[Indic-dev] Indic scripts and English in LaTeX: What is the way ahead?

R (Chandra) Chandrasekhar chandra at ee.uwa.edu.au
Wed Oct 4 11:14:11 CEST 2006

CV Radhakrishnan wrote:
> On Tue, 2006-10-03 at 15:50 +0800, R (Chandra) Chandrasekhar wrote:
>> Dear Folks,
>> I do not know if this is the right forum for this question: pardon me if it is not.
>> I am a TUG member who has been using LaTeX for the last 12 years and have been 
>> involved in ongoing work that requires Indic scripts and English in the same 
>> document.  I have been and am still using the excellent Itrans package from 
>> Avinash Chopde.  He has, however, ceased maintenance of Itrans because of the 
>> advent of Unicode, as stated on the Itrans web site.
> Have you tried XeTeX?  If not, please take a look at:
>  http://scripts.sil.org/cms/scripts/page.php?site_id=nrsi&item_id=xetex
> [...]
> Best
> Radhakrishnan

Thank you for this suggestion.

Indeed, I am now trying XeTeX (as XeLaTeX) after installing it on my Linux PC. I 
have had much help from Jonathan Kew and have been able to successfully compile 
a minimal test file that incorporates Tamil, Devanagari (and for a right to left 
script), Arabic.  I have also processed the .tex file through TTH and HeVeA to 
see how the automatic HTML output fares.

Kindly take a look at the following files:

1. http://members.westnet.com.au/chandra/unicodetest.tex: .tex file

2. http://members.westnet.com.au/chandra/unicodetest.pdf: PDF output of .tex file

3. http://members.westnet.com.au/chandra/unicodetest.htm: HTML output after 
running .tex file through tth

4. http://members.westnet.com.au/chandra/unicodetest.html: HTML output after 
running .tex file through HeVeA

It would appear that the XeTeX path along with direct Unicode input is very 
promising because there is no tedious "metafontization" of system-recognized 
fonts.  This means that new and comprehensive TTF fonts, recognized by the 
operating system, may be used as soon as they become available and with little 

One issue that is unclear is how Unicoded Romanized Sanskrit may be input 
directly into the .tex file for use with XeLaTeX, although this is primarily a 
keyboarding problem.  I would be grateful for your suggestions or pointers on this.

I will be writing a comprehensive document that looks at the pros and cons of 
the XeTeX approach to using Indic scripts in LaTeX.  Hopefully, it will cover 
the entire document preparation chain, beginning with keyboarding, and ending 
with finished documents in multiple formats like PDF and HTML, from the same 
source .tex document.  I will post it to this list on completion.

Thank you.

04 Oct 06

More information about the Indic-dev mailing list