[azopenai] Updating to the 2023-07-01 API surface

sdk/cognitiveservices/azopenai/assets.json

@@ -2,5 +2,5 @@
   "AssetsRepo": "Azure/azure-sdk-assets",
   "AssetsRepoPrefixPath": "go",
   "TagPrefix": "go/cognitiveservices/azopenai",
-  "Tag": "go/cognitiveservices/azopenai_2b6f93a94d"
+  "Tag": "go/cognitiveservices/azopenai_63852f374c"
 }

sdk/cognitiveservices/azopenai/autorest.md

@@ -16,7 +16,8 @@ go: true
 use: "@autorest/go@4.0.0-preview.52"
 title: "OpenAI"
 slice-elements-byval: true
-remove-non-reference-schema: true
+# can't use this since it removes an innererror type that we want ()
+# remove-non-reference-schema: true
 ```
 
 ## Transformations
@@ -81,7 +82,7 @@ directive:
     where: $.components.schemas["ImageOperation"].properties.status
     transform: $["$ref"] = $.anyOf[0]["$ref"];delete $.anyOf;
   - from: openapi-document
-    where: $.components.schemas["ImageGenerationOptions"].properties
+    where: $.components.schemas.ImageGenerationOptions.properties
     transform: |
       $.size["$ref"] = "#/components/schemas/ImageSize"; delete $.allOf;
       $.response_format["$ref"] = "#/components/schemas/ImageGenerationResponseFormat"; delete $.allOf;
@@ -93,11 +94,12 @@ directive:
   - from: openapi-document
     where: $.components.schemas["ImageOperationStatus"].properties.status
     transform: $["$ref"] = "#/components/schemas/State"; delete $.allOf; 
+  - from: openapi-document
+    where: $.components.schemas["ContentFilterResult"].properties.severity
+    transform: $["$ref"] = "#/components/schemas/ContentFilterSeverity"; delete $.allOf;
   - from: openapi-document
     where: $.components.schemas["ChatChoice"].properties.finish_reason
-    transform: >
-      delete $.oneOf;
-      $["$ref"] = "#/components/schemas/CompletionsFinishReason";
+    transform: $["$ref"] = "#/components/schemas/CompletionsFinishReason"; delete $.oneOf;
   # Fix "AutoGenerated" models
   - from: openapi-document
     where: $.components.schemas["ChatCompletions"].properties.usage
@@ -163,7 +165,7 @@ directive:
       - client.go
       - models.go
       - options.go
-      - response_types.go      
+      - response_types.go
     where: $
     transform: return $.replace(/Client(\w+)((?:Options|Response))/g, "$1$2");
 
@@ -172,10 +174,19 @@ directive:
     where: $
     transform: return $.replace(/runtime\.JoinPaths\(client.endpoint, urlPath\)/g, "client.formatURL(urlPath)");
 
+  # Some ImageGenerations hackery to represent the ImageLocation/ImagePayload polymorphism.
+  # - Remove the auto-generated ImageGenerationsDataItem.
+  # - Replace the ImageGenerations.Data type with []ImageGenerationDataItem
+  # - from: models.go
+  #   where: $
+  #   transform: |
+  #     return $.replace(/type ImageGenerationsDataItem struct {[^}]+}/, "// ImageGenerationsDataItem represents an image URL or payload\ntype ImageGenerationsDataItem struct{\nImageLocation\nImagePayload\n}")
+  #       $.replace(/(type ImageGenerations struct.+?)Data any/g, "$1Data []ImageGenerationsDataItem")
+
   - from: models.go
     where: $
     transform: |
-      return $.replace(/type ImageGenerationsDataItem struct {[^}]+}/, "// ImageGenerationsDataItem represents an image URL or payload\ntype ImageGenerationsDataItem struct{\nImageLocation\nImagePayload\n}");
+      return $.replace(/(type ImageGenerations struct.+?)Data any/sg, "$1Data []ImageGenerationsDataItem")
 
   # delete the auto-generated ImageGenerationsDataItem, we handle that custom
   - from: models.go
@@ -218,6 +229,17 @@ directive:
         .replace(/BeginAzureBatchImageGenerationInternal/g, "beginAzureBatchImageGeneration")
         .replace(/BatchImageGenerationOperationResponse/g, "batchImageGenerationOperationResponse");
 
+  # BUG: ChatCompletionsOptionsFunctionCall is another one of those "here's mutually exclusive values" options...
+  - from: 
+    - models.go
+    - models_serde.go
+    where: $
+    transform: |
+      return $
+        .replace(/populateAny\(objectMap, "function_call", c.FunctionCall\)/, 'populate(objectMap, "function_call", c.FunctionCall)')
+        .replace(/\/\/ ChatCompletionsOptionsFunctionCall.+?\n}/, "")
+        .replace(/FunctionCall any/, "FunctionCall *ChatCompletionsOptionsFunctionCall");
+
   # fix some casing
   - from: 
     - client.go
@@ -228,8 +250,30 @@ directive:
     where: $
     transform: return $.replace(/Logprobs/g, "LogProbs")
 
-  # remove PossibleazureOpenAIOperationStateValues, since we don't expose the poller
+  # delete ContentFilterResult in favor of our custom representation.
+  - from: 
+    - models.go
+    - models_serde.go
+    where: $
+    transform: |
+      return $.replace(/\/\/ ContentFilterResult.+?\n}/s, "")
+        .replace(/\/\/ MarshalJSON implements the json.Marshaller interface for type ContentFilterResult.+?\n}/s, "")
+        .replace(/\/\/ UnmarshalJSON implements the json.Unmarshaller interface for type ContentFilterResult.+?\n}/s, "");
+
   - from: constants.go
     where: $
     transform: return $.replace(/\/\/ PossibleazureOpenAIOperationStateValues returns.+?\n}/s, "");
+
+  # fix incorrect property name for content filtering
+  # TODO: I imagine we should able to fix this in the tsp?
+  - from: models_serde.go
+    where: $
+    transform: |
+      return $
+        .replace(/		case "selfHarm":/g, '		case "self_harm":')
+        .replace(/populate\(objectMap, "selfHarm", c.SelfHarm\)/g, 'populate(objectMap, "self_harm", c.SelfHarm)');
+
+  - from: client.go
+    where: $
+    transform: return $.replace(/runtime\.NewResponseError/sg, "client.newError");
 ```

sdk/cognitiveservices/azopenai/client_chat_completions_test.go

@@ -44,30 +44,33 @@ func TestClient_GetChatCompletions(t *testing.T) {
 	chatClient, err := azopenai.NewClientWithKeyCredential(endpoint, cred, chatCompletionsModelDeployment, newClientOptionsForTest(t))
 	require.NoError(t, err)
 
-	testGetChatCompletions(t, chatClient)
+	testGetChatCompletions(t, chatClient, true)
 }
 
 func TestClient_GetChatCompletionsStream(t *testing.T) {
-	cred, err := azopenai.NewKeyCredential(apiKey)
-	require.NoError(t, err)
-
-	chatClient, err := azopenai.NewClientWithKeyCredential(endpoint, cred, chatCompletionsModelDeployment, newClientOptionsForTest(t))
-	require.NoError(t, err)
-
-	testGetChatCompletionsStream(t, chatClient, true)
+	chatClient := newAzureOpenAIClientForTest(t, canaryChatCompletionsModelDeployment, true)
+	testGetChatCompletionsStream(t, chatClient)
 }
 
 func TestClient_OpenAI_GetChatCompletions(t *testing.T) {
+	if testing.Short() {
+		t.Skip("Skipping OpenAI tests when attempting to do quick tests")
+	}
+
 	chatClient := newOpenAIClientForTest(t)
-	testGetChatCompletions(t, chatClient)
+	testGetChatCompletions(t, chatClient, false)
 }
 
 func TestClient_OpenAI_GetChatCompletionsStream(t *testing.T) {
+	if testing.Short() {
+		t.Skip("Skipping OpenAI tests when attempting to do quick tests")
+	}
+
 	chatClient := newOpenAIClientForTest(t)
-	testGetChatCompletionsStream(t, chatClient, false)
+	testGetChatCompletionsStream(t, chatClient)
 }
 
-func testGetChatCompletions(t *testing.T, client *azopenai.Client) {
+func testGetChatCompletions(t *testing.T, client *azopenai.Client, isAzure bool) {
 	expected := azopenai.ChatCompletions{
 		Choices: []azopenai.ChatChoice{
 			{
@@ -91,6 +94,15 @@ func testGetChatCompletions(t *testing.T, client *azopenai.Client) {
 	resp, err := client.GetChatCompletions(context.Background(), chatCompletionsRequest, nil)
 	require.NoError(t, err)
 
+	if isAzure {
+		// Azure also provides content-filtering. This particular prompt and responses
+		// will be considered safe.
+		expected.PromptAnnotations = []azopenai.PromptFilterResult{
+			{PromptIndex: to.Ptr[int32](0), ContentFilterResults: (*azopenai.PromptFilterResultContentFilterResults)(safeContentFilter)},
+		}
+		expected.Choices[0].ContentFilterResults = safeContentFilter
+	}
+
 	require.NotEmpty(t, resp.ID)
 	require.NotEmpty(t, resp.Created)
 
@@ -100,17 +112,10 @@ func testGetChatCompletions(t *testing.T, client *azopenai.Client) {
 	require.Equal(t, expected, resp.ChatCompletions)
 }
 
-func testGetChatCompletionsStream(t *testing.T, client *azopenai.Client, isAzure bool) {
+func testGetChatCompletionsStream(t *testing.T, client *azopenai.Client) {
 	streamResp, err := client.GetChatCompletionsStream(context.Background(), chatCompletionsRequest, nil)
 	require.NoError(t, err)
 
-	if isAzure {
-		// there's a bug right now where the first event comes back empty
-		// Issue: https://github.com/Azure/azure-sdk-for-go/issues/21086
-		_, err := streamResp.ChatCompletionsStream.Read()
-		require.NoError(t, err)
-	}
-
 	// the data comes back differently for streaming
 	// 1. the text comes back in the ChatCompletion.Delta field
 	// 2. the role is only sent on the first streamed ChatCompletion
@@ -125,6 +130,18 @@ func testGetChatCompletionsStream(t *testing.T, client *azopenai.Client, isAzure
 		}
 
 		require.NoError(t, err)
+
+		if completion.PromptAnnotations != nil {
+			require.Equal(t, []azopenai.PromptFilterResult{
+				{PromptIndex: to.Ptr[int32](0), ContentFilterResults: (*azopenai.PromptFilterResultContentFilterResults)(safeContentFilter)},
+			}, completion.PromptAnnotations)
+		}
+
+		if len(completion.Choices) == 0 {
+			// you can get empty entries that contain just metadata (ie, prompt annotations)
+			continue
+		}
+
 		require.Equal(t, 1, len(completion.Choices))
 		choices = append(choices, completion.Choices[0])
 	}
@@ -140,7 +157,6 @@ func testGetChatCompletionsStream(t *testing.T, client *azopenai.Client, isAzure
 	}
 
 	require.Equal(t, expectedContent, message, "Ultimately, the same result as GetChatCompletions(), just sent across the .Delta field instead")
-
 	require.Equal(t, azopenai.ChatRoleAssistant, expectedRole)
 }
 
@@ -167,7 +183,7 @@ func TestClient_GetChatCompletions_DefaultAzureCredential(t *testing.T) {
 	})
 	require.NoError(t, err)
 
-	testGetChatCompletions(t, chatClient)
+	testGetChatCompletions(t, chatClient, true)
 }
 
 func TestClient_GetChatCompletions_InvalidModel(t *testing.T) {