Rename DOMPath IDL interface to Path to match the specification

Rename DOMPath IDL interface to Path to match the specification

Rename DOMPath IDL interface to Path to match its name on JS side and use
[ImplementedAs=DOMPath] extended attribute instead.

This is an initial step towards removing [InterfaceName] extended attribute
and consistently use [ImplementedAs] instead.

This patch also adds a IDL test for this as [ImplementedAs] was not usable
on interfaces before, only attributes and operations.

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=150154

[haraken, 2013-05-10 09:08:24]

LGTM, (though I don't fully understand why #include lines of binding tests have
to be changed in this CL.)

[do-not-use, 2013-05-10 09:10:04]

On 2013/05/10 09:08:24, haraken wrote:
> LGTM, (though I don't fully understand why #include lines of binding tests
have
> to be changed in this CL.)

This is because someone forgot to rebaseline the results after a recent change.
Those are unrelated to my patch and the bindings tests are currently failing on
master.

Do you want me to rebaseline the results in another CL?

[haraken, 2013-05-10 09:11:13]

> This is because someone forgot to rebaseline the results after a recent
change.
> Those are unrelated to my patch and the bindings tests are currently failing
on
> master.
> 
> Do you want me to rebaseline the results in another CL?

The changes are minor, so I think it's OK to include the changes in this CL.

[commit-bot: I haz the power, 2013-05-10 10:01:40]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/ch.dumez@sisa.samsung.com/14988008/1

[commit-bot: I haz the power, 2013-05-10 13:42:43]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/ch.dumez@sisa.samsung.com/14988008/1

[arv (Not doing code reviews), 2013-05-10 13:52:14]

LGTM

[commit-bot: I haz the power, 2013-05-10 20:03:03]

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

[Rik, 2013-05-10 20:28:50]

very nice!
I made a proposal to fix architectural issue with canvas paths. would you be
interested in adding them to blink?

[Rik, 2013-05-10 20:29:40]

Also, you forgot to send an 'intent to implement' message to the list.

[arv (Not doing code reviews), 2013-05-10 20:32:03]

This does not need an 'intent to implement'. It just a bug fix.


On Fri, May 10, 2013 at 4:29 PM, <cabanier@adobe.com> wrote:

> Also, you forgot to send an 'intent to implement' message to the list.
>
>
>
https://chromiumcodereview.**appspot.com/14988008/<https://chromiumcodereview...
>



-- 
erik

[eseidel, 2013-05-10 22:21:56]

Do you have a plan for code-generation of String.h and Window.h?  These may have
conflicts with other headers, depending on how we do include paths...

[eseidel, 2013-05-10 22:29:33]

Adam reminds me that DOMString is a primitive and does not have an IDL. :)

[do-not-use, 2013-05-10 23:58:56]

On 2013/05/10 22:29:33, Eric Seidel wrote:
> Adam reminds me that DOMString is a primitive and does not have an IDL. :)

I'm not renaming the headers, only the IDL interfaces / files. The idea is to
get rid of [InterfaceName] extended attribute and use consistently
[ImplementedAs] instead. So I do plan to rename DOMWindow.idl to Window.idl.
However, I have no plan to rename DOMWindow.h to Window.h at this stage.

[commit-bot: I haz the power, 2013-05-10 23:59:53]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/ch.dumez@sisa.samsung.com/14988008/1

[commit-bot: I haz the power, 2013-05-11 00:00:42]

Change committed as 150154

[eseidel, 2013-05-11 00:12:02]

Oh, and nevermind me.  The generated file is not Window.h, it's V8Window.h. :)

[haraken, 2013-05-11 05:08:02]

Yes, we might want to rename DOMXXX to XXX in the future, but for the time being
our objective is to remove [InterfaceName] without changing any behavior nor any
generated code.

[alancutter (OOO until 2018), 2013-05-13 04:46:43]

This patch is failing run-bindings-tests with:

--- bindings/tests/results/V8TestInterfaceImplementedAs.cpp	2013-05-13
14:44:51.472341309 +1000
+++ /tmp/tmpxsK6k_/V8TestInterfaceImplementedAs.cpp	2013-05-13
14:44:56.222328968 +1000
@@ -87,7 +87,7 @@
 #endif
 WrapperTypeInfo V8TestInterfaceImplementedAs::info = {
V8TestInterfaceImplementedAs::GetTemplate,
V8TestInterfaceImplementedAs::derefObject, 0, 0, 0,
V8TestInterfaceImplementedAs::installPerContextPrototypeProperties, 0,
WrapperTypeObjectPrototype };
 
-namespace TestInterfaceImplementedAsV8Internal {
+namespace RealClassV8Internal {
 
 template <typename T> void V8_USE(T) { }
 
@@ -167,7 +167,7 @@
     return
RealClassV8Internal::funcTestInterfaceImplementedAsParamMethod(args);
 }
 
-} // namespace TestInterfaceImplementedAsV8Internal
+} // namespace RealClassV8Internal
 
 static const V8DOMConfiguration::BatchedAttribute
V8TestInterfaceImplementedAsAttrs[] = {
     // Attribute 'a' (Type: 'attribute' ExtAttr: '')


Should these tests be rebaselined?

[Rik, 2013-05-14 13:55:50]

There was some concern about adding Path to the DOM:
https://chromiumcodereview.appspot.com/14988008/

paper.js websites might break with this change

[abarth-chromium, 2013-05-14 16:41:15]

I thought this change didn't have an observable behavior change.

[arv (Not doing code reviews), 2013-05-14 17:30:05]

On 2013/05/14 16:41:15, abarth wrote:
> I thought this change didn't have an observable behavior change.

Yeah, this doesn't change the APIs to JavaScript in any way.