Reading the documentation I was expecting the Athena cache to work also with the unload_approach=True parameter, but looking at the code I see that it is not intended to work by design.
Specifically here I see that the csv file format is imposed to all DML queries https://github.com/aws/aws-sdk-pandas/blob/3.9.1/awswrangler/athena/_cache.py#L222

I was wondering if there's a specific reason for this, or is it something that is planned to be introduced in the future?
Thanks!