[ISSUE] catalog_alias Field in CleanRoomCollaborator Not Mapped Correctly

[markdegroat]

Description
The databricks-sdk does not correctly map the collaborator_alias field in the CleanRoomCollaborator class to the required catalog_alias field for the Databricks Clean Room API. This causes the following error when attempting to create a clean room:

Missing required fields: clean_room.remote_detailed_info.collaborators.catalog_alias, clean_room.remote_detailed_info.collaborators.catalog_alias

This issue prevents the SDK from being used to create clean rooms without manual intervention or subclassing.

Reproduction

1. Attempt to create a clean room using the SDK:

from databricks.sdk import WorkspaceClient
from databricks.sdk.service.cleanrooms import CleanRoom, CleanRoomRemoteDetail, CleanRoomCollaborator

client = WorkspaceClient()

creator = CleanRoomCollaborator(
    collaborator_alias="creator_catalog",
    global_metastore_id="aws:us-west-2:creator-metastore-id"
)

collaborators = [
    creator,
    CleanRoomCollaborator(
        collaborator_alias="collaborator_catalog",
        global_metastore_id="aws:us-east-2:collaborator-metastore-id"
    )
]

remote_detail = CleanRoomRemoteDetail(
    cloud_vendor="aws",
    region="us-west-2",
    collaborators=collaborators
)

clean_room = CleanRoom(
    owner="[email protected]",
    name="test_clean_room",
    remote_detailed_info=remote_detail
)

response = client.clean_rooms.create(clean_room=clean_room)

2. Observe the error:

Missing required fields: clean_room.remote_detailed_info.collaborators.catalog_alias

Expected behavior
The CleanRoomCollaborator class should map collaborator_alias to the API-required catalog_alias field during serialization.

Is it a regression?
Did this work in a previous version of the SDK? If so, which versions did you try?
No, this is a new feature.

Debug Logs
The SDK logs helpful debugging information when debug logging is enabled. Set the log level to debug by adding logging.basicConfig(level=logging.DEBUG) to your program, and include the logs here.
Not necessary for this I don't believe but can add more information if needed.

Other Information

• SDK Version: 0.40.0

Additional context

Workaround:
Subclass the CleanRoomCollaborator class to correctly map the collaborator_alias field to catalog_alias. The workaround implementation looks like this:

class PatchedCleanRoomCollaborator(CleanRoomCollaborator):
    def as_dict(self):
        """Override to map 'collaborator_alias' to 'catalog_alias'."""
        body = super().as_dict()
        if 'collaborator_alias' in body:
            body['catalog_alias'] = body['collaborator_alias']
        return body

# Updated usage
creator = PatchedCleanRoomCollaborator(
    global_metastore_id="aws:us-west-2:creator-metastore-id",
    collaborator_alias="creator_catalog"
)

collaborators = [
    creator,
    PatchedCleanRoomCollaborator(
        global_metastore_id="aws:us-east-2:collaborator-metastore-id",
        collaborator_alias="collaborator_catalog"
    )
]

remote_detail = CleanRoomRemoteDetail(
    cloud_vendor="aws",
    region="us-west-2",
    collaborators=collaborators
)

clean_room = CleanRoom(
    owner="[email protected]",
    name="test_clean_room",
    remote_detailed_info=remote_detail
)

response = client.clean_rooms.create(clean_room=clean_room)

I'm currently using this workaround and confirmed this successfully creates a clean room utilizing the SDK.

Proposed Fix:
Update the SDK to automatically map the collaborator_alias field to catalog_alias during serialization in the CleanRoomCollaborator class.
Ensure SDK tests verify compatibility with the latest API requirements for clean room collaborators.