Clean up some code around RenderElement::addLayers()

• Component: Layout and Rendering
• Status: RESOLVED FIXED

[Simon Fraser (smfr)]

Clean up some code around RenderElement::addLayers()

[Simon Fraser (smfr)]

Created attachment 449267 [details]
Patch

[Darin Adler]

Comment on attachment 449267 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=449267&action=review

> Source/WebCore/rendering/RenderElement.cpp:640
> +        if (!beforeChild.has_value())

You can omit the ".has_value()" if you like. But if you think including it makes the code clearer, that’s fine.

> Source/WebCore/rendering/RenderElement.cpp:643
> +        parentLayer.addChild(*downcast<RenderLayerModelObject>(currentRenderer).layer(), beforeChild.value());

I often prefer *beforeChild, treating an optional as if it was a pointer, but if you like value() better, I’m OK with it.

> Source/WebCore/rendering/RenderElement.cpp:699
> +        for (auto* child = siblingToTraverseFrom ? siblingToTraverseFrom->nextSibling() : firstChild(); child; child = child->nextSibling()) {

Seems like childAfter helper function, either member or non-member, that returns first if the argument is nullptr and nextSibling if it’s not would make this nicer to read.

[Simon Fraser (smfr)]

When dealing with optionals of bools or pointers I like using the explicit has_value() and value() functions.

[EWS]

Committed r288127 (246137@main): <https://commits.webkit.org/246137@main>

All reviewed patches have been landed. Closing bug and clearing flags on attachment 449267 [details].

[Radar WebKit Bug Importer]

<rdar://problem/87722144>