[XeTeX] Polyglossia and Sanskrit

[Yves Codet]

Hello.

Le 5 oct. 08 à 12:55, François Patte a écrit :

> Le 05.10.2008 10:29, Daniel Stender a écrit :
>> Yves Codet wrote:
>>
>>> I wonder if this behaviour of Polyglossia should not be changed.
>>> Actually Sanskrit has no proper script, or it has many. If
>>> Devanagari is often preferred nowadays if does not mean that it is
>>> *the* Sanskrit script. I would be interested to know what members
>>> of the list think.
>>
>> I think principally that's right. I don't know what is the case with
>> other scripts than Devanagari, but it could be argued that there are
>> even differences in hyphentation between Devanagari and
>> transliteration (that means romanized Sanskrit), while Devanagari
>> hyphens Akshara-wise that could be handled more freely in
>> Romanization: "satyam eva jayate" is possible instead of "satyameva
>> jayate" which results in "sa-tyam e-va ja-ya-te" instead of
>> "sa-tya-me-va ja-ya-te"
>>
>> Furthermore I think a separated hyphenation pattern for romanization
>> is needed, for example nobody hyphens "ga-ccha-ti" but "gac-cha-ti",
>> isn't it?
>
> This problem has been argued a long time ago (Indology, Wujastik  
> and co).
>
> Practice of Sanskrit manuscripts shows that there is absolutely no
> pattern to hyphenate a text in Indian scripts, the scribe writes his
> text continuously and goes to the following line anywhere inside a  
> word.
> So the akshara model seems to be a good (and I think easy)  
> solution. But
> I am wondering if the same file can be use to automatically  
> hyphenate a
> Sanskrit text whatever the script. I you have a text in grantha  
> script,
> as there are less ligatures in this script than in denanagari, can we
> use the same hyphen-file with xe(la)tex?

I think the same file could be used for all scripts deriving from  
Brahmi, even Grantha. If we prohibit a break between two consonants  
which are graphically realised as a conjunct, can we allow it between  
two consonants to which no conjunct corresponds? The same issue  
arises in Devanagari as well. Some conjuncts are very rare and do not  
exist in all fonts. Should there be a special way to handle such  
cases? I would rather be inclined to stick to the same rule for all  
sequences of consonants.

Kharoshthi will perhaps need separate hyphenation patterns, but we  
only have a proposal of encoding at the moment and we must wait and  
see how it will be encoded in the end.

> Moreover, for grantha, there is not yet an unicode scheme for this
> script and the only fonts I found uses the unicode Bengali scheme!

As far as I know Grantha has not even been submitted to the Unicode  
Consortium.

Best wishes,

Yves