Add [TypeChecking=Unrestricted] on Device Orientation interfaces

Add [TypeChecking=Unrestricted] on Device Orientation interfaces

The only definitions that this actually affects are methods with custom
implementations, so the IDL change is just for show.

Also adjust the custom implementations of these methods to be closer to
normal generated code in handling their float arguments, thereby actually
making them restricted and thus honoring the [TypeChecking=Unrestricted]
flag.

BUG=354298
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=190994

[Jens Widell, 2015-02-18 08:56:54]

jl@opera.com changed reviewers:
+ haraken@chromium.org

[Jens Widell, 2015-02-18 08:56:55]

PTAL

[haraken, 2015-02-18 09:44:47]

LGTM, though I wonder why these methods need to be written in custom bindings.

[Jens Widell, 2015-02-18 12:56:01]

On 2015/02/18 09:44:47, haraken wrote:
> LGTM, though I wonder why these methods need to be written in custom bindings.

Thanks. I'll try to do something about the custom bindings in another CL.

In the case of DeviceOrientationEvent, the only out-of-the-ordinary thing is
that all the arguments are optional, and the implementation is told explicitly
whether the individual arguments were present or not. This is somewhat
inconvenient with our current handling of optional arguments (without defaults.)

In the case of DeviceMotionEvent, some of the arguments are dictionaries, except
the specification defines them as interfaces (without constructors) which I
don't quite see how it's supposed to be usable. We can probably just implement
those as proper dictionaries instead.

OTOH, neither of these init*Event() methods are actually in the specification;
the specification instead has constructors and *EventInit dictionaries. (The
most recent WD has the init methods, but the one in DeviceMotionEvent is named
initAccelerometerEvent() there.)

Specification:

http://w3c.github.io/deviceorientation/spec-source-orientation.html#devicemotion
http://w3c.github.io/deviceorientation/spec-source-orientation.html#deviceori...

[Jens Widell, 2015-02-18 12:56:10]

The CQ bit was checked by jl@opera.com

[commit-bot: I haz the power, 2015-02-18 12:56:47]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/934203002/1

[commit-bot: I haz the power, 2015-02-18 14:45:56]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-02-18 14:45:59]

Try jobs failed on following builders:
  linux_blink_rel on tryserver.blink (JOB_FAILED,
http://build.chromium.org/p/tryserver.blink/builders/linux_blink_rel/builds/4...)

[Jens Widell, 2015-02-27 11:27:30]

On 2015/02/18 14:45:59, I haz the power (commit-bot) wrote:
> Try jobs failed on following builders:
>   linux_blink_rel on tryserver.blink (JOB_FAILED,
>
http://build.chromium.org/p/tryserver.blink/builders/linux_blink_rel/builds/4...)

My original fix here was pretty stupid. The arguments were implemented to be
optional with [Default=Undefined] logic. But undefined => NaN => TypeError with
my original change, so the argument's were no longer optional, effectively.

Luckily, this showed up as test failures (and then along the way I completely
forgot about this CL).

Improved fix now implements the arguments as (restricted) doubles only if they
were actually specified. Also gets rid of some conversion macro usage.

PTAL?

[haraken, 2015-02-27 11:32:32]

LGTM

[Jens Widell, 2015-02-27 11:36:44]

The CQ bit was checked by jl@opera.com

[commit-bot: I haz the power, 2015-02-27 11:37:35]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/934203002/20002

[commit-bot: I haz the power, 2015-02-27 12:51:17]

Committed patchset #3 (id:20002) as
https://src.chromium.org/viewvc/blink?view=rev&revision=190994