IDL compiler: avoid include for WindowMediaSource.idl

IDL compiler: avoid include for WindowMediaSource.idl

WindowMediaSource.idl is a partial interface for Window.
Thus normally we should include WindowMediaSource.h.
However, this is only constructor attributes, so there's no implementation
and no header.

Python determines the includes at the dependency resolution stage,
just checking the partial interface and if there's an [ImplementedAs].

Perl instead computes the includes at every member, hence b/c there
are only constructors it doesn't trigger it here.

Now, we don't want to have to parse all member of a partial interface
to determine if we need a header!

Conveniently, we have an existing extended attribute that says
"no header", namely [LegacyImplementedInBaseClass].

There are thus 3 clean solutions to this:
1. Add an empty header, WindowMediaSource.h;
   thus we can always include the header (unless implemented elsewhere).
2. Add a new extended attribute, [NoHeader], used only here.
3. Use the existing [LegacyImplementInBaseClass];
if we get rid of other cases, can rename to [NoHeader].

I think 1. (always include the header) is cleanest, as it minimizes logic.
However, this CL is simplest (it does 3), as it uses existing logic.

WDYT?

R=haraken

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=166672

[Nils Barth (inactive), 2014-02-06 08:21:33]

BTW, no changes to generated bindings;
just prevents Python from including a non-existent header.

[haraken, 2014-02-06 08:34:04]

On 2014/02/06 08:21:33, Nils Barth wrote:
> BTW, no changes to generated bindings;
> just prevents Python from including a non-existent header.

What does [LegacyImplementInBaseClass] mean currently?

[Nils Barth (inactive), 2014-02-06 08:39:50]

On 2014/02/06 08:34:04, haraken wrote:
> On 2014/02/06 08:21:33, Nils Barth wrote:
> > BTW, no changes to generated bindings;
> > just prevents Python from including a non-existent header.
> 
> What does [LegacyImplementInBaseClass] mean currently?

Use interface_name of main interface instead of file_name of dependency for
cpp_class
(changes namespace use and prevents header inclusion).

[haraken, 2014-02-06 08:44:00]

On 2014/02/06 08:39:50, Nils Barth wrote:
> On 2014/02/06 08:34:04, haraken wrote:
> > On 2014/02/06 08:21:33, Nils Barth wrote:
> > > BTW, no changes to generated bindings;
> > > just prevents Python from including a non-existent header.
> > 
> > What does [LegacyImplementInBaseClass] mean currently?
> 
> Use interface_name of main interface instead of file_name of dependency for
> cpp_class
> (changes namespace use and prevents header inclusion).

hmm, understood. Then this CL looks not that hacky. LGTM.

- Can we rename [LegacyImplementInBaseClass] to [NoHeader] once we switch to
Python?
- How hard would it be to remove [LegacyImplementInBaseClass] completely?

[Nils Barth (inactive), 2014-02-06 08:50:29]

+Chris

On 2014/02/06 08:44:00, haraken wrote:
> - Can we rename [LegacyImplementedInBaseClass] to [NoHeader] once we switch to
> Python?

We could, but because it also means "use interface_name, not file_name of
dependency",
which is why Chris added it.

> - How hard would it be to remove [LegacyImplementInBaseClass] completely?

I don't know -- I think it's just shuffling around code from one .cpp file to
another,
but Chris may have found some reason to keep it around;
it was planned to be temporary.

Chris?

[Nils Barth (inactive), 2014-02-06 08:50:33]

The CQ bit was checked by nbarth@chromium.org

[commit-bot: I haz the power, 2014-02-06 08:50:45]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/nbarth@chromium.org/156573002/30001

[commit-bot: I haz the power, 2014-02-06 10:06:32]

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

[commit-bot: I haz the power, 2014-02-06 10:06:32]

Retried try job too often on win_layout for step(s) webkit_lint
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=win_layout...

[Nils Barth (inactive), 2014-02-07 00:30:22]

The CQ bit was checked by nbarth@chromium.org

[commit-bot: I haz the power, 2014-02-07 00:30:34]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/nbarth@chromium.org/156573002/30001

[commit-bot: I haz the power, 2014-02-07 01:45:41]

Change committed as 166672