Enable the client to specify a custom csv delimiter for parsing documents

[MixusMinimax]

Pull Request

What does this PR do?

Fixes #2806

This PR adds the ability to specify custom csv delimiters, which I think is a very common use case, especially with large amounts of pre-existing data.

Description

In meilisearch-lib, I added a u8 parameter to the DocumentAdditionFormat::Csv document type.

In meilisearch-http, I added the field pub csv_delimiter: Option<char>, to UpdateDocumentsQuery (the actual query parameter is in camelCase thanks to serde: csvDelimiter)

The default delimiter is the comma (44), just like before. For example, the query for data delimited with semicolons:

POST localhost:7700/indexes/example/documents?csvDelimiter=%3B

cargo test succeeded on my system, and I was able to add documents using different delimiters using postman.

PR checklist

Please check if your PR fulfills the following requirements:

• Does this PR fix an existing issue?
• Have you read the contributing guidelines?
• Have you made sure that the title is accurate and descriptive of the changes?

[curquiza]

Hello @MixusMinimax thanks for your PR, since it impacts the product we need to discuss with the team first, and we will review it as soon as possible 😄

[curquiza]

For traceability, more information about the state of this PR here: #2806 (comment) 😇

[MixusMinimax]

Hi!
Should I update this now to be up to date with main? I think it could work like this: have a struct of options, implement FromRequest for it, and pass it to add_documents. This could include the csv delimiter, and what quotation marks to use (csv fields that include the delimiter themselves are put in quotation marks in the csv standard)

Update: Ignore this, this PR is replaced by #3505

[curquiza]

Hello @MixusMinimax
Sorry for the late answer! Thanks for your previous work, it helped us implement #3505, you will be put as a contributor in the next release changelog of course 😇

[MixusMinimax]

No worries, thank you!