[webkit-reviews] review granted: [Bug 65902] Spell-checking doesn't recognize word boundaries on contests inserted by execCommand('insertHTML') : [Attachment 104156] Patch

Wed Aug 17 00:42:16 PDT 2011

Ryosuke Niwa <rniwa at webkit.org> has granted MORITA Hajime
<morrita at google.com>'s request for review:
Bug 65902: Spell-checking doesn't recognize word boundaries on contests
inserted by execCommand('insertHTML')
https://bugs.webkit.org/show_bug.cgi?id=65902

Attachment 104156: Patch
https://bugs.webkit.org/attachment.cgi?id=104156&action=review

------- Additional Comments from Ryosuke Niwa <rniwa at webkit.org>
View in context: https://bugs.webkit.org/attachment.cgi?id=104156&action=review

> LayoutTests/editing/spelling/spelling-insert-html.html:1
> +<html>

missing DOCTYPE?

> LayoutTests/editing/spelling/spelling-insert-html.html:34
> +    // The first "foo" isn't checked because it crosses the pasted and base
html.

Is this a bug?	If also, you should say that "this demonstrates a bug X" where
X is the bug number in the description above.

> Source/WebCore/editing/Editor.cpp:1945
>      bool markSpelling = isContinuousSpellCheckingEnabled();
>      bool markGrammar = markSpelling && isGrammarCheckingEnabled();

It appears that we don't need these booleans anymore, especially markSpelling.

> Source/WebCore/testing/Internals.cpp:184
> +unsigned Internals::markerCountOf(Node* node, ExceptionCode& ec)

I think we should include "Node" in the name since this is a ES5 function and
ES5 is duck-typed; i.e. can't tell what the argument is.

How about markerCountForNode?

> Source/WebCore/testing/Internals.cpp:194
> +PassRefPtr<Range> Internals::markerRangeAt(Node* node, unsigned index,
ExceptionCode& ec)

"RangeAt" sounds like you're obtaining a marker at a particular DOM position. 
How about just markerForNode?