Refactor the last refactor :)

Removed HandlePartialResponse, add LLMRequest which handles all common
logic of making LLM requests and returning/showing their response.

pkg/cli/cmd.go

@@ -53,19 +53,28 @@ func SystemPrompt() string {
 	return systemPrompt
 }
 
-// HandlePartialResponse accepts a response and an err. If err is nil, it does
+// LLMRequest prompts the LLM with the given Message, writes the (partial)
-// nothing and returns nil. If response != "" and err != nil, it prints a
+// response to stdout, and returns the (partial) response or any errors.
-// warning and returns nil. If response == "" and err != nil, it returns the
+func LLMRequest(messages []Message) (string, error) {
-// error.
+	// receiver receives the reponse from LLM
-func HandlePartialResponse(response string, err error) (e error) {
+	receiver := make(chan string)
-	if err != nil {
+	defer close(receiver)
-		if response != "" {
+
-			Warn("Received partial response. Error: %v\n", err)
+	// start HandleDelayedContent goroutine to print received data to stdout
-		} else {
+	go HandleDelayedContent(receiver)
-			e = err
+
-		}
+	response, err := CreateChatCompletionStream(model, messages, maxTokens, receiver)
+	if err != nil && response != "" {
+		Warn("Received partial response. Error: %v\n", err)
+		err = nil // ignore partial response error
 	}
-	return
+
+	if response != "" {
+		// there was some content, so break to a new line after it
+		fmt.Println()
+	}
+
+	return response, err
 }
 
 // InputFromArgsOrEditor returns either the provided input from the args slice
@@ -314,17 +323,12 @@ var replyCmd = &cobra.Command{
 		}
 		assistantReply.RenderTTY()
 
-		receiver := make(chan string)
+		response, err := LLMRequest(messages)
-		go HandleDelayedContent(receiver)
-		response, err := CreateChatCompletionStream(model, messages, maxTokens, receiver)
-
-		err = HandlePartialResponse(response, err)
 		if err != nil {
-			Fatal("Error while receiving response: %v\n", err)
+			Fatal("Error fetching LLM response: %v\n", err)
 		}
 
 		assistantReply.OriginalContent = response
-		fmt.Println()
 
 		err = store.SaveMessage(&assistantReply)
 		if err != nil {
@@ -382,16 +386,11 @@ var newCmd = &cobra.Command{
 		}
 		reply.RenderTTY()
 
-		receiver := make(chan string)
+		response, err := LLMRequest(messages)
-		go HandleDelayedContent(receiver)
-		response, err := CreateChatCompletionStream(model, messages, maxTokens, receiver)
-
-		err = HandlePartialResponse(response, err)
 		if err != nil {
-			Fatal("Error while receiving response: %v\n", err)
+			Fatal("Error fetching LLM response: %v\n", err)
 		}
 
-		fmt.Println()
 		reply.OriginalContent = response
 
 		err = store.SaveMessage(&reply)
@@ -432,16 +431,10 @@ var promptCmd = &cobra.Command{
 			},
 		}
 
-		receiver := make(chan string)
+		_, err := LLMRequest(messages)
-		go HandleDelayedContent(receiver)
-		response, err := CreateChatCompletionStream(model, messages, maxTokens, receiver)
-
-		err = HandlePartialResponse(response, err)
 		if err != nil {
-			Fatal("%v\n", err)
+			Fatal("Error fetching LLM response: %v\n", err)
 		}
-
-		fmt.Println()
 	},
 }
 
@@ -485,25 +478,17 @@ var retryCmd = &cobra.Command{
 		}
 		assistantReply.RenderTTY()
 
-		receiver := make(chan string)
+		response, err := LLMRequest(messages)
-		go HandleDelayedContent(receiver)
-		response, err := CreateChatCompletionStream(model, messages, maxTokens, receiver)
-
-		err = HandlePartialResponse(response, err)
 		if err != nil {
-			Fatal("Error while receiving response: %v\n", err)
+			Fatal("Error fetching LLM response: %v\n", err)
 		}
 
-		fmt.Println()
 		assistantReply.OriginalContent = response
 
 		err = store.SaveMessage(&assistantReply)
 		if err != nil {
 			Fatal("Could not save assistant reply: %v\n", err)
 		}
-
-		fmt.Println()
-
 	},
 	ValidArgsFunction: func(cmd *cobra.Command, args []string, toComplete string) ([]string, cobra.ShellCompDirective) {
 		compMode := cobra.ShellCompDirectiveNoFileComp
@@ -544,25 +529,17 @@ var continueCmd = &cobra.Command{
 		}
 		assistantReply.RenderTTY()
 
-		receiver := make(chan string)
+		response, err := LLMRequest(messages)
-		go HandleDelayedContent(receiver)
-		response, err := CreateChatCompletionStream(model, messages, maxTokens, receiver)
-
-		err = HandlePartialResponse(response, err)
 		if err != nil {
-			Fatal("Error while receiving response: %v\n", err)
+			Fatal("Error fetching LLM response: %v\n", err)
 		}
 
-		fmt.Println()
 		assistantReply.OriginalContent = response
 
 		err = store.SaveMessage(&assistantReply)
 		if err != nil {
 			Fatal("Could not save assistant reply: %v\n", err)
 		}
-
-		fmt.Println()
-
 	},
 	ValidArgsFunction: func(cmd *cobra.Command, args []string, toComplete string) ([]string, cobra.ShellCompDirective) {
 		compMode := cobra.ShellCompDirectiveNoFileComp

pkg/cli/openai.go

@@ -45,8 +45,6 @@ func CreateChatCompletionStream(model string, messages []Message, maxTokens int,
 	client := openai.NewClient(*config.OpenAI.APIKey)
 	req := CreateChatCompletionRequest(model, messages, maxTokens)
 
-	defer close(output)
-
 	stream, err := client.CreateChatCompletionStream(context.Background(), req)
 	if err != nil {
 		return "", err