Locale is per-user - file systems on Unix are multi-user and there is
no meta-data to say which locale a user was in when they wrote the
file.
Thanks for underlining this, I meant to mention it last night but
forgot.
Many locale() C libraries don't give access to what the encoding
_is_ - and when they do there is next to no standardization of the
names used.
Incidentally, if anyone is interested in helping in getting a new locale
standard (one can never have too many :-), the CLDR project can always
use extra eyeballs. CLDR? Common Locale Data Repository:
http://oss.software.ibm.com/cvs/icu/~checkout~/locale/CLDR_status.html
It's the "locale data" part of the ICU project
(http://oss.software.ibm.com/icu/)
ICU will hopefully be used for Parrot / Perl 6.
CLDR has just entered 1.0 Release Candidate phase, but never mind,
checking
(and filling in) the data still is useful. The locale data contains
the usual
things like time and date formats, day and month names, collation order,
language and country names. So if you know your language / national
standards,
go ahead and take a look, does CLDR agree with you.
For example to see the big 'comparison chart' for "French French":
http://oss.software.ibm.com/cvs/icu/~checkout~/locale/diff/main/
index.html
and click on "fr_FR". Try the various 'Demo' buttons.
NOTE! It only makes sense to report bugs / send additions for the
"common" data,
the JDK, AIX, OpenOffice etc. locale data are listed there only for
reference.
In other words: the "common" data is the CLDR data.
--
Jarkko Hietaniemi <jhi(_at_)iki(_dot_)fi> http://www.iki.fi/jhi/ "There is this
special
biologist word we use for 'stable'. It is 'dead'." -- Jack Cohen
