missing some point. Why this is "maybe" but on line 145,
VP9Probably is "" if sdk < 19?

[amogh.bihani, 2014-04-30 03:47:32]

Comments are in patchset 5.

(Please note that I will be out of office till 11th May.)

https://codereview.chromium.org/230413003/diff/70001/content/browser/media/me...
File content/browser/media/media_canplaytype_browsertest.cc (right):

https://codereview.chromium.org/230413003/diff/70001/content/browser/media/me...
content/browser/media/media_canplaytype_browsertest.cc:506: PropAndVP9Probably =
"maybe";
On 2014/04/29 17:13:42, xhwang wrote:
> I think I am missing some point. Why this is "maybe" but on line 145,
> VP9Probably is "" if sdk < 19? 

That is because webm have been added in the list of strict media mime types[1].
If it finds any codec that is not supported, it returns ""[2].
However, mp4 and HLS are yet to be added in that list. So if they find any
incompatible codec they return a "maybe"[3]. My other CL will fix this[4].
------------
[1]
https://code.google.com/p/chromium/codesearch#chromium/src/net/base/mime_util...

[2]
https://code.google.com/p/chromium/codesearch#chromium/src/content/renderer/r...

[3]
https://code.google.com/p/chromium/codesearch#chromium/src/content/renderer/r...

[4] https://codereview.chromium.org/254983006/

[xhwang, 2014-04-30 16:55:49]

On 2014/04/30 03:47:32, amogh.bihani(ooo till 11may) wrote:
> Comments are in patchset 5.
> 
> (Please note that I will be out of office till 11th May.)
> 
>
https://codereview.chromium.org/230413003/diff/70001/content/browser/media/me...
> File content/browser/media/media_canplaytype_browsertest.cc (right):
> 
>
https://codereview.chromium.org/230413003/diff/70001/content/browser/media/me...
> content/browser/media/media_canplaytype_browsertest.cc:506: PropAndVP9Probably
=
> "maybe";
> On 2014/04/29 17:13:42, xhwang wrote:
> > I think I am missing some point. Why this is "maybe" but on line 145,
> > VP9Probably is "" if sdk < 19? 
> 
> That is because webm have been added in the list of strict media mime
types[1].
> If it finds any codec that is not supported, it returns ""[2].
> However, mp4 and HLS are yet to be added in that list. So if they find any
> incompatible codec they return a "maybe"[3]. My other CL will fix this[4].
> ------------
> [1]
>
https://code.google.com/p/chromium/codesearch#chromium/src/net/base/mime_util...
> 
> [2]
>
https://code.google.com/p/chromium/codesearch#chromium/src/content/renderer/r...
> 
> [3]
>
https://code.google.com/p/chromium/codesearch#chromium/src/content/renderer/r...
> 
> [4] https://codereview.chromium.org/254983006/

Thanks for the explanation. The test lgtm.

[amogh.bihani, 2014-05-12 03:41:49]

The CQ bit was checked by amogh.bihani@samsung.com

[commit-bot: I haz the power, 2014-05-12 03:42:04]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/amogh.bihani@samsung.com/230413003/90001

[commit-bot: I haz the power, 2014-05-12 07:39:30]

Change committed as 269748