Apply TypeChecking to MediaSource IDLs

Apply TypeChecking to MediaSource IDLs

This allows cleaning up a number of null-checks, since the generated IDL
code now assures non-null where indicated in the IDL.
Also drop explicit throwTypeError() in MediaSourceBase::endOfStream, since
the bindings code would've thrown that already.
Add 'unrestricted' to the IDLs where the spec has them.

This updates to match spec. revision ba5233c6654d.
https://dvcs.w3.org/hg/html-media/rev/ba5233c6654d
...per W3 bug:
Bug 25347 - [MSE] 'nullable' arguments and type conversions and
INVALID_ACCESS_ERR (vs. WebIDL)
https://www.w3.org/Bugs/Public/show_bug.cgi?id=25347

One additional non-specced change is to mark arguments to
SourceBuffer.remove() as 'unrestricted' since that's the
semantics they expect.

BUG=368634
BUG=354298
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=173197

[fs, 2014-04-30 13:04:23]

In response to fix of W3C bug 25347. (No test added for createObjectURL because
of the overloads catching the null-case.)

(@nbarth: This might be possible to use for TypeChecking=Unrestricted on
attributes - see for example SourceBuffer.timestampOffset)

[acolwell GONE FROM CHROMIUM, 2014-04-30 17:28:28]

lgtm % comment

https://codereview.chromium.org/266453005/diff/20001/Source/modules/mediasour...
File Source/modules/mediasource/SourceBuffer.idl (right):

https://codereview.chromium.org/266453005/diff/20001/Source/modules/mediasour...
Source/modules/mediasource/SourceBuffer.idl:69: [RaisesException] void
remove(unrestricted double start, unrestricted double end);
start should not be unrestricted. end could be unrestricted to allow
POSITIVE_INFINITY but other non-finite values should result in an error. For
now, I'd prefer to keep the IDL as is so that it matches the spec.

[acolwell GONE FROM CHROMIUM, 2014-04-30 17:35:11]

On 2014/04/30 17:28:28, acolwell wrote:
> lgtm % comment
> 
>
https://codereview.chromium.org/266453005/diff/20001/Source/modules/mediasour...
> File Source/modules/mediasource/SourceBuffer.idl (right):
> 
>
https://codereview.chromium.org/266453005/diff/20001/Source/modules/mediasour...
> Source/modules/mediasource/SourceBuffer.idl:69: [RaisesException] void
> remove(unrestricted double start, unrestricted double end);
> start should not be unrestricted. end could be unrestricted to allow
> POSITIVE_INFINITY but other non-finite values should result in an error. For
> now, I'd prefer to keep the IDL as is so that it matches the spec.

FYI, I filed https://www.w3.org/Bugs/Public/show_bug.cgi?id=25518 . We'll see if
I get any pushback.

[Nils Barth (inactive), 2014-05-01 01:07:44]

On 2014/04/30 13:04:23, fs wrote:
> In response to fix of W3C bug 25347. (No test added for createObjectURL
because
> of the overloads catching the null-case.)

For reference:
Bug 25347 - [MSE] 'nullable' arguments and type conversions and
INVALID_ACCESS_ERR (vs. WebIDL)
https://www.w3.org/Bugs/Public/show_bug.cgi?id=25347

> (@nbarth: This might be possible to use for TypeChecking=Unrestricted on
> attributes - see for example SourceBuffer.timestampOffset)

Yup, I'll add float type checking for attributes soon, hopefully today!

[Nils Barth (inactive), 2014-05-01 01:08:46]

Thanks Fredrik!
Agree with Aaron's comment re: spec change/bug, and a few minor nits/requests,
but LGTM!

https://codereview.chromium.org/266453005/diff/20001/Source/modules/mediasour...
File Source/modules/mediasource/SourceBuffer.cpp (right):

https://codereview.chromium.org/266453005/diff/20001/Source/modules/mediasour...
Source/modules/mediasource/SourceBuffer.cpp:175: // (Needs
TypeChecking=Unrestricted to apply to attributes before it can be
Could you add a link to the bug here so I can grep for this?
http://crbug.com/354298

https://codereview.chromium.org/266453005/diff/20001/Source/modules/mediasour...
Source/modules/mediasource/SourceBuffer.cpp:214: // (Needs
TypeChecking=Unrestricted to apply to attributes before it can be
http://crbug.com/354298

https://codereview.chromium.org/266453005/diff/20001/Source/modules/mediasour...
File Source/modules/mediasource/SourceBuffer.idl (right):

https://codereview.chromium.org/266453005/diff/20001/Source/modules/mediasour...
Source/modules/mediasource/SourceBuffer.idl:30: 
While we're here, could you add a link to the spec?
https://dvcs.w3.org/hg/html-media/raw-file/tip/media-source/media-source.html...

If you prefer, the published one:
http://www.w3.org/TR/media-source/#idl-def-SourceBuffer
...but we're generally tracking ED, not CR, right?

https://codereview.chromium.org/266453005/diff/20001/Source/modules/mediasour...
Source/modules/mediasource/SourceBuffer.idl:37: NoInterfaceObject,
While you're here, could you alphabetize this?
Thanks!

https://codereview.chromium.org/266453005/diff/20001/Source/modules/mediasour...
Source/modules/mediasource/SourceBuffer.idl:69: [RaisesException] void
remove(unrestricted double start, unrestricted double end);
On 2014/04/30 17:28:28, acolwell wrote:
> start should not be unrestricted. end could be unrestricted to allow
> POSITIVE_INFINITY but other non-finite values should result in an error. For
> now, I'd prefer to keep the IDL as is so that it matches the spec.

Fredrik, could you link to the W3 bug?
https://www.w3.org/Bugs/Public/show_bug.cgi?id=25518

[Nils Barth (inactive), 2014-05-01 02:25:05]

Two other nits.

https://codereview.chromium.org/266453005/diff/20001/Source/modules/mediasour...
File Source/modules/mediasource/MediaSource.idl (right):

https://codereview.chromium.org/266453005/diff/20001/Source/modules/mediasour...
Source/modules/mediasource/MediaSource.idl:30: 
Could you add a link to the spec:
https://dvcs.w3.org/hg/html-media/raw-file/tip/media-source/media-source.html...
?

Thanks!

https://codereview.chromium.org/266453005/diff/20001/Source/modules/mediasour...
Source/modules/mediasource/MediaSource.idl:41: TypeChecking=Interface|Nullable,
Want to add Unrestricted and make this
TypeChecking=Interface|Nullable|Unrestricted
...just to be safe?

[Nils Barth (inactive), 2014-05-01 03:14:52]

On 2014/04/30 13:04:23, fs wrote:
> (@nbarth: This might be possible to use for TypeChecking=Unrestricted on
> attributes - see for example SourceBuffer.timestampOffset)

On 2014/05/01 01:07:44, Nils Barth wrote:
> Yup, I'll add float type checking for attributes soon, hopefully today!

Done!
I've added type checking in a separate CL, using SourceBuffer as example:
Add support for type checking of floating point attributes
https://codereview.chromium.org/265753003/

Once that CL lands, could you sync to it and then you can just put
[TypeChecking=Unrestricted] on the interface definition!

[Nils Barth (inactive), 2014-05-02 05:14:28]

Hi Fredrik, everything's landed for float type checking, so just need to sync
this and land!

[fs, 2014-05-02 09:26:26]

https://codereview.chromium.org/266453005/diff/20001/Source/modules/mediasour...
File Source/modules/mediasource/MediaSource.idl (right):

https://codereview.chromium.org/266453005/diff/20001/Source/modules/mediasour...
Source/modules/mediasource/MediaSource.idl:30: 
On 2014/05/01 02:25:05, Nils Barth wrote:
> Could you add a link to the spec:
>
https://dvcs.w3.org/hg/html-media/raw-file/tip/media-source/media-source.html...
> ?
> 
> Thanks!

Done.

https://codereview.chromium.org/266453005/diff/20001/Source/modules/mediasour...
Source/modules/mediasource/MediaSource.idl:41: TypeChecking=Interface|Nullable,
On 2014/05/01 02:25:05, Nils Barth wrote:
> Want to add Unrestricted and make this
> TypeChecking=Interface|Nullable|Unrestricted
> ...just to be safe?

Done.

https://codereview.chromium.org/266453005/diff/20001/Source/modules/mediasour...
File Source/modules/mediasource/SourceBuffer.idl (right):

https://codereview.chromium.org/266453005/diff/20001/Source/modules/mediasour...
Source/modules/mediasource/SourceBuffer.idl:69: [RaisesException] void
remove(unrestricted double start, unrestricted double end);
On 2014/05/01 01:08:46, Nils Barth wrote:
> On 2014/04/30 17:28:28, acolwell wrote:
> > start should not be unrestricted. end could be unrestricted to allow
> > POSITIVE_INFINITY but other non-finite values should result in an error. For
> > now, I'd prefer to keep the IDL as is so that it matches the spec.

Ok. Removed 'unrestricted' again (for the time being.) I'll do minimal "damage"
to the tests (the change/semantics doesn't seem entirely unreasonable.)

> 
> Fredrik, could you link to the W3 bug?
> https://www.w3.org/Bugs/Public/show_bug.cgi?id=25518

Added.

[fs, 2014-05-02 11:31:35]

The CQ bit was checked by fs@opera.com

[commit-bot: I haz the power, 2014-05-02 11:31:48]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/fs@opera.com/266453005/60001

[commit-bot: I haz the power, 2014-05-02 12:16:02]

Change committed as 173197