LaodingBot/internal/transport/feishu/bot.go

package feishu

import (
	"context"
	"encoding/json"
	"fmt"
	"io"
	"mime"
	"os"
	"path/filepath"
	"strings"
	"sync"
	"time"

	"laodingbot/internal/logger"

	lark "github.com/larksuite/oapi-sdk-go/v3"
	larkcore "github.com/larksuite/oapi-sdk-go/v3/core"
	"github.com/larksuite/oapi-sdk-go/v3/event/dispatcher"
	larkim "github.com/larksuite/oapi-sdk-go/v3/service/im/v1"
	larkws "github.com/larksuite/oapi-sdk-go/v3/ws"
)

type Bot struct {
	appID       string
	appSecret   string
	verifyToken string

	apiClient *lark.Client
	log       *logger.Logger

	dedupTTL  time.Duration
	dedupMu   sync.Mutex
	dedupSeen map[string]time.Time
}

type IncomingMessage struct {
	MessageID string
	ChatID    string
	UserID    string
	MsgType   string
	Text      string
	FileName  string
	FileKey   string
	FileMime  string
	FileBytes []byte
	FilePath  string
}

const maxFeishuFileBytes = 20 * 1024 * 1024

func NewBot(appID, appSecret, verifyToken, _ string, _ string, log *logger.Logger) (*Bot, error) {
	if appID == "" || appSecret == "" {
		return nil, fmt.Errorf("empty feishu app credentials")
	}

	return &Bot{
		appID:       appID,
		appSecret:   appSecret,
		verifyToken: verifyToken,
		apiClient: lark.NewClient(appID, appSecret,
			lark.WithLogLevel(toLarkLogLevel(log)),
			lark.WithReqTimeout(10*time.Second),
			lark.WithEnableTokenCache(true),
		),
		log:       log,
		dedupTTL:  10 * time.Minute,
		dedupSeen: make(map[string]time.Time),
	}, nil
}

func (b *Bot) Run(ctx context.Context, handler func(context.Context, IncomingMessage) (string, error)) error {
	if b.log != nil {
		b.log.Infof("feishu websocket transport started")
	}
	eventHandler := dispatcher.NewEventDispatcher(b.verifyToken, "").
		OnP2MessageReceiveV1(func(evtCtx context.Context, event *larkim.P2MessageReceiveV1) error {
			incoming, ok := parseIncoming(event)
			if !ok {
				if b.log != nil {
					b.log.Debugf("skip unsupported or invalid feishu event")
				}
				return nil
			}
			if !b.shouldProcessMessage(incoming.MessageID) {
				if b.log != nil {
					b.log.Warnf("skip duplicated feishu message message_id=%s chat_id=%s", incoming.MessageID, incoming.ChatID)
				}
				return nil
			}
			if incoming.MsgType == "file" {
				b.enrichFileIncoming(evtCtx, &incoming)
			}
			if b.log != nil {
				b.log.Infof("feishu message received message_id=%s chat_id=%s user_id=%s msg_type=%s text=%s", incoming.MessageID, incoming.ChatID, incoming.UserID, incoming.MsgType, incoming.Text)
			}
			reply, err := handler(evtCtx, incoming)
			if err != nil {
				if b.log != nil {
					b.log.Errorf("feishu handler failed chat_id=%s err=%v", incoming.ChatID, err)
				}
				return err
			}
			if strings.TrimSpace(reply) == "" {
				if b.log != nil {
					b.log.Debugf("feishu empty reply chat_id=%s", incoming.ChatID)
				}
				return nil
			}
			return b.sendText(evtCtx, incoming.ChatID, reply)
		})

	wsClient := larkws.NewClient(
		b.appID,
		b.appSecret,
		larkws.WithEventHandler(eventHandler),
		larkws.WithLogLevel(toLarkLogLevel(b.log)),
	)

	errCh := make(chan error, 1)
	go func() {
		errCh <- wsClient.Start(ctx)
	}()

	select {
	case <-ctx.Done():
		if b.log != nil {
			b.log.Infof("feishu websocket transport stopped: %v", ctx.Err())
		}
		return ctx.Err()
	case err := <-errCh:
		if err != nil && b.log != nil {
			b.log.Errorf("feishu websocket transport failed err=%v", err)
		}
		return err
	}
}

func (b *Bot) sendText(ctx context.Context, chatID, text string) error {
	resp, err := b.apiClient.Im.Message.Create(ctx, larkim.NewCreateMessageReqBuilder().
		ReceiveIdType("chat_id").
		Body(larkim.NewCreateMessageReqBodyBuilder().
			ReceiveId(chatID).
			MsgType("text").
			Content(fmt.Sprintf(`{"text":%q}`, text)).
			Uuid(fmt.Sprintf("%d", time.Now().UnixNano())).
			Build()).
		Build())
	if err != nil {
		if b.log != nil {
			b.log.Errorf("feishu send message request failed chat_id=%s err=%v", chatID, err)
		}
		return err
	}
	if !resp.Success() {
		if b.log != nil {
			b.log.Warnf("feishu send message unsuccessful chat_id=%s code=%d msg=%s", chatID, resp.Code, resp.Msg)
		}
		return fmt.Errorf("feishu send message failed: code=%d msg=%s log_id=%s", resp.Code, resp.Msg, resp.LogId())
	}
	if b.log != nil {
		b.log.Debugf("feishu message sent chat_id=%s text_len=%d", chatID, len(text))
	}
	return nil
}

func extractText(content string) (string, error) {
	var parsed struct {
		Text string `json:"text"`
	}
	if err := json.Unmarshal([]byte(content), &parsed); err != nil {
		return "", err
	}
	return parsed.Text, nil
}

func extractFileMeta(content string) (fileName string, fileKey string, err error) {
	var parsed map[string]any
	if err := json.Unmarshal([]byte(content), &parsed); err != nil {
		return "", "", err
	}

	readString := func(keys ...string) string {
		for _, key := range keys {
			if v, ok := parsed[key]; ok {
				s, ok := v.(string)
				if ok {
					trimmed := strings.TrimSpace(s)
					if trimmed != "" {
						return trimmed
					}
				}
			}
		}
		return ""
	}

	fileName = readString("file_name", "fileName", "name", "filename")
	fileKey = readString("file_key", "fileKey", "key")
	return fileName, fileKey, nil
}

func buildFileRecognitionText(fileName, fileKey string) string {
	if strings.TrimSpace(fileName) == "" {
		fileName = "(unknown)"
	}
	if strings.TrimSpace(fileKey) == "" {
		fileKey = "(unknown)"
	}

	return strings.Join([]string{
		"用户发送了一条飞书文件消息。",
		"文件名: " + fileName,
		"文件Key: " + fileKey,
		"系统将先上传该文件到 LLM Provider，再由模型完成文档解析。若上传失败，本次请求将直接中止。",
	}, "\n")
}

func (b *Bot) enrichFileIncoming(ctx context.Context, incoming *IncomingMessage) {
	if incoming == nil {
		return
	}
	if strings.TrimSpace(incoming.MessageID) == "" || strings.TrimSpace(incoming.FileKey) == "" {
		incoming.Text = buildFileRecognitionText(incoming.FileName, incoming.FileKey)
		incoming.Text += "\n\n未找到完整 file_key 或 message_id，暂时无法下载文件内容。"
		return
	}

	content, fileName, err := b.downloadFileContent(ctx, incoming.MessageID, incoming.FileKey)
	if err != nil {
		if b.log != nil {
			b.log.Warnf("feishu download file content failed message_id=%s file_key=%s err=%v", incoming.MessageID, incoming.FileKey, err)
		}
		incoming.Text = buildFileRecognitionText(incoming.FileName, incoming.FileKey)
		incoming.Text += "\n\n文件下载失败: " + err.Error()
		return
	}
	if strings.TrimSpace(fileName) != "" {
		incoming.FileName = fileName
	}
	incoming.FileBytes = content
	incoming.FileMime = detectMimeByName(incoming.FileName)
	localPath, saveErr := saveIncomingFile("files", incoming.FileName, incoming.FileBytes)
	if saveErr != nil {
		if b.log != nil {
			b.log.Warnf("save incoming feishu file failed name=%s err=%v", incoming.FileName, saveErr)
		}
		incoming.Text = buildFileRecognitionText(incoming.FileName, incoming.FileKey)
		incoming.Text += "\n\n文件已下载但本地保存失败: " + saveErr.Error()
		return
	}
	incoming.FilePath = localPath
	incoming.Text = buildFileRecognitionText(incoming.FileName, incoming.FileKey)
	incoming.Text += fmt.Sprintf("\n\n文件已下载并保存到本地，路径=%s，大小=%d bytes，mime=%s。", incoming.FilePath, len(content), incoming.FileMime)
}

func (b *Bot) downloadFileContent(ctx context.Context, messageID, fileKey string) ([]byte, string, error) {
	req := larkim.NewGetMessageResourceReqBuilder().
		MessageId(messageID).
		FileKey(fileKey).
		Type("file").
		Build()

	resp, err := b.apiClient.Im.MessageResource.Get(ctx, req)
	if err != nil {
		return nil, "", err
	}
	if resp == nil || resp.File == nil {
		if resp != nil {
			return nil, "", fmt.Errorf("empty file stream code=%d msg=%s", resp.Code, resp.Msg)
		}
		return nil, "", fmt.Errorf("empty file stream")
	}

	bts, err := io.ReadAll(resp.File)
	if err != nil {
		return nil, "", err
	}
	if len(bts) > maxFeishuFileBytes {
		return nil, "", fmt.Errorf("file too large: %d bytes, max=%d", len(bts), maxFeishuFileBytes)
	}
	return bts, strings.TrimSpace(resp.FileName), nil
}

func detectMimeByName(fileName string) string {
	ext := strings.ToLower(strings.TrimSpace(filepath.Ext(fileName)))
	if ext == "" {
		return "application/octet-stream"
	}
	m := strings.TrimSpace(mime.TypeByExtension(ext))
	if m == "" {
		return "application/octet-stream"
	}
	return m
}

func saveIncomingFile(baseDir, fileName string, content []byte) (string, error) {
	if len(content) == 0 {
		return "", fmt.Errorf("empty file content")
	}
	if strings.TrimSpace(baseDir) == "" {
		baseDir = "files"
	}
	if err := os.MkdirAll(baseDir, 0o755); err != nil {
		return "", err
	}
	safeName := sanitizeFileName(fileName)
	if safeName == "" {
		safeName = "upload.bin"
	}
	finalName := fmt.Sprintf("%d_%s", time.Now().UnixNano(), safeName)
	target := filepath.Join(baseDir, finalName)
	if err := os.WriteFile(target, content, 0o644); err != nil {
		return "", err
	}
	abs, err := filepath.Abs(target)
	if err != nil {
		return target, nil
	}
	return abs, nil
}

func sanitizeFileName(fileName string) string {
	name := strings.TrimSpace(filepath.Base(fileName))
	if name == "" || name == "." || name == ".." {
		return ""
	}
	var b strings.Builder
	for _, r := range name {
		if (r >= 'a' && r <= 'z') || (r >= 'A' && r <= 'Z') || (r >= '0' && r <= '9') || r == '.' || r == '_' || r == '-' {
			b.WriteRune(r)
			continue
		}
		b.WriteByte('_')
	}
	out := strings.TrimSpace(b.String())
	if out == "" || out == "." || out == ".." {
		return ""
	}
	if strings.HasPrefix(out, ".") {
		out = "file" + out
	}
	return out
}

func parseIncoming(event *larkim.P2MessageReceiveV1) (IncomingMessage, bool) {
	if event == nil || event.Event == nil || event.Event.Message == nil || event.Event.Sender == nil || event.Event.Sender.SenderId == nil {
		return IncomingMessage{}, false
	}
	if event.Event.Sender.SenderType != nil && *event.Event.Sender.SenderType != "user" {
		return IncomingMessage{}, false
	}

	msg := event.Event.Message
	if msg.MessageType == nil || msg.ChatId == nil || msg.Content == nil || msg.MessageId == nil {
		return IncomingMessage{}, false
	}
	msgType := strings.TrimSpace(*msg.MessageType)
	if msgType == "" {
		return IncomingMessage{}, false
	}

	userID := ""
	if event.Event.Sender.SenderId.OpenId != nil {
		userID = *event.Event.Sender.SenderId.OpenId
	} else if event.Event.Sender.SenderId.UserId != nil {
		userID = *event.Event.Sender.SenderId.UserId
	} else if event.Event.Sender.SenderId.UnionId != nil {
		userID = *event.Event.Sender.SenderId.UnionId
	}

	incoming := IncomingMessage{
		MessageID: *msg.MessageId,
		ChatID:    *msg.ChatId,
		UserID:    userID,
		MsgType:   msgType,
	}

	switch msgType {
	case "text":
		text, err := extractText(*msg.Content)
		if err != nil {
			return IncomingMessage{}, false
		}
		incoming.Text = text
		return incoming, true
	case "file":
		fileName, fileKey, err := extractFileMeta(*msg.Content)
		if err != nil {
			return IncomingMessage{}, false
		}
		incoming.FileName = fileName
		incoming.FileKey = fileKey
		incoming.Text = buildFileRecognitionText(fileName, fileKey)
		return incoming, true
	default:
		return IncomingMessage{}, false
	}
}

func (b *Bot) shouldProcessMessage(messageID string) bool {
	messageID = strings.TrimSpace(messageID)
	if messageID == "" {
		if b.log != nil {
			b.log.Warnf("feishu message without message_id; skip idempotency check")
		}
		return true
	}

	now := time.Now()
	b.dedupMu.Lock()
	defer b.dedupMu.Unlock()

	for id, seenAt := range b.dedupSeen {
		if now.Sub(seenAt) > b.dedupTTL {
			delete(b.dedupSeen, id)
		}
	}

	if _, exists := b.dedupSeen[messageID]; exists {
		return false
	}
	b.dedupSeen[messageID] = now
	return true
}

func toLarkLogLevel(log *logger.Logger) larkcore.LogLevel {
	if log == nil {
		return larkcore.LogLevelInfo
	}
	switch log.Level() {
	case logger.LevelDebug:
		return larkcore.LogLevelDebug
	case logger.LevelWarn:
		return larkcore.LogLevelWarn
	case logger.LevelError:
		return larkcore.LogLevelError
	default:
		return larkcore.LogLevelInfo
	}
}