opencode-cursor-agent/internal/server/handlers.go

package server

import (
	"context"
	"encoding/json"
	"fmt"
	"log/slog"
	"math"
	"net/http"
	"strings"
	"sync"
	"time"

	"github.com/daniel/cursor-adapter/internal/converter"
	"github.com/daniel/cursor-adapter/internal/sanitize"
	"github.com/daniel/cursor-adapter/internal/types"
)

var (
	modelCacheMu    sync.Mutex
	modelCacheData  []string
	modelCacheAt    time.Time
	modelCacheTTL   = 5 * time.Minute
)

func (s *Server) handleListModels(w http.ResponseWriter, r *http.Request) {
	models, err := s.cachedListModels(r.Context())
	if err != nil {
		writeJSON(w, http.StatusInternalServerError, types.NewErrorResponse(err.Error(), "internal_error", ""))
		return
	}

	ts := time.Now().Unix()
	data := make([]types.ModelInfo, 0, len(models)*2)
	for _, m := range models {
		data = append(data, types.ModelInfo{ID: m, Object: "model", Created: ts, OwnedBy: "cursor"})
	}
	aliases := converter.GetAnthropicModelAliases(models)
	for _, a := range aliases {
		data = append(data, types.ModelInfo{ID: a.ID, Object: "model", Created: ts, OwnedBy: "cursor"})
	}
	writeJSON(w, http.StatusOK, types.ModelList{Object: "list", Data: data})
}

func (s *Server) cachedListModels(ctx context.Context) ([]string, error) {
	modelCacheMu.Lock()
	defer modelCacheMu.Unlock()

	if modelCacheData != nil && time.Since(modelCacheAt) < modelCacheTTL {
		return modelCacheData, nil
	}

	models, err := s.br.ListModels(ctx)
	if err != nil {
		return nil, err
	}
	modelCacheData = models
	modelCacheAt = time.Now()
	return models, nil
}

func (s *Server) handleChatCompletions(w http.ResponseWriter, r *http.Request) {
	var req types.ChatCompletionRequest
	if err := json.NewDecoder(r.Body).Decode(&req); err != nil {
		writeJSON(w, http.StatusBadRequest, types.NewErrorResponse("invalid request body: "+err.Error(), "invalid_request_error", ""))
		return
	}
	defer r.Body.Close()

	if len(req.Messages) == 0 {
		writeJSON(w, http.StatusBadRequest, types.NewErrorResponse("messages must not be empty", "invalid_request_error", ""))
		return
	}

	// --- Pure brain: only our system prompt, drop the client's ---
	var parts []string
	if s.cfg.SystemPrompt != "" {
		parts = append(parts, "system: "+s.cfg.SystemPrompt)
	}
	for _, m := range req.Messages {
		// Drop client system messages (mode descriptions, tool schemas).
		if m.Role == "system" {
			continue
		}
		text := sanitize.Text(string(m.Content))
		// Strip <system-reminder> blocks embedded in messages.
		text = systemReminderRe.ReplaceAllString(text, "")
		text = strings.TrimSpace(text)
		if text == "" {
			continue
		}
		parts = append(parts, fmt.Sprintf("%s: %s", m.Role, text))
	}
	prompt := strings.Join(parts, "\n")

	model := req.Model
	if model == "" {
		model = s.cfg.DefaultModel
	}
	cursorModel := converter.ResolveToCursorModel(model)
	sessionKey := ensureSessionHeader(w, r)

	if r.Header.Get(workspaceHeaderName) == "" {
		if detected := detectOpenAICwd(req); detected != "" {
			slog.Debug("workspace detected from prompt", "path", detected)
			r.Header.Set(workspaceHeaderName, detected)
		}
	}

	chatID := fmt.Sprintf("chatcmpl-%d", time.Now().UnixNano())
	created := time.Now().Unix()

	if req.Stream {
		s.streamChat(w, r, prompt, cursorModel, model, chatID, created, sessionKey)
		return
	}

	s.nonStreamChat(w, r, prompt, cursorModel, model, chatID, created, sessionKey)
}

func (s *Server) streamChat(w http.ResponseWriter, r *http.Request, prompt, cursorModel, displayModel, chatID string, created int64, sessionKey string) {
	sse := NewSSEWriter(w)
	parser := converter.NewStreamParser(chatID)

	ctx, cancel := context.WithTimeout(requestContext(r), time.Duration(s.cfg.Timeout)*time.Second)
	defer cancel()
	go func() {
		<-r.Context().Done()
		cancel()
	}()

	outputChan, errChan := s.br.Execute(ctx, prompt, cursorModel, sessionKey)

	roleAssistant := "assistant"
	initChunk := types.NewChatCompletionChunk(chatID, created, displayModel, types.Delta{
		Role: &roleAssistant,
	})
	if err := sse.WriteChunk(initChunk); err != nil {
		return
	}

	var accumulated strings.Builder

	for line := range outputChan {
		result := parser.Parse(line)

		if result.Skip {
			continue
		}

		if result.Error != nil {
			if strings.Contains(result.Error.Error(), "unmarshal error") {
				result = parser.ParseRawText(line)
				if result.Skip {
					continue
				}
				if result.Chunk != nil {
					result.Chunk.Created = created
					result.Chunk.Model = displayModel
					if c := result.Chunk.Choices[0].Delta.Content; c != nil {
						accumulated.WriteString(*c)
					}
					if err := sse.WriteChunk(*result.Chunk); err != nil {
						return
					}
					continue
				}
			}
			sse.WriteError(result.Error.Error())
			return
		}

		if result.Chunk != nil {
			result.Chunk.Created = created
			result.Chunk.Model = displayModel
			if len(result.Chunk.Choices) > 0 {
				if c := result.Chunk.Choices[0].Delta.Content; c != nil {
					accumulated.WriteString(*c)
				}
			}
			if err := sse.WriteChunk(*result.Chunk); err != nil {
				return
			}
		}

		if result.Done {
			break
		}
	}

	promptTokens := maxInt(1, int(math.Round(float64(len(prompt))/4.0)))
	completionTokens := maxInt(1, int(math.Round(float64(accumulated.Len())/4.0)))
	usage := &types.Usage{
		PromptTokens:     promptTokens,
		CompletionTokens: completionTokens,
		TotalTokens:      promptTokens + completionTokens,
	}

	select {
	case err := <-errChan:
		if err != nil {
			slog.Error("stream bridge error", "err", err)
			sse.WriteError(err.Error())
			return
		}
	default:
	}

	stopReason := "stop"
	finalChunk := types.NewChatCompletionChunk(chatID, created, displayModel, types.Delta{})
	finalChunk.Choices[0].FinishReason = &stopReason
	finalChunk.Usage = usage
	sse.WriteChunk(finalChunk)
	sse.WriteDone()
}

func (s *Server) nonStreamChat(w http.ResponseWriter, r *http.Request, prompt, cursorModel, displayModel, chatID string, created int64, sessionKey string) {
	ctx, cancel := context.WithTimeout(requestContext(r), time.Duration(s.cfg.Timeout)*time.Second)
	defer cancel()
	go func() {
		<-r.Context().Done()
		cancel()
	}()

	content, err := s.br.ExecuteSync(ctx, prompt, cursorModel, sessionKey)
	if err != nil {
		writeJSON(w, http.StatusInternalServerError, types.NewErrorResponse(err.Error(), "internal_error", ""))
		return
	}

	usage := estimateUsage(prompt, content)
	stopReason := "stop"

	resp := types.ChatCompletionResponse{
		ID:      chatID,
		Object:  "chat.completion",
		Created: created,
		Model:   displayModel,
		Choices: []types.Choice{
			{
				Index:        0,
				Message:      types.ChatMessage{Role: "assistant", Content: types.ChatMessageContent(content)},
				FinishReason: &stopReason,
			},
		},
		Usage: usage,
	}
	writeJSON(w, http.StatusOK, resp)
}

func estimateUsage(prompt, content string) types.Usage {
	promptTokens := maxInt(1, int(math.Round(float64(len(prompt))/4.0)))
	completionTokens := maxInt(1, int(math.Round(float64(len(content))/4.0)))
	return types.Usage{
		PromptTokens:     promptTokens,
		CompletionTokens: completionTokens,
		TotalTokens:      promptTokens + completionTokens,
	}
}

func maxInt(a, b int) int {
	if a > b {
		return a
	}
	return b
}

func (s *Server) handleHealth(w http.ResponseWriter, r *http.Request) {
	ctx, cancel := context.WithTimeout(r.Context(), 5*time.Second)
	defer cancel()

	status := "ok"
	cliStatus := "available"
	if err := s.br.CheckHealth(ctx); err != nil {
		cliStatus = fmt.Sprintf("unavailable: %v", err)
	}

	writeJSON(w, http.StatusOK, map[string]string{
		"status":     status,
		"cursor_cli": cliStatus,
		"version":    "0.2.0",
	})
}

func writeJSON(w http.ResponseWriter, status int, v interface{}) {
	w.Header().Set("Content-Type", "application/json")
	w.WriteHeader(status)
	json.NewEncoder(w).Encode(v)
}