[Bug localedata/31411] Add Latgalian locale

[Bug localedata/31411] Add Latgalian locale

[maiku.fabian at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=31411

Mike FABIAN <maiku.fabian at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2024-06-11
     Ever confirmed|0                           |1
          Component|locale                      |localedata
           Assignee|unassigned at sourceware dot org   |maiku.fabian at gmail dot com
             Status|UNCONFIRMED                 |ASSIGNED
                 CC|                            |libc-locales at sourceware dot org
                   |                            |, maiku.fabian at gmail dot com
   Target Milestone|---                         |2.40

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug localedata/31411] Add Latgalian locale

[maiku.fabian at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=31411

--- Comment #1 from Mike FABIAN <maiku.fabian at gmail dot com> ---
(In reply to Rudolfs Mazurs from comment #0)
> Created attachment 15377 [details]
> Possible Latgalian locale file demo
> 
> Glibc is missing the locale description for Latgalian (ltg).
> 
> Latgalian locale is in process of being added to the CLDR:
> https://unicode-org.atlassian.net/browse/CLDR-2671
> 
> I have attached a possible implementation for the locale file. It is
> somewhat based on the Latvian locale file. I have not verified that the
> ltg_LV file is valid.
> 
> Some things that would be nice to have, but I have no idea how to describe
> it in code:
> 
> - Collation rules for Latvian (lv_LV) and Latgalian are the same. Is it
> possible to just use the Latvian rules, so that if someone reports a bug for
> the Latvian collation rules, they would automatically apply to the Latgalian?\

Yes.

LC_COLLATE
copy "lv_LV"
END LC_COLLATE

Several others can be copied from lv_LV as well as they are identical:

LC_CTYPE
copy "lv_LV"
END LC_CTYPE

LC_MONETARY
copy "lv_LV"
END LC_MONETARY

LC_NUMERIC
copy "lv_LV"
END LC_NUMERIC

LC_TELEPHONE
copy "lv_LV"
END LC_TELEPHONE

LC_NAME
copy "lv_LV"
END LC_NAME

> - Is there a way to designate Latvian locale as the fallback to the
> Latgalian locale? I mean, if the user chooses to use Latgalian translations,
> but for a given application there is no Latgalian translation, the system
> should try to use the Latvian translation before falling back to English.

I don't know of any way to do that except using the LANGUAGE environment
variable, i.e. you can use these settings:

LANGUAGE=ltg:lv
LANG=ltg_LV.UTF-8

See:
https://www.gnu.org/software/gettext/manual/html_node/The-LANGUAGE-variable.html

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug localedata/31411] Add Latgalian locale

[maiku.fabian at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=31411

--- Comment #2 from Mike FABIAN <maiku.fabian at gmail dot com> ---
[error] LC_ADDRESS: field `lang_ab' must not be defined

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug localedata/31411] Add Latgalian locale

[maiku.fabian at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=31411

--- Comment #3 from Mike FABIAN <maiku.fabian at gmail dot com> ---
Why are the names for months in you  comments different from those in the
variables below?:

% Months, LVS 24:1993 6.3.3.
% janvars, februars, marts, apreļs, majs, juņs, juļs, augusts, septembris,
oktobris, novembris, decembris
% LV:  jan, feb, mar, apr, mai, jūn, jūl, aug, sep, okt, nov, dec
% LTG: jan, feb, mar, apr, maj, juņ, juļ, aug, sep, okt, nov, dec

% Dienas, LVS 24:1993 6.3.1.
% pyrmūdīne, ūtardīne, trešdīne, catūrtdīne, pīktdīne, sastdīne, svātdīne
% LV:  P, O, T, C, Pk, S, Sv
% LTG: P, Ū, T, C, Pk, S, Sv
% LV:  Pr, Ot, Tr, Ce, Pk, Se, Sv
% LTG: Pr, Ūt, Tr, Ca, Pk, Sa, Sv

LC_TIME
abday    "Sv";"P ";/
         "Ū ";"T ";/
         "C ";"Pk";/
         "S "
day      "svātdīná";/
         "pyrmūdīná";/
         "ūtornīks";/
         "trešdīná";/
         "catūrgs";/
         "pīktdīná";/
         "sastdīná"
abmon    "jan";"feb";/
         "mar";"apr";/
         "maj";"jun";/
         "jul";"aug";/
         "sep";"okt";/
         "nov";"dec"
mon      "janvarș";/
         "febralș";/
         "marts";/
         "aprelș";/
         "majș";/
         "junș";/
         "julș";/
         "augusts";/
         "septembris";/
         "oktobris";/
         "novembris";/
         "decembris"

For example the comment says in LTG “juņ” (last letter is ņ U+0146 LATIN SMALL
LETTER N WITH CEDILLA) but abmon contains just “jun”.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug localedata/31411] Add Latgalian locale

[maiku.fabian at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=31411

--- Comment #4 from Mike FABIAN <maiku.fabian at gmail dot com> ---
Is this patch OK?:

https://patchwork.sourceware.org/project/glibc/patch/20240611115705.144363-1-mfabian@redhat.com/

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug localedata/31411] Add Latgalian locale

[rudolfs.mazurs at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=31411

--- Comment #5 from Rudolfs Mazurs <rudolfs.mazurs at gmail dot com> ---
Good catch about the day and month names. Originally I copied the day and month
names from ltg_LV file that was shipped in Austrumi Linux distribution, where
the developer was using archaic word forms. The forms in the comments are the
standardized modern ones, I just forgot to update the variables.

I would rather not encode the unicode code points. But if you rather I did, let
me know.

(days from sunday to saturday)
day      "svātdīne";/
         "pyrmūdīne";/
         "ūtardīne";/
         "trešdīne";/
         "catūrtdīne";/
         "pīktdīne";/
         "sastdīne"

abmon    "jan";"feb";/
         "mar";"apr";/
         "maj";"juņ";/
         "juļ";"aug";/
         "sep";"okt";/
         "nov";"dec"

mon      "janvars";/
         "februars";/
         "marts";/
         "apreļs";/
         "majs";/
         "juņs";/
         "juļs";/
         "augusts";/
         "septembris";/
         "oktobris";/
         "novembris";/
         "decembris"

Other than that, the commit looks good to me.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug localedata/31411] Add Latgalian locale

[maiku.fabian at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=31411

--- Comment #6 from Mike FABIAN <maiku.fabian at gmail dot com> ---
New patch with your day/abmon/mon update:

https://patchwork.sourceware.org/project/glibc/patch/20240611205635.385690-1-mfabian@redhat.com/

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug localedata/31411] Add Latgalian locale

[maiku.fabian at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=31411

--- Comment #7 from Mike FABIAN <maiku.fabian at gmail dot com> ---
(In reply to Rudolfs Mazurs from comment #5)

> I would rather not encode the unicode code points. But if you rather I did,
> let me know.

I am not sure what you mean.

In the past we had to use <Uxxxx> for non-ASCII code points like this:

lang_name    "latgal<U012B><U0161>u"

But this has been improved and now it is OK to write:

lang_name    "latgalīšu"

which looks nicer and is easier to edit.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug localedata/31411] Add Latgalian locale

[maiku.fabian at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=31411

--- Comment #8 from Mike FABIAN <maiku.fabian at gmail dot com> ---
Please also have a look at this mail:

https://sourceware.org/pipermail/libc-alpha/2024-June/157382.html

I think we should change abday a bit in lv_LV **and** the new ltg_LV to agree
better with CLDR.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug localedata/31411] Add Latgalian locale

[rudolfs.mazurs at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=31411

--- Comment #9 from Rudolfs Mazurs <rudolfs.mazurs at gmail dot com> ---
It looks like the current lv_LV and ltg_LV formats are "narrow" and what you
propose is to use the "short" version. The LVS 1993:24 standard lists both
formats as equally valid, so it's up to you.


> So probably I should sync this part of the lv_LV locale to CLDR

> abday    "Sv";"Pr";/
>          "Ot";"Tr";/
>          "Ce";"Pk";/
>          "Se"

> and in the ltg_LV locale use

> abday    "Sv";"Pi";/
>          "Ūt";"Tr";/
>          "Ca";"Pk";/
>          "Sa"

This looks good to me.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug localedata/31411] Add Latgalian locale

[maiku.fabian at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=31411

--- Comment #10 from Mike FABIAN <maiku.fabian at gmail dot com> ---
Ah, but you are right that the current format for abday in lv_LV is the same as
in CLDR "stand-alone" "narrow":

<dayContext type="stand-alone">
...
<dayWidth type="narrow">
        <day type="sun">↑↑↑</day>
        <day type="mon">P</day>
        <day type="tue">O</day>
        <day type="wed">T</day>
        <day type="thu">C</day>
        <day type="fri">P</day>
        <day type="sat">↑↑↑</day>
</dayWidth>

I had overlooked that. So it already is in sync with CLDR. I.e. for lv_LV the
current

abday    "Sv";"P ";/
         "O ";"T ";/
         "C ";"Pk";/
         "S "

and for ltg_LV

abday    "Sv";"P ";/
         "Ū ";"T ";/
         "C ";"Pk";/
         "S "

is also fine. I'll take that then, then I don't have to change lv_LV.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug localedata/31411] Add Latgalian locale

[cvs-commit at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=31411

--- Comment #11 from Sourceware Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Mike Fabian <mfabian@sourceware.org>:

https://sourceware.org/git/gitweb.cgi?p=glibc.git;h=3ea79f50853afcbe17d6a4e2537e1bd5a2d17e7d

commit 3ea79f50853afcbe17d6a4e2537e1bd5a2d17e7d
Author: Mike FABIAN <mfabian@redhat.com>
Date:   Mon Jun 10 19:54:42 2024 +0200

    Define ISO 639-3 "ltg" (Latgalian) and add ltg_LV locale

    Resolves: BZ # 31411

    References:
    https://iso639-3.sil.org/code/ltg
    https://en.wikipedia.org/wiki/Latgalian_language
    https://github.com/unicode-org/cldr/blob/main/common/main/ltg.xml

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug localedata/31411] Add Latgalian locale

[maiku.fabian at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=31411

Mike FABIAN <maiku.fabian at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |FIXED
             Status|ASSIGNED                    |RESOLVED

--- Comment #12 from Mike FABIAN <maiku.fabian at gmail dot com> ---
Fixed in glibc master.

-- 
You are receiving this mail because:
You are on the CC list for the bug.