Method handleOpenAIWithLabels

scaletest/llmmock/server.go:210–279

(w http.ResponseWriter, r *http.Request)

}

func (s *Server) handleOpenAIWithLabels(w http.ResponseWriter, r *http.Request) {
	s.logger.Debug(r.Context(), "handling OpenAI request")
	defer s.logger.Debug(r.Context(), "handled OpenAI request")

	ctx := r.Context()
	requestID := uuid.New()
	now := time.Now()

	var req llmRequest
	if err := json.NewDecoder(r.Body).Decode(&req); err != nil {
		s.logger.Error(ctx, "failed to parse OpenAI request", slog.Error(err))
		http.Error(w, "invalid request body", http.StatusBadRequest)
		return
	}

	if s.artificialLatency > 0 {
		time.Sleep(s.artificialLatency)
	}

	var resp openAIResponse
	resp.ID = fmt.Sprintf("chatcmpl-%s", requestID.String()[:8])
	resp.Object = "chat.completion"
	resp.Created = now.Unix()
	resp.Model = req.Model

	var responseContent string
	if s.responsePayloadSize > 0 {
		pattern := "x"
		repeated := strings.Repeat(pattern, s.responsePayloadSize)
		responseContent = repeated[:s.responsePayloadSize]
	} else {
		responseContent = "This is a mock response from OpenAI."
	}

	resp.Choices = []struct {
		Index        int           `json:"index"`
		Message      openAIMessage `json:"message"`
		FinishReason string        `json:"finish_reason"`
	}{
		{
			Index: 0,
			Message: openAIMessage{
				Role:    "assistant",
				Content: responseContent,
			},
			FinishReason: "stop",
		},
	}

	resp.Usage.PromptTokens = 10
	resp.Usage.CompletionTokens = 5
	resp.Usage.TotalTokens = 15

	responseBody, _ := json.Marshal(resp)

	if req.Stream {
		s.sendOpenAIStream(ctx, w, resp)
	} else {