[webkit-reviews] review granted: [Bug 220904] Clean up CharacterData : [Attachment 418241] Patch

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Mon Jan 25 17:56:19 PST 2021


Ryosuke Niwa <rniwa at webkit.org> has granted Yusuke Suzuki <ysuzuki at apple.com>'s
request for review:
Bug 220904: Clean up CharacterData
https://bugs.webkit.org/show_bug.cgi?id=220904

Attachment 418241: Patch

https://bugs.webkit.org/attachment.cgi?id=418241&action=review




--- Comment #3 from Ryosuke Niwa <rniwa at webkit.org> ---
Comment on attachment 418241
  --> https://bugs.webkit.org/attachment.cgi?id=418241
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=418241&action=review

> Source/WebCore/dom/CharacterData.cpp:179
> +    NodeType nodeType = this->nodeType();

nodeType is a virtual function. Let's not use that?

> Source/WebCore/dom/CharacterData.cpp:187
> -    if (is<Text>(*this) && parentNode())
> +    if (nodeType == TEXT_NODE && parentNode())

Keep the old code. is<Text>(*this) is a node flag check (bit flag).

> Source/WebCore/dom/CharacterData.cpp:190
> -    if (is<ProcessingInstruction>(*this))
> +    if (nodeType == PROCESSING_INSTRUCTION_NODE)

Let's just add ProcessingInstruction as a node flag too. We've got quite a few
bits available in node flags now.

> Source/WebCore/dom/Text.h:61
> +    // Like appendData, but optimized for the parser (e.g., no mutation
events).
> +    // Returns how much could be added before length limit was met.

I don't think this comment is necessary.


More information about the webkit-reviews mailing list