Upload data to dataset using readable object

[aaraney]

Adds an async upload_data_to_dataset method to dataset clients that polls an async readable object that implements async def read(self, size: Optional[int] = None) -> bytes. This adds flexibility, so it is not required to treat to-be uploaded items as files.

Additions

• Adds an async upload_data_to_dataset method to dataset clients that polls an async readable object that implements async def read(self, size: Optional[int] = None) -> bytes.

Notes

• Not sure the correct internal equivalent type to MaaSDatasetManagementMessage

Todos

• unit tests
• implement upload_data_to_dataset for DatasetInternalClient

[robertbartel]

I had a minor question and then a requested change. We should be able to add an implementation for DatasetInternalClient without too much trouble, so let's include it here.

[robertbartel]

Any particular reason for Union[int, None] instead of Optional[int]?

[aaraney]

Nothing in particular, I just took inspiration from python's BufferedReader interface:

def read(self, __size: int | None = ...) -> bytes: ...

[robertbartel]

The correct internal equivalent to MaaSDatasetManagementMessage is just DatasetManagementMessage. MaaSDatasetManagementMessage is separate only (as far as I remember) to give us a separate subtype that also inherits from ExternalRequest, to get things related to authentication, etc. I don't exactly know why MaaSDatasetManagementMessage has those two additional attributes.

We really should include an implementation for this at the same time. You should be able to copy the other for the most part, making a few adjustments and removals for things related to external auth.

[aaraney]

Thanks for the guidance. Ill make that addition and let you know if I run into any roadblocks.

[aaraney]

Closing this in favor of #540.