Update SVG IDL interfaces to stop inheriting from SVGTests

Update SVG IDL interfaces to stop inheriting from SVGTests

Update SVG IDL interfaces to stop inheriting from SVGTests and use
implements statements instead, as per the latest specification:
http://www.w3.org/TR/SVG2/types.html#InterfaceSVGTests

This is one step towards getting rid of multiple inheritence in
SVG which is no longer supported by Web IDL. There is no web
exposed behavior change.

BUG=252181
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=152883

[haraken, 2013-06-21 11:26:13]

LGTM

[commit-bot: I haz the power, 2013-06-21 11:31:05]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/ch.dumez@sisa.samsung.com/17549002/1

[commit-bot: I haz the power, 2013-06-21 13:09:33]

Change committed as 152883

[arv (Not doing code reviews), 2013-06-21 13:45:42]

LGTM

https://codereview.chromium.org/17549002/diff/1/Source/core/svg/SVGTests.idl
File Source/core/svg/SVGTests.idl (right):

https://codereview.chromium.org/17549002/diff/1/Source/core/svg/SVGTests.idl#...
Source/core/svg/SVGTests.idl:37: SVGAElement implements SVGTests;
Maybe we should start putting these in the idl file that implements the
interface so that we keep more relevant information in one place?

[do-not-use, 2013-06-21 13:52:52]

On 2013/06/21 13:45:42, arv wrote:
> LGTM
> 
> https://codereview.chromium.org/17549002/diff/1/Source/core/svg/SVGTests.idl
> File Source/core/svg/SVGTests.idl (right):
> 
>
https://codereview.chromium.org/17549002/diff/1/Source/core/svg/SVGTests.idl#...
> Source/core/svg/SVGTests.idl:37: SVGAElement implements SVGTests;
> Maybe we should start putting these in the idl file that implements the
> interface so that we keep more relevant information in one place?

I was starting to feel the same :) The bindings generator does not currently
support it but it should be easy to do. I'll look into this soon.

[do-not-use, 2013-06-21 14:46:05]

On 2013/06/21 13:52:52, Christophe Dumez wrote:
> On 2013/06/21 13:45:42, arv wrote:
> > LGTM
> > 
> > https://codereview.chromium.org/17549002/diff/1/Source/core/svg/SVGTests.idl
> > File Source/core/svg/SVGTests.idl (right):
> > 
> >
>
https://codereview.chromium.org/17549002/diff/1/Source/core/svg/SVGTests.idl#...
> > Source/core/svg/SVGTests.idl:37: SVGAElement implements SVGTests;
> > Maybe we should start putting these in the idl file that implements the
> > interface so that we keep more relevant information in one place?
> 
> I was starting to feel the same :) The bindings generator does not currently
> support it but it should be easy to do. I'll look into this soon.

Note that it currently behaves very similarly to partial interfaces (but can
extend several interfaces instead of just 1). It is impossible already to have
all the relevant information in one place because of partial interfaces.

That said, for implements statements it is possible to have them in the IDL file
that implements the interface. A lot of specs seem to follow this convention as
well.

One question though, do you think we should allow both? Or should we force the
developers to write the implements statements in the IDL file that implements
the interface?

[arv (Not doing code reviews), 2013-06-21 14:51:46]

On 2013/06/21 14:46:05, Christophe Dumez wrote:
> One question though, do you think we should allow both? Or should we force the
> developers to write the implements statements in the IDL file that implements
> the interface?

I don't think it is that important to allow both if it makes things harder.

[do-not-use, 2013-06-21 14:53:44]

On 2013/06/21 14:51:46, arv wrote:
> On 2013/06/21 14:46:05, Christophe Dumez wrote:
> > One question though, do you think we should allow both? Or should we force
the
> > developers to write the implements statements in the IDL file that
implements
> > the interface?
> 
> I don't think it is that important to allow both if it makes things harder.

It is not really much harder I believe. I was just wondering if we would want to
enforce this style to avoid a mix and maybe confusion.

[haraken, 2013-06-21 15:27:31]

On 2013/06/21 14:53:44, Christophe Dumez wrote:
> On 2013/06/21 14:51:46, arv wrote:
> > On 2013/06/21 14:46:05, Christophe Dumez wrote:
> > > One question though, do you think we should allow both? Or should we force
> the
> > > developers to write the implements statements in the IDL file that
> implements
> > > the interface?
> > 
> > I don't think it is that important to allow both if it makes things harder.
> 
> It is not really much harder I believe. I was just wondering if we would want
to
> enforce this style to avoid a mix and maybe confusion.

I'd say we don't want to mix the two styles. Forcing developers to write the
implements statements in the IDL file that implements the interface would be
less confusing.