On 2010-11-08, Jürgen Spitzmüller wrote:
> Uwe Stöhr wrote:
>> We should consider do do this for 2.0 because the users-list is full
>> with requests of XeTeX and we already implemented basic support.
>> Jürgen, what do you think, will you have time to implement support for
>> polyglossia?
> No. And I would rather postpone it, since the language handling is deeply
> entangled with the code.
The language switching code in the document is nearly the same. It's the
preamble code that differs:
-\usepackage[british,french,ngerman,english]{babel}
+\usepackage{polyglossia}
+\setdefaultlanguage{english}
+\setotherlanguages{german,british,french}
OTOH, there is a different set of supported languages and some languages
have different names:
# Additionally supported or differently named languages:
language_codes.update({
# code Polyglossia-name comment
'cop': 'coptic',
'de': 'german', # new spelling (de_1996)
'de_1901': 'ogerman', # old spelling
'dsb': 'lsorbian',
'el_polyton': 'polygreek',
'fa': 'farsi',
'grc': 'ancientgreek',
'hsb': 'usorbian',
'sh-cyrl': 'serbian', # Serbo-Croatian, Cyrillic script
'sh-latn': 'croatian', # Serbo-Croatian, Latin script
'sq': 'albanian',
'sr': 'serbian', # Cyrillic script (sr-cyrl)
'th': 'thai',
# zh-latn: ??? # Chinese Pinyin
})
# Languages without Polyglossia support:
for key in ('af', # 'afrikaans',
'de_at', # 'naustrian',
'de_at_1901', # 'austrian',
'fr_ca', # 'canadien',
'grc_ibycus', # 'ibycus', (Greek Ibycus encoding)
'sr-latn', # 'serbian script=latin'
'vi'): # 'vietnam',
del(language_codes[key])
Also, polyglossia is said to be working with XeTeX only, not with
LuaTeX while the newest fontspec supports both engines.
As an alternative, we could consider "Unicode-safe" Babel *.ldf files
for e.g. Russian, Greek, ...
Unfortunately, babel is not very actively maintained but also not free
for adoption.
Günter
