Bug 65737 – Remove LegacyDefaultOptionalArguments flag from canvas IDL files

• Component: New Bugs
• Status: RESOLVED FIXED

[Mark Pilgrim (Google)]

As discussed in IRC, we are migrating our IDL files away from the interface-level "LegacyDefaultOptionalArguments" flag and onto argument-level [Optional] or [Optional=CallWithDefaultValue] flags. This patch migrates the remaining canvas-related IDL files. It does not change any behavior, i.e. it does not make any arguments required that were previously optional, nor vice-versa.

All existing tests pass.

[Mark Pilgrim (Google)]

Created attachment 103021 [details]
Patch

[Adam Barth]

Comment on attachment 103021 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=103021&action=review

> Source/WebCore/html/canvas/OESVertexArrayObject.idl:38
> -        [StrictTypeChecking] void         deleteVertexArrayOES(in WebGLVertexArrayObjectOES arrayObject);
> -        [StrictTypeChecking] boolean      isVertexArrayOES(in WebGLVertexArrayObjectOES arrayObject);
> -        [StrictTypeChecking] void         bindVertexArrayOES(in WebGLVertexArrayObjectOES arrayObject) raises(DOMException);
> +        [StrictTypeChecking] void         deleteVertexArrayOES(in [Optional=CallWithDefaultValue] WebGLVertexArrayObjectOES arrayObject);
> +        [StrictTypeChecking] boolean      isVertexArrayOES(in [Optional=CallWithDefaultValue] WebGLVertexArrayObjectOES arrayObject);
> +        [StrictTypeChecking] void         bindVertexArrayOES(in [Optional=CallWithDefaultValue] WebGLVertexArrayObjectOES arrayObject) raises(DOMException);

Is there any bad interaction between StrictTypeChecking and CallWithDefaultValue?

[Mark Pilgrim (Google)]

(In reply to comment #2)
> (From update of attachment 103021 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=103021&action=review
> 
> > Source/WebCore/html/canvas/OESVertexArrayObject.idl:38
> > -        [StrictTypeChecking] void         deleteVertexArrayOES(in WebGLVertexArrayObjectOES arrayObject);
> > -        [StrictTypeChecking] boolean      isVertexArrayOES(in WebGLVertexArrayObjectOES arrayObject);
> > -        [StrictTypeChecking] void         bindVertexArrayOES(in WebGLVertexArrayObjectOES arrayObject) raises(DOMException);
> > +        [StrictTypeChecking] void         deleteVertexArrayOES(in [Optional=CallWithDefaultValue] WebGLVertexArrayObjectOES arrayObject);
> > +        [StrictTypeChecking] boolean      isVertexArrayOES(in [Optional=CallWithDefaultValue] WebGLVertexArrayObjectOES arrayObject);
> > +        [StrictTypeChecking] void         bindVertexArrayOES(in [Optional=CallWithDefaultValue] WebGLVertexArrayObjectOES arrayObject) raises(DOMException);
> 
> Is there any bad interaction between StrictTypeChecking and CallWithDefaultValue?

No. StrictTypeChecking adds some additional type checking after the function declaration is constructed. See line 936 of Source/WebCore/bindings/scripts/CodeGeneratorV8.pm for details. It does not interact with [Optional=CallWithDefaultValue].

[Adam Barth]

Comment on attachment 103021 [details]
Patch

So what happens if you don't supply the argument?  It throws a different sort of exception?

[WebKit Review Bot]

Comment on attachment 103021 [details]
Patch

Clearing flags on attachment: 103021

Committed r92443: <http://trac.webkit.org/changeset/92443>

[WebKit Review Bot]

All reviewed patches have been landed.  Closing bug.

[Kenichi Ishibashi]

Created attachment 103035 [details]
Patch V1(fix mac build failure)

[Kenichi Ishibashi]

Comment on attachment 103035 [details]
Patch V1(fix mac build failure)

Sorry I uploaded the patch to wrong bug id.  Please ignore it.