Looks fine to me.
On 10/12/12 16:43, Alexander Scherbatiy wrote:
Could you review the updated fix:
http://cr.openjdk.java.net/~alexsch/8000626/webrev.01/
The only changes is that the key typed event is not generated for the
dead keys.
Thanks,
Alexandr.
On 10/9/2012 7:51 PM, Alexander Scherbatiy wrote:
bug: http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=8000626
webrev: http://cr.openjdk.java.net/~alexsch/8000626/webrev.00
- The Java Key code should keep modifiers for the dead key detection.
For other keys the Shift modifier usually is cleaned.
- Only extended Java key code keep information about dead keys.
Pass extended java key to postKeyEvent method as the key code
argument.
- Assign unicode chars for the dead key symbols according to table:
http://www.unicode.org/Public/UNIDATA/UnicodeData.txt
This is done to be compatible with Windows and Mac OS X where there
are dead key chars in the key pressed and key released methods.
- The XKeysym.java file is generated from the UnicodeData.txt file
and keysym2ucs.h .
Some code that was only in the XKeysym.java file (like
XKeySymConstants.XK_Shift_Lock) is moved to keysym2ucs.h so to not
miss it after XKeysym.java regeneration.
Thanks,
Alexandr.
