[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Fwd: Fwd: [Issue 14590] - LTR text ending with punctuation in RTL paragraph misordered]
- To: developer at arabeyes dot org, karouri at openoffice dot org
- Subject: Re: [Fwd: Fwd: [Issue 14590] - LTR text ending with punctuation in RTL paragraph misordered]
- From: Munzir Taha Obeid <munzirtaha at myrealbox dot com>
- Date: Tue, 3 Jun 2003 16:23:03 +0000
- Cc: karouri at uofk dot edu
- Organization: New Horizons CLC
- User-agent: KMail/1.5
First, I welcome you on the list and happy to see you here dear karouri, I am
not going to introduce you because I know you won't like it but let your
deeds speak loudly.
Second, I expected some one who is more knowledgeable than me to respond to
this but since all of them are busy let me try.
I agree with ssa at openoffice dot org dot The UNICODE specification assigns
directionality to Unicode characters and defines a (complex) algorithm for
determining the proper directionality of text. The Unicode bidirectional
algorithm requires an initial text direction.
Think of it the other way around. Suppose I want to write
البرمجة بلغة C + البرمجة ب .Net تعني أنك ربما تكون مبرمجاً
In this case you want the plus to appear after the C to the left (so how can
the application differentiate between my case and yours?) whereas I
want the dot to appear to the right of the letter N (which is not the normal
case). The point is the application couldn't tell what the user intention is
unless explicitly defined by the user.
OK, now comes the question of how to solve it. Simply using Unicode control
characters. RLM and LRM will solve it. I am not sure how to insert it in
OpenOffice since I haven't it installed but there should be away. I tried it
in Katoob and it works.
An enhancement to this is to use a high level protocol that will insert these
characters automatically via a shortcut key like Ctrl+Right Shift and
Ctrl+Left Shift as it's the case with M$ Word but this is another issue and
it is surely an enhanement rather than a bug.
There are useful info on these two docs
http://www.unicode.org/unicode/reports/tr9/
Behdad Esfahbod (بهداد اسفهبد), Fribidi maintainer,
and Roozbeh Pournader (روزبه پورنادر) who contributed to this Annex are
themselves on the Arabeyes list so let's hope they have time to confirm.
http://www.abisource.com/information/news/2002/awn95.phtml
which is very useful in case you are in a hurry.
> - Can anyone test this in Linux?
> - Is it possible that they are right in this issue? I believe that it
> should be possible to control the placement of weak characters, based on
> the writing direction last specified not on the paragraph direction..
>
> -----Forwarded Message-----
>
> >
> > > ------- Additional Comments From ssa at openoffice dot org
> > > 2003-05-28 07:38 PDT -------
> > > This is the correct behaviour. Weak characters (like
> > > '+') are always
> > > written in the current writing direction, ie, to the
> > > left when you are
> > > in an LTR paragraph. As soon as they are enclosed in
> > > unique scripts,
> > > the direction is known and they are reversed. First
> > > type C++ and then
> > > D, as a result ++C will change to C++D.
> > > Because there is no portable way to obtain the
> > > current input language,
> > > there is no hint available to decide what the user
> > > might have meant.
--
__/ __/ _/ _/ _/
_/ _\ _/ _\ _/
_/ \ / _\ unzir _/ aha
Telecommunications Engineer
Linux Registered User #279362 at http://counter.li.org
CIW, ICDL, MOUS.
New Horizons CLC
Riyadh, SA