[Webkit-unassigned] [Bug 13136] Spurious glyphs in Google Israel and Gmail (all languages)

[bugzilla-daemon at webkit.org]

http://bugs.webkit.org/show_bug.cgi?id=13136





------- Comment #12 from jungshik.shin at gmail.com  2007-03-22 03:27 PDT -------
David and mitz,

Thanks a lot for your detailed comments.  I'll try to come up with a html file
with test cases and attach it here.

As for group A and group B. My explanation was not very clear. 

Simply put, characters in group A should be rendered with zero-width glyph (no
visible effect at all) *no matter what*. Therefore, we should block them higher
up instead of passing them down. LRM and RLM belong to this group (especially
considering that they don't have any more role once they're taken care of by 
WebCore BiDi algorithm as you explained). So, your concern about Windows
version of Arial having a bogus glyph can be addressed.

As for characters in group B, we expect fonts and drawing/measuring routines to
do a 'reasonable' (not perfect) job. Even if they're not combined well with a
base character or rendered with a little-off glyph, it's OK. 

In case of ZWNJ and ZWJ, I believe they should go down along the complex path
(in most cases). My understanding of complex path  in WebKit (and ATSUI) is
rather weak. 

> A "last resort" glyph (or a glyph from a different font) is used only in cases > where the original font does not have a  mapping for the character. 

I don't have  the source code at the moment and am writing from the memory. If
I remember correctly and read the code correctly, there's some like this in
complex code path:

 1. even a system fallback fails
 2. do something as the last resort

What I'm saying is for group B, we should not take step 2 and just ignore them. 

If no other font covers a given character, on Mac OS X, I believe 'last resort'
font comes into play. That 'font' (is it a real font?) has glyphs for script
blocks. For characters like Hebrew vowel signs, using that glyph representing
Hebrew script is not so good as just ignoring them. 


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