Toki Pona was added as a recognised 3-letter-code in 70f021e66ae73a99bff9bcc5b71eefb63fc74320, but still needs localedata. A patch was posted to https://patchwork.sourceware.org/project/glibc/patch/20240107023848.372587-2-uhhadd@gmail.com/.
This new locale contains: +LC_MESSAGES +yesexpr "^[+1wW]" +noexpr "^[-0aA]" +yesstr "wile" +nostr "ala" +END LC_MESSAGES Most locales have yY and nN added to teh yesexpr and noexpr, except when there is a conflict. For example the sw_KE locale does not have nN in noexpr because there seems to be a conflict: $ grep expr sw_KE yesexpr "^[+1nNyY]" noexpr "^[-0hHlL]" What do you think in the case of Toki Pona, should yY and nN be in the yesexpr and noexpr or not?
yY/nN in the yes/noexprs is fine.
The master branch has been updated by Mike Fabian <mfabian@sourceware.org>: https://sourceware.org/git/gitweb.cgi?p=glibc.git;h=d3a2aecc1ce78d512dd5fb5e30bd39d7c5947eb8 commit d3a2aecc1ce78d512dd5fb5e30bd39d7c5947eb8 Author: Janet Blackquill <uhhadd@gmail.com> Date: Sat Jan 6 21:38:49 2024 -0500 localedata: add data for tok (Toki Pona) Resolves: BZ # 31221 glibc can recognise its code, but does not have its data. This patch remedies that. Signed-off-by: Janet Blackquill <uhhadd@gmail.com>
Fixed in glibc master.