IDL: Add [LegacyInterfaceTypeChecking] extended attribute

IDL: Add [LegacyInterfaceTypeChecking] extended attribute

This attribute is the opposite of [TypeChecking=Interface], and is
intended to replace it in all IDL files. By doing that, we'll explicitly
flag attributes and methods with incorrect behavior, instead of having
incorrect behavior by default.

We should be able to later switch individual attribute and methods over
to the correct behavior on a case-by-case basis by dropping the
[LegacyInterfaceTypeChecking] attribute from them, and eventually drop
support for the attribute altogether.

BUG=462561
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=191185

[Jens Widell, 2015-03-03 07:38:34]

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

[Jens Widell, 2015-03-03 07:38:34]

PTAL

[haraken, 2015-03-03 08:25:31]

BTW, how many [LegacyInterfaceTypeChecking] will we need to add? If it's a
thousand, it will mess up the code base.

[Jens Widell, 2015-03-03 08:31:27]

On 2015/03/03 08:25:31, haraken wrote:
> BTW, how many [LegacyInterfaceTypeChecking] will we need to add? If it's a
> thousand, it will mess up the code base.

We'll need to add at most 179 instances.

[haraken, 2015-03-03 08:36:38]

On 2015/03/03 08:31:27, Jens Widell wrote:
> On 2015/03/03 08:25:31, haraken wrote:
> > BTW, how many [LegacyInterfaceTypeChecking] will we need to add? If it's a
> > thousand, it will mess up the code base.
> 
> We'll need to add at most 179 instances.

If you have a plan to remove the majority of them in a reasonable timeline,
LGTM.

[Jens Widell, 2015-03-03 08:45:49]

On 2015/03/03 08:36:38, haraken wrote:
> On 2015/03/03 08:31:27, Jens Widell wrote:
> > On 2015/03/03 08:25:31, haraken wrote:
> > > BTW, how many [LegacyInterfaceTypeChecking] will we need to add? If it's a
> > > thousand, it will mess up the code base.
> > 
> > We'll need to add at most 179 instances.
> 
> If you have a plan to remove the majority of them in a reasonable timeline,
> LGTM.

That's the plan, yes. Thanks.

[Jens Widell, 2015-03-03 08:45:52]

The CQ bit was checked by jl@opera.com

[commit-bot: I haz the power, 2015-03-03 08:46:14]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/961073004/1

[philipj_slow, 2015-03-03 09:23:36]

LGTM, no more wondering in review if the interface needs more type checking.
This will probably reveal some not-per-spec instances that were previously not
obvious, like in partial interfaces.

[commit-bot: I haz the power, 2015-03-03 10:17:58]

Committed patchset #1 (id:1) as
https://src.chromium.org/viewvc/blink?view=rev&revision=191185