[Bug localedata/33952] brh_PK locale for Brahui language: New Locale

hammadalo99 at gmail dot com via Glibc-bugs <[email protected]> Mon, 20 Jul 2026 08:04:48 +0000
Newsgroups gmane.comp.lib.glibc.bugs
Message-ID <[email protected]/bugzilla/>
https://sourceware.org/bugzilla/show_bug.cgi?id=3D33952

Hammad <hammadalo99 at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
  Attachment #16718|0                           |1
        is obsolete|                            |

--- Comment #4 from Hammad <hammadalo99 at gmail dot com> ---
Created attachment 16851
  --> https://sourceware.org/bugzilla/attachment.cgi?id=3D16851&action=3Ded=
it
Patch v4: rebased on master; first_weekday, date-time format, and LC_MONETA=
RY
fixes

v4 sent to libc-alpha and threaded under the v3 message:
https://inbox.sourceware.org/libc-alpha/CALoFT+5VNbvTXCnPr41UGNoUy7bW=3DfFa=
[email protected]/T/

Rebased on current master.  Applies cleanly.

Changes since v3:

- iso-639.def: restored alphabetical order (Brahui before Braj).
- LC_TIME: first_weekday corrected from 7 to 1.  The old value
  selected Saturday; CLDR territory data for PK gives firstDay =3D sun,
  and ur_PK and sd_PK both use 1.
- LC_TIME: d_t_fmt and date_fmt reordered so the date precedes the
  time, matching the CLDR brh combination pattern and sd_PK, with
  U+060C ARABIC COMMA as the separator.
- LC_TIME: abday now repeats the full day names; Brahui has no
  distinct abbreviated forms and CLDR brh gives identical values at
  the abbreviated and wide widths, as ur_PK, pa_PK and fa_IR do.
- LC_TIME: am_pm aligned with the CLDR brh day-period values.
- LC_MONETARY: n_sep_by_space now matches p_sep_by_space, so positive
  and negative amounts are spaced alike.
- LC_MESSAGES: yesexpr accepts U+062C and U+0647, noexpr accepts
  U+0627, covering attested spellings of the affirmative and negative
  words.
- LC_CTYPE: transliterations added for U+06C1 and U+06B7.
- LC_TELEPHONE: tel_int_fmt uses the "+%c %a %l" form.

Tested with localedef on glibc 2.39.  The only diagnostics are the two
ISO 639 lookups that this patch's iso-639.def hunk resolves; confirmed
by substituting an already-registered 639-3 code (brx) into the same
file, which compiles with no diagnostics at all.  LC_TIME, LC_COLLATE,
LC_MESSAGES, LC_MONETARY, LC_CTYPE and LC_ADDRESS output verified at
runtime through strftime, strcoll, rpmatch, strfmon, iconv and
nl_langinfo.

The retroflex lateral U+06B7 that LC_COLLATE addresses is attested in
a 100,744-sentence Brahui corpus: 362 example sentences and 16
dictionary entries (=D8=AA=D9=88=DA=B7, =DB=81=DA=B7, =D8=AE=DA=B7, =D9=85=
=DB=8C=DA=B7, =D9=85=D8=A7=DA=B7, =D9=BE=D8=A7=DA=B7, =D9=85=D9=88=DA=B7 an=
d others).

--=20
You are receiving this mail because:
You are on the CC list for the bug.=