Hello,
can someone help him? I have not much time at the moment.
Please CC him since he may not yet be on the list.
(I reduced a bit the number of attached files, you miss those:
> ------------------------------------------------------------------------
> Name: il5hyp.dvi
> il5hyp.dvi Type: TeX DVI Data (application/x-dvi)
> Encoding: base64
>
> Name: il5hyp.tex
> il5hyp.tex Type: TeX Document (application/x-tex)
> Encoding: quoted-printable
>
> Name: latex-hyp.dvi
> latex-hyp.dvi Type: TeX DVI Data (application/x-dvi)
> Encoding: base64
>
> Name: latex-hyp.tex
> latex-hyp.tex Type: TeX Document (application/x-tex)
> Encoding: quoted-printable)
Tobias
-------- Original Message --------
Subject: Turkish language issues with ConTeXt
Date: Sat, 1 Feb 2003 08:48:10 +0200
From: Alper Ersoy <dirt(a)gtk.org>
To: Tobias Burnus <tobias.burnus(a)physik.fu-berlin.de>
Hello Tobias,
A few days ago I have installed ConTeXt and since then I am
consistently trying to achieve the
\usepackage{ae} % calls \usepackage[T1]{fontenc}
\usepackage[latin5]{inputenc}
effect known from LaTeX.
Since it is doable in LaTeX, I am positive that it is also doable in
ConTeXt, I just do not know how yet.
I have composed two TeX files for ConTeXt to address the input
encoding issue. (Those files are regi-il5.tex and enco-il5.tex.)
With regi-il5.tex I am able to type accented characters without any
problems, if you disregard hyphenation. When hyphenation is taken
into account, I am completely lost. Not that the words with accented
characters are completely unhyphenated, but only some of them are.
I am attaching my regi-il5.tex and a test file (and its DVI form, to
save you from tinkering with your setup just to see the problems
I describe below). Please note that I have tried many variations of
both the macro I call (such as \definecharacter, etc.), and the way
I type the substituted accents (\c c vs. \ccedilla). Here are the
problems:
- Even though the hyphenation for a\u{g}ra is correct, I cannot
redefine it using ð (\u{g} in Latin5 encoding). I can redefine it
using \u{g}, but then I lose all of its hyphenation.
- Words which have one of \i (dotless i) or \.I characters are not
hyphenated at all. This is most problematic in the second column
of the sample text.
- Some of the hyphenations are wrong. For example da-\u{g}-lar, it
should be da\u{g}-lar. Similarly ta-\c{s}-lar should be
ta\c{s}-lar. Also \c{c}o-rap\c{c}\i should have an hyphen before
the second ccedilla (though I have an intuition that it is due to
the right syllable being short, can it be?).
I am also attaching a similar file for LaTeX (again, with related DVI
output). Please notice how words containing dotless i characters are
correctly hyphenated. Since ConTeXt and LaTeX share the same
hyphenation patterns, I cannot see why they are generating different
outputs. From texexec's output:
language : patterns 2:2-tr-2 3:2-tr-2 4:2-en-2 5:2-uk-2 6:2-de-2
2-fr-2 9:2-it-2 2-nl-2 loaded
Possible clash with hyphenation pattern files? Or do I have to
explicitly state that I want tr hyphenation in my test file? (tr is
the default language.)
What can be the source of my problems? I am willing to provide you
any information regarding my TeX/ConTeXt/LaTeX setup you may need.
Once I have all going, I will send everything to Hans Hagen to be
included in the future ConTeXt releases.
By the way, I am a total newbie with TeX. This can be the source of
all my problems ;) Also, tomorrow is my birthday, so I am looking
forward to seeing a few presents, such as a fix ;)
Thanks.
Regards,
PS. In one of Hans Hagen's earlier messages to ntg-context list (dated
around 1998), he mentions your name as the language guru, that is why
I am sending this to you. Also I am having problems with the address
I used when subscribing to ntg-context list, and I do not want to
delay my ConTeXt adventure anymore. I hope I am not being rude.
--
Alper Ersoy