Bug 136484 – Add support for the initial-letter CSS property to first-letter

• Component: CSS
• Status: RESOLVED FIXED

[Dave Hyatt]

Bug for adding initial-letter support to WebKit.

[Dave Hyatt]

Created attachment 237561 [details]
Patch

[WebKit Commit Bot]

Attachment 237561 [details] did not pass style-queue:


ERROR: Source/WebCore/rendering/RenderBlockFlow.cpp:2350:  Weird number of spaces at line-start.  Are you using a 4-space indent?  [whitespace/indent] [3]
ERROR: Source/WebCore/css/CSSLineBoxContainValue.h:38:  Weird number of spaces at line-start.  Are you using a 4-space indent?  [whitespace/indent] [3]
Total errors found: 2 in 33 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[Dean Jackson]

Comment on attachment 237561 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=237561&action=review

> Source/WebCore/css/CSSValueKeywords.in:974
> +initial-letter

Do you actually ever use this?

> Source/WebCore/platform/graphics/FontMetrics.h:102
> -
> +    

whoopsy

> Source/WebCore/rendering/RenderBlock.cpp:3484
> +        while (actualCapHeight > desiredCapHeight) {

Would this ever be slow?

> Source/WebCore/rendering/RenderBlock.cpp:3487
> +            newFontDescription.setComputedSize(newFontDescription.computedSize() -1);

Nit: missing space between - and 1

> Source/WebCore/rendering/style/RenderStyle.h:1087
> +    const IntSize& initialLetter() const { return rareNonInheritedData->m_initialLetter; }
> +    int initialLetterDrop() const { return initialLetter().width(); }
> +    int initialLetterHeight() const { return initialLetter().height(); }

This is a bit weird - maybe initialLetter() should be private, so that people won't ever think that the IntSize.width() is an actual width.

> Source/WebCore/rendering/style/RenderStyle.h:1620
> +    void setInitialLetter(const IntSize& size) { SET_VAR(rareNonInheritedData, m_initialLetter, size); }

Again, a bit weird that you need to know what is width and what is height.

[Dave Hyatt]

Landed in r173217.

[Tim Horton]

Build fix that I'm not 100% sure about in http://trac.webkit.org/changeset/173221; Dave, if you could check it that would be great.

[Bem Jones-Bey]

Comment on attachment 237561 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=237561&action=review

> Source/WebCore/css/CSSParser.cpp:2436
> +            addProperty(propId, createPrimitiveValuePair(parsedValue1.release(), parsedValue2.release()), important);

This looks dangerous to me: in the case of only 1 value, you're release()ing it twice. Why is this ok here?

[Tim Horton]

Comment on attachment 237561 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=237561&action=review

>> Source/WebCore/css/CSSParser.cpp:2436
>> +            addProperty(propId, createPrimitiveValuePair(parsedValue1.release(), parsedValue2.release()), important);
> 
> This looks dangerous to me: in the case of only 1 value, you're release()ing it twice. Why is this ok here?

There are two RefPtrs; even if they point at the same thing they'll both own one reference apiece, and each release will move said reference into a PassRefPtr.

[Radar WebKit Bug Importer]

<rdar://problem/18941449>