Define languagues in settings

[curquiza]

Following a discussion with @ManyTheFish

We encounter issues where letting the users set the language(s) of their dataset would solve completely or partially their issues

• Japanese users: Incorrect Results for phrase in Japanese meilisearch#4162 -> currently our Japanese users don't use Meilisearch but the alternative image Many provides in Japanese specialized Meilisearch Docker Image meilisearch#3882
• Nordic characters are sorted out of order meilisearch#4133 -> Meilisearch could behave for sorting based on language detection

[uintaam]

This would also help you if you decided to add Japanese Transliteration, as it's the only language I know where the written reading is essential to search. i.e., If the value in the dictionary is 日本語は楽しい, it would be ideal for someone searching with the hiragana or katakana versions of 日本語 (にほんご, ニホンゴ) or 楽しい (たのしい,タノシイ) to see the results

[curquiza]

More information about a user having different language in the same document: meilisearch/meilisearch#4162 (comment)

[ManyTheFish]

Related to A user Mixing several Languages in the same index.

[uintaam]

Another related issue meilisearch/meilisearch#4561