[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: lynx-dev gettext() question #2

From: Kari E. Hurtta
Subject: Re: lynx-dev gettext() question #2
Date: Thu, 19 Nov 1998 11:36:39 +0200 (EET)

Leonid Pauzner:
> > $LANG includes implicity character set, because it specifies
> > also LC_CTYPE locale.
> not always, but should to.
> > So if user is using some remote terminal, he needs adjust
> > $LANG also to get other programs (than lynx) to work correctly.
> > Some systems there is several possible $LANG values for same
> > language (which just differ by used caharacter-set). For example:
> > en_US
> > en_US.ISO8859-1
> > It may make do sense to map $LANG (or result of setlocale(LC_CTYPE,NULL)
> > actually) to character set with some (OS dependent) mapping table.
> probably Yes.
> It was proposed just before the releasing 2.8,
> concerning case-insensitive search, but nobody
> have contributed the mapping table (we find another workaround).

Problem is that different OSes use different charster-set for
same language.

(For example _if I remember correclty_,
 'fi_FI' on HPes is roman8 and 'fi_FI.88591' was ISO-8859-1.
 Other unixes use ISO-8859-1 for 'fi'.)

Following mapping does sense on IRIX:

POSIX           ISO-8859-1
C               US-ASCII (or ISO-8859-1)
ar              ISO-8859-6
ar_AA           ISO-8859-6
bg              ISO-8859-5
bg_BG           ISO-8859-5
cs              ISO-8859-2
cs_CS           ISO-8859-2
cz              ISO-8859-2
cz_CZ           ISO-8859-2
da              ISO-8859-1
da_DK           ISO-8859-1
de              ISO-8859-1
de_DE           ISO-8859-1
de_AT           ISO-8859-1
de_CH           ISO-8859-1
el              ISO-8859-7
el_GR           ISO-8859-7
en              ISO-8859-1
en_AU           ISO-8859-1
en_CA           ISO-8859-1
en_US           ISO-8859-1
es              ISO-8859-1
es_ES           ISO-8859-1
es_AR           ISO-8859-1
es_MX           ISO-8859-1
fi              ISO-8859-1
fi_FI           ISO-8859-1
fr              ISO-8859-1
fr_BE           ISO-8859-1
fr_CA           ISO-8859-1
fr_CH           ISO-8859-1
fr_FR           ISO-8859-1
hr              ISO-8859-2
hr_HR           ISO-8859-2
hu              ISO-8859-2
hu_HU           ISO-8859-2
is              ISO-8859-1
is_IS           ISO-8859-1
it              ISO-8859-1
it_IT           ISO-8859-1
it_CH           ISO-8859-1
iw              ISO-8859-8
iw_IL           ISO-8859-8
mk              ISO-8859-5
mk_MK           ISO-8859-5
nl              ISO-8859-1
nl_BE           ISO-8859-1
nl_NL           ISO-8859-1
no              ISO-8859-1
no_NO           ISO-8859-1
pl              ISO-8859-2
pl_PL           ISO-8859-2
pt              ISO-8859-1
pt_PT           ISO-8859-1
pt_BR           ISO-8859-1
ro              ISO-8859-2
ro_RO           ISO-8859-2
ru              ISO-8859-5
ru_SU           ISO-8859-5
sh              ISO-8859-2
sh_YU           ISO-8859-2
sk              ISO-8859-2
sk_SK           ISO-8859-2
sl              ISO-8859-2
sl_CS           ISO-8859-2
sp              ISO-8859-5
sp_YU           ISO-8859-5
sv              ISO-8859-1
sv_SE           ISO-8859-1
tr              ISO-8859-9
tr_TR           ISO-8859-9


> I personaly use PC (cp866 for display charset) and the remote terminal
> running lynx have LOCALE at "ru_RU.KOI8-R" what should lynx do here?
> Probably set language support only if locale match display charset?

Hmm. Locale using KOI8-R gots probably gives wrong result for ctype(3)
routines if character set iused is actually cp866.

/ Kari Hurtta

(*) I have omitted following from this list:

ja              eucJP
ja_JP           eucJP
ja_JP.EUC       eucJP
ja_JP.ujis      eucJP
ja_JP.eucJP     eucJP
ko              eucKR
ko_KR           eucKR
ko_KR.euc       eucKR
th_TH           TIS620
zh              eucCN
zh_CN           eucCN
zh_CN.ugb       eucCN
zh_TW           eucTW
zh_TW.ucns      eucTW

reply via email to

[Prev in Thread] Current Thread [Next in Thread]