i18next plurals Weblate sync inconsistencies

[danielweck]

• ru: Weblate removed _other JSON key/value pairs ( https://github.com/edrlab/thorium-reader/pull/3361/changes#diff-85983fce3f5958e2102aa82b8a5101dd326b04835bc8abda71b9e60e326fae30 )
• lt: Weblate removed _many ( https://github.com/edrlab/thorium-reader/pull/3361/changes#diff-ab35202fb7c9716bef1477c727bde2e804df1a099280bedbf70d4918ee3be75c )
• et: Weblate added _one and _other which were missing in our canonical JSON ( https://github.com/edrlab/thorium-reader/pull/3361/changes#diff-f7adebe4d32e65124f071d88bb7baeb078010007d029ae7d469b943daaf16e57 )

For et, the bug is in our "secondary" locales list:

thorium-reader/package.json

Lines 76 to 81 in fc7a466

	"i18n-scan_": "node ./scripts/translate-scan.js \"src/resources/locales/temp.json\" && sync-i18n --files 'src/resources/locales/*.json' --primary temp --languages en ta tr ar cs sl bg ca da de el es eu fi fr gl hr it ja ka ko lt nl pt-br pt-pt ru sv zh-cn zh-tw --space \" \" --finalnewline --newkeysempty && rimraf \"src/resources/locales/temp.json\"",
	"i18n-check_": "sync-i18n --files 'src/resources/locales/*.json' --primary en --languages ta tr ar cs sl bg ca da de el es eu fi fr gl hr it ja ka ko lt nl pt-br pt-pt ru sv zh-cn zh-tw --space \" \" --finalnewline --newkeysempty",
	"i18n-scan": "node ./scripts/translate-scan.js \"src/resources/locales/und.json\" && i18next-locales-sync --compatibilityJSON v4 --localesFolder ./src/resources/locales/ --primaryLanguage und --secondaryLanguages en ta tr ar cs sl bg ca da de el es eu fi fr gl hr it ja ka ko lt nl pt-br pt-pt ru sv zh-cn zh-tw --spaces 4 --useEmptyString true && rimraf \"src/resources/locales/und.json\"",
	"posti18n-scan": "npm run i18n-sort",
	"i18n-check": "i18next-locales-sync --compatibilityJSON v4 --localesFolder ./src/resources/locales/ --primaryLanguage en --secondaryLanguages ta tr ar cs sl bg ca da de el es eu fi fr gl hr it ja ka ko lt nl pt-br pt-pt ru sv zh-cn zh-tw --spaces 4 --useEmptyString true",
	"posti18n-check": "npm run i18n-sort",

For ru and lt, I will run Thorium Desktop's own tooling (canonical JSON formatting, key ordering, and automatic key injection + plural expansion), after merging the PR #3361 (I will lock Weblate prior to pushing changes, in order to avoid Git conflicts)

[danielweck]

To quickly check supported plurals:

https://jsfiddle.net/6bpxsgd4

(for example for fr try 1000000 which will trigger _many)

[danielweck]

Suggestion: alphabetical sort of locale codes

https://github.com/edrlab/thorium-reader/tree/v3.3.0/src/resources/locales

Excluding en (primary), the secondary locales are currently:

ta tr ar cs sl bg ca da de el es eu fi fr gl hr it ja ka ko lt nl pt-br pt-pt ru sv zh-cn zh-tw

...even if et is currently not integrated/activated at runtime, we still need to run the i18next tooling on all available locale JSON files in the folder

https://github.com/edrlab/thorium-reader/blob/v3.3.0/src/common/services/translator.ts#L9-L37

Concrete case hi and sa: #3176

[danielweck]

Fixed via:

• c3bd42b
• 6ea1d73