Locale tags "sr_SR@latin" and "sr_SR@Latn". Not sure if we correctly handle at-sign tags. Need to.
Created attachment 18390 [details]
Serbian Latin orth file
Created attachment 18391 [details]
This is also a fixed-up sr.orth file (to include only the really relevant Serbian letters).
Behdad asked me to provide firstname.lastname@example.org, so here it is. Depending on what this is used for: if it's used to detect if a font supports a language, this should be enough. Otherwise, there are Unicode digraphs 01c4-01cc which are commonly typed as two separate characters by Serbian people.
Also, note that email@example.com is likely exactly the same as hr.orth. I have not included accented letters (Serbian allows for two tonal accents, but they are rarely used except with a few words which are ambiguous).
A fixed-up sr.orth is included as well.
I've committed the new sr.orth to my tree.
Danilo, if I understand correctly both the Cyrillic and the Latin writing system are used in Serbia, right? It's not like sr_SR only uses Latin? Neither of fontconfig and Pango currently support @ tags. Should be easier if we could get away without it.
What needs to be done here is to be able to derive "sr@latin" from "sr_SR@latin". *If* Uli had accepted "sr_SR@Latn", we could at least recognize the script names (it's still possible to do with latin, assuming Unicode script names instead of ISO15924 ones) and match the best orthography available.
Anyway, this bug kinda relates to adding support for scripts to fontconfig.
According to RFC 1766 the language tag for sr@latin would be sr-latin. That will immediate work if we just fold @ to '-'. Will do that for pango too. Or does fontconfig already do that?
Please note that the correct base locale tag for Serbian is sr_RS (and variants), not sr_SR (see bug #11456).
Also, according to RFC 4646 (1766 was long obsoleted by 3066, which in turn got obsoleted by 4646), the tag for the Latin script variant of the locale should be sr-Latn, not sr-latin. However, this doesn't matter, as glibc doesn't follow RFC 4646 anyway and (wrongly IMHO) uses @latin modifiers, hence the knock on effect on everything else.
BTW fontconfig doesn't support modifiers yet. need to fix it first.
-- GitLab Migration Automatic Message --
This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.
You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/fontconfig/fontconfig/issues/31.