Move code that maps a CompositeOperator and BlendMode to a CGBlendMode into a helper function

[Simon Fraser (smfr)]

Move code that maps a CompositeOperator and BlendMode to a CGBlendMode into a helper function

[Simon Fraser (smfr)]

Created attachment 326050 [details]
Patch

[Simon Fraser (smfr)]

Created attachment 326056 [details]
Patch

[Simon Fraser (smfr)]

Created attachment 326057 [details]
Patch

[Darin Adler]

Comment on attachment 326057 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=326057&action=review

> Source/WebCore/platform/graphics/cg/GraphicsContextCG.cpp:140
> +static CGBlendMode compositeAndBlendToCGBlendMode(CompositeOperator compositeOperator, BlendMode blendMode)

Since this is C++ I don’t think we really need to restate what the arguments are in the function name. I think this should be called something like selectCGBlendMode.

I think it’s a little strange that it ignores the composite operator when the blend mode is something other than normal. It’s hard to understand why that’s OK. I’m sure there is a good reason. I would love to have a comment explaining it.

[Simon Fraser (smfr)]

(In reply to Darin Adler from comment #4)
> Comment on attachment 326057 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=326057&action=review
> 
> > Source/WebCore/platform/graphics/cg/GraphicsContextCG.cpp:140
> > +static CGBlendMode compositeAndBlendToCGBlendMode(CompositeOperator compositeOperator, BlendMode blendMode)
> 
> Since this is C++ I don’t think we really need to restate what the arguments
> are in the function name. I think this should be called something like
> selectCGBlendMode.

I'll change it.

> I think it’s a little strange that it ignores the composite operator when
> the blend mode is something other than normal. It’s hard to understand why
> that’s OK. I’m sure there is a good reason. I would love to have a comment
> explaining it.

Yeah. I'm not sure why we distinguish between CompositeOperator and BlendMode throughout the WebKit code; it seems like some spec purity modeled after https://www.w3.org/TR/compositing-1. CG smooshes them into one enum.

[Simon Fraser (smfr)]

https://trac.webkit.org/changeset/224465/webkit

[Radar WebKit Bug Importer]

<rdar://problem/35567092>