Many the fish

I've tried to reproduce it on v1.15.0-rc.0 and the bug doesn't appear anymore. Thank you @MattExact, I close this issue

Hello @LukasKalbertodt, I understand your report, and I hesitated to merge phrase matches when I implemented the feature. However, this way of highlighting allows us to retrieve the matches word-by-word,...

Hello! After discussing this with the team, we agree to change the behavior of the match positions when computing a phrase search. However, it's not a priority for the team,...

Hello @flevi29, I may help you for this, don't hesitate to create a draft PR with your changes, this way I can suggest changes more easily. Don't hesitate to ping...

Hello @h-morales, Which request are you doing? Are you using an SDK to make the query? Sorry for the delay. 🙏

Ok, but, what is exactly the request? for example, what is the CURL request? This will help me to reproduce your bug.

Hello @dashi6174, this issue is related to the tokenizer used by Meilisearch internally: Jieba. This tokenizer chooses how the words are segmented based on probabilities, and Meilisearch only supports prefix...

Hey @RichwithPoor, Could you please specify the version you are using and how many documents you are indexing in the same batch?

@Rorke76753, > From my understanding, the "same batch" refers to a single write operation calling meili search, meaning each batch contains around 2 documents. Not really. The batches list can...

Hello @CodeMan62, Thank you for this PR. Can you fix the failing test? In this case, it is quite simple: 1) run `cargo insta test -- --test-threads=4` to run the...