Hi, I’m new to Typesense. Since v0.24.1, it seems that Typesense supports Japanese tokenization. However, when I try to search using half-width and full-width characters, it doesn’t seem to be working at all. any idea on this? example: 1. create an index with kanji:

個人情報

2. try to search with kata full-size:

コジンジョウホウ

and kata half-size:

ｺｼﾞﾝｼﾞｮｳﾎｳ

👋 Can you please confirm that you have set the

locale

property of the field in the collection schema to

ja

?

yes i did, but let me correct my question: create index with kanji :

個人情報

test case: • Hiragara：`こじんじょうほう` - OK • Kata fullsize：`コジンジョウホウ` - OK • Kata halfsize:

ｺｼﾞﾝｼﾞｮｳﾎｳ

- NOK

Are those half-width hiragana?

no, only half-width katakana

Ok, let me check into what's happening here and get back to you.