This is the mail archive of the libc-alpha@sources.redhat.com mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Updated/new locales for sr_CS (Serbia and Montenegro)


Regards everyone,

I have made updated locale definitions for the Serbian language, as used in Serbia and Montenegro, following the country name change (in February 2003), and ISO 3166 country code assignment (July 24, 2003).

The new country code is "CS", but I updated much more content than just the country code, fixed a couple of bugs in the previous (sr_YU) locales and improved it a bit.

Because official script of Serbian language, as used in Serbia and Montenegro, is cyrillic, I decided to use cyrillic script in "sr_CS" locale, and latin script in "sr_CS@Latn". This is different from sr_YU which was latin, and sr_YU@cyrillic which was cyrillic in previous GNU libc releases.

Because there was country code change, I don't believe this will cause any compatibility problems.

The LC_COLLATE category in sr_YU was incorrect (used both "copy" keyword and other keywords besides "reorder*" stuff, which is not according to ISO 14652), so I'll delay that a bit (I'm just updating it, but I need to gather some official informations prior to this).

The reasoning behind "Latn", instead of the "latin" or something similar, is that there exists a standard (ISO 15924[1]) which can help in deciding on the code for any script in use.

There was some discussion on relevant mailing lists (concerning l10n and i18n), and it was agreed that this is the best way to go.


The locale definitions are available at http://www.srpski.org/locale/ in files sr_CS and sr_CS@Latn.


Concerning collation, it's now using ISO 14651 rules, which is basicaly equivalent to what sr_YU did (because it was broken). Of course, a "localedef" program would need to be updated too, because the country_num is the same as for the "YU", while the other codes have changed (two letter code is "CS" and three letter code is "SCG") -- currently, it reports errors so I must use "-c" (force) option to build a locale.


I'm hoping this can get integrated before the next GNU libc release.


Cheers, Danilo


[1]http://www.evertype.com/standards/iso15924/



Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]