In amazonka-schedule, the client Tokens are not autogenerated while creating /updating new schedules

[kushagarr]

In amazonka-schedule library(there might be others with this problem though), while creating and updating a schedule, we require to pass a random string for client token generation, though the corresponding documentation says otherwise

[endgame]

Can you explain a bit more about what you mean here? According to that documentation, and the most recent AWS API docs, that field is optional and the API will do something sensible if it's not provided.

Can you please provide a dump of the response from EventBridge Schedules if you attempt a request with no token?

Is this an instance of #739 ?

[kushagarr]

@endgame i don't think it is about #739
However, it is about all the fields in botocore service definitions marked as
"idempotencyToken": true
https://github.com/boto/botocore/blob/develop/botocore/data/scheduler/2021-06-30/service-2.json#L373

From what i could gather, both google and AI

Botocore models expose those members via OperationModel.idempotent_members by scanning the input shape for members with the idempotencyToken trait. See botocore/model.py (idempotent_members property).

The docs generator uses that same metadata to append the “This field is autopopulated if not provided.” note you’re seeing. See botocore/docs/params.py (_add_special_trait_documentation/_append_idempotency_documentation).

Botocore has a request lifecycle handler that, just before a call is made, checks OperationModel.idempotent_members and injects a token for any such member that wasn’t provided by the caller.

I don't know if we have a mechanism to handle this or where to handle this.
https://github.com/search?q=repo%3Aboto%2Fbotocore+idempotencyToken&type=code

https://github.com/search?q=repo%3Aboto%2Fbotocore%20generate_idempotent_uuid&type=code