[webkit-reviews] review granted: [Bug 191532] ASSERTION FAILED: !m_embeddedObjectsToUpdate->contains(nullptr) in WebCore::FrameView::updateEmbeddedObjects : [Attachment 390903] Patch

[bugzilla-daemon at webkit.org]

Darin Adler <darin at apple.com> has granted Jack <shihchieh_lee at apple.com>'s
request for review:
Bug 191532: ASSERTION FAILED: !m_embeddedObjectsToUpdate->contains(nullptr) in
WebCore::FrameView::updateEmbeddedObjects
https://bugs.webkit.org/show_bug.cgi?id=191532

Attachment 390903: Patch

https://bugs.webkit.org/attachment.cgi?id=390903&action=review




--- Comment #16 from Darin Adler <darin at apple.com> ---
Comment on attachment 390903
  --> https://bugs.webkit.org/attachment.cgi?id=390903
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=390903&action=review

Looks good

> Source/WebCore/html/HTMLAppletElement.cpp:99
> +    return canEmbedJava()? HTMLPlugInElement::renderWidgetLoadingPlugin() :
nullptr;

WebKit coding style: Need a space before the "?" here.

When calling through to "super", I think it’s usually better style to write the
actual base class, HTMLPlugInImageElement, rather than the class that has the
function in it skipping a level of inheritance HTMLPlugInElement.

> Source/WebCore/html/HTMLEmbedElement.cpp:85
> +    RenderWidget* widget = HTMLPlugInElement::renderWidgetLoadingPlugin();
> +    
> +    return widget? widget : findWidgetRenderer(this);

Ditto, same two comments here.

> Source/WebCore/html/HTMLPlugInElement.cpp:169
> +    return renderWidget(); // This will return 0 if the renderer is not a
RenderWidget.

I think we should say nullptr rather than "0" in this comment. I know we are
just moving it.