[Webkit-unassigned] [Bug 127324] REGRESSION (r160259?): text-combine glyphs are not rendered

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Tue Feb 18 02:57:58 PST 2014


https://bugs.webkit.org/show_bug.cgi?id=127324





--- Comment #10 from Yuki Sekiguchi <yuki.sekiguchi at access-company.com>  2014-02-18 02:55:08 PST ---
Hi Koivisto,

Thank you for reviewing.

(In reply to comment #9)
> (From update of attachment 224467 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=224467&action=review
> 
> I think at minimum this needs to have a good test case.
> 
> > Source/WebCore/ChangeLog:16
> > +        No new tests.
> > +        Image test of fast/text/international/text-combine-image-test.html covers this.
> 
> It shouldn't be hard to add a proper test for this that doesn't require running in pixel mode. Reftest (that is test.html/test-expected.html pair) might be easiest.

I tried to create reftest using text-orientation: upright, transform: rotate(-90deg) or writing-mode: vertical, but all of them draw horizontal text to different position from text-combine: horizontal.
Do you have any idea to create reftests?

> > Source/WebCore/rendering/svg/RenderSVGInlineText.h:55
> > +    virtual void setTextInternal(const String&, bool replace = false) override;
> 
> It would be better to add an explicit API, RenderText::overrideOriginalTextForTextCombine or similar instead of adding a confusing new parameter. (Also I don't fully understand why RenderCombineText needs this special behavior.)

r77153 says "If a new font found, we replace the text with 0xFFFC. This is needed for a combined text block to be able to behave like a single character against text decorations."
However, I don't think this replacement is not needed in the current code.

I'm creating removing replacement patch. This also fix odd line break in red border block in fast/text/international/text-combine-image-test.html. I'm investigating why this bug is fixed.

After the investigation is finished, I'll upload a patch and ask review. If you think this method is wrong, please correct me.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.


More information about the webkit-unassigned mailing list