emacs-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Emacs-24 Hangs When Displaying Unicode #x6c0 (ARABIC HEH WITH YEH ABOVE)


From: Mohsen BANAN
Subject: Emacs-24 Hangs When Displaying Unicode #x6c0 (ARABIC HEH WITH YEH ABOVE)
Date: Tue, 06 Mar 2012 10:33:25 -0800
User-agent: Gnus/5.13 (Gnus v5.13) Emacs/24.0.50 (gnu/linux)


Emacs-24 hangs when displaying unicode #x6c0
(arabic heh with yeh above) -- at least in my
environments. The display hangs but the emacs process
stays alive.

Before trying to reproduce the problem, save all
your buffers.

You can reproduce the problem by eval-ing
    (ucs-insert 1728)

I am not including the actual character; as a whole
lot of people would likely be unhappy if I did.

This problem is reproducible in all my
environments Debian and Ubuntu emacs-24.0.50 through
emacs-24.0.90.4.

I have looked some into this problematic character.

On emacs23, describe-char on that character produces:

    character: (1728, #o3300, #x6c0)
preferred charset: unicode (Unicode (ISO10646))
       code point: 0x06C0
           syntax: w    which means: word
         category: .:Base, b:Arabic
      buffer code: #xDB #x80
        file code: #xDB #x80 (encoded by coding system utf-8-unix)
          display: by this font (glyph code)
    xft:-unknown-B Compset-normal-normal-normal-*-13-*-*-*-*-0-iso10646-1 (#xC7)

Character code properties: customize what to show
  name: ARABIC LETTER HEH WITH YEH ABOVE
  old-name: ARABIC LETTER HAMZAH ON HA
  general-category: Lo (Letter, Other)
  decomposition: (1749 1620)

According to isiri-6219, in Persian it is a
forbiden character. In other words persian input
methods should not produce this character --
because it can be produced by 1749+1620.

But, Microsoft does produce that character. So, I
have been receiving various emails that were/are
hanging my emacs display. It took a while to
narrow it down to this character.

If this problem is not specific to my
environement, then perhaps we should also look for
other problematic characters.

Perhaps we can have a buffer that contains all
unicode characters and have the display go through it
all as a regression tester.

Please let me know if this problem is reproducible in your
environements.

Thanks.

...Mohsen





reply via email to

[Prev in Thread] Current Thread [Next in Thread]