[SchemaRegistry] handle unknown content types (Azure#37166)

* [SchemaRegistry] handle unknown content types

* remove extra space in Accept headers

* fix license in tests

* fix spacing in accept header async

* add tests with mocked unknown content type

* mypy/lint

* cspell

Author: swathipil

sdk/schemaregistry/azure-schemaregistry/CHANGELOG.md

@@ -6,15 +6,16 @@ This version and all future versions will require Python 3.8+. Python 3.7 is no
 
 ### Features Added
 
-- `V2022_10` has been added to `ApiVersion` and set as the default API version.
-  - `Json` and `Custom` have been added to supported formats in `SchemaFormat`.
 - Sync and async `JsonSchemaEncoder` have been added under `azure.schemaregistry.encoder.jsonencoder`.
 - `InvalidContentError` have been added under `azure.schemaregistry.encoder.jsonencoder` for use with the `JsonSchemaEncoder`.
 - `MessageContent`, `OutboundMessageContent`,`InboundMessageContent`, and `SchemaContentValidate` have been added under `azure.schemaregistry` as protocols for use with the `JsonSchemaEncoder` and/or future encoder implementations.
+- `Json` and `Custom` have been added to supported formats in `SchemaFormat`.
+- `V2022_10` has been added to `ApiVersion` and set as the default API version.
 
 ### Bugs Fixed
 
 - Fixed a bug in sync/async `register_schema` and `get_schema_properties` that did not accept case insensitive strings as an argument to the `format` parameter.
+- Fixed a bug where unknown content type strings from the service raised a client error, rather than being returned as a string in the SchemaProperties `format` property.
 
 ### Other Changes
 

sdk/schemaregistry/azure-schemaregistry/assets.json

@@ -2,5 +2,5 @@
   "AssetsRepo": "Azure/azure-sdk-assets",
   "AssetsRepoPrefixPath": "python",
   "TagPrefix": "python/schemaregistry/azure-schemaregistry",
-  "Tag": "python/schemaregistry/azure-schemaregistry_a1c9d18bfd"
+  "Tag": "python/schemaregistry/azure-schemaregistry_0a27c7561c"
 }

sdk/schemaregistry/azure-schemaregistry/azure/schemaregistry/_patch.py

@@ -22,12 +22,13 @@
     overload,
     IO,
 )
+from enum import Enum
 from typing_extensions import Protocol, TypedDict, Self
-
+from azure.core import CaseInsensitiveEnumMeta
 from azure.core.tracing.decorator import distributed_trace
 
 from ._client import SchemaRegistryClient as GeneratedServiceClient
-from .models._patch import SchemaFormat
+from .models._patch import SchemaFormat, NormalizedSchemaContentTypes
 
 if TYPE_CHECKING:
     from azure.core.credentials import TokenCredential
@@ -57,22 +58,24 @@ def _parse_schema_properties_dict(response_headers: Mapping[str, Union[str, int]
         "version": int(response_headers["Schema-Version"]),
     }
 
+def _normalize_content_type(content_type: str) -> str:
+    return content_type.replace(" ", "").lower()
 
-def _get_format(content_type: str) -> SchemaFormat:
+def _get_format(content_type: str) -> Union[SchemaFormat, str]:
     # pylint:disable=redefined-builtin
     # Exception cases may be due to forward compatibility.
     # i.e. Getting a schema with a content type from a future API version.
-    # In this case, we default to CUSTOM format.
-    try:
-        format = content_type.split("serialization=")[1]
-        try:
-            return SchemaFormat(format.capitalize())
-        except ValueError:
-            pass
-    except IndexError:
-        pass
-    return SchemaFormat.CUSTOM
-
+    # In this case, we default to returning the content type string.
+
+    # remove whitespace and case from string
+    normalized_content_type = _normalize_content_type(content_type)
+    if normalized_content_type == NormalizedSchemaContentTypes.AVRO.value:
+        return SchemaFormat.AVRO
+    if normalized_content_type == NormalizedSchemaContentTypes.JSON.value:
+        return SchemaFormat.JSON
+    if normalized_content_type == NormalizedSchemaContentTypes.CUSTOM.value:
+        return SchemaFormat.CUSTOM
+    return content_type
 
 def prepare_schema_properties_result(  # pylint:disable=unused-argument,redefined-builtin
     format: str,
@@ -220,10 +223,62 @@ def register_schema(  # pylint:disable=arguments-differ
         return SchemaProperties(**properties)
 
     @overload
-    def get_schema(self, schema_id: str, **kwargs: Any) -> Schema: ...
+    def get_schema(self, schema_id: str, **kwargs: Any) -> Schema:
+        """Gets a registered schema.
+
+        To get a registered schema by its unique ID, pass the `schema_id` parameter and any optional
+        keyword arguments. Azure Schema Registry guarantees that ID is unique within a namespace.
+
+        WARNING: If retrieving a schema format that is unsupported by this client version, upgrade to a client
+         version that supports the schema format. Otherwise, the content MIME type string will be returned as
+         the `format` value in the `properties` of the returned Schema.
+
+        :param str schema_id: References specific schema in registry namespace. Required if `group_name`,
+         `name`, and `version` are not provided.
+        :return: The schema stored in the registry associated with the provided arguments.
+        :rtype: ~azure.schemaregistry.Schema
+        :raises: :class:`~azure.core.exceptions.HttpResponseError`
+
+        .. admonition:: Example:
+
+            .. literalinclude:: ../samples/sync_samples/sample_code_schemaregistry.py
+                :start-after: [START get_schema_sync]
+                :end-before: [END get_schema_sync]
+                :language: python
+                :dedent: 4
+                :caption: Get schema by id.
+
+        """
+        ...
 
     @overload
-    def get_schema(self, *, group_name: str, name: str, version: int, **kwargs: Any) -> Schema: ...
+    def get_schema(self, *, group_name: str, name: str, version: int, **kwargs: Any) -> Schema:
+        """Gets a registered schema.
+
+        To get a specific version of a schema within the specified schema group, pass in the required
+        keyword arguments `group_name`, `name`, and `version` and any optional keyword arguments.
+
+        WARNING: If retrieving a schema format that is unsupported by this client version, upgrade to a client
+         version that supports the schema format. Otherwise, the content MIME type string will be returned as
+         the `format` value in the `properties` of the returned Schema.
+
+        :keyword str group_name: Name of schema group that contains the registered schema.
+        :keyword str name: Name of schema which should be retrieved.
+        :keyword int version: Version of schema which should be retrieved.
+        :return: The schema stored in the registry associated with the provided arguments.
+        :rtype: ~azure.schemaregistry.Schema
+        :raises: :class:`~azure.core.exceptions.HttpResponseError`
+
+        .. admonition:: Example:
+
+            .. literalinclude:: ../samples/sync_samples/sample_code_schemaregistry.py
+                :start-after: [START get_schema_by_version_sync]
+                :end-before: [END get_schema_by_version_sync]
+                :language: python
+                :dedent: 4
+                :caption: Get schema by version.
+        """
+        ...
 
     @distributed_trace
     def get_schema(  # pylint: disable=docstring-missing-param,docstring-should-be-keyword
@@ -237,6 +292,10 @@ def get_schema(  # pylint: disable=docstring-missing-param,docstring-should-be-k
         2) To get a specific version of a schema within the specified schema group, pass in the required
         keyword arguments `group_name`, `name`, and `version` and any optional keyword arguments.
 
+        WARNING: If retrieving a schema format that is unsupported by this client version, upgrade to a client
+         version that supports the schema format. Otherwise, the content MIME type string will be returned as
+         the `format` value in the `properties` of the returned Schema.
+
         :param str schema_id: References specific schema in registry namespace. Required if `group_name`,
          `name`, and `version` are not provided.
         :keyword str group_name: Name of schema group that contains the registered schema.
@@ -281,8 +340,8 @@ def get_schema(  # pylint: disable=docstring-missing-param,docstring-should-be-k
                 id=schema_id,
                 cls=prepare_schema_result,
                 headers={  # TODO: remove when multiple content types in response are supported
-                    "Accept": """application/json; serialization=Avro, application/json; \
-                        serialization=json, text/plain; charset=utf-8"""
+                    "Accept": """application/json; serialization=Avro, application/json; """
+                        """serialization=json, text/plain; charset=utf-8"""
                 },
                 stream=True,
                 **http_request_kwargs,
@@ -305,8 +364,8 @@ def get_schema(  # pylint: disable=docstring-missing-param,docstring-should-be-k
                 schema_version=version,
                 cls=prepare_schema_result,
                 headers={  # TODO: remove when multiple content types in response are supported
-                    "Accept": """application/json; serialization=Avro, application/json; \
-                        serialization=json, text/plain; charset=utf-8"""
+                    "Accept": """application/json; serialization=Avro, application/json; """
+                        """serialization=json, text/plain; charset=utf-8"""
                 },
                 stream=True,
                 **http_request_kwargs,
@@ -414,6 +473,16 @@ def __init__(self, **kwargs: Any) -> None:
     def __repr__(self) -> str:
         return f"Schema(definition={self.definition}, properties={self.properties})"[:1024]
 
+# ApiVersion was added to a previously GA'd version. However, newer libraries should not
+# accept ApiVersion enums and only take strings. Leaving this here for backwards compatibility.
+class ApiVersion(str, Enum, metaclass=CaseInsensitiveEnumMeta):
+    """
+    Represents the Schema Registry API version to use for requests.
+    """
+
+    V2021_10 = "2021-10"
+    V2022_10 = "2022-10"
+    """This is the default version."""
 
 ###### Encoder Protocols ######
 

sdk/schemaregistry/azure-schemaregistry/azure/schemaregistry/aio/_patch.py

@@ -143,10 +143,62 @@ async def register_schema(
         return SchemaProperties(**properties)
 
     @overload
-    async def get_schema(self, schema_id: str, **kwargs: Any) -> Schema: ...
+    async def get_schema(self, schema_id: str, **kwargs: Any) -> Schema:
+        """Gets a registered schema.
+
+        To get a registered schema by its unique ID, pass the `schema_id` parameter and any optional
+        keyword arguments. Azure Schema Registry guarantees that ID is unique within a namespace.
+
+        WARNING: If retrieving a schema format that is unsupported by this client version, upgrade to a client
+         version that supports the schema format. Otherwise, the content MIME type string will be returned as
+         the `format` value in the `properties` of the returned Schema.
+
+        :param str schema_id: References specific schema in registry namespace. Required if `group_name`,
+         `name`, and `version` are not provided.
+        :return: The schema stored in the registry associated with the provided arguments.
+        :rtype: ~azure.schemaregistry.Schema
+        :raises: :class:`~azure.core.exceptions.HttpResponseError`
+
+        .. admonition:: Example:
+
+            .. literalinclude:: ../samples/async_samples/sample_code_schemaregistry_async.py
+                :start-after: [START get_schema_async]
+                :end-before: [END get_schema_async]
+                :language: python
+                :dedent: 4
+                :caption: Get schema by id.
+
+        """
+        ...
 
     @overload
-    async def get_schema(self, *, group_name: str, name: str, version: int, **kwargs: Any) -> Schema: ...
+    async def get_schema(self, *, group_name: str, name: str, version: int, **kwargs: Any) -> Schema:
+        """Gets a registered schema.
+
+        To get a specific version of a schema within the specified schema group, pass in the required
+        keyword arguments `group_name`, `name`, and `version` and any optional keyword arguments.
+
+        WARNING: If retrieving a schema format that is unsupported by this client version, upgrade to a client
+         version that supports the schema format. Otherwise, the content MIME type string will be returned as
+         the `format` value in the `properties` of the returned Schema.
+
+        :keyword str group_name: Name of schema group that contains the registered schema.
+        :keyword str name: Name of schema which should be retrieved.
+        :keyword int version: Version of schema which should be retrieved.
+        :return: The schema stored in the registry associated with the provided arguments.
+        :rtype: ~azure.schemaregistry.Schema
+        :raises: :class:`~azure.core.exceptions.HttpResponseError`
+
+        .. admonition:: Example:
+
+            .. literalinclude:: ../samples/async_samples/sample_code_schemaregistry_async.py
+                :start-after: [START get_schema_by_version_async]
+                :end-before: [END get_schema_by_version_async]
+                :language: python
+                :dedent: 4
+                :caption: Get schema by version.
+        """
+        ...
 
     @distributed_trace_async
     async def get_schema(  # pylint: disable=docstring-missing-param,docstring-should-be-keyword
@@ -160,6 +212,10 @@ async def get_schema(  # pylint: disable=docstring-missing-param,docstring-shoul
         2) To get a specific version of a schema within the specified schema group, pass in the required
         keyword arguments `group_name`, `name`, and `version` and any optional keyword arguments.
 
+        WARNING: If retrieving a schema format that is unsupported by this client version, upgrade to a client
+         version that supports the schema format. Otherwise, the content MIME type string will be returned as
+         the `format` value in the `properties` of the returned Schema.
+
         :param str schema_id: References specific schema in registry namespace. Required if `group_name`,
          `name`, and `version` are not provided.
         :keyword str group_name: Name of schema group that contains the registered schema.
@@ -204,8 +260,8 @@ async def get_schema(  # pylint: disable=docstring-missing-param,docstring-shoul
                 id=schema_id,
                 cls=prepare_schema_result,
                 headers={  # TODO: remove when multiple content types are supported
-                    "Accept": """application/json; serialization=Avro, application/json; \
-                        serialization=json, text/plain; charset=utf-8"""
+                    "Accept": """application/json; serialization=Avro, application/json; """
+                        """serialization=json, text/plain; charset=utf-8"""
                 },
                 stream=True,
                 **http_request_kwargs,
@@ -229,8 +285,8 @@ async def get_schema(  # pylint: disable=docstring-missing-param,docstring-shoul
                 schema_version=version,
                 cls=prepare_schema_result,
                 headers={  # TODO: remove when multiple content types are supported
-                    "Accept": """application/json; serialization=Avro, application/json; \
-                        serialization=json, text/plain; charset=utf-8"""
+                    "Accept": """application/json; serialization=Avro, application/json; """
+                        """serialization=json, text/plain; charset=utf-8"""
                 },
                 stream=True,
                 **http_request_kwargs,

sdk/schemaregistry/azure-schemaregistry/azure/schemaregistry/models/_patch.py

@@ -9,7 +9,9 @@
 from typing import List
 
 from enum import Enum
+
 from azure.core import CaseInsensitiveEnumMeta
+from ._enums import SchemaContentTypeValues
 
 
 class SchemaFormat(str, Enum, metaclass=CaseInsensitiveEnumMeta):
@@ -22,18 +24,16 @@ class SchemaFormat(str, Enum, metaclass=CaseInsensitiveEnumMeta):
     CUSTOM = "Custom"
     """Represents a custom schema format."""
 
+# Normalizing the schema content type strings for whitespace and case insensitive comparison.
+class NormalizedSchemaContentTypes(str, Enum, metaclass=CaseInsensitiveEnumMeta):
+    """Describes closed list of normalized schema content type values."""
 
-class ApiVersion(str, Enum, metaclass=CaseInsensitiveEnumMeta):
-    """
-    Represents the Schema Registry API version to use for requests.
-    """
-
-    V2021_10 = "2021-10"
-    V2022_10 = "2022-10"
-    """This is the default version."""
-
-
-DEFAULT_VERSION = ApiVersion.V2022_10
+    AVRO = SchemaContentTypeValues.AVRO.value.replace(" ", "").lower()
+    """Avro encoding."""
+    JSON = SchemaContentTypeValues.JSON.value.replace(" ", "").lower()
+    """JSON encoding"""
+    CUSTOM = SchemaContentTypeValues.CUSTOM.value.replace(" ", "").lower()
+    """Plain text custom encoding."""
 
 
 __all__: List[str] = []  # Add all objects you want publicly available to users at this package level