Questions around API design for Client's `generate_tenant_token` method's UID parameter

[moy2010]

Description I would like to open a discussion on the API design decision around the Client's generate_tenant_token method's UID parameter. Since the Discussions feature is not enabled for this repo, I had to open it as an issue instead.

Currently in the SDK v0.24.1 the Client's generate_tenant_token methos is implemented as follows:

pub fn generate_tenant_token(
        &self,
        api_key_uid: String,
       ...
}

By accepting a String, as an API consumer what this signature tells me is that I can pass any string which would uniquely identify the tenant token to be generated.

Unfortunately, this assumption is broken downstream by the following implementation detail in the tenant_tokens module:

// Validate uuid format
pub fn generate_tenant_token(
    api_key_uid: String,
    ...
) -> Result<String, Error> {
    let uid = Uuid::try_parse(&api_key_uid)?;

Here the api_key_uid is parsed into a Uuid and, if the parsing fails, the whole tenant token generation will fail with a Uuid error.

Expected behavior A more coherent API experience. The API should clearly communicate to the consumers what does it expect from them.

Current behavior If a non-Uuid string is passed, an error is returned.

Environment (please complete the following information):

• OS: [e.g. Debian GNU/Linux]
• Meilisearch version: n/a
• meilisearch-rust version: v0.24.1