Add client.datasets.create_dataset() to create a new dataset with uploaded file.

sdk/ai/azure-ai-projects-dp1/azure/ai/projects/dp1/operations/_patch.py

@@ -6,9 +6,101 @@
 
 Follow our quickstart for examples: https://aka.ms/azsdk/python/dpcodegen/python/customize
 """
-from typing import List
+from typing import List, Optional
+from pathlib import Path
+from azure.storage.blob import ContainerClient, BlobClient, BlobType
 
-__all__: List[str] = []  # Add all objects you want publicly available to users at this package level
+from ._operations import DatasetsOperations as DatasetsOperationsGenerated
+from ..models._models import (
+    DatasetVersion,
+    PendingUploadRequest,
+    PendingUploadType,
+    PendingUploadResponse,
+    CredentialType,
+)
+from ..models._enums import (
+    DatasetType,
+)
+
+class DatasetsOperations(DatasetsOperationsGenerated):
+
+    def create_and_upload(
+        self,
+        *,
+        name: str,
+        version: Optional[str] = None,
+        file: str,
+        connection_name: Optional[str] = None, # TODO: Use me.
+    ) -> DatasetVersion:
+        """Upload file to an internal blob storage, and create a dataset that references this file.
+
+        :param name: The name of the dataset. Required.
+        :type name: str
+        :param version: The version identifier for the dataset. Optional.
+        :type version: str or None
+        :param file: The file name (including optional path) to be uploaded. Required.
+        :type file: str
+        :param connection_name: The name of the AI Project Connection to be used. Optional.
+        :type connection_name: str or None
+        :return: The created dataset version.
+        :rtype: ~azure.ai.projects.dp1.models.DatasetVersion
+        :raises ~azure.core.exceptions.HttpResponseError: If an error occurs during the HTTP request.
+        """
+
+        pending_upload_response: PendingUploadResponse = self.create_or_get_start_pending_upload(
+            name=name,
+            version=version,
+            pending_upload_request=PendingUploadRequest(
+                pending_upload_type=PendingUploadType.TEMPORARY_BLOB_REFERENCE
+            ),
+        )
+
+        if (True): # Debugging. Remove me.
+            print(pending_upload_response.pending_upload_id)
+            print(pending_upload_response.pending_upload_type) # == PendingUploadType.TEMPORARY_BLOB_REFERENCE
+            print(pending_upload_response.blob_reference_for_consumption.blob_uri) # Hosted on behalf of (HOBO) not visible to the user. If the form of: "https://<account>.blob.core.windows.net/<container>?<sasToken>"
+            print(pending_upload_response.blob_reference_for_consumption.storage_account_arm_id) # /subscriptions/<>/resourceGroups/<>/Microsoft.Storage/accounts/<>
+            print(pending_upload_response.blob_reference_for_consumption.credential.sas_token)
+            print(pending_upload_response.blob_reference_for_consumption.credential.type) # == CredentialType.SAS
+
+        if pending_upload_response.blob_reference_for_consumption.credential.type != CredentialType.SAS:
+            raise ValueError("Credential type is not SAS")
+
+        # For overview on Blob storage SDK in Python see: 
+        # https://learn.microsoft.com/azure/storage/blobs/storage-quickstart-blobs-python
+        # https://learn.microsoft.com/azure/storage/blobs/storage-blob-upload-python
+
+        # See https://learn.microsoft.com/python/api/azure-storage-blob/azure.storage.blob.containerclient?view=azure-python#azure-storage-blob-containerclient-from-container-url
+        container_client: ContainerClient = ContainerClient.from_container_url(
+            container_url=pending_upload_response.blob_reference_for_consumption.blob_uri, # Of the form: "https://<account>.blob.core.windows.net/<container>?<sasToken>"
+        )
+
+        with open(file=file, mode="rb") as data:
+
+            # See https://learn.microsoft.com/en-us/python/api/azure-storage-blob/azure.storage.blob.containerclient?view=azure-python#azure-storage-blob-containerclient-upload-blob
+            blob_client: BlobClient = container_client.upload_blob(
+                name=Path(file).name, # Extract the file name from the path.
+                data=data,
+                blob_type=BlobType.BLOCKBLOB,
+            )
+
+        dataset_version = self.create_or_update(
+            name=name,
+            version=version, 
+            dataset_version=DatasetVersion(
+                # See https://learn.microsoft.com/python/api/azure-storage-blob/azure.storage.blob.blobclient?view=azure-python#azure-storage-blob-blobclient-url
+                # Per above doc the ".url" contains SAS token... should this be stripped away?
+                dataset_uri=blob_client.url, # "<account>.blob.windows.core.net/<container>/<file_name>"
+                dataset_type=DatasetType.URI_FILE
+            ),
+        )
+
+        return dataset_version
+
+
+__all__: List[str] = [
+    "DatasetsOperations"
+]  # Add all objects you want publicly available to users at this package level
 
 
 def patch_sdk():

sdk/ai/azure-ai-projects-dp1/setup.py

@@ -67,6 +67,7 @@
         "isodate>=0.6.1",
         "azure-core>=1.30.0",
         "typing-extensions>=4.6.0",
+        "azure-storage-blob>=12.15.0",
     ],
     python_requires=">=3.8",
 )