fix: remove unnecessary error handling in token counting functions

relay/audio_handler.go

@@ -67,9 +67,6 @@ func AudioHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode) {
 	preConsumedTokens := common.PreConsumedQuota
 	preConsumedTokens := common.PreConsumedQuota
 	if relayInfo.RelayMode == relayconstant.RelayModeAudioSpeech {
 	if relayInfo.RelayMode == relayconstant.RelayModeAudioSpeech {
 		promptTokens = service.CountTTSToken(audioRequest.Input, audioRequest.Model)
 		promptTokens = service.CountTTSToken(audioRequest.Input, audioRequest.Model)
-		if err != nil {
-			return service.OpenAIErrorWrapper(err, "count_audio_token_failed", http.StatusInternalServerError)
-		}
 		preConsumedTokens = promptTokens
 		preConsumedTokens = promptTokens
 		relayInfo.PromptTokens = promptTokens
 		relayInfo.PromptTokens = promptTokens
 	}
 	}

relay/channel/claude/relay-claude.go

@@ -619,9 +619,6 @@ func HandleClaudeResponseData(c *gin.Context, info *relaycommon.RelayInfo, claud
 	}
 	}
 	if requestMode == RequestModeCompletion {
 	if requestMode == RequestModeCompletion {
 		completionTokens := service.CountTextToken(claudeResponse.Completion, info.OriginModelName)
 		completionTokens := service.CountTextToken(claudeResponse.Completion, info.OriginModelName)
-		if err != nil {
-			return service.OpenAIErrorWrapper(err, "count_token_text_failed", http.StatusInternalServerError)
-		}
 		claudeInfo.Usage.PromptTokens = info.PromptTokens
 		claudeInfo.Usage.PromptTokens = info.PromptTokens
 		claudeInfo.Usage.CompletionTokens = completionTokens
 		claudeInfo.Usage.CompletionTokens = completionTokens
 		claudeInfo.Usage.TotalTokens = info.PromptTokens + completionTokens
 		claudeInfo.Usage.TotalTokens = info.PromptTokens + completionTokens

relay/channel/coze/relay-coze.go

@@ -106,7 +106,7 @@ func cozeChatStreamHandler(c *gin.Context, resp *http.Response, info *relaycommo
 
 
 	var currentEvent string
 	var currentEvent string
 	var currentData string
 	var currentData string
-	var usage dto.Usage
+	var usage = &dto.Usage{}
 
 
 	for scanner.Scan() {
 	for scanner.Scan() {
 		line := scanner.Text()
 		line := scanner.Text()
@@ -114,7 +114,7 @@ func cozeChatStreamHandler(c *gin.Context, resp *http.Response, info *relaycommo
 		if line == "" {
 		if line == "" {
 			if currentEvent != "" && currentData != "" {
 			if currentEvent != "" && currentData != "" {
 				// handle last event
 				// handle last event
-				handleCozeEvent(c, currentEvent, currentData, &responseText, &usage, id, info)
+				handleCozeEvent(c, currentEvent, currentData, &responseText, usage, id, info)
 				currentEvent = ""
 				currentEvent = ""
 				currentData = ""
 				currentData = ""
 			}
 			}
@@ -134,7 +134,7 @@ func cozeChatStreamHandler(c *gin.Context, resp *http.Response, info *relaycommo
 
 
 	// Last event
 	// Last event
 	if currentEvent != "" && currentData != "" {
 	if currentEvent != "" && currentData != "" {
-		handleCozeEvent(c, currentEvent, currentData, &responseText, &usage, id, info)
+		handleCozeEvent(c, currentEvent, currentData, &responseText, usage, id, info)
 	}
 	}
 
 
 	if err := scanner.Err(); err != nil {
 	if err := scanner.Err(); err != nil {
@@ -143,12 +143,10 @@ func cozeChatStreamHandler(c *gin.Context, resp *http.Response, info *relaycommo
 	helper.Done(c)
 	helper.Done(c)
 
 
 	if usage.TotalTokens == 0 {
 	if usage.TotalTokens == 0 {
-		usage.PromptTokens = info.PromptTokens
-		usage.CompletionTokens = service.CountTextToken("gpt-3.5-turbo", responseText)
-		usage.TotalTokens = usage.PromptTokens + usage.CompletionTokens
+		usage = service.ResponseText2Usage(responseText, info.UpstreamModelName, c.GetInt("coze_input_count"))
 	}
 	}
 
 
-	return nil, &usage
+	return nil, usage
 }
 }
 
 
 func handleCozeEvent(c *gin.Context, event string, data string, responseText *string, usage *dto.Usage, id string, info *relaycommon.RelayInfo) {
 func handleCozeEvent(c *gin.Context, event string, data string, responseText *string, usage *dto.Usage, id string, info *relaycommon.RelayInfo) {

relay/channel/dify/relay-dify.go

@@ -243,15 +243,8 @@ func difyStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Re
 		return true
 		return true
 	})
 	})
 	helper.Done(c)
 	helper.Done(c)
-	err := resp.Body.Close()
-	if err != nil {
-		// return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
-		common.SysError("close_response_body_failed: " + err.Error())
-	}
 	if usage.TotalTokens == 0 {
 	if usage.TotalTokens == 0 {
-		usage.PromptTokens = info.PromptTokens
-		usage.CompletionTokens = service.CountTextToken("gpt-3.5-turbo", responseText)
-		usage.TotalTokens = usage.PromptTokens + usage.CompletionTokens
+		usage = service.ResponseText2Usage(responseText, info.UpstreamModelName, info.PromptTokens)
 	}
 	}
 	usage.CompletionTokens += nodeToken
 	usage.CompletionTokens += nodeToken
 	return nil, usage
 	return nil, usage

relay/responses_handler.go

@@ -73,9 +73,6 @@ func ResponsesHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode)
 		relayInfo.SetPromptTokens(promptTokens)
 		relayInfo.SetPromptTokens(promptTokens)
 	} else {
 	} else {
 		promptTokens := getInputTokens(req, relayInfo)
 		promptTokens := getInputTokens(req, relayInfo)
-		if err != nil {
-			return service.OpenAIErrorWrapperLocal(err, "count_input_tokens_error", http.StatusBadRequest)
-		}
 		c.Set("prompt_tokens", promptTokens)
 		c.Set("prompt_tokens", promptTokens)
 	}
 	}