[Webkit-unassigned] [Bug 50147] [Qt] 4 fast/text tests fail with Qt 4.7.1

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=50147


Jocelyn Turcotte <jocelyn.turcotte at nokia.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jocelyn.turcotte at nokia.com




--- Comment #5 from Jocelyn Turcotte <jocelyn.turcotte at nokia.com>  2011-04-01 04:31:00 PST ---
(In reply to comment #4)
> Do you have any advice on how to best correct?  Should we be taking the
> complex path for zero-width chars?  Or is there something deeper wrong?

As far as I know the main difference between the fast and the complex path is that the complex path will use all the OpenType information to position the glyphs, while the simple path only use horizontal "advance widths".

What seems to happen in your case is that Qt do not get the right advance width value when loading the test font, which is calculated correctly somehow when using harfbuzz in the complex path. At this level this code in Qt is quite different among platforms, so you could also try to see if you correctly get the zero-width on other platforms for these fonts.

A workaround would be to modify Font::codePath to force the complex path for these characters, but that would be a rather ugly idea.

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