Bug 99731 – Add 8-bit path to RenderBlock::handleTrailingSpaces()

• Status: RESOLVED
• Resolution: FIXED
• Component: Layout and Rendering

[Michael Saboff]

RenderBlock::handleTrailingSpaces() uses the characters() method to look for the range of space characters at the end of a text run.  This should be updated to use characters8() or characters16() as appropriate.

[Michael Saboff]

Created attachment 169430 [details]
Patch

Comment on attachment 169430 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=169430&action=review

> Source/WebCore/rendering/RenderBlockLineLayout.cpp:1030
> +        UChar current = characters[firstSpace - 1];
> +        if (!isCollapsibleSpace(current, lastText))

Have you tried making current a CharacterType and making isCollapsibleSpace() a function template as well?

[Michael Saboff]

(In reply to comment #2)
> (From update of attachment 169430 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=169430&action=review
> 
> > Source/WebCore/rendering/RenderBlockLineLayout.cpp:1030
> > +        UChar current = characters[firstSpace - 1];
> > +        if (!isCollapsibleSpace(current, lastText))
> 
> Have you tried making current a CharacterType and making isCollapsibleSpace() a function template as well?

In most cases I haven't done this.  The implicit conversion from an LChar (unsigned char) to a UChar (unsigned short) is typically a nop.  If you think it is more readable I can make that change.

Comment on attachment 169430 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=169430&action=review

>>> Source/WebCore/rendering/RenderBlockLineLayout.cpp:1030
>>> +        if (!isCollapsibleSpace(current, lastText))
>> 
>> Have you tried making current a CharacterType and making isCollapsibleSpace() a function template as well?
> 
> In most cases I haven't done this.  The implicit conversion from an LChar (unsigned char) to a UChar (unsigned short) is typically a nop.  If you think it is more readable I can make that change.

If it doesn’t change the generated code, then no, I don’t think it’s necessary.

[WebKit Review Bot]

Comment on attachment 169430 [details]
Patch

Clearing flags on attachment: 169430

Committed r131776: <http://trac.webkit.org/changeset/131776>

[WebKit Review Bot]

All reviewed patches have been landed.  Closing bug.