Skip to content

Update azure blob SDK version #1515

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 8 commits into from
Feb 17, 2021
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
179 changes: 128 additions & 51 deletions libraries/botbuilder-azure/botbuilder/azure/blob_storage.py
Original file line number Diff line number Diff line change
Expand Up @@ -3,13 +3,34 @@

from jsonpickle import encode
from jsonpickle.unpickler import Unpickler
from azure.storage.blob import BlockBlobService, Blob, PublicAccess
from azure.core import MatchConditions
from azure.core.exceptions import (
HttpResponseError,
ResourceExistsError,
ResourceNotFoundError,
)
from azure.storage.blob.aio import (
BlobServiceClient,
BlobClient,
StorageStreamDownloader,
)
from botbuilder.core import Storage

# TODO: sanitize_blob_name


class BlobStorageSettings:
"""The class for Azure Blob configuration for the Azure Bot Framework.

:param container_name: Name of the Blob container.
:type container_name: str
:param account_name: Name of the Blob Storage account. Required if not using connection_string.
:type account_name: str
:param account_key: Key of the Blob Storage account. Required if not using connection_string.
:type account_key: str
:param connection_string: Connection string of the Blob Storage account.
Required if not using account_name and account_key.
:type connection_string: str
"""

def __init__(
self,
container_name: str,
Expand All @@ -23,56 +44,105 @@ def __init__(
self.connection_string = connection_string


# New Azure Blob SDK only allows connection strings, but our SDK allows key+name.
# This is here for backwards compatibility.
def convert_account_name_and_key_to_connection_string(settings: BlobStorageSettings):
if not settings.account_name or not settings.account_key:
raise Exception(
"account_name and account_key are both required for BlobStorageSettings if not using a connections string."
)
return (
f"DefaultEndpointsProtocol=https;AccountName={settings.account_name};"
f"AccountKey={settings.account_key};EndpointSuffix=core.windows.net"
)


class BlobStorage(Storage):
"""An Azure Blob based storage provider for a bot.

This class uses a single Azure Storage Blob Container.
Each entity or StoreItem is serialized into a JSON string and stored in an individual text blob.
Each blob is named after the store item key, which is encoded so that it conforms a valid blob name.
If an entity is an StoreItem, the storage object will set the entity's e_tag
property value to the blob's e_tag upon read. Afterward, an match_condition with the ETag value
will be generated during Write. New entities start with a null e_tag.

:param settings: Settings used to instantiate the Blob service.
:type settings: :class:`botbuilder.azure.BlobStorageSettings`
"""

def __init__(self, settings: BlobStorageSettings):
if not settings.container_name:
raise Exception("Container name is required.")

if settings.connection_string:
client = BlockBlobService(connection_string=settings.connection_string)
elif settings.account_name and settings.account_key:
client = BlockBlobService(
account_name=settings.account_name, account_key=settings.account_key
blob_service_client = BlobServiceClient.from_connection_string(
settings.connection_string
)
else:
raise Exception(
"Connection string should be provided if there are no account name and key"
blob_service_client = BlobServiceClient.from_connection_string(
convert_account_name_and_key_to_connection_string(settings)
)

self.client = client
self.settings = settings
self.__container_client = blob_service_client.get_container_client(
settings.container_name
)

self.__initialized = False

async def _initialize(self):
if self.__initialized is False:
# This should only happen once - assuming this is a singleton.
# ContainerClient.exists() method is available in an unreleased version of the SDK. Until then, we use:
try:
await self.__container_client.create_container()
except ResourceExistsError:
pass
self.__initialized = True
return self.__initialized

async def read(self, keys: List[str]) -> Dict[str, object]:
"""Retrieve entities from the configured blob container.

:param keys: An array of entity keys.
:type keys: Dict[str, object]
:return dict:
"""
if not keys:
raise Exception("Keys are required when reading")

self.client.create_container(self.settings.container_name)
self.client.set_container_acl(
self.settings.container_name, public_access=PublicAccess.Container
)
await self._initialize()

items = {}

for key in keys:
if self.client.exists(
container_name=self.settings.container_name, blob_name=key
):
items[key] = self._blob_to_store_item(
self.client.get_blob_to_text(
container_name=self.settings.container_name, blob_name=key
)
)
blob_client = self.__container_client.get_blob_client(key)

try:
items[key] = await self._inner_read_blob(blob_client)
except HttpResponseError as err:
if err.status_code == 404:
continue

return items

async def write(self, changes: Dict[str, object]):
"""Stores a new entity in the configured blob container.

:param changes: The changes to write to storage.
:type changes: Dict[str, object]
:return:
"""
if changes is None:
raise Exception("Changes are required when writing")
if not changes:
return

self.client.create_container(self.settings.container_name)
self.client.set_container_acl(
self.settings.container_name, public_access=PublicAccess.Container
Copy link
Contributor Author

@mdrichardson mdrichardson Feb 11, 2021

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I didn't see anything regarding public access referenced in .NET or JS and tests all passed fine without it. I can add it back, if necessary.

)
await self._initialize()

for (name, item) in changes.items():
blob_reference = self.__container_client.get_blob_client(name)

e_tag = None
if isinstance(item, dict):
e_tag = item.get("e_tag", None)
Expand All @@ -81,39 +151,46 @@ async def write(self, changes: Dict[str, object]):
e_tag = None if e_tag == "*" else e_tag
if e_tag == "":
raise Exception("blob_storage.write(): etag missing")

item_str = self._store_item_to_str(item)
try:
self.client.create_blob_from_text(
container_name=self.settings.container_name,
blob_name=name,
text=item_str,
if_match=e_tag,

if e_tag:
await blob_reference.upload_blob(
item_str, match_condition=MatchConditions.IfNotModified, etag=e_tag
)
except Exception as error:
raise error
else:
await blob_reference.upload_blob(item_str, overwrite=True)
Comment on lines +157 to +162
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Match conditions and eTags changed quite a bit and I'm not 100% confident that I implemented this correctly. I tried some other approaches and this was the one that made all of our tests pass. Please review this part closely.


async def delete(self, keys: List[str]):
"""Deletes entity blobs from the configured container.

:param keys: An array of entity keys.
:type keys: Dict[str, object]
"""
if keys is None:
raise Exception("BlobStorage.delete: keys parameter can't be null")

self.client.create_container(self.settings.container_name)
self.client.set_container_acl(
self.settings.container_name, public_access=PublicAccess.Container
)
await self._initialize()

for key in keys:
if self.client.exists(
container_name=self.settings.container_name, blob_name=key
):
self.client.delete_blob(
container_name=self.settings.container_name, blob_name=key
)

def _blob_to_store_item(self, blob: Blob) -> object:
item = json.loads(blob.content)
item["e_tag"] = blob.properties.etag
result = Unpickler().restore(item)
return result
blob_client = self.__container_client.get_blob_client(key)
try:
await blob_client.delete_blob()
# We can't delete what's already gone.
except ResourceNotFoundError:
pass

def _store_item_to_str(self, item: object) -> str:
return encode(item)

async def _inner_read_blob(self, blob_client: BlobClient):
blob = await blob_client.download_blob()

return await self._blob_to_store_item(blob)

@staticmethod
async def _blob_to_store_item(blob: StorageStreamDownloader) -> object:
item = json.loads(await blob.content_as_text())
item["e_tag"] = blob.properties.etag.replace('"', "")
result = Unpickler().restore(item)
return result
2 changes: 1 addition & 1 deletion libraries/botbuilder-azure/setup.py
Original file line number Diff line number Diff line change
Expand Up @@ -6,7 +6,7 @@

REQUIRES = [
"azure-cosmos==3.2.0",
"azure-storage-blob==2.1.0",
"azure-storage-blob==12.7.0",
"botbuilder-schema==4.12.0",
"botframework-connector==4.12.0",
"jsonpickle==1.2",
Expand Down
14 changes: 8 additions & 6 deletions libraries/botbuilder-azure/tests/test_blob_storage.py
Original file line number Diff line number Diff line change
Expand Up @@ -2,6 +2,8 @@
# Licensed under the MIT License.

import pytest
from azure.core.exceptions import ResourceNotFoundError
from azure.storage.blob.aio import BlobServiceClient
from botbuilder.core import StoreItem
from botbuilder.azure import BlobStorage, BlobStorageSettings
from botbuilder.testing import StorageBaseTests
Expand All @@ -26,12 +28,12 @@ def get_storage():


async def reset():
storage = get_storage()
storage = BlobServiceClient.from_connection_string(
BLOB_STORAGE_SETTINGS.connection_string
)
try:
await storage.client.delete_container(
container_name=BLOB_STORAGE_SETTINGS.container_name
)
except Exception:
await storage.delete_container(BLOB_STORAGE_SETTINGS.container_name)
except ResourceNotFoundError:
pass


Expand All @@ -44,7 +46,7 @@ def __init__(self, counter=1, e_tag="*"):

class TestBlobStorageConstructor:
@pytest.mark.asyncio
async def test_blob_storage_init_should_error_without_cosmos_db_config(self):
async def test_blob_storage_init_should_error_without_blob_config(self):
try:
BlobStorage(BlobStorageSettings()) # pylint: disable=no-value-for-parameter
except Exception as error:
Expand Down