[Webkit-unassigned] [Bug 53696] Caret is rendered at an incorrect position at the boundary of Arabic number in a LTR context

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=53696





--- Comment #27 from Ryosuke Niwa <rniwa at webkit.org>  2011-03-07 18:41:23 PST ---
After some discussion with Xiaomei, I'm little more convinced that my current approach (at least offset assignments) is indeed correct.  We considered several possibilites for offsets in CBA123 (arabic numbers followed by arabic letters in a LTR block) here:

First, the following is the offsets assigned by Firefox/IE.
Internet Explorer: (0)(1)(7)C(6)B(5)A(4)1(2)2(3)3
Firefox: (0)(7)C(6)B(5)A(1)(4)1(2)2(3)3

Note that neither assignments seem ideal in this case and both of them have offset 7 even though there are only 6 letters.  Now, let us consider 4 offset assignments:

(0)C(5)B(4)A(1)1(2)2(3)3(6) - The offsets between ABC contradict with those of Firefox/IE.  Also it's odd that even though (1) corresponds to the position before 2 but it appears between A and 1.

(0)C(6)B(5)A(1)1(2)2(3)3(4) - This is a little better in that offsets between CBA match with those of Firefox/IE.

(6)C(5)B(4)A(0)1(1)2(2)3(3) - This is similar to what TextEdit, Firefox, Internet Explorer does but contradicts with the offset assignments in WebKit when CBA appeared in a LTR block: (0)C(2)B(1)A(3).  Also, the offsets between ABC contradict with those of Firefox/IE.

(0)C(5)B(4)A(3)1(1)2(2)3(6) - This is similar to the above but all offsets except (0) and (6) flows in the same order as those of Firefox and Internet Explorer except (1) and we're off by 1.

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