A Semetic Aporia

[Idris S Hamid <ishamid@attbi.com>]

Dear Hans,
I have a serious problem: 

In Classical Arabic numerals are pronounced from single digits to tens to 
hundreds, etc., e.g. 

1234== four, thirty, two hundred, and a thousand. This makes transcribing 
Arabic numerals easy: they run the same as in roman languages. So in my otp's 
(Omega-Gamma) there is a command to transcribe any left-to-right L-R string 
numeral into an L-R string in Arabic (so I don't have to type the numeral 
backwards in the L-R transcription; i.e., typing 1234 in L-R comes out the 
same in R-L).

But there is a complicated scenario: For purposes of sectioning, we get (as in 
a TOC) for e.g. Section 3.12

in roman it should be (L-R):
3.12-------------------------------------------<page no.>

in Arabic it should be (R-L):
<page no.>-------------------------------------------12.3

Now my Gamma configuration is such that everything is flipped R-L except for 
numerals, so I _actually_ get
<page no.>-------------------------------------------3.12

in my Arabic sections and TOC. Is there any way to reverse the order of 
subsections so that subsection 3.12 comes out as 12.3? I assume there is some 
code that commands ConTeXt to print the section number, followed by the 
subsection number. Can that be hacked to reverse the sequence for Semetic 
documents (Arabic and Hebrew)? So we could have an option say,  
[style=semetic]. I did something similar in EDMAC once for Arabic critical 
editions.

I am finishing up a treatise in Arabic to be submitted later this week so I'm 
desparate. I guess I could replace all the numerals using the manual 
mechanism but there must be a better way!

Best wishes
Idris
-- 
Dr. Idris S Hamid
Department of Philosophy
Colorado State University
Fort Collins, CO 80523