On Aug 1, 2023, 10:09 AM -0600, Hans Hagen <j.hagen@xs4all.nl>, wrote:
** Caution: EXTERNAL Sender **

On 8/1/2023 4:18 PM, Hamid,Idris wrote:
Dear gang,

It appears that \defineactivecharacter does not work in lmtx:

\starttext
\defineactivecharacter Ḥ {\d{H}}
\defineactivecharacter ḥ {\d{h}}
\defineactivecharacter Ṣ {\d{S}}
\defineactivecharacter ṣ {\d{s}}
\stoptext

This should not produce any typeset output but in lmtx it does.

Background: The following characters are essential for Arabic transliteration (assuming your email client has the chars):

Ṯṯ Ḥḥ Ḫḫ Ḏḏ Šš Ṣṣ Ḍḍ Ṭṭ Ẓẓ Ġġ Āā Īī Ūū ʿ ʾ
I'd use a different font but as you want to go cheap ... attached gets
you going and will give you a few hours playing around with fonts.

Many thanks. It would be great to see this developed into a module; see below:
Using a scaled c for some c like shape is kind of bad anyway, kind of
using a comma for an ogonek.

The issue is this: The use of 02BE and 02BF (ʿ ʾ) to transliterate the two glottal stops in Arabic is now standard in scholarship, but the default shape of these two characters is unattractive. So some typesetters began replacing the default half-ring looks with superscript 'c' and superscript mirrored c. Here is a article about this:

http://andreasmhallberg.github.io/typographyofaynandhamza/

Now, as the author notices, the TeX Gyre fonts render 02BE and 02BF as superscript 'c'. (That was added to TeX Gyre at our request many years ago.)

Linked below is a pdf that shows some font rendering of Latin Modern, Pagella, Brill - a nice gratis font for download but with a strange license, so dangerous to use for real work -, and Minion. It also includes a table of composition analysis for each of the transliteration characters.

https://www.dropbox.com/scl/fi/m1e3quht30auzi3rxyhzb/test-transliteration.pdf?rlkey=o18dlm85q2ie80iv0004lynlv&dl=0

Here is the test string:

Ṯṯ Ḥḥ Ḫḫ Ḏḏ Šš Ṣṣ Ḍḍ Ṭṭ Ẓẓ Ġġ Āā Īī Ūū ʿ ʾ

ʿAyn QURʾĀN

The TeX Gyre fonts use superscript c; Brill tries to make the half-rings elegant, and Minion is missing many characters.

Another approach might be to use a breve shape (rotated 90 degrees in either direction) instead of superscript c to represent the half rings. In any case, a good module would facilitate exploration of such possibilities.

Thanks again, Hans.

Idris
--
Professor Idris Samawi Hamid
Department of Philosophy
Colorado State University
Fort Collins, CO 80523