Fix v3 appservice ci

common/smoketest/README.md

@@ -74,7 +74,7 @@ pip install -r requiriments.txt
 pip install -r requiriments_async.txt
 ```
 
-If a python version below 3.5 is being used, it is still possible to run the samples. When it gets to the async tests a message `'Async not suported'` will be displayed.
+If a python version below 3.5 is being used, it is still possible to run the samples. When it gets to the async tests a message `'Async not supported'` will be displayed.
 
 ## Key concepts
 

doc/dev/mgmt/generating-integration-test.md

@@ -127,7 +127,7 @@ now you can run live integration test:
 
 ## Fixing Test
 
-It's obvious that when running test for the first time someting is not going to work.
+It's obvious that when running test for the first time something is not going to work.
 
 The best approach is to:
 - fix the test manually

doc/dev/mgmt/generation.md

@@ -8,7 +8,7 @@ IMPORTANT NOTE: All the commands prefixed by `python` in this page assumes you h
 
 ### Autorest versioning
 
-A few notes on [Autorest for Python versionning](https://github.com/Azure/autorest.python/blob/master/ChangeLog.md):
+A few notes on [Autorest for Python versioning](https://github.com/Azure/autorest.python/blob/master/ChangeLog.md):
 - Autorest for Python v2.x is deprecated, and should not be used anymore for any generation under any circumstances.
 - Autorest for Python v3.x is the most currently used one. Should not be used, but still ok if service team are still in v3.x and they want to avoid breaking changes for a given version (rare).
 - Autorest for Python v4.x is the current recommendation. This generator can generates async code, but this should be disabled with --no-async. No package should be shipped with async based on v4

doc/dev/mgmt/multiapi.md

@@ -14,7 +14,7 @@ Because there is different flavors of Azure that are not necessarly provided wit
 
 ### Why a multi-api package?
 
-Indeed, a simple solution would be to write down explictly what version of SDK supports what API version. Example: 1.0 supports 2015-06-01, 2.0 supports 2017-07-01, etc. The story for customers then would be to pin the specific SDK version for the specific API version they need. However, this was considered unacceptable in an end-to-end scenario:
+Indeed, a simple solution would be to write down explicitly what version of SDK supports what API version. Example: 1.0 supports 2015-06-01, 2.0 supports 2017-07-01, etc. The story for customers then would be to pin the specific SDK version for the specific API version they need. However, this was considered unacceptable in an end-to-end scenario:
 - It means you cannot install in the same Python environment packages that would target different cloud (Python doesn't allow installation of different versions of the same package together). Azure CLI or Ansible supports for different clouds would then be extremely complicated.
 - This forces customers to use old SDK, that might have been fixed on different axis than API version (security fixes, new SDK features like async, etc.)
 - Customers rarely needs only one package, but a set of them (storage, compute, network, etc.) and having to keep track of the correct list of packages is challenging.
@@ -44,7 +44,7 @@ Network interfaces operations are defines in a [network interface file](https://
 
 **Python multi-api packaging is based on the assumptions that it's true.** If it's not, it's usually ok but requires a little more subtle packaging (see final section here)
 
-Being that a given Swagger defines only *one* fixed API version, doing multi-api version in one package implies shipping several Swagger files into one package. This is achived by the `batch` directive of Autorest. More details on how to write Readme for Swagger in the specific page for it [swagger_conf.md](./swagger_conf.md).
+Being that a given Swagger defines only *one* fixed API version, doing multi-api version in one package implies shipping several Swagger files into one package. This is archived by the `batch` directive of Autorest. More details on how to write Readme for Swagger in the specific page for it [swagger_conf.md](./swagger_conf.md).
 
 Python SDK team is responsible to design the correct set of tags to set for the `batch` node. Each line of the batch directive should contains only *one* api version to match the folder name used. this might require adding new tags in the readme.md that are specific to only one API version. These tags are usually suffixed by "-only" ([example with compute](https://github.com/Azure/azure-rest-api-specs/tree/master/specification/compute/resource-manager#tag-package-2019-03-01-only))
 

doc/dev/mgmt/swagger_conf.md

@@ -14,7 +14,7 @@ In practical terms, we want to control the version of Autorest used, the output
 
 ## Writing the readme
 
-Writing the readme is the responsability of the Python SDK team. There is currently two types of templates for Python readmes:
+Writing the readme is the responsibility of the Python SDK team. There is currently two types of templates for Python readmes:
 - Readme that handles only one API version, and generates packages that handle one API version only
 - Readme that handles several API versions, and generates packages with multiples API and profile supports
 

eng/pipelines/templates/jobs/update_pr.yml

@@ -26,4 +26,4 @@ steps:
 - script: python3  -m packaging_tools.update_pr -v --pr-number $(System.PullRequest.PullRequestNumber) --repo $(Build.Repository.Name)
   displayName: 'Update packaging of PR'
   env:
-    GH_TOKEN: $(python-mgmt-update-pr-token)
+    GH_TOKEN: $(azuresdk-github-pat)

sdk/appservice/ci.yml

@@ -20,7 +20,7 @@ trigger:
     - restapi*
   paths:
     include:
-    - sdk/web/
+    - sdk/appservice/
 
 pr:
   branches:
@@ -32,7 +32,7 @@ pr:
     - restapi*
   paths:
     include:
-    - sdk/web/
+    - sdk/appservice/
 
 
 stages:

sdk/cognitiveservices/azure-cognitiveservices-vision-customvision/CHANGELOG.md

@@ -19,6 +19,10 @@
   - Added operation CustomVisionPredictionClientOperationsMixin.detect_image
   - Added operation group CustomVisionTrainingClientOperationsMixin
 
+**Breaking changes**
+
+- Credentials are now longer a simple string, but a `msrest.authentication.ApiKeyCredentials` instance instead  
+
 **General Breaking changes**
 
 This version uses a next-generation code generator that *might*

sdk/core/azure-common/azure/common/client_factory.py

@@ -14,9 +14,6 @@
 except ImportError:
     from inspect import getargspec as get_arg_spec
 
-import adal
-from msrestazure.azure_active_directory import AdalAuthentication
-
 from .credentials import get_azure_cli_credentials
 from .cloud import get_cli_active_cloud
 
@@ -153,6 +150,9 @@ def get_client_from_json_dict(client_class, config_dict, **kwargs):
     :param dict config_dict: A config dict.
     :return: An instantiated client
     """
+    import adal
+    from msrestazure.azure_active_directory import AdalAuthentication
+
     is_graphrbac = client_class.__name__ == 'GraphRbacManagementClient'
     is_keyvault = client_class.__name__ == 'KeyVaultClient'
     parameters = {

sdk/core/azure-core/CHANGELOG.md

@@ -1,7 +1,10 @@
 
 # Release History
 
-## 1.6.0 (Unreleased)
+## 1.6.1 (Unreleased)
+
+
+## 1.6.0 (2020-06-03)
 
 ### Bug fixes
 
@@ -13,7 +16,7 @@
 - Added support for changesets as part of multipart message support #10485
 - Add AsyncLROPoller in azure.core.polling #10801
 - Add get_continuation_token/from_continuation_token/polling_method methods in pollers (sync and async) #10801
-- HttpResponse objects are now pickable #10801
+- HttpResponse and PipelineContext objects are now pickable #10801
 
 ## 1.5.0 (2020-05-04)
 

sdk/core/azure-core/azure/core/_version.py

@@ -9,4 +9,4 @@
 # regenerated.
 # --------------------------------------------------------------------------
 
-VERSION = "1.6.0"
+VERSION = "1.6.1"

sdk/core/azure-core/azure/core/pipeline/__init__.py

@@ -63,6 +63,9 @@ class PipelineContext(dict):
     :param transport: The HTTP transport type.
     :param kwargs: Developer-defined keyword arguments.
     """
+    _PICKLE_CONTEXT = {
+        'deserialized_data'
+    }
 
     def __init__(self, transport, **kwargs):  # pylint: disable=super-init-not-called
         self.transport = transport
@@ -75,6 +78,21 @@ def __getstate__(self):
         del state['transport']
         return state
 
+    def __reduce__(self):
+        reduced = super(PipelineContext, self).__reduce__()
+        saved_context = {}
+        for key, value in self.items():
+            if key in self._PICKLE_CONTEXT:
+                saved_context[key] = value
+        # 1 is for from __reduce__ spec of pickle (generic args for recreation)
+        # 2 is how dict is implementing __reduce__ (dict specific)
+        # tuple are read-only, we use a list in the meantime
+        reduced = list(reduced)
+        dict_reduced_result = list(reduced[1])
+        dict_reduced_result[2] = saved_context
+        reduced[1] = tuple(dict_reduced_result)
+        return tuple(reduced)
+
     def __setstate__(self, state):
         self.__dict__.update(state)
         # Re-create the unpickable entries

sdk/core/azure-core/azure/core/polling/_async_poller.py

@@ -84,7 +84,7 @@ async def async_poller(client, initial_response, deserialization_callback, polli
     :param client: A pipeline service client.
     :type client: ~azure.core.PipelineClient
     :param initial_response: The initial call response
-    :type initial_response: ~azure.core.pipeline.transport.AsyncHttpResponse
+    :type initial_response: ~azure.core.pipeline.PipelineResponse
     :param deserialization_callback: A callback that takes a Response and return a deserialized object.
                                      If a subclass of Model is given, this passes "deserialize" as callback.
     :type deserialization_callback: callable or msrest.serialization.Model
@@ -101,8 +101,7 @@ class AsyncLROPoller(Generic[PollingReturnType], Awaitable):
     :param client: A pipeline service client
     :type client: ~azure.core.PipelineClient
     :param initial_response: The initial call response
-    :type initial_response:
-     ~azure.core.pipeline.transport.HttpResponse or ~azure.core.pipeline.transport.AsyncHttpResponse
+    :type initial_response: ~azure.core.pipeline.PipelineResponse
     :param deserialization_callback: A callback that takes a Response and return a deserialized object.
                                      If a subclass of Model is given, this passes "deserialize" as callback.
     :type deserialization_callback: callable or msrest.serialization.Model

sdk/core/azure-core/azure/core/polling/_poller.py

@@ -32,7 +32,6 @@
     from urllib.parse import urlparse
 
 from typing import Any, Callable, Union, List, Optional, Tuple, TypeVar, Generic
-from azure.core.pipeline.transport._base import HttpResponse
 from azure.core.tracing.decorator import distributed_trace
 from azure.core.tracing.common import with_current_context
 
@@ -140,8 +139,7 @@ class LROPoller(Generic[PollingReturnType]):
     :param client: A pipeline service client
     :type client: ~azure.core.PipelineClient
     :param initial_response: The initial call response
-    :type initial_response:
-     ~azure.core.pipeline.transport.HttpResponse or ~azure.core.pipeline.transport.AsyncHttpResponse
+    :type initial_response: ~azure.core.pipeline.PipelineResponse
     :param deserialization_callback: A callback that takes a Response and return a deserialized object.
                                      If a subclass of Model is given, this passes "deserialize" as callback.
     :type deserialization_callback: callable or msrest.serialization.Model
@@ -150,7 +148,7 @@ class LROPoller(Generic[PollingReturnType]):
     """
 
     def __init__(self, client, initial_response, deserialization_callback, polling_method):
-        # type: (Any, HttpResponse, Callable, PollingMethod[PollingReturnType]) -> None
+        # type: (Any, Any, Callable, PollingMethod[PollingReturnType]) -> None
         self._callbacks = []  # type: List[Callable]
         self._polling_method = polling_method
 

sdk/core/azure-core/tests/test_universal_pipeline.py

@@ -25,6 +25,7 @@
 #
 #--------------------------------------------------------------------------
 import logging
+import pickle
 try:
     from unittest import mock
 except ImportError:
@@ -56,6 +57,37 @@
     HTTPPolicy,
 )
 
+def test_pipeline_context():
+    kwargs={
+        'stream':True,
+        'cont_token':"bla"
+    }
+    context = PipelineContext('transport', **kwargs)
+    context['foo'] = 'bar'
+    context['xyz'] = '123'
+    context['deserialized_data'] = 'marvelous'
+
+    assert context['foo'] == 'bar'
+    assert context.options == kwargs
+
+    with pytest.raises(TypeError):
+        context.clear()
+
+    with pytest.raises(TypeError):
+        context.update({})
+
+    assert context.pop('foo') == 'bar'
+    assert 'foo' not in context
+
+    serialized = pickle.dumps(context)
+
+    revived_context = pickle.loads(serialized)
+    assert revived_context.options == kwargs
+    assert revived_context.transport is None
+    assert 'deserialized_data' in revived_context
+    assert len(revived_context) == 1
+
+
 def test_request_history():
     class Non_deep_copiable(object):
         def __deepcopy__(self, memodict={}):

sdk/cosmos/azure-cosmos/CHANGELOG.md

@@ -1,3 +1,8 @@
+## 4.0.1 (Unreleased)
+
+- Fixed error raised when a non string ID is used in an item. It now raises TypeError rather than AttributeError. Issue 11793 - thank you @Rabbit994.
+
+
 ## 4.0.0 (2020-05-20)
 
 - Stable release.

sdk/cosmos/azure-cosmos/azure/cosmos/_cosmos_client_connection.py

@@ -31,6 +31,7 @@
 from urllib3.util.retry import Retry
 from azure.core.paging import ItemPaged  # type: ignore
 from azure.core import PipelineClient  # type: ignore
+from azure.core.exceptions import raise_with_traceback  # type: ignore
 from azure.core.pipeline.policies import (  # type: ignore
     HTTPPolicy,
     ContentDecodePolicy,
@@ -2480,11 +2481,14 @@ def __CheckAndUnifyQueryFormat(self, query_body):
     def __ValidateResource(resource):
         id_ = resource.get("id")
         if id_:
-            if id_.find("/") != -1 or id_.find("\\") != -1 or id_.find("?") != -1 or id_.find("#") != -1:
-                raise ValueError("Id contains illegal chars.")
-
-            if id_[-1] == " ":
-                raise ValueError("Id ends with a space.")
+            try:
+                if id_.find("/") != -1 or id_.find("\\") != -1 or id_.find("?") != -1 or id_.find("#") != -1:
+                    raise ValueError("Id contains illegal chars.")
+
+                if id_[-1] == " ":
+                    raise ValueError("Id ends with a space.")
+            except AttributeError:
+                raise_with_traceback(TypeError, message="Id type must be a string.")
 
     # Adds the partition key to options
     def _AddPartitionKey(self, collection_link, document, options):

sdk/cosmos/azure-cosmos/azure/cosmos/_version.py

@@ -19,4 +19,4 @@
 # OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
 # SOFTWARE.
 
-VERSION = "4.0.0"
+VERSION = "4.0.1"

sdk/cosmos/azure-cosmos/samples/access_cosmos_with_resource_token.py

@@ -14,7 +14,7 @@
 #    https://docs.microsoft.com/azure/cosmos-db/create-sql-api-python#create-a-database-account
 #
 # 2. Microsoft Azure Cosmos
-#    pip install azure-cosmos==4.0.0
+#    pip install azure-cosmos>=4.0.0
 # ----------------------------------------------------------------------------------------------------------
 # Sample - how to get and use resource token that allows restricted access to data
 # ----------------------------------------------------------------------------------------------------------

sdk/cosmos/azure-cosmos/test/test_crud.py

@@ -207,7 +207,7 @@ def test_sql_query_crud(self):
         self.assertEqual(0, len(databases), 'Unexpected number of query results.')
 
         # query with a string.
-        databases = list(self.client.query_databases('SELECT * FROM root r WHERE r.id="' + db2.id + '"'))
+        databases = list(self.client.query_databases('SELECT * FROM root r WHERE r.id="' + db2.id + '"'))   #nosec
         self.assertEqual(1, len(databases), 'Unexpected number of query results.')
         self.client.delete_database(db1.id)
         self.client.delete_database(db2.id)
@@ -507,30 +507,30 @@ def test_partitioned_collection_document_crud_and_query(self):
         # query document on the partition key specified in the predicate will pass even without setting enableCrossPartitionQuery or passing in the partitionKey value
         documentlist = list(created_collection.query_items(
             {
-                'query': 'SELECT * FROM root r WHERE r.id=\'' + replaced_document.get('id') + '\''
+                'query': 'SELECT * FROM root r WHERE r.id=\'' + replaced_document.get('id') + '\''  #nosec
             }))
         self.assertEqual(1, len(documentlist))
 
         # query document on any property other than partitionKey will fail without setting enableCrossPartitionQuery or passing in the partitionKey value
         try:
             list(created_collection.query_items(
                 {
-                    'query': 'SELECT * FROM root r WHERE r.key=\'' + replaced_document.get('key') + '\''
+                    'query': 'SELECT * FROM root r WHERE r.key=\'' + replaced_document.get('key') + '\''    #nosec
                 }))
         except Exception:
             pass
 
         # cross partition query
         documentlist = list(created_collection.query_items(
-            query='SELECT * FROM root r WHERE r.key=\'' + replaced_document.get('key') + '\'',
+            query='SELECT * FROM root r WHERE r.key=\'' + replaced_document.get('key') + '\'',  #nosec
             enable_cross_partition_query=True
         ))
 
         self.assertEqual(1, len(documentlist))
 
         # query document by providing the partitionKey value
         documentlist = list(created_collection.query_items(
-            query='SELECT * FROM root r WHERE r.key=\'' + replaced_document.get('key') + '\'',
+            query='SELECT * FROM root r WHERE r.key=\'' + replaced_document.get('key') + '\'',  #nosec
             partition_key=replaced_document.get('id')
         ))
 
@@ -746,14 +746,14 @@ def test_partitioned_collection_conflict_crud_and_query(self):
         # query conflicts on any property other than partitionKey will fail without setting enableCrossPartitionQuery or passing in the partitionKey value
         try:
             list(created_collection.query_conflicts(
-                    query='SELECT * FROM root r WHERE r.resourceType=\'' + conflict_definition.get(
+                    query='SELECT * FROM root r WHERE r.resourceType=\'' + conflict_definition.get( #nosec
                         'resourceType') + '\''
                 ))
         except Exception:
             pass
 
         conflictlist = list(created_collection.query_conflicts(
-                query='SELECT * FROM root r WHERE r.resourceType=\'' + conflict_definition.get('resourceType') + '\'',
+                query='SELECT * FROM root r WHERE r.resourceType=\'' + conflict_definition.get('resourceType') + '\'',  #nosec
                 enable_cross_partition_query=True
         ))
 
@@ -762,7 +762,7 @@ def test_partitioned_collection_conflict_crud_and_query(self):
         # query conflicts by providing the partitionKey value
         options = {'partitionKey': conflict_definition.get('id')}
         conflictlist = list(created_collection.query_conflicts(
-            query='SELECT * FROM root r WHERE r.resourceType=\'' + conflict_definition.get('resourceType') + '\'',
+            query='SELECT * FROM root r WHERE r.resourceType=\'' + conflict_definition.get('resourceType') + '\'',  #nosec
             partition_key=conflict_definition['id']
         ))
 
@@ -939,6 +939,11 @@ def test_document_upsert(self):
         self.assertEqual(created_document['id'],
                          document_definition['id'])
 
+        # test error for non-string id
+        with pytest.raises(TypeError):
+            document_definition['id'] = 7
+            created_collection.upsert_item(body=document_definition)
+
         # read documents after creation and verify updated count
         documents = list(created_collection.read_all_items())
         self.assertEqual(