[XeTeX] polyglossia and hebrew: arabic numerals and hyphenation

Peter Dyballa Peter_Dyballa at Web.DE
Sat Sep 26 18:08:16 CEST 2009

Am 26.09.2009 um 17:12 schrieb David Perry:

>>> The xunicode package is useful when you try to use LaTeX  
>>> constructs  like \"u instead of Unicode characters. It also can  
>>> have bad  influence when used in an environment with bidi. In  
>>> that case it's  recommended to load bidi before xunicode. This  
>>> means in your case to  load polyglossia before xunicode.
> In fact, current versions of bidi will generate a warning if you  
> don't load it before xunicode.

And the problem is not but the *problems are* that polyglossia loads  
xunicode and if bidi is loaded before polyglossia then the latter  
undoes bidi fixes to AmS math... (which isn't so bad when you're only  
interested in text)

>> language.dat and language.def don't have an hebrew pattern and babel
>> does not list it. Does this hyphenation support for hebrew exist at
>> all? Am I doing it wrong, very wrong? Is the TL installation
>> misconfigured?
> I don't believe it exists; I have never seen it.  I am far from an  
> expert in Hebrew, but my understanding is that hyphenation (to  
> split words at ends of lines, as used in Latin-script languages) is  
> generally avoided in high-quality Hebrew printing and is forbidden  
> in biblical texts, so it's not surprising that there is no set of  
> hyphenation rules around.

Which then would rather recommend to use a pdfTeX based version  
(IvriTeX?) because of the use of micro-typographical fixes to the  



Bake pizza not war!

More information about the XeTeX mailing list