feat: 完善爬虫去重、上传进度和源文件删除

为脚本爬虫增加候选预算、重复 source 记录和默认爬虫标签，避免重复视频占满目标新增数量。

新增爬虫上传迁移进度上报和管理页上传卡片，让每个爬虫可以展示本轮上传处理情况。

为视频删除增加可选删除云盘源文件能力，补齐播放页、管理页交互，并为多个网盘驱动实现 Remove 接口。

补充相关测试并更新爬虫协议文档。

.gitignore

@@ -41,4 +41,5 @@ __pycache__/
 /image003.jpg
 /image004.jpg
 /image005.png
+/image006.png
 /image02.png

backend/cmd/server/main.go

@@ -86,6 +86,7 @@ func main() {
 		Registry:         app.registry,
 		GetTargetDriveID: func() string { return app.Spider91UploadDriveID() },
 		CommonThumbDir:   app.commonThumbsDir(),
+		OnUploadProgress: app.updateCrawlerUploadProgress,
 	})
 
 	// 初始化本地内置盘；外部云盘放到 HTTP 服务启动后异步挂载，避免上游
@@ -217,8 +218,8 @@ func main() {
 		OnRegenFailedFingerprints: func(driveID string) {
 			go app.regenFailedFingerprints(ctx, driveID)
 		},
-		OnDeleteVideo: func(reqCtx context.Context, videoID string) (api.DeleteVideoResult, error) {
-			return app.deleteVideo(reqCtx, videoID)
+		OnDeleteVideo: func(reqCtx context.Context, videoID string, deleteSource bool) (api.DeleteVideoResult, error) {
+			return app.deleteVideo(reqCtx, videoID, deleteSource)
 		},
 		GetDriveGenerationStatuses: func() map[string]api.DriveGenerationStatuses {
 			return app.driveGenerationStatuses()
@@ -363,6 +364,10 @@ type App struct {
 	// crawlerUploadRunning 去重"保存上传目标后检查本地未上传文件"的后台任务。
 	crawlerUploadMu      sync.Mutex
 	crawlerUploadRunning map[string]bool
+
+	// uploadProgress 跟踪脚本爬虫迁移到云盘时的实时上传状态。
+	uploadProgressMu sync.Mutex
+	uploadProgress   map[string]driveUploadProgress
 }
 
 type driveScanProgress struct {
@@ -370,6 +375,14 @@ type driveScanProgress struct {
 	Added   int
 }
 
+type driveUploadProgress struct {
+	State        string
+	CurrentTitle string
+	QueueLength  int
+	DoneCount    int
+	TotalCount   int
+}
+
 type spider91MigrationRunner interface {
 	RunOnce(ctx context.Context) error
 }
@@ -522,6 +535,13 @@ func (a *App) driveGenerationStatuses() map[string]api.DriveGenerationStatuses {
 	}
 	a.scanQueueMu.Unlock()
 
+	a.uploadProgressMu.Lock()
+	uploadProgresses := make(map[string]driveUploadProgress, len(a.uploadProgress))
+	for id, progress := range a.uploadProgress {
+		uploadProgresses[id] = progress
+	}
+	a.uploadProgressMu.Unlock()
+
 	a.mu.Lock()
 	previewWorkers := make(map[string]*preview.Worker, len(a.workers))
 	for id, worker := range a.workers {
@@ -537,7 +557,7 @@ func (a *App) driveGenerationStatuses() map[string]api.DriveGenerationStatuses {
 	}
 	a.mu.Unlock()
 
-	out := make(map[string]api.DriveGenerationStatuses, len(scanningDrives)+len(previewWorkers)+len(thumbWorkers)+len(fingerprintWorkers))
+	out := make(map[string]api.DriveGenerationStatuses, len(scanningDrives)+len(previewWorkers)+len(thumbWorkers)+len(fingerprintWorkers)+len(uploadProgresses))
 	for id, running := range scanningDrives {
 		if !running {
 			continue
@@ -566,9 +586,75 @@ func (a *App) driveGenerationStatuses() map[string]api.DriveGenerationStatuses {
 		status.Fingerprint = generationStatusFromFingerprint(worker.Status())
 		out[id] = status
 	}
+	for id, progress := range uploadProgresses {
+		state := progress.State
+		if state == "" {
+			state = "idle"
+		}
+		status := out[id]
+		status.Upload = api.GenerationStatus{
+			State:        state,
+			CurrentTitle: progress.CurrentTitle,
+			QueueLength:  progress.QueueLength,
+			DoneCount:    progress.DoneCount,
+			TotalCount:   progress.TotalCount,
+		}
+		out[id] = status
+	}
 	return out
 }
 
+func (a *App) updateCrawlerUploadProgress(progress spider91migrate.UploadProgress) {
+	driveID := strings.TrimSpace(progress.DriveID)
+	if driveID == "" {
+		return
+	}
+	state := strings.TrimSpace(progress.State)
+	if state == "" {
+		state = "idle"
+	}
+	a.uploadProgressMu.Lock()
+	if a.uploadProgress == nil {
+		a.uploadProgress = make(map[string]driveUploadProgress)
+	}
+	if state == "idle" {
+		delete(a.uploadProgress, driveID)
+		a.uploadProgressMu.Unlock()
+		return
+	}
+	a.uploadProgress[driveID] = driveUploadProgress{
+		State:        state,
+		CurrentTitle: strings.TrimSpace(progress.CurrentTitle),
+		QueueLength:  progress.QueueLength,
+		DoneCount:    progress.DoneCount,
+		TotalCount:   progress.TotalCount,
+	}
+	a.uploadProgressMu.Unlock()
+}
+
+func (a *App) clearCrawlerUploadProgress(driveID string) bool {
+	driveID = strings.TrimSpace(driveID)
+	if driveID == "" {
+		return false
+	}
+	a.uploadProgressMu.Lock()
+	_, ok := a.uploadProgress[driveID]
+	delete(a.uploadProgress, driveID)
+	a.uploadProgressMu.Unlock()
+	return ok
+}
+
+func (a *App) clearAllCrawlerUploadProgress() []string {
+	a.uploadProgressMu.Lock()
+	ids := make([]string, 0, len(a.uploadProgress))
+	for id := range a.uploadProgress {
+		ids = append(ids, id)
+	}
+	a.uploadProgress = nil
+	a.uploadProgressMu.Unlock()
+	return ids
+}
+
 func generationStatusFromPreview(status preview.TaskStatus) api.GenerationStatus {
 	state := status.State
 	if state == "" {
@@ -905,6 +991,7 @@ func (a *App) attachScriptCrawler(d *catalog.Drive, drv *scriptcrawler.Driver) {
 	c := scriptcrawler.NewCrawler(scriptcrawler.CrawlerConfig{
 		Driver:         drv,
 		Catalog:        a.cat,
+		CrawlerName:    d.Name,
 		SourceKind:     sourceKind,
 		PythonPath:     pythonPath,
 		ScriptPath:     scriptPath,
@@ -929,6 +1016,7 @@ func (a *App) attachScriptCrawler(d *catalog.Drive, drv *scriptcrawler.Driver) {
 	a.scriptCrawlers[driveID] = c
 	a.mu.Unlock()
 
+	a.ensureScriptCrawlerNameTag(driveID, sourceKind, d.Name)
 	if sourceKind == spider91.Kind {
 		a.ensureSpider91SourceTag(driveID)
 	}
@@ -959,6 +1047,24 @@ func (a *App) ensureSpider91SourceTag(driveID string) {
 	}()
 }
 
+func (a *App) ensureScriptCrawlerNameTag(driveID, sourceKind, crawlerName string) {
+	tagName := strings.TrimSpace(crawlerName)
+	if tagName == "" {
+		tagName = strings.TrimSpace(driveID)
+	}
+	if tagName == "" {
+		return
+	}
+	bgCtx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
+	go func() {
+		defer cancel()
+		prefix := scriptcrawler.BuildVideoIDForKind(sourceKind, driveID, "")
+		if _, err := a.cat.EnsureTagForVideoIDPrefix(bgCtx, prefix, tagName, nil, "legacy"); err != nil {
+			log.Printf("[scriptcrawler] drive=%s ensure crawler tag %q: %v", driveID, tagName, err)
+		}
+	}()
+}
+
 func (a *App) registerPreviewWorkers(ctx context.Context, driveID string, worker *preview.Worker, thumbWorker *preview.ThumbWorker, fingerprintWorker *fingerprint.Worker, cancel context.CancelFunc) {
 	a.registerPreviewWorkersWithOptions(ctx, driveID, worker, thumbWorker, fingerprintWorker, cancel, true)
 }
@@ -1185,6 +1291,13 @@ func (a *App) driveHasActiveWork(driveID string) bool {
 		return true
 	}
 
+	a.uploadProgressMu.Lock()
+	uploading := a.uploadProgress[driveID].State != ""
+	a.uploadProgressMu.Unlock()
+	if uploading {
+		return true
+	}
+
 	a.mu.Lock()
 	previewWorker := a.workers[driveID]
 	thumbWorker := a.thumbWorkers[driveID]
@@ -1304,10 +1417,11 @@ func (a *App) stopDriveTasks(ctx context.Context, driveID string) bool {
 	canceled := a.cancelDriveTaskContexts(driveID)
 	queued := a.clearQueuedDriveTask(driveID)
 	fingerprintQueued := a.clearFingerprintQueueing(driveID)
+	uploading := a.clearCrawlerUploadProgress(driveID)
 	hadWorkers := a.resetDriveGenerationWorkers(ctx, driveID)
-	stopped := canceled > 0 || queued || fingerprintQueued || hadWorkers
-	log.Printf("[tasks] stop drive=%s stopped=%v canceled_tasks=%d queued=%v fingerprint_queue=%v workers=%v",
-		driveID, stopped, canceled, queued, fingerprintQueued, hadWorkers)
+	stopped := canceled > 0 || queued || fingerprintQueued || uploading || hadWorkers
+	log.Printf("[tasks] stop drive=%s stopped=%v canceled_tasks=%d queued=%v fingerprint_queue=%v uploading=%v workers=%v",
+		driveID, stopped, canceled, queued, fingerprintQueued, uploading, hadWorkers)
 	return stopped
 }
 
@@ -1325,6 +1439,9 @@ func (a *App) stopAllDriveTasks(ctx context.Context) int {
 	for _, id := range a.clearAllFingerprintQueueing() {
 		stoppedIDs[id] = struct{}{}
 	}
+	for _, id := range a.clearAllCrawlerUploadProgress() {
+		stoppedIDs[id] = struct{}{}
+	}
 	for _, id := range a.resetAllDriveGenerationWorkers(ctx) {
 		stoppedIDs[id] = struct{}{}
 	}
@@ -1679,7 +1796,7 @@ func (a *App) cleanupMissingDriveVideos(ctx context.Context, driveID string, liv
 	return removed, nil
 }
 
-func (a *App) deleteVideo(ctx context.Context, videoID string) (api.DeleteVideoResult, error) {
+func (a *App) deleteVideo(ctx context.Context, videoID string, deleteSource bool) (api.DeleteVideoResult, error) {
 	if a == nil || a.cat == nil {
 		return api.DeleteVideoResult{}, sql.ErrNoRows
 	}
@@ -1688,6 +1805,14 @@ func (a *App) deleteVideo(ctx context.Context, videoID string) (api.DeleteVideoR
 		return api.DeleteVideoResult{}, err
 	}
 
+	deletedSource := false
+	if deleteSource {
+		deletedSource, err = a.removeVideoSourceFile(ctx, v)
+		if err != nil {
+			return api.DeleteVideoResult{}, err
+		}
+	}
+
 	localDir := ""
 	if a.cfg != nil {
 		localDir = a.cfg.Storage.LocalPreviewDir
@@ -1695,16 +1820,61 @@ func (a *App) deleteVideo(ctx context.Context, videoID string) (api.DeleteVideoR
 	if err := removeLocalVideoAssets(localDir, v); err != nil {
 		return api.DeleteVideoResult{}, fmt.Errorf("remove local assets for %s: %w", v.ID, err)
 	}
-	deletedSource, err := a.removeSpider91SourceFile(ctx, v)
-	if err != nil {
-		return api.DeleteVideoResult{}, err
-	}
 	if err := a.cat.DeleteVideoWithTombstone(ctx, v.ID); err != nil {
 		return api.DeleteVideoResult{}, err
 	}
 	return api.DeleteVideoResult{OK: true, DeletedSource: deletedSource}, nil
 }
 
+func (a *App) removeVideoSourceFile(ctx context.Context, v *catalog.Video) (bool, error) {
+	if v == nil {
+		return false, errors.New("remove video source: empty video")
+	}
+	if a == nil {
+		return false, fmt.Errorf("remove video source %s: app unavailable: %w", v.ID, drives.ErrNotSupported)
+	}
+	if strings.HasPrefix(v.ID, "spider91-") {
+		deleted, err := a.removeSpider91SourceFile(ctx, v)
+		if err != nil || deleted {
+			return deleted, err
+		}
+		if a.cat != nil {
+			if drive, driveErr := a.cat.GetDrive(ctx, v.DriveID); driveErr == nil && drive.Kind == spider91.Kind {
+				return false, nil
+			}
+		} else if strings.HasPrefix(v.ID, "spider91-"+v.DriveID+"-") {
+			return false, nil
+		}
+	}
+	fileID := strings.TrimSpace(v.FileID)
+	if fileID == "" {
+		return false, fmt.Errorf("remove video source %s: empty file id", v.ID)
+	}
+	if a == nil || a.registry == nil {
+		return false, fmt.Errorf("remove video source %s: drive registry unavailable: %w", v.ID, drives.ErrNotSupported)
+	}
+	if _, ok := a.registry.Get(v.DriveID); !ok {
+		if a.cat == nil {
+			return false, fmt.Errorf("remove video source %s: drive %s not attached: %w", v.ID, v.DriveID, drives.ErrNotSupported)
+		}
+		if err := a.ensureDriveAttached(ctx, v.DriveID); err != nil {
+			return false, fmt.Errorf("remove video source %s: attach drive %s: %w", v.ID, v.DriveID, err)
+		}
+	}
+	drv, ok := a.registry.Get(v.DriveID)
+	if !ok {
+		return false, fmt.Errorf("remove video source %s: drive %s not attached: %w", v.ID, v.DriveID, drives.ErrNotSupported)
+	}
+	remover, ok := drv.(drives.Remover)
+	if !ok {
+		return false, fmt.Errorf("remove video source %s: drive %s (%s) does not support source deletion: %w", v.ID, v.DriveID, drv.Kind(), drives.ErrNotSupported)
+	}
+	if err := remover.Remove(ctx, fileID); err != nil {
+		return false, fmt.Errorf("remove video source %s from drive %s: %w", v.ID, v.DriveID, err)
+	}
+	return true, nil
+}
+
 func (a *App) removeSpider91SourceFile(ctx context.Context, v *catalog.Video) (bool, error) {
 	if a == nil || a.cfg == nil || v == nil || !strings.HasPrefix(v.ID, "spider91-") {
 		return false, nil
@@ -2642,8 +2812,8 @@ func (a *App) runScriptCrawlerCrawlWithTaskContext(ctx context.Context, driveID
 	if runErr != nil {
 		log.Printf("[scriptcrawler] drive=%s crawl failed: %v", driveID, runErr)
 	} else if res != nil {
-		log.Printf("[scriptcrawler] drive=%s crawl done target=%d total=%d new=%d skipped=%d failed=%d seen_snapshot=%d",
-			driveID, res.TargetNew, res.TotalEntries, res.NewVideos, res.Skipped, res.Failed, res.SeenSnapshot)
+		log.Printf("[scriptcrawler] drive=%s crawl done target=%d candidate_budget=%d total=%d new=%d skipped=%d failed=%d seen_snapshot=%d",
+			driveID, res.TargetNew, res.CandidateBudget, res.TotalEntries, res.NewVideos, res.Skipped, res.Failed, res.SeenSnapshot)
 	}
 
 	if d.Credentials == nil {

backend/cmd/server/main_test.go

@@ -1243,7 +1243,7 @@ func TestDeleteVideoRemovesGeneratedAssetsKeepsLocalOriginalAndTombstones(t *tes
 		cfg: &config.Config{Storage: config.Storage{LocalPreviewDir: localDir}},
 		cat: cat,
 	}
-	result, err := app.deleteVideo(ctx, "localstorage-local-main-file")
+	result, err := app.deleteVideo(ctx, "localstorage-local-main-file", false)
 	if err != nil {
 		t.Fatalf("delete video: %v", err)
 	}
@@ -1270,6 +1270,73 @@ func TestDeleteVideoRemovesGeneratedAssetsKeepsLocalOriginalAndTombstones(t *tes
 	}
 }
 
+func TestDeleteVideoRemovesSourceFileWhenRequested(t *testing.T) {
+	ctx := context.Background()
+	root := t.TempDir()
+	localDir := filepath.Join(root, "previews")
+	cat, err := catalog.Open(filepath.Join(t.TempDir(), "catalog.db"))
+	if err != nil {
+		t.Fatalf("open catalog: %v", err)
+	}
+	t.Cleanup(func() { _ = cat.Close() })
+
+	previewPath := filepath.Join(localDir, "video-with-source.mp4")
+	thumbPath := filepath.Join(localDir, "thumbs", "video-with-source.jpg")
+	for _, path := range []string{previewPath, thumbPath} {
+		if err := os.MkdirAll(filepath.Dir(path), 0o755); err != nil {
+			t.Fatalf("mkdir %s: %v", path, err)
+		}
+		if err := os.WriteFile(path, []byte("file"), 0o644); err != nil {
+			t.Fatalf("write %s: %v", path, err)
+		}
+	}
+
+	now := time.Now()
+	if err := cat.UpsertVideo(ctx, &catalog.Video{
+		ID:            "video-with-source",
+		DriveID:       "source-drive",
+		FileID:        "source-file",
+		FileName:      "clip.mp4",
+		Title:         "Source File",
+		PreviewLocal:  previewPath,
+		PreviewStatus: "ready",
+		ThumbnailURL:  "/p/thumb/video-with-source",
+		Size:          123,
+		PublishedAt:   now,
+		CreatedAt:     now,
+		UpdatedAt:     now,
+	}); err != nil {
+		t.Fatalf("seed video: %v", err)
+	}
+
+	registry := proxy.NewRegistry()
+	drv := &serverRemovableFakeDrive{id: "source-drive"}
+	registry.Set(drv.ID(), drv)
+	app := &App{
+		cfg:      &config.Config{Storage: config.Storage{LocalPreviewDir: localDir}},
+		cat:      cat,
+		registry: registry,
+	}
+	result, err := app.deleteVideo(ctx, "video-with-source", true)
+	if err != nil {
+		t.Fatalf("delete video: %v", err)
+	}
+	if !result.OK || !result.DeletedSource {
+		t.Fatalf("delete result = %#v, want source deleted", result)
+	}
+	if got, want := drv.removedFileID, "source-file"; got != want {
+		t.Fatalf("removed source fileID = %q, want %q", got, want)
+	}
+	if _, err := cat.GetVideo(ctx, "video-with-source"); err != sql.ErrNoRows {
+		t.Fatalf("deleted video lookup error = %v, want sql.ErrNoRows", err)
+	}
+	for _, path := range []string{previewPath, thumbPath} {
+		if _, err := os.Stat(path); !os.IsNotExist(err) {
+			t.Fatalf("generated asset %s still exists, stat err=%v", path, err)
+		}
+	}
+}
+
 func TestDeleteVideoRemovesSpider91SourceFile(t *testing.T) {
 	ctx := context.Background()
 	root := t.TempDir()
@@ -1326,7 +1393,7 @@ func TestDeleteVideoRemovesSpider91SourceFile(t *testing.T) {
 		t.Fatalf("seed video: %v", err)
 	}
 
-	result, err := app.deleteVideo(ctx, "spider91-spider-main-source")
+	result, err := app.deleteVideo(ctx, "spider91-spider-main-source", true)
 	if err != nil {
 		t.Fatalf("delete spider video: %v", err)
 	}
@@ -1740,6 +1807,22 @@ type serverFakeKindDrive struct {
 func (d *serverFakeKindDrive) Kind() string { return d.kind }
 func (d *serverFakeKindDrive) ID() string   { return d.id }
 
+type serverRemovableFakeDrive struct {
+	serverFakeDrive
+	id            string
+	removedFileID string
+}
+
+func (d *serverRemovableFakeDrive) Kind() string { return "fake-removable" }
+func (d *serverRemovableFakeDrive) ID() string   { return d.id }
+func (d *serverRemovableFakeDrive) Remove(ctx context.Context, fileID string) error {
+	if err := ctx.Err(); err != nil {
+		return err
+	}
+	d.removedFileID = fileID
+	return nil
+}
+
 type serverFakeSpider91MigrationRunner struct {
 	called int
 }

backend/internal/api/admin.go

@@ -57,7 +57,7 @@ type AdminServer struct {
 	OnRegenFailedPreviews      func(driveID string)
 	OnRegenFailedThumbnails    func(driveID string)
 	OnRegenFailedFingerprints  func(driveID string)
-	OnDeleteVideo              func(ctx context.Context, videoID string) (DeleteVideoResult, error)
+	OnDeleteVideo              func(ctx context.Context, videoID string, deleteSource bool) (DeleteVideoResult, error)
 	GetDriveGenerationStatuses func() map[string]DriveGenerationStatuses
 	// OnTeaserEnabledChanged 在 per-drive 预览视频开关被切换后调用。
 	// enabled=true 时上层应该重新把 pending 预览视频入队（类似旧的全局开关从关到开）；
@@ -103,6 +103,8 @@ type GenerationStatus struct {
 	CooldownUntil string `json:"cooldownUntil,omitempty"`
 	ScannedCount  int    `json:"scannedCount"`
 	AddedCount    int    `json:"addedCount"`
+	DoneCount     int    `json:"doneCount"`
+	TotalCount    int    `json:"totalCount"`
 }
 
 type DriveGenerationStatuses struct {
@@ -110,6 +112,7 @@ type DriveGenerationStatuses struct {
 	Thumbnail   GenerationStatus `json:"thumbnail"`
 	Preview     GenerationStatus `json:"preview"`
 	Fingerprint GenerationStatus `json:"fingerprint"`
+	Upload      GenerationStatus `json:"upload"`
 }
 
 type NightlyJobStatus struct {
@@ -127,6 +130,10 @@ type DeleteVideoResult struct {
 	DeletedSource bool `json:"deletedSource"`
 }
 
+type deleteVideoReq struct {
+	DeleteSource bool `json:"deleteSource"`
+}
+
 func (a *AdminServer) Register(r chi.Router) {
 	r.Route("/admin/api", func(r chi.Router) {
 		// 登录、登出和首次部署初始化不需要鉴权
@@ -637,6 +644,7 @@ type crawlerDTO struct {
 	ThumbnailGenerationStatus   GenerationStatus `json:"thumbnailGenerationStatus"`
 	PreviewGenerationStatus     GenerationStatus `json:"previewGenerationStatus"`
 	FingerprintGenerationStatus GenerationStatus `json:"fingerprintGenerationStatus"`
+	UploadGenerationStatus      GenerationStatus `json:"uploadGenerationStatus"`
 	ThumbnailReadyCount         int              `json:"thumbnailReadyCount"`
 	ThumbnailPendingCount       int              `json:"thumbnailPendingCount"`
 	ThumbnailFailedCount        int              `json:"thumbnailFailedCount"`
@@ -698,6 +706,9 @@ func (a *AdminServer) crawlerDTOForDrive(d *catalog.Drive, assets catalog.Crawle
 	if generation.Fingerprint.State == "" {
 		generation.Fingerprint.State = "idle"
 	}
+	if generation.Upload.State == "" {
+		generation.Upload.State = "idle"
+	}
 	lastCrawlAt := int64(0)
 	if raw := strings.TrimSpace(d.Credentials["last_crawl_at"]); raw != "" {
 		if v, err := strconv.ParseInt(raw, 10, 64); err == nil {
@@ -719,6 +730,7 @@ func (a *AdminServer) crawlerDTOForDrive(d *catalog.Drive, assets catalog.Crawle
 		ThumbnailGenerationStatus:   generation.Thumbnail,
 		PreviewGenerationStatus:     generation.Preview,
 		FingerprintGenerationStatus: generation.Fingerprint,
+		UploadGenerationStatus:      generation.Upload,
 		ThumbnailReadyCount:         assets.Thumbnail.Ready,
 		ThumbnailPendingCount:       assets.Thumbnail.Pending,
 		ThumbnailFailedCount:        assets.Thumbnail.Failed,
@@ -1824,12 +1836,21 @@ func (a *AdminServer) handleDeleteVideo(w http.ResponseWriter, r *http.Request)
 		writeErr(w, http.StatusBadRequest, errors.New("invalid video id"))
 		return
 	}
+	var body deleteVideoReq
+	if r.Body != nil {
+		defer r.Body.Close()
+		decoder := json.NewDecoder(r.Body)
+		if err := decoder.Decode(&body); err != nil && !errors.Is(err, io.EOF) {
+			writeErr(w, http.StatusBadRequest, err)
+			return
+		}
+	}
 	var (
 		result DeleteVideoResult
 		err    error
 	)
 	if a.OnDeleteVideo != nil {
-		result, err = a.OnDeleteVideo(r.Context(), id)
+		result, err = a.OnDeleteVideo(r.Context(), id, body.DeleteSource)
 	} else {
 		err = a.Catalog.DeleteVideoWithTombstone(r.Context(), id)
 		result = DeleteVideoResult{OK: err == nil}

backend/internal/api/admin_test.go

@@ -121,6 +121,63 @@ func TestHandleSetupStoresCredentialsAndCreatesSession(t *testing.T) {
 	}
 }
 
+func TestHandleDeleteVideoDefaultsDeleteSourceFalse(t *testing.T) {
+	req := httptest.NewRequest(http.MethodDelete, "/admin/api/videos/video-1", nil)
+	rctx := chi.NewRouteContext()
+	rctx.URLParams.Add("id", "video-1")
+	req = req.WithContext(context.WithValue(req.Context(), chi.RouteCtxKey, rctx))
+	rr := httptest.NewRecorder()
+
+	called := false
+	(&AdminServer{
+		OnDeleteVideo: func(ctx context.Context, videoID string, deleteSource bool) (DeleteVideoResult, error) {
+			called = true
+			if videoID != "video-1" {
+				t.Fatalf("videoID = %q, want video-1", videoID)
+			}
+			if deleteSource {
+				t.Fatal("deleteSource defaulted to true")
+			}
+			return DeleteVideoResult{OK: true}, nil
+		},
+	}).handleDeleteVideo(rr, req)
+
+	if rr.Code != http.StatusOK {
+		t.Fatalf("status = %d, want 200; body = %s", rr.Code, rr.Body.String())
+	}
+	if !called {
+		t.Fatal("OnDeleteVideo was not called")
+	}
+}
+
+func TestHandleDeleteVideoPassesDeleteSourceOption(t *testing.T) {
+	req := httptest.NewRequest(http.MethodDelete, "/admin/api/videos/video-1", strings.NewReader(`{"deleteSource":true}`))
+	rctx := chi.NewRouteContext()
+	rctx.URLParams.Add("id", "video-1")
+	req = req.WithContext(context.WithValue(req.Context(), chi.RouteCtxKey, rctx))
+	rr := httptest.NewRecorder()
+
+	(&AdminServer{
+		OnDeleteVideo: func(ctx context.Context, videoID string, deleteSource bool) (DeleteVideoResult, error) {
+			if !deleteSource {
+				t.Fatal("deleteSource = false, want true")
+			}
+			return DeleteVideoResult{OK: true, DeletedSource: true}, nil
+		},
+	}).handleDeleteVideo(rr, req)
+
+	if rr.Code != http.StatusOK {
+		t.Fatalf("status = %d, want 200; body = %s", rr.Code, rr.Body.String())
+	}
+	var got DeleteVideoResult
+	if err := json.Unmarshal(rr.Body.Bytes(), &got); err != nil {
+		t.Fatalf("decode response: %v", err)
+	}
+	if !got.DeletedSource {
+		t.Fatalf("DeletedSource = false, want true; response = %s", rr.Body.String())
+	}
+}
+
 func TestHandleCheckUpdateReportsNewRelease(t *testing.T) {
 	dir := t.TempDir()
 	versionFile := filepath.Join(dir, ".version")

backend/internal/catalog/catalog.go

@@ -88,6 +88,11 @@ type Video struct {
 func (c *Catalog) UpsertVideo(ctx context.Context, v *Video) error {
 	existed := c.videoExists(ctx, v.ID)
 	v.ContentHash = normalizeContentHash(v.ContentHash)
+	v.SampledSHA256 = normalizeContentHash(v.SampledSHA256)
+	fingerprintStatus := nullableStatus(v.FingerprintStatus)
+	if v.SampledSHA256 != "" && (v.FingerprintStatus == "" || v.FingerprintStatus == "pending") {
+		fingerprintStatus = "ready"
+	}
 	tagsJSON, _ := json.Marshal(v.Tags)
 	badgesJSON, _ := json.Marshal(v.Badges)
 	now := time.Now().UnixMilli()
@@ -98,13 +103,13 @@ func (c *Catalog) UpsertVideo(ctx context.Context, v *Video) error {
 
 	_, err := c.db.ExecContext(ctx, `
 INSERT INTO videos (
-  id, drive_id, file_id, file_name, content_hash, parent_id, title, author, tags,
+  id, drive_id, file_id, file_name, content_hash, sampled_sha256, fingerprint_status, fingerprint_error, parent_id, title, author, tags,
   duration_seconds, size_bytes, ext, quality, thumbnail_url, thumbnail_status,
   preview_file_id, preview_local, preview_status,
   views, favorites, comments, likes, dislikes,
   category, hidden, badges, description, published_at, created_at, updated_at
 ) VALUES (
-  ?, ?, ?, ?, ?, ?, ?, ?, ?,
+  ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?,
   ?, ?, ?, ?, ?, CASE WHEN COALESCE(?, '') != '' THEN 'ready' ELSE 'pending' END,
   ?, ?, ?,
   ?, ?, ?, ?, ?,
@@ -123,15 +128,18 @@ ON CONFLICT(id) DO UPDATE SET
                       ELSE videos.content_hash
                     END,
   sampled_sha256  = CASE
-                      WHEN videos.size_bytes != excluded.size_bytes THEN ''
+                      WHEN videos.size_bytes != excluded.size_bytes THEN excluded.sampled_sha256
+                      WHEN excluded.sampled_sha256 != '' THEN excluded.sampled_sha256
                       ELSE videos.sampled_sha256
                     END,
   fingerprint_status = CASE
-                      WHEN videos.size_bytes != excluded.size_bytes THEN 'pending'
+                      WHEN videos.size_bytes != excluded.size_bytes THEN COALESCE(excluded.fingerprint_status, 'pending')
+                      WHEN excluded.sampled_sha256 != '' THEN COALESCE(excluded.fingerprint_status, 'ready')
                       ELSE COALESCE(videos.fingerprint_status, 'pending')
                     END,
   fingerprint_error = CASE
-                      WHEN videos.size_bytes != excluded.size_bytes THEN ''
+                      WHEN videos.size_bytes != excluded.size_bytes THEN COALESCE(excluded.fingerprint_error, '')
+                      WHEN excluded.sampled_sha256 != '' THEN COALESCE(excluded.fingerprint_error, '')
                       ELSE COALESCE(videos.fingerprint_error, '')
                     END,
   duration_seconds= excluded.duration_seconds,
@@ -152,7 +160,7 @@ ON CONFLICT(id) DO UPDATE SET
   description     = excluded.description,
   updated_at      = excluded.updated_at
 `,
-		v.ID, v.DriveID, v.FileID, v.FileName, v.ContentHash, v.ParentID, v.Title, v.Author, string(tagsJSON),
+		v.ID, v.DriveID, v.FileID, v.FileName, v.ContentHash, v.SampledSHA256, fingerprintStatus, v.FingerprintError, v.ParentID, v.Title, v.Author, string(tagsJSON),
 		v.DurationSeconds, v.Size, v.Ext, v.Quality, v.ThumbnailURL, v.ThumbnailURL,
 		v.PreviewFileID, v.PreviewLocal, nullableStatus(v.PreviewStatus),
 		v.Views, v.Favorites, v.Comments, v.Likes, v.Dislikes,
@@ -731,8 +739,11 @@ func (c *Catalog) ListCrawlerSourceIDs(ctx context.Context, kind, driveID string
 	rows, err := c.db.QueryContext(ctx,
 		`SELECT SUBSTR(id, ?) FROM videos WHERE id LIKE ? || '%'
 		 UNION
-		 SELECT SUBSTR(id, ?) FROM deleted_videos WHERE id LIKE ? || '%'`,
-		len(prefix)+1, prefix, len(prefix)+1, prefix)
+		 SELECT SUBSTR(id, ?) FROM deleted_videos WHERE id LIKE ? || '%'
+		 UNION
+		 SELECT source_id FROM crawler_seen_sources
+		  WHERE kind = ? AND drive_id = ? AND status IN ('imported', 'duplicate')`,
+		len(prefix)+1, prefix, len(prefix)+1, prefix, kind, driveID)
 	if err != nil {
 		return nil, err
 	}
@@ -750,6 +761,47 @@ func (c *Catalog) ListCrawlerSourceIDs(ctx context.Context, kind, driveID string
 	return out, rows.Err()
 }
 
+// MarkCrawlerSourceSeen records the outcome for a crawler source item. Duplicate
+// source IDs are included in future seen files so scripts can skip them before
+// the backend downloads the same duplicate content again.
+func (c *Catalog) MarkCrawlerSourceSeen(ctx context.Context, kind, driveID, sourceID, status, canonicalVideoID, sampledSHA256 string, size int64) error {
+	kind = strings.TrimSpace(kind)
+	driveID = strings.TrimSpace(driveID)
+	sourceID = strings.TrimSpace(sourceID)
+	status = strings.TrimSpace(status)
+	if kind == "" || driveID == "" || sourceID == "" {
+		return nil
+	}
+	switch status {
+	case "imported", "duplicate":
+	default:
+		return fmt.Errorf("catalog: unsupported crawler source status %q", status)
+	}
+	sampledSHA256 = normalizeContentHash(sampledSHA256)
+	if size < 0 {
+		size = 0
+	}
+	now := time.Now().UnixMilli()
+	_, err := c.db.ExecContext(ctx, `
+INSERT INTO crawler_seen_sources (
+  kind, drive_id, source_id, status, canonical_video_id, sampled_sha256, size_bytes, first_seen_at, last_seen_at
+) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?)
+ON CONFLICT(kind, drive_id, source_id) DO UPDATE SET
+  status = excluded.status,
+  canonical_video_id = excluded.canonical_video_id,
+  sampled_sha256 = CASE
+                    WHEN excluded.sampled_sha256 != '' THEN excluded.sampled_sha256
+                    ELSE crawler_seen_sources.sampled_sha256
+                   END,
+  size_bytes = CASE
+                WHEN excluded.size_bytes > 0 THEN excluded.size_bytes
+                ELSE crawler_seen_sources.size_bytes
+               END,
+  last_seen_at = excluded.last_seen_at`,
+		kind, driveID, sourceID, status, strings.TrimSpace(canonicalVideoID), sampledSHA256, size, now, now)
+	return err
+}
+
 // DeleteVideoWithTombstone records that an administrator explicitly deleted a
 // video, then removes the visible catalog row. The tombstone is used by
 // scanners/crawlers to avoid importing the same source file again.
@@ -922,6 +974,29 @@ func (c *Catalog) FindVideoByFileSignature(ctx context.Context, fileName string,
 	return scanVideo(row)
 }
 
+// FindEquivalentVideo returns the earliest visible video that represents the
+// same content as source by strong hash or sampled fingerprint, regardless of
+// which drive currently owns it.
+func (c *Catalog) FindEquivalentVideo(ctx context.Context, source *Video) (*Video, error) {
+	if source == nil {
+		return nil, sql.ErrNoRows
+	}
+	where, args, ok := equivalentVideoLookupWhere(source)
+	if !ok {
+		return nil, sql.ErrNoRows
+	}
+	args = append([]any{source.ID}, args...)
+	row := c.db.QueryRowContext(ctx,
+		`SELECT `+allVideoCols+` FROM videos
+		 WHERE id != ?
+		   AND COALESCE(hidden, 0) = 0
+		   AND COALESCE(file_id, '') != ''
+		   AND (`+where+`)
+		 ORDER BY created_at ASC, id ASC
+		 LIMIT 1`, args...)
+	return scanVideo(row)
+}
+
 // FindEquivalentVideoOnDrive returns a visible video on driveID that represents
 // the same content as source by strong hash or sampled fingerprint.
 func (c *Catalog) FindEquivalentVideoOnDrive(ctx context.Context, source *Video, driveID string) (*Video, error) {

backend/internal/catalog/schema.sql

@@ -89,6 +89,24 @@ CREATE INDEX IF NOT EXISTS idx_deleted_videos_drive_hash
 CREATE INDEX IF NOT EXISTS idx_deleted_videos_drive_signature
     ON deleted_videos(drive_id, file_name, size_bytes);
 
+-- 爬虫来源记录。用于把已确认重复的 source_id 写回 seen 列表，
+-- 避免后续爬虫反复下载同一个候选视频。
+CREATE TABLE IF NOT EXISTS crawler_seen_sources (
+    kind               TEXT NOT NULL,
+    drive_id           TEXT NOT NULL,
+    source_id          TEXT NOT NULL,
+    status             TEXT NOT NULL DEFAULT 'imported', -- imported / duplicate
+    canonical_video_id TEXT NOT NULL DEFAULT '',
+    sampled_sha256     TEXT NOT NULL DEFAULT '',
+    size_bytes         INTEGER NOT NULL DEFAULT 0,
+    first_seen_at      INTEGER NOT NULL,
+    last_seen_at       INTEGER NOT NULL,
+    PRIMARY KEY (kind, drive_id, source_id)
+);
+
+CREATE INDEX IF NOT EXISTS idx_crawler_seen_sources_drive
+    ON crawler_seen_sources(kind, drive_id, status);
+
 -- 网盘账户
 CREATE TABLE IF NOT EXISTS drives (
     id            TEXT PRIMARY KEY,

backend/internal/drives/googledrive/driver.go

@@ -593,6 +593,17 @@ func (d *Driver) Rename(ctx context.Context, fileID, newName string) error {
 	return nil
 }
 
+func (d *Driver) Remove(ctx context.Context, fileID string) error {
+	fileID = strings.TrimSpace(fileID)
+	if fileID == "" {
+		return errors.New("googledrive remove: empty file id")
+	}
+	if err := d.request(ctx, d.fileURL(fileID), http.MethodDelete, nil, nil); err != nil {
+		return fmt.Errorf("googledrive remove: %w", err)
+	}
+	return nil
+}
+
 func (d *Driver) findUploadedFileID(ctx context.Context, parentID, name, md5Hex string) (string, error) {
 	entries, err := d.List(ctx, parentID)
 	if err != nil {
@@ -624,6 +635,8 @@ func (d *Driver) findUploadedFileID(ctx context.Context, parentID, name, md5Hex
 	return "", fmt.Errorf("googledrive upload: uploaded file %q not found in parent %q", name, parentID)
 }
 
+var _ drives.Remover = (*Driver)(nil)
+
 func isGoogleUploadHTTPRateLimit(status int, header http.Header, body []byte, apiErr apiErrorBody) bool {
 	if status == http.StatusTooManyRequests {
 		return true

backend/internal/drives/iface.go

@@ -40,6 +40,12 @@ type Drive interface {
 	RootID() string
 }
 
+// Remover is an optional drive capability. It mirrors OpenList's optional
+// Remove interface: callers must type-assert before deleting a source file.
+type Remover interface {
+	Remove(ctx context.Context, fileID string) error
+}
+
 type Entry struct {
 	ID       string
 	Name     string

backend/internal/drives/localstorage/driver.go

@@ -257,6 +257,39 @@ func (d *Driver) EnsureDir(context.Context, string) (string, error) {
 	return "", drives.ErrNotSupported
 }
 
+func (d *Driver) Remove(ctx context.Context, fileID string) error {
+	if err := ctx.Err(); err != nil {
+		return err
+	}
+	p, rel, err := d.pathForID(fileID)
+	if err != nil {
+		if os.IsNotExist(err) {
+			return nil
+		}
+		return err
+	}
+	if rel == "" {
+		return errors.New("localstorage: refusing to remove root")
+	}
+	info, err := os.Stat(p)
+	if err != nil {
+		if os.IsNotExist(err) {
+			return nil
+		}
+		return err
+	}
+	if info.IsDir() {
+		return errors.New("localstorage: refusing to remove directory")
+	}
+	if !info.Mode().IsRegular() {
+		return errors.New("localstorage: refusing to remove non-regular file")
+	}
+	if err := os.Remove(p); err != nil && !os.IsNotExist(err) {
+		return err
+	}
+	return nil
+}
+
 func (d *Driver) root() (string, error) {
 	raw := strings.TrimSpace(d.rootPath)
 	if raw == "" {
@@ -276,6 +309,8 @@ func (d *Driver) root() (string, error) {
 	return filepath.Abs(raw)
 }
 
+var _ drives.Remover = (*Driver)(nil)
+
 func (d *Driver) pathForID(id string) (string, string, error) {
 	root, err := d.root()
 	if err != nil {

backend/internal/drives/localupload/driver.go

@@ -78,12 +78,38 @@ func (d *Driver) EnsureDir(context.Context, string) (string, error) {
 	return "", drives.ErrNotSupported
 }
 
+func (d *Driver) Remove(ctx context.Context, fileID string) error {
+	if err := ctx.Err(); err != nil {
+		return err
+	}
+	path, err := d.uploadPath(fileID)
+	if err != nil {
+		return err
+	}
+	info, err := os.Stat(path)
+	if err != nil {
+		if os.IsNotExist(err) {
+			return nil
+		}
+		return err
+	}
+	if info.IsDir() {
+		return errors.New("localupload: refusing to remove directory")
+	}
+	if err := os.Remove(path); err != nil && !os.IsNotExist(err) {
+		return err
+	}
+	return nil
+}
+
 func (d *Driver) RootID() string { return d.uploadDir() }
 
 func (d *Driver) uploadDir() string {
 	return d.uploadDirPath
 }
 
+var _ drives.Remover = (*Driver)(nil)
+
 func (d *Driver) uploadPath(fileID string) (string, error) {
 	if strings.TrimSpace(fileID) == "" || filepath.Base(fileID) != fileID {
 		return "", errors.New("invalid upload file id")

backend/internal/drives/onedrive/driver.go

@@ -501,6 +501,17 @@ func (d *Driver) Rename(ctx context.Context, fileID, newName string) error {
 	return nil
 }
 
+func (d *Driver) Remove(ctx context.Context, fileID string) error {
+	fileID = strings.TrimSpace(fileID)
+	if fileID == "" {
+		return errors.New("onedrive remove: empty file id")
+	}
+	if err := d.request(ctx, d.itemURL(fileID), http.MethodDelete, nil, nil); err != nil {
+		return fmt.Errorf("onedrive remove: %w", err)
+	}
+	return nil
+}
+
 func (d *Driver) request(ctx context.Context, rawURL, method string, configure func(*resty.Request), out any) error {
 	return d.requestOnce(ctx, rawURL, method, configure, out, true)
 }
@@ -741,3 +752,4 @@ func guessMime(name string) string {
 }
 
 var _ drives.Drive = (*Driver)(nil)
+var _ drives.Remover = (*Driver)(nil)

backend/internal/drives/p115/driver.go

@@ -461,6 +461,23 @@ func (d *Driver) Rename(ctx context.Context, fileID, newName string) error {
 	return nil
 }
 
+func (d *Driver) Remove(ctx context.Context, fileID string) error {
+	if d.client == nil {
+		return errors.New("p115 remove: driver not initialized")
+	}
+	if err := ctx.Err(); err != nil {
+		return err
+	}
+	fileID = strings.TrimSpace(fileID)
+	if fileID == "" {
+		return errors.New("p115 remove: empty fileID")
+	}
+	if err := d.client.Delete(fileID); err != nil {
+		return fmt.Errorf("p115 remove: %w", err)
+	}
+	return nil
+}
+
 // bufferAndHashSha1 把 r 全量复制到一个临时文件，同时计算 SHA1。
 // 返回临时文件（位置在末尾，需调用方 Seek 回 0）、SHA1 hex 大写、实际字节数。
 //
@@ -563,3 +580,4 @@ func guessMime(name string) string {
 }
 
 var _ drives.Drive = (*Driver)(nil)
+var _ drives.Remover = (*Driver)(nil)

backend/internal/drives/p123/driver.go

@@ -42,6 +42,7 @@ const (
 	endpointDownloadInfo = "/file/download_info"
 	endpointMkdir        = "/file/upload_request"
 	endpointRename       = "/file/rename"
+	endpointTrash        = "/file/trash"
 	endpointUpload       = "/file/upload_request"
 	endpointS3Auth       = "/file/s3_upload_object/auth"
 	endpointS3Parts      = "/file/s3_repare_upload_parts_batch"
@@ -545,6 +546,32 @@ func (d *Driver) Rename(ctx context.Context, fileID, newName string) error {
 	return nil
 }
 
+func (d *Driver) Remove(ctx context.Context, fileID string) error {
+	fileID = strings.TrimSpace(fileID)
+	if fileID == "" {
+		return errors.New("123pan remove: empty file id")
+	}
+	f, _, err := d.findFile(ctx, fileID)
+	if err != nil {
+		if strings.Contains(strings.ToLower(err.Error()), "not found") {
+			return nil
+		}
+		return fmt.Errorf("123pan remove metadata: %w", err)
+	}
+	body := map[string]any{
+		"driveId":           0,
+		"operation":         true,
+		"fileTrashInfoList": []panFile{f},
+	}
+	if _, err := d.request(ctx, endpointTrash, http.MethodPost, func(req *resty.Request) {
+		req.SetBody(body)
+	}, nil); err != nil {
+		return fmt.Errorf("123pan remove: %w", err)
+	}
+	d.removeCachedFile(fileID)
+	return nil
+}
+
 func (d *Driver) EnsureDir(ctx context.Context, pathFromRoot string) (string, error) {
 	parts := splitPath(pathFromRoot)
 	currentID := d.rootID
@@ -942,6 +969,12 @@ func (d *Driver) renameCachedFile(fileID, newName string) {
 	}
 }
 
+func (d *Driver) removeCachedFile(fileID string) {
+	d.fileMu.Lock()
+	delete(d.files, fileID)
+	d.fileMu.Unlock()
+}
+
 func (d *Driver) cachedFile(fileID string) (panFile, string, bool) {
 	d.fileMu.RLock()
 	defer d.fileMu.RUnlock()
@@ -1111,3 +1144,4 @@ func guessMime(name string) string {
 }
 
 var _ drives.Drive = (*Driver)(nil)
+var _ drives.Remover = (*Driver)(nil)

backend/internal/drives/pikpak/driver.go

@@ -356,6 +356,19 @@ func (d *Driver) Rename(ctx context.Context, fileID, newName string) error {
 	return nil
 }
 
+func (d *Driver) Remove(ctx context.Context, fileID string) error {
+	fileID = strings.TrimSpace(fileID)
+	if fileID == "" {
+		return errors.New("pikpak remove: empty file id")
+	}
+	if err := d.request(ctx, filesURL+":batchTrash", http.MethodPost, func(req *resty.Request) {
+		req.SetBody(map[string]any{"ids": []string{fileID}})
+	}, nil); err != nil {
+		return fmt.Errorf("pikpak remove: %w", err)
+	}
+	return nil
+}
+
 func (d *Driver) EnsureDir(ctx context.Context, pathFromRoot string) (string, error) {
 	currentID := d.rootID
 	for _, name := range splitPath(pathFromRoot) {
@@ -565,3 +578,4 @@ func ParseBoolDefault(raw string, def bool) bool {
 }
 
 var _ drives.Drive = (*Driver)(nil)
+var _ drives.Remover = (*Driver)(nil)

backend/internal/drives/quark/driver.go

@@ -269,6 +269,22 @@ func (d *Driver) Upload(ctx context.Context, parentID, name string, r io.Reader,
 	return "", drives.ErrNotSupported
 }
 
+func (d *Driver) Remove(ctx context.Context, fileID string) error {
+	fileID = strings.TrimSpace(fileID)
+	if fileID == "" {
+		return errors.New("quark remove: empty file id")
+	}
+	body := map[string]any{
+		"action_type":  1,
+		"exclude_fids": []string{},
+		"filelist":     []string{fileID},
+	}
+	if err := d.request(ctx, "/file/delete", http.MethodPost, nil, body, nil); err != nil {
+		return fmt.Errorf("quark remove: %w", err)
+	}
+	return nil
+}
+
 // ---------- helpers ----------
 
 func fileToEntry(f *file, parentID string) drives.Entry {
@@ -343,3 +359,4 @@ func setCookieValue(cookie, key, value string) string {
 }
 
 var _ drives.Drive = (*Driver)(nil)
+var _ drives.Remover = (*Driver)(nil)

backend/internal/drives/scriptcrawler/crawler.go

@@ -4,6 +4,7 @@ import (
 	"bufio"
 	"context"
 	"crypto/sha256"
+	"database/sql"
 	"encoding/hex"
 	"encoding/json"
 	"errors"
@@ -23,6 +24,7 @@ import (
 	"time"
 
 	"github.com/video-site/backend/internal/catalog"
+	"github.com/video-site/backend/internal/fingerprint"
 	"github.com/video-site/backend/internal/mediaasset"
 	"golang.org/x/net/proxy"
 )
@@ -30,11 +32,15 @@ import (
 const (
 	DefaultTargetNew           = 10
 	defaultUserAgent           = "Mozilla/5.0 (compatible; video-site-91-scriptcrawler/1.0)"
+	defaultCandidateMultiplier = 10
+	defaultCandidateFloorExtra = 50
+	defaultCandidateBudgetMax  = 500
 )
 
 type CrawlerConfig struct {
 	Driver          *Driver
 	Catalog         *catalog.Catalog
+	CrawlerName     string
 	SourceKind      string
 	PythonPath      string
 	ScriptPath      string
@@ -76,6 +82,7 @@ func NewCrawler(cfg CrawlerConfig) *Crawler {
 
 type CrawlResult struct {
 	TargetNew       int
+	CandidateBudget int
 	TotalEntries    int
 	NewVideos       int
 	Skipped         int
@@ -105,6 +112,8 @@ type Job struct {
 	RunID             string          `json:"run_id"`
 	CrawlerID         string          `json:"crawler_id"`
 	TargetNew         int             `json:"target_new"`
+	UniqueTarget      int             `json:"unique_target,omitempty"`
+	CandidateBudget   int             `json:"candidate_budget,omitempty"`
 	SeenSourceIDsFile string          `json:"seen_source_ids_file"`
 	OutputDir         string          `json:"output_dir"`
 	Config            json.RawMessage `json:"config"`
@@ -253,11 +262,12 @@ func (c *Crawler) RunOnce(ctx context.Context, targetNew int) (*CrawlResult, err
 	if targetNew <= 0 {
 		targetNew = DefaultTargetNew
 	}
+	candidateBudget := candidateBudgetForTarget(targetNew)
 	if err := c.cfg.Driver.Init(ctx); err != nil {
 		return nil, fmt.Errorf("scriptcrawler: driver init: %w", err)
 	}
 
-	result := &CrawlResult{TargetNew: targetNew, StartedAt: time.Now()}
+	result := &CrawlResult{TargetNew: targetNew, CandidateBudget: candidateBudget, StartedAt: time.Now()}
 	defer func() { result.FinishedAt = time.Now() }()
 	emit := func(p CrawlProgress) {
 		if c.cfg.OnProgress == nil {
@@ -293,11 +303,11 @@ func (c *Crawler) RunOnce(ctx context.Context, targetNew int) (*CrawlResult, err
 	result.SeenSnapshot = seenCount
 	emit(CrawlProgress{})
 
-	if err := c.writeJobFile(jobPath, runID, targetNew, seenPath); err != nil {
+	if err := c.writeJobFile(jobPath, runID, targetNew, candidateBudget, seenPath); err != nil {
 		return result, fmt.Errorf("scriptcrawler: write job: %w", err)
 	}
 
-	cmd, stdout, err := c.startScript(ctx, jobPath, targetNew)
+	cmd, stdout, err := c.startScript(ctx, jobPath, targetNew, candidateBudget)
 	if err != nil {
 		return result, fmt.Errorf("scriptcrawler: start: %w", err)
 	}
@@ -408,7 +418,7 @@ func (c *Crawler) writeSeenSourceIDs(ctx context.Context, path string) (int, err
 	return len(seen), nil
 }
 
-func (c *Crawler) writeJobFile(path, runID string, targetNew int, seenPath string) error {
+func (c *Crawler) writeJobFile(path, runID string, targetNew, candidateBudget int, seenPath string) error {
 	cfg := json.RawMessage([]byte("{}"))
 	if raw := strings.TrimSpace(c.cfg.ConfigJSON); raw != "" {
 		if !json.Valid([]byte(raw)) {
@@ -425,7 +435,9 @@ func (c *Crawler) writeJobFile(path, runID string, targetNew int, seenPath strin
 		Mode:              "crawl",
 		RunID:             runID,
 		CrawlerID:         c.cfg.Driver.ID(),
-		TargetNew:         targetNew,
+		TargetNew:         candidateBudget,
+		UniqueTarget:      targetNew,
+		CandidateBudget:   candidateBudget,
 		SeenSourceIDsFile: seenPath,
 		OutputDir:         outputDir,
 		Config:            cfg,
@@ -442,7 +454,7 @@ func (c *Crawler) writeJobFile(path, runID string, targetNew int, seenPath strin
 	return os.Rename(tmp, path)
 }
 
-func (c *Crawler) startScript(ctx context.Context, jobPath string, targetNew int) (*exec.Cmd, io.ReadCloser, error) {
+func (c *Crawler) startScript(ctx context.Context, jobPath string, targetNew, candidateBudget int) (*exec.Cmd, io.ReadCloser, error) {
 	cmd := exec.CommandContext(ctx, c.cfg.PythonPath, c.cfg.ScriptPath, "--job", jobPath)
 	if strings.TrimSpace(c.cfg.WorkDir) != "" {
 		cmd.Dir = c.cfg.WorkDir
@@ -466,7 +478,7 @@ func (c *Crawler) startScript(ctx context.Context, jobPath string, targetNew int
 		_ = stdout.Close()
 		return nil, nil, err
 	}
-	log.Printf("[scriptcrawler] drive=%s exec %s --job=%s target_new=%d", c.cfg.Driver.ID(), c.cfg.ScriptPath, jobPath, targetNew)
+	log.Printf("[scriptcrawler] drive=%s exec %s --job=%s unique_target=%d candidate_budget=%d", c.cfg.Driver.ID(), c.cfg.ScriptPath, jobPath, targetNew, candidateBudget)
 	if err := cmd.Start(); err != nil {
 		_ = stdout.Close()
 		_ = stderr.Close()
@@ -493,7 +505,8 @@ func (c *Crawler) processItem(ctx context.Context, item Item) (bool, error) {
 	if err != nil {
 		return false, err
 	}
-	videoID := BuildVideoIDForKind(c.sourceKind(), c.cfg.Driver.ID(), sourceID)
+	sourceKind := c.sourceKind()
+	videoID := BuildVideoIDForKind(sourceKind, c.cfg.Driver.ID(), sourceID)
 	if deleted, err := c.cfg.Catalog.IsVideoDeleted(ctx, videoID); err != nil {
 		return false, err
 	} else if deleted {
@@ -513,25 +526,6 @@ func (c *Crawler) processItem(ctx context.Context, item Item) (bool, error) {
 		return false, fmt.Errorf("video: %w", err)
 	}
 
-	thumbReady := false
-	if item.Thumbnail.URL != "" || item.Thumbnail.LocalFile != "" {
-		thumbFile := sourceID + detectThumbExt(item.Thumbnail.URL, item.Thumbnail.LocalFile)
-		thumbPath, err := c.cfg.Driver.ThumbPath(thumbFile)
-		if err == nil {
-			if _, err := c.materializeMedia(ctx, item.Thumbnail, thumbPath, item.DetailURL, false); err != nil {
-				log.Printf("[scriptcrawler] drive=%s source_id=%s thumbnail failed: %v", c.cfg.Driver.ID(), sourceID, err)
-			} else if c.cfg.CommonThumbDir != "" {
-				if err := os.MkdirAll(c.cfg.CommonThumbDir, 0o755); err != nil {
-					log.Printf("[scriptcrawler] drive=%s common thumbs mkdir: %v", c.cfg.Driver.ID(), err)
-				} else if err := copyFileAtomic(thumbPath, mediaasset.ThumbnailPathInDir(c.cfg.CommonThumbDir, videoID)); err != nil {
-					log.Printf("[scriptcrawler] drive=%s source_id=%s copy thumbnail: %v", c.cfg.Driver.ID(), sourceID, err)
-				} else {
-					thumbReady = true
-				}
-			}
-		}
-	}
-
 	now := time.Now()
 	title := strings.TrimSpace(item.Title)
 	if title == "" {
@@ -545,6 +539,9 @@ func (c *Crawler) processItem(ctx context.Context, item Item) (bool, error) {
 	if matched, err := c.cfg.Catalog.MatchTags(ctx, title+" "+author+" "+strings.Join(tags, " ")); err == nil {
 		tags = mergeStringLists(tags, matched)
 	}
+	if crawlerTag := c.crawlerTagName(); crawlerTag != "" {
+		tags = mergeStringLists(tags, []string{crawlerTag})
+	}
 	publishedAt := now
 	if parsed := parsePublishedAt(item.PublishedAt); !parsed.IsZero() {
 		publishedAt = parsed
@@ -572,6 +569,43 @@ func (c *Crawler) processItem(ctx context.Context, item Item) (bool, error) {
 		CreatedAt:       now,
 		UpdatedAt:       now,
 	}
+	sampled, err := fingerprint.Compute(ctx, c.cfg.Driver, v, fingerprint.Config{}, c.cfg.HTTPClient)
+	if err != nil {
+		_ = os.Remove(videoPath)
+		return false, fmt.Errorf("fingerprint: %w", err)
+	}
+	v.SampledSHA256 = sampled
+	v.FingerprintStatus = "ready"
+	if duplicate, err := c.cfg.Catalog.FindEquivalentVideo(ctx, v); err == nil && duplicate != nil {
+		_ = os.Remove(videoPath)
+		if markErr := c.cfg.Catalog.MarkCrawlerSourceSeen(ctx, sourceKind, c.cfg.Driver.ID(), sourceID, "duplicate", duplicate.ID, sampled, size); markErr != nil {
+			log.Printf("[scriptcrawler] drive=%s source_id=%s mark duplicate seen: %v", c.cfg.Driver.ID(), sourceID, markErr)
+		}
+		log.Printf("[scriptcrawler] drive=%s source_id=%s duplicate_of=%s title=%q size=%d", c.cfg.Driver.ID(), sourceID, duplicate.ID, title, size)
+		return false, nil
+	} else if err != nil && !errors.Is(err, sql.ErrNoRows) {
+		_ = os.Remove(videoPath)
+		return false, fmt.Errorf("duplicate lookup: %w", err)
+	}
+
+	thumbReady := false
+	if item.Thumbnail.URL != "" || item.Thumbnail.LocalFile != "" {
+		thumbFile := sourceID + detectThumbExt(item.Thumbnail.URL, item.Thumbnail.LocalFile)
+		thumbPath, err := c.cfg.Driver.ThumbPath(thumbFile)
+		if err == nil {
+			if _, err := c.materializeMedia(ctx, item.Thumbnail, thumbPath, item.DetailURL, false); err != nil {
+				log.Printf("[scriptcrawler] drive=%s source_id=%s thumbnail failed: %v", c.cfg.Driver.ID(), sourceID, err)
+			} else if c.cfg.CommonThumbDir != "" {
+				if err := os.MkdirAll(c.cfg.CommonThumbDir, 0o755); err != nil {
+					log.Printf("[scriptcrawler] drive=%s common thumbs mkdir: %v", c.cfg.Driver.ID(), err)
+				} else if err := copyFileAtomic(thumbPath, mediaasset.ThumbnailPathInDir(c.cfg.CommonThumbDir, videoID)); err != nil {
+					log.Printf("[scriptcrawler] drive=%s source_id=%s copy thumbnail: %v", c.cfg.Driver.ID(), sourceID, err)
+				} else {
+					thumbReady = true
+				}
+			}
+		}
+	}
 	if thumbReady {
 		v.ThumbnailURL = "/p/thumb/" + v.ID
 	}
@@ -579,6 +613,9 @@ func (c *Crawler) processItem(ctx context.Context, item Item) (bool, error) {
 		_ = os.Remove(videoPath)
 		return false, err
 	}
+	if err := c.cfg.Catalog.MarkCrawlerSourceSeen(ctx, sourceKind, c.cfg.Driver.ID(), sourceID, "imported", v.ID, sampled, size); err != nil {
+		log.Printf("[scriptcrawler] drive=%s source_id=%s mark imported seen: %v", c.cfg.Driver.ID(), sourceID, err)
+	}
 	log.Printf("[scriptcrawler] drive=%s source_id=%s ok title=%q size=%d", c.cfg.Driver.ID(), sourceID, title, size)
 	return true, nil
 }
@@ -898,6 +935,36 @@ func (c *Crawler) sourceKind() string {
 	return Kind
 }
 
+func (c *Crawler) crawlerTagName() string {
+	if c == nil {
+		return ""
+	}
+	if v := strings.TrimSpace(c.cfg.CrawlerName); v != "" {
+		return v
+	}
+	if c.cfg.Driver != nil {
+		return strings.TrimSpace(c.cfg.Driver.ID())
+	}
+	return ""
+}
+
+func candidateBudgetForTarget(targetNew int) int {
+	if targetNew <= 0 {
+		targetNew = DefaultTargetNew
+	}
+	budget := targetNew * defaultCandidateMultiplier
+	if floor := targetNew + defaultCandidateFloorExtra; budget < floor {
+		budget = floor
+	}
+	if budget > defaultCandidateBudgetMax {
+		budget = defaultCandidateBudgetMax
+	}
+	if budget < targetNew {
+		return targetNew
+	}
+	return budget
+}
+
 func BuildVideoID(driveID, sourceID string) string {
 	return BuildVideoIDForKind(Kind, driveID, sourceID)
 }

backend/internal/drives/scriptcrawler/crawler_test.go

@@ -10,8 +10,15 @@ import (
 	"path/filepath"
 	"strings"
 	"testing"
+	"time"
 
 	"github.com/video-site/backend/internal/catalog"
+	"github.com/video-site/backend/internal/fingerprint"
+)
+
+const (
+	scriptCrawlerDuplicateBytes = "duplicate-video-bytes"
+	scriptCrawlerUniqueBytes    = "unique-video-bytes"
 )
 
 func TestCrawlerRunOnceImportsLocalFileAndSkipsExisting(t *testing.T) {
@@ -44,6 +51,7 @@ func TestCrawlerRunOnceImportsLocalFileAndSkipsExisting(t *testing.T) {
 	c := NewCrawler(CrawlerConfig{
 		Driver:      drv,
 		Catalog:     cat,
+		CrawlerName: "Demo Crawler",
 		PythonPath:  wrapper,
 		ScriptPath:  dummyScript,
 	})
@@ -61,6 +69,9 @@ func TestCrawlerRunOnceImportsLocalFileAndSkipsExisting(t *testing.T) {
 	if v.Title != "Imported From Helper" || v.FileID != "abc-123.mp4" || v.Size == 0 {
 		t.Fatalf("video = title:%q file:%q size:%d", v.Title, v.FileID, v.Size)
 	}
+	if !hasString(v.Tags, "Demo Crawler") {
+		t.Fatalf("video tags = %#v, want crawler name tag", v.Tags)
+	}
 	if _, err := os.Stat(filepath.Join(drv.VideosDir(), "abc-123.mp4")); err != nil {
 		t.Fatalf("video file not copied: %v", err)
 	}
@@ -267,6 +278,113 @@ func TestCrawlerRunOnceImportsSimpleMediaURLWithoutSourceID(t *testing.T) {
 	}
 }
 
+func TestCrawlerRunOnceSkipsFingerprintDuplicateAndContinues(t *testing.T) {
+	ctx := context.Background()
+	tmp := t.TempDir()
+	cat, err := catalog.Open(filepath.Join(tmp, "catalog.db"))
+	if err != nil {
+		t.Fatalf("open catalog: %v", err)
+	}
+	t.Cleanup(func() {
+		if err := cat.Close(); err != nil {
+			t.Fatalf("close catalog: %v", err)
+		}
+	})
+	drv := New(Config{ID: "demo", RootDir: filepath.Join(tmp, "crawler")})
+	if err := drv.Init(ctx); err != nil {
+		t.Fatalf("driver init: %v", err)
+	}
+
+	seedFile := "seed-canonical.mp4"
+	if err := os.WriteFile(filepath.Join(drv.VideosDir(), seedFile), []byte(scriptCrawlerDuplicateBytes), 0o644); err != nil {
+		t.Fatalf("write seed video: %v", err)
+	}
+	seed := &catalog.Video{
+		ID:          "seed-for-hash",
+		DriveID:     drv.ID(),
+		FileID:      seedFile,
+		Title:       "Seed",
+		Size:        int64(len(scriptCrawlerDuplicateBytes)),
+		PublishedAt: time.Now(),
+	}
+	sampled, err := fingerprint.Compute(ctx, drv, seed, fingerprint.Config{}, nil)
+	if err != nil {
+		t.Fatalf("compute seed fingerprint: %v", err)
+	}
+	_ = os.Remove(filepath.Join(drv.VideosDir(), seedFile))
+
+	now := time.Now()
+	if err := cat.UpsertVideo(ctx, &catalog.Video{
+		ID:                "existing-canonical",
+		DriveID:           "other-drive",
+		FileID:            "existing.mp4",
+		FileName:          "existing.mp4",
+		Title:             "Existing Canonical",
+		Size:              int64(len(scriptCrawlerDuplicateBytes)),
+		Ext:               "mp4",
+		SampledSHA256:     sampled,
+		FingerprintStatus: "ready",
+		PublishedAt:       now,
+		CreatedAt:         now,
+		UpdatedAt:         now,
+	}); err != nil {
+		t.Fatalf("seed canonical video: %v", err)
+	}
+
+	dummyScript := filepath.Join(tmp, "helper-script")
+	if err := os.WriteFile(dummyScript, []byte("helper"), 0o755); err != nil {
+		t.Fatalf("write dummy script: %v", err)
+	}
+	wrapper := filepath.Join(tmp, "helper-wrapper.sh")
+	wrapperScript := fmt.Sprintf("#!/bin/sh\nexec %q -test.run=TestScriptCrawlerHelperProcess \"$@\"\n", os.Args[0])
+	if err := os.WriteFile(wrapper, []byte(wrapperScript), 0o755); err != nil {
+		t.Fatalf("write helper wrapper: %v", err)
+	}
+
+	t.Setenv("GO_WANT_SCRIPTCRAWLER_HELPER", "1")
+	t.Setenv("GO_WANT_SCRIPTCRAWLER_DUP_UNIQUE", "1")
+	c := NewCrawler(CrawlerConfig{
+		Driver:     drv,
+		Catalog:    cat,
+		PythonPath: wrapper,
+		ScriptPath: dummyScript,
+	})
+	res, err := c.RunOnce(ctx, 1)
+	if err != nil {
+		t.Fatalf("run once: %v", err)
+	}
+	if res.NewVideos != 1 || res.Skipped != 1 || res.Failed != 0 || res.TotalEntries != 2 {
+		t.Fatalf("result = total:%d new:%d skipped:%d failed:%d, want 2/1/1/0", res.TotalEntries, res.NewVideos, res.Skipped, res.Failed)
+	}
+	if res.CandidateBudget <= res.TargetNew {
+		t.Fatalf("candidate budget = %d, target = %d; want expanded budget", res.CandidateBudget, res.TargetNew)
+	}
+	if _, err := cat.GetVideo(ctx, BuildVideoID("demo", "dup-source")); err == nil {
+		t.Fatal("duplicate candidate should not be imported")
+	}
+	if _, err := os.Stat(filepath.Join(drv.VideosDir(), "dup-source.mp4")); !os.IsNotExist(err) {
+		t.Fatalf("duplicate local file stat = %v, want removed", err)
+	}
+	v, err := cat.GetVideo(ctx, BuildVideoID("demo", "unique-source"))
+	if err != nil {
+		t.Fatalf("unique video should be imported: %v", err)
+	}
+	if v.SampledSHA256 == "" || v.FingerprintStatus != "ready" {
+		t.Fatalf("unique fingerprint = %q status=%q, want ready sampled fingerprint", v.SampledSHA256, v.FingerprintStatus)
+	}
+	seen, err := cat.ListCrawlerSourceIDs(ctx, Kind, "demo")
+	if err != nil {
+		t.Fatalf("list seen source ids: %v", err)
+	}
+	seenSet := map[string]bool{}
+	for _, id := range seen {
+		seenSet[id] = true
+	}
+	if !seenSet["dup-source"] || !seenSet["unique-source"] {
+		t.Fatalf("seen ids = %#v, want duplicate and imported source ids", seen)
+	}
+}
+
 func TestScriptCrawlerHelperProcess(t *testing.T) {
 	if os.Getenv("GO_WANT_SCRIPTCRAWLER_HELPER") != "1" {
 		return
@@ -307,6 +425,41 @@ func TestScriptCrawlerHelperProcess(t *testing.T) {
 		_ = json.NewEncoder(os.Stdout).Encode(event)
 		os.Exit(0)
 	}
+	if os.Getenv("GO_WANT_SCRIPTCRAWLER_DUP_UNIQUE") == "1" {
+		duplicateFile := filepath.Join(job.OutputDir, "duplicate.mp4")
+		if err := os.WriteFile(duplicateFile, []byte(scriptCrawlerDuplicateBytes), 0o644); err != nil {
+			fmt.Fprintln(os.Stderr, err)
+			os.Exit(2)
+		}
+		uniqueFile := filepath.Join(job.OutputDir, "unique.mp4")
+		if err := os.WriteFile(uniqueFile, []byte(scriptCrawlerUniqueBytes), 0o644); err != nil {
+			fmt.Fprintln(os.Stderr, err)
+			os.Exit(2)
+		}
+		for _, event := range []Event{
+			{
+				Type: "item",
+				Item: Item{
+					SourceID: "dup-source",
+					Title:    "Duplicate Candidate",
+					Author:   "helper",
+					Media:    MediaRef{LocalFile: duplicateFile},
+				},
+			},
+			{
+				Type: "item",
+				Item: Item{
+					SourceID: "unique-source",
+					Title:    "Unique Candidate",
+					Author:   "helper",
+					Media:    MediaRef{LocalFile: uniqueFile},
+				},
+			},
+		} {
+			_ = json.NewEncoder(os.Stdout).Encode(event)
+		}
+		os.Exit(0)
+	}
 	localFile := filepath.Join(job.OutputDir, "helper.mp4")
 	if err := os.WriteFile(localFile, []byte("helper-video"), 0o644); err != nil {
 		fmt.Fprintln(os.Stderr, err)
@@ -324,3 +477,12 @@ func TestScriptCrawlerHelperProcess(t *testing.T) {
 	_ = json.NewEncoder(os.Stdout).Encode(event)
 	os.Exit(0)
 }
+
+func hasString(values []string, want string) bool {
+	for _, value := range values {
+		if value == want {
+			return true
+		}
+	}
+	return false
+}

backend/internal/drives/scriptcrawler/driver.go

@@ -147,6 +147,46 @@ func (d *Driver) EnsureDir(context.Context, string) (string, error) {
 	return "", drives.ErrNotSupported
 }
 
+func (d *Driver) Remove(ctx context.Context, fileID string) error {
+	if err := ctx.Err(); err != nil {
+		return err
+	}
+	videoPath, err := d.VideoPath(fileID)
+	if err != nil {
+		return err
+	}
+	info, err := os.Stat(videoPath)
+	if err != nil {
+		if os.IsNotExist(err) {
+			removeThumbCandidates(d.ThumbPath, strings.TrimSuffix(fileID, filepath.Ext(fileID)))
+			return nil
+		}
+		return err
+	}
+	if info.IsDir() {
+		return errors.New("scriptcrawler: refusing to remove directory")
+	}
+	if err := os.Remove(videoPath); err != nil && !os.IsNotExist(err) {
+		return err
+	}
+	removeThumbCandidates(d.ThumbPath, strings.TrimSuffix(fileID, filepath.Ext(fileID)))
+	return nil
+}
+
+func removeThumbCandidates(pathFor func(string) (string, error), stem string) {
+	stem = strings.TrimSpace(stem)
+	if stem == "" {
+		return
+	}
+	for _, ext := range []string{".jpg", ".jpeg", ".png", ".webp"} {
+		path, err := pathFor(stem + ext)
+		if err != nil {
+			continue
+		}
+		_ = os.Remove(path)
+	}
+}
+
 func safeJoin(root, fileID string) (string, error) {
 	id := strings.TrimSpace(fileID)
 	if id == "" || filepath.Base(id) != id {
@@ -170,3 +210,4 @@ func safeJoin(root, fileID string) (string, error) {
 }
 
 var _ drives.Drive = (*Driver)(nil)
+var _ drives.Remover = (*Driver)(nil)

backend/internal/drives/spider91/driver.go

@@ -167,6 +167,46 @@ func (d *Driver) EnsureDir(ctx context.Context, pathFromRoot string) (string, er
 	return "", drives.ErrNotSupported
 }
 
+func (d *Driver) Remove(ctx context.Context, fileID string) error {
+	if err := ctx.Err(); err != nil {
+		return err
+	}
+	videoPath, err := d.VideoPath(fileID)
+	if err != nil {
+		return err
+	}
+	info, err := os.Stat(videoPath)
+	if err != nil {
+		if os.IsNotExist(err) {
+			removeThumbCandidates(d.ThumbPath, strings.TrimSuffix(fileID, filepath.Ext(fileID)))
+			return nil
+		}
+		return err
+	}
+	if info.IsDir() {
+		return errors.New("spider91: refusing to remove directory")
+	}
+	if err := os.Remove(videoPath); err != nil && !os.IsNotExist(err) {
+		return err
+	}
+	removeThumbCandidates(d.ThumbPath, strings.TrimSuffix(fileID, filepath.Ext(fileID)))
+	return nil
+}
+
+func removeThumbCandidates(pathFor func(string) (string, error), stem string) {
+	stem = strings.TrimSpace(stem)
+	if stem == "" {
+		return
+	}
+	for _, ext := range []string{".jpg", ".jpeg", ".png", ".webp"} {
+		path, err := pathFor(stem + ext)
+		if err != nil {
+			continue
+		}
+		_ = os.Remove(path)
+	}
+}
+
 // safeJoin 把 fileID 拼到 root 下，保证最终路径不会逃出 root。
 // fileID 必须是单纯的文件名（不含 / 或 .. 等组件）。
 func safeJoin(root, fileID string) (string, error) {
@@ -192,3 +232,4 @@ func safeJoin(root, fileID string) (string, error) {
 }
 
 var _ drives.Drive = (*Driver)(nil)
+var _ drives.Remover = (*Driver)(nil)

backend/internal/drives/wopan/driver.go

@@ -11,6 +11,7 @@ import (
 	"time"
 
 	sdk "github.com/OpenListTeam/wopan-sdk-go"
+	"github.com/go-resty/resty/v2"
 	"github.com/video-site/backend/internal/drives"
 )
 
@@ -145,6 +146,22 @@ func (d *Driver) Upload(ctx context.Context, parentID, name string, r io.Reader,
 	return fid, nil
 }
 
+func (d *Driver) Remove(ctx context.Context, fileID string) error {
+	if d.client == nil {
+		return fmt.Errorf("wopan remove: driver not initialized")
+	}
+	fileID = strings.TrimSpace(fileID)
+	if fileID == "" {
+		return fmt.Errorf("wopan remove: empty file id")
+	}
+	if err := d.client.DeleteFile(d.spaceType(), nil, []string{fileID}, func(req *resty.Request) {
+		req.SetContext(ctx)
+	}); err != nil {
+		return fmt.Errorf("wopan remove: %w", err)
+	}
+	return nil
+}
+
 func (d *Driver) EnsureDir(ctx context.Context, pathFromRoot string) (string, error) {
 	parts := splitPath(pathFromRoot)
 	currentID := d.rootID
@@ -229,3 +246,4 @@ func guessMime(name string) string {
 
 // 确保实现接口
 var _ drives.Drive = (*Driver)(nil)
+var _ drives.Remover = (*Driver)(nil)

backend/internal/spider91migrate/migrator.go

@@ -82,6 +82,15 @@ type UploadResult struct {
 	Size   int64
 }
 
+type UploadProgress struct {
+	DriveID      string
+	State        string
+	CurrentTitle string
+	QueueLength  int
+	DoneCount    int
+	TotalCount   int
+}
+
 const (
 	spider91UploadDirName          = "91 Spider"
 	scriptCrawlerUploadRootDirName = "Script Crawlers"
@@ -257,6 +266,7 @@ type Config struct {
 	CaptchaCooldown  time.Duration
 	CommonThumbDir   string
 	OnMigrated       func(videoID string)
+	OnUploadProgress func(UploadProgress)
 }
 
 type Migrator struct {
@@ -444,6 +454,20 @@ func (m *Migrator) runOnce(ctx context.Context) {
 	}
 }
 
+func (m *Migrator) reportUploadProgress(progress UploadProgress) {
+	if m == nil || m.cfg.OnUploadProgress == nil {
+		return
+	}
+	progress.DriveID = strings.TrimSpace(progress.DriveID)
+	if progress.DriveID == "" {
+		return
+	}
+	if progress.State == "" {
+		progress.State = "idle"
+	}
+	m.cfg.OnUploadProgress(progress)
+}
+
 // targetKindForLog 把当前目标盘 kind 转成对人友好的简称，用于日志。
 // 解析失败时回退 "target"。
 func (m *Migrator) targetKindForLog() string {
@@ -669,8 +693,17 @@ func (m *Migrator) migrateDrive(ctx context.Context, plan migrationPlan) (int, e
 	if skip < len(files) {
 		candidates = files[skip:]
 	} else {
+		m.reportUploadProgress(UploadProgress{DriveID: src.ID(), State: "idle"})
 		return 0, nil
 	}
+	totalCandidates := len(candidates)
+	m.reportUploadProgress(UploadProgress{
+		DriveID:     src.ID(),
+		State:       "uploading",
+		QueueLength: totalCandidates,
+		TotalCount:  totalCandidates,
+	})
+	defer m.reportUploadProgress(UploadProgress{DriveID: src.ID(), State: "idle"})
 
 	localVideos, err := m.cfg.Catalog.ListVideosByDriveID(ctx, src.ID(), 100000)
 	if err != nil {
@@ -684,7 +717,8 @@ func (m *Migrator) migrateDrive(ctx context.Context, plan migrationPlan) (int, e
 	}
 
 	migrated := 0
-	for _, f := range candidates {
+	processed := 0
+	for index, f := range candidates {
 		if err := ctx.Err(); err != nil {
 			return migrated, err
 		}
@@ -694,11 +728,35 @@ func (m *Migrator) migrateDrive(ctx context.Context, plan migrationPlan) (int, e
 
 		v := m.findVideoForLocalFile(ctx, plan, f.name, byFileID)
 		if v == nil {
+			processed++
+			m.reportUploadProgress(UploadProgress{
+				DriveID:     src.ID(),
+				State:       "uploading",
+				QueueLength: maxInt(totalCandidates-processed, 0),
+				DoneCount:   processed,
+				TotalCount:  totalCandidates,
+			})
 			continue
 		}
+		m.reportUploadProgress(UploadProgress{
+			DriveID:      src.ID(),
+			State:        "uploading",
+			CurrentTitle: v.Title,
+			QueueLength:  maxInt(totalCandidates-index-1, 0),
+			DoneCount:    processed,
+			TotalCount:   totalCandidates,
+		})
 
 		if v.DriveID != src.ID() {
 			CleanupSpider91Local(src, f.name)
+			processed++
+			m.reportUploadProgress(UploadProgress{
+				DriveID:     src.ID(),
+				State:       "uploading",
+				QueueLength: maxInt(totalCandidates-processed, 0),
+				DoneCount:   processed,
+				TotalCount:  totalCandidates,
+			})
 			continue
 		}
 
@@ -718,6 +776,14 @@ func (m *Migrator) migrateDrive(ctx context.Context, plan migrationPlan) (int, e
 					m.cfg.OnMigrated(v.ID)
 				}
 			}
+			processed++
+			m.reportUploadProgress(UploadProgress{
+				DriveID:     src.ID(),
+				State:       "uploading",
+				QueueLength: maxInt(totalCandidates-processed, 0),
+				DoneCount:   processed,
+				TotalCount:  totalCandidates,
+			})
 			continue
 		}
 
@@ -728,6 +794,14 @@ func (m *Migrator) migrateDrive(ctx context.Context, plan migrationPlan) (int, e
 				continue
 			}
 			if !ready {
+				processed++
+				m.reportUploadProgress(UploadProgress{
+					DriveID:     src.ID(),
+					State:       "uploading",
+					QueueLength: maxInt(totalCandidates-processed, 0),
+					DoneCount:   processed,
+					TotalCount:  totalCandidates,
+				})
 				continue
 			}
 		}
@@ -752,10 +826,25 @@ func (m *Migrator) migrateDrive(ctx context.Context, plan migrationPlan) (int, e
 				m.cfg.OnMigrated(v.ID)
 			}
 		}
+		processed++
+		m.reportUploadProgress(UploadProgress{
+			DriveID:     src.ID(),
+			State:       "uploading",
+			QueueLength: maxInt(totalCandidates-processed, 0),
+			DoneCount:   processed,
+			TotalCount:  totalCandidates,
+		})
 	}
 	return migrated, nil
 }
 
+func maxInt(a, b int) int {
+	if a > b {
+		return a
+	}
+	return b
+}
+
 func (m *Migrator) findVideoForLocalFile(ctx context.Context, plan migrationPlan, localFile string, byFileID map[string]*catalog.Video) *catalog.Video {
 	if v := byFileID[localFile]; v != nil {
 		return v

docs/crawler-protocol.md

@@ -28,7 +28,9 @@ python3 /path/to/crawler.py --job /path/to/job.json
   "mode": "crawl",
   "run_id": "20260609T120000Z",
   "crawler_id": "example",
-  "target_new": 10,
+  "target_new": 100,
+  "unique_target": 10,
+  "candidate_budget": 100,
   "seen_source_ids_file": "/data/scriptcrawlers/example/.crawl/seen.txt",
   "output_dir": "/data/scriptcrawlers/example/output",
   "config": {
@@ -40,6 +42,14 @@ python3 /path/to/crawler.py --job /path/to/job.json
 }
 ```
 
+`unique_target` is the user's requested number of content-unique new videos.
+`candidate_budget` is how many candidate items the script should emit at most.
+For backward compatibility, `target_new` is set to the same value as
+`candidate_budget`, because older scripts only read `target_new`.
+
+The backend may skip candidates that are already present by sampled content
+fingerprint. Skipped duplicate candidates do not count toward `unique_target`.
+
 ## Importing Scripts
 
 Crawler scripts are configured from the admin crawler page. A script can be
@@ -118,4 +128,6 @@ Optional progress/done events:
   `output_dir`.
 - Scripts can read `seen_source_ids_file` and skip known IDs when they provide
   stable `source_id` values. The backend still dedupes every item.
-- The backend stops the process after `target_new` new videos are imported.
+- The backend stops the process after `unique_target` content-unique new videos
+  are imported. Scripts should stop after emitting `candidate_budget` candidates
+  even if the backend has not reached `unique_target`.

src/admin/ConfirmModal.tsx

@@ -1,3 +1,4 @@
+import type { ReactNode } from "react";
 import { AlertTriangle } from "lucide-react";
 import { Modal } from "./Modal";
 
@@ -12,6 +13,7 @@ type ConfirmModalProps = {
   centerMessage?: boolean;
   modalClassName?: string;
   loading?: boolean;
+  children?: ReactNode;
   onCancel: () => void;
   onConfirm: () => void;
 };
@@ -27,6 +29,7 @@ export function ConfirmModal({
   centerMessage = false,
   modalClassName = "",
   loading = false,
+  children,
   onCancel,
   onConfirm,
 }: ConfirmModalProps) {
@@ -65,6 +68,7 @@ export function ConfirmModal({
               ))}
             </ul>
           )}
+          {children}
         </div>
       </div>
     </Modal>

src/admin/CrawlersPage.tsx

@@ -30,7 +30,7 @@ import { generationStateClass, generationStateLabel } from "./drive/constants";
 import { Spider91UploadTargetField } from "./drive/Spider91UploadTargetField";
 import { SpiderIcon } from "./icons/SpiderIcon";
 
-const BUSY_STATES = new Set(["scanning", "generating", "queued"]);
+const BUSY_STATES = new Set(["scanning", "generating", "uploading", "queued"]);
 const POLL_INTERVAL_MS = 5000;
 const UPLOAD_TARGET_KINDS = new Set(["p115", "pikpak", "p123", "googledrive", "onedrive"]);
 
@@ -43,7 +43,8 @@ function crawlerBusy(crawler: api.AdminCrawler) {
     statusBusy(crawler.scanGenerationStatus) ||
     statusBusy(crawler.thumbnailGenerationStatus) ||
     statusBusy(crawler.previewGenerationStatus) ||
-    statusBusy(crawler.fingerprintGenerationStatus)
+    statusBusy(crawler.fingerprintGenerationStatus) ||
+    statusBusy(crawler.uploadGenerationStatus)
   );
 }
 
@@ -273,12 +274,14 @@ function crawlerStages(crawler: api.AdminCrawler): StageInfo[] {
     { key: "thumbnail", label: "封面", status: crawler.thumbnailGenerationStatus },
     { key: "preview", label: "预览", status: crawler.previewGenerationStatus },
     { key: "fingerprint", label: "指纹", status: crawler.fingerprintGenerationStatus },
+    { key: "upload", label: "上传", status: crawler.uploadGenerationStatus },
   ];
 }
 
 function stageStateLabel(stage: StageInfo): string {
   const state = stage.status?.state || "idle";
   if (stage.key === "scan" && state === "scanning") return "抓取中";
+  if (stage.key === "upload" && state === "uploading") return "上传中";
   return generationStateLabel(state);
 }
 
@@ -364,6 +367,7 @@ function CrawlerRow({
 
 function CrawlerDetail({ crawler }: { crawler: api.AdminCrawler }) {
   const scan = crawler.scanGenerationStatus;
+  const upload = crawlerUploadDisplayStatus(crawler);
   return (
     <div className="admin-crawler-detail">
       <div className="admin-crawler-detail__grid">
@@ -373,12 +377,21 @@ function CrawlerDetail({ crawler }: { crawler: api.AdminCrawler }) {
           stateText={scan?.state === "scanning" ? "抓取中" : generationStateLabel(scan?.state || "idle")}
           counts={[
             { label: "累计爬取", value: crawler.totalCrawledCount ?? 0 },
-            { label: "本地保留", value: crawler.localVideoCount ?? 0 },
-            { label: "已上传", value: crawler.migratedVideoCount ?? 0 },
             { label: "本轮检查", value: scan?.scannedCount ?? 0 },
             { label: "本轮新增", value: scan?.addedCount ?? 0 },
           ]}
         />
+        <GenStageCard
+          label="上传"
+          status={upload.status}
+          stateText={upload.text}
+          counts={[
+            { label: "已上传", value: crawler.migratedVideoCount ?? 0 },
+            { label: crawler.uploadDriveId ? "待上传" : "本地保留", value: crawler.localVideoCount ?? 0 },
+            { label: "本轮处理", value: upload.status.doneCount ?? 0 },
+            { label: "本轮总数", value: upload.status.totalCount ?? 0 },
+          ]}
+        />
         <GenStageCard
           label="封面"
           status={crawler.thumbnailGenerationStatus}
@@ -417,6 +430,49 @@ function CrawlerDetail({ crawler }: { crawler: api.AdminCrawler }) {
   );
 }
 
+function crawlerUploadDisplayStatus(crawler: api.AdminCrawler): {
+  status: api.DriveGenerationStatus;
+  text: string;
+} {
+  const live = crawler.uploadGenerationStatus;
+  const state = live?.state || "idle";
+  const localCount = crawler.localVideoCount ?? 0;
+  const totalCount = crawler.totalCrawledCount ?? 0;
+  const base: api.DriveGenerationStatus = {
+    state,
+    currentTitle: live?.currentTitle,
+    queueLength: live?.queueLength ?? 0,
+    cooldownUntil: live?.cooldownUntil,
+    scannedCount: live?.scannedCount ?? 0,
+    addedCount: live?.addedCount ?? 0,
+    doneCount: live?.doneCount ?? 0,
+    totalCount: live?.totalCount ?? 0,
+  };
+
+  if (!crawler.uploadDriveId) {
+    return {
+      status: base,
+      text: localCount > 0 ? "本地保存" : generationStateLabel(state),
+    };
+  }
+  if (state === "uploading") {
+    return { status: base, text: "上传中" };
+  }
+  if (state === "queued") {
+    return { status: base, text: "排队中" };
+  }
+  if (localCount > 0) {
+    return {
+      status: { ...base, state: "queued", queueLength: localCount },
+      text: "待上传",
+    };
+  }
+  if (totalCount > 0) {
+    return { status: base, text: "完成" };
+  }
+  return { status: base, text: generationStateLabel(state) };
+}
+
 function GenStageCard({
   label,
   status,

src/admin/VideosPage.tsx

@@ -27,8 +27,10 @@ export function VideosPage() {
   const [batchRegening, setBatchRegening] = useState(false);
   const [batchDeleteOpen, setBatchDeleteOpen] = useState(false);
   const [batchDeleting, setBatchDeleting] = useState(false);
+  const [batchDeleteSource, setBatchDeleteSource] = useState(false);
   const [deleteTarget, setDeleteTarget] = useState<api.AdminVideo | null>(null);
   const [deleting, setDeleting] = useState(false);
+  const [deleteSource, setDeleteSource] = useState(false);
   const pageSize = useVideosPageSize();
   const { show } = useToast();
 
@@ -100,6 +102,7 @@ export function VideosPage() {
 
   async function handleBatchDelete() {
     if (selectedIds.size === 0) return;
+    setBatchDeleteSource(false);
     setBatchDeleteOpen(true);
   }
 
@@ -127,14 +130,15 @@ export function VideosPage() {
     const target = deleteTarget;
     setDeleting(true);
     try {
-      const result = await api.deleteVideo(target.id);
+      const result = await api.deleteVideo(target.id, { deleteSource });
       setDeleteTarget(null);
+      setDeleteSource(false);
       setSelectedIds((ids) => {
         const next = new Set(ids);
         next.delete(target.id);
         return next;
       });
-      show(result.deletedSource ? "已删除视频，并清理 91Spider 源文件" : "已删除视频", "success");
+      show(result.deletedSource ? "已删除视频，并清理源文件" : "已删除视频", "success");
       if (listItems.length === 1 && page > 1) {
         setPage((p) => Math.max(1, p - 1));
       } else {
@@ -156,7 +160,7 @@ export function VideosPage() {
       let deletedSources = 0;
       for (const id of ids) {
         try {
-          const result = await api.deleteVideo(id);
+          const result = await api.deleteVideo(id, { deleteSource: batchDeleteSource });
           success++;
           if (result.deletedSource) deletedSources++;
         } catch {
@@ -165,13 +169,14 @@ export function VideosPage() {
       }
       const failed = ids.length - success;
       if (failed === 0) {
-        const extra = deletedSources > 0 ? `，其中 ${deletedSources} 个清理了 91Spider 源文件` : "";
+        const extra = deletedSources > 0 ? `，其中 ${deletedSources} 个清理了源文件` : "";
         show(`批量删除完成，成功 ${success} 个${extra}`, "success");
       } else {
         show(`批量删除完成，成功 ${success} / ${ids.length} 个，失败 ${failed} 个`, success > 0 ? "info" : "error");
       }
       setSelectedIds(new Set());
       setBatchDeleteOpen(false);
+      setBatchDeleteSource(false);
       if (success >= listItems.length && page > 1) {
         setPage((p) => Math.max(1, p - 1));
       } else {
@@ -363,7 +368,15 @@ export function VideosPage() {
                     <button type="button" className="admin-btn" onClick={() => handleRegen(v)} title="重生预览视频">
                       <RefreshCw size={13} />
                     </button>{" "}
-                    <button type="button" className="admin-btn is-danger" onClick={() => setDeleteTarget(v)} title="删除视频">
+                    <button
+                      type="button"
+                      className="admin-btn is-danger"
+                      onClick={() => {
+                        setDeleteSource(false);
+                        setDeleteTarget(v);
+                      }}
+                      title="删除视频"
+                    >
                       <Trash2 size={13} />
                     </button>
                   </td>
@@ -443,10 +456,26 @@ export function VideosPage() {
         modalClassName="admin-modal--delete-confirm"
         loading={deleting}
         onCancel={() => {
-          if (!deleting) setDeleteTarget(null);
+          if (!deleting) {
+            setDeleteTarget(null);
+            setDeleteSource(false);
+          }
         }}
         onConfirm={confirmDeleteVideo}
+      >
+        <label className="admin-delete-source-option">
+          <input
+            type="checkbox"
+            checked={deleteSource}
+            disabled={deleting}
+            onChange={(e) => setDeleteSource(e.target.checked)}
           />
+          <span>
+            <strong>同时删除网盘中的源文件</strong>
+            <small>开启后会先删除源文件，失败则不会删除管理库记录。</small>
+          </span>
+        </label>
+      </ConfirmModal>
       <ConfirmModal
         open={batchDeleteOpen}
         title="批量删除视频"
@@ -457,10 +486,26 @@ export function VideosPage() {
         modalClassName="admin-modal--delete-confirm"
         loading={batchDeleting}
         onCancel={() => {
-          if (!batchDeleting) setBatchDeleteOpen(false);
+          if (!batchDeleting) {
+            setBatchDeleteOpen(false);
+            setBatchDeleteSource(false);
+          }
         }}
         onConfirm={confirmBatchDelete}
+      >
+        <label className="admin-delete-source-option">
+          <input
+            type="checkbox"
+            checked={batchDeleteSource}
+            disabled={batchDeleting}
+            onChange={(e) => setBatchDeleteSource(e.target.checked)}
           />
+          <span>
+            <strong>同时删除网盘中的源文件</strong>
+            <small>开启后会先删除源文件，失败的视频会保留管理库记录。</small>
+          </span>
+        </label>
+      </ConfirmModal>
     </section>
   );
 }

src/admin/api.ts

@@ -121,6 +121,8 @@ export type DriveGenerationStatus = {
   cooldownUntil?: string;
   scannedCount: number;
   addedCount: number;
+  doneCount: number;
+  totalCount: number;
 };
 
 export function listDrives() {
@@ -206,6 +208,7 @@ export type AdminCrawler = {
   thumbnailGenerationStatus?: DriveGenerationStatus;
   previewGenerationStatus?: DriveGenerationStatus;
   fingerprintGenerationStatus?: DriveGenerationStatus;
+  uploadGenerationStatus?: DriveGenerationStatus;
   thumbnailReadyCount: number;
   thumbnailPendingCount: number;
   thumbnailFailedCount: number;
@@ -483,10 +486,13 @@ export function updateVideo(id: string, body: UpdateVideoInput) {
   });
 }
 
-export function deleteVideo(id: string) {
+export function deleteVideo(id: string, options: { deleteSource?: boolean } = {}) {
   return request<{ ok: boolean; deletedSource: boolean }>(
     `/videos/${encodeURIComponent(id)}`,
-    { method: "DELETE" }
+    {
+      method: "DELETE",
+      body: JSON.stringify({ deleteSource: !!options.deleteSource }),
+    }
   );
 }
 

src/admin/drive/constants.ts

@@ -72,6 +72,7 @@ export function nightlyBusyText(status: { running: boolean; queued: boolean }) {
 
 export function generationStateLabel(state: string): string {
   if (state === "scanning") return "扫盘中";
+  if (state === "uploading") return "上传中";
   if (state === "generating") return "生成中";
   if (state === "cooling") return "冷却中";
   if (state === "queued") return "排队中";
@@ -79,8 +80,8 @@ export function generationStateLabel(state: string): string {
 }
 
 export function generationStateClass(state: string): string {
-  if (state === "scanning" || state === "generating" || state === "cooling" || state === "queued") {
-    if (state === "scanning") return "generating";
+  if (state === "scanning" || state === "uploading" || state === "generating" || state === "cooling" || state === "queued") {
+    if (state === "scanning" || state === "uploading") return "generating";
     return state;
   }
   return "idle";

src/components/VideoActions.tsx

@@ -1,12 +1,14 @@
 import { useEffect, useState } from "react";
-import { EyeOff, ThumbsDown, ThumbsUp } from "lucide-react";
+import { EyeOff, ThumbsDown, ThumbsUp, Trash2 } from "lucide-react";
 import type { VideoDetail } from "@/types";
 import { formatCount } from "@/lib/format";
 
 type Props = {
   video: VideoDetail;
   onHideVideo: () => void;
+  onDeleteVideo: () => void;
   hideSaving?: boolean;
+  deleteSaving?: boolean;
 };
 
 /**
@@ -19,7 +21,13 @@ type Props = {
  * - 后端只有点赞计数接口，点踩仅本地 state。
  * - 失败回滚已经处理。
  */
-export function VideoActions({ video, onHideVideo, hideSaving }: Props) {
+export function VideoActions({
+  video,
+  onHideVideo,
+  onDeleteVideo,
+  hideSaving,
+  deleteSaving,
+}: Props) {
   const [likes, setLikes] = useState(video.likes ?? 0);
   const [dislikes, setDislikes] = useState(video.dislikes ?? 0);
   const [bursting, setBursting] = useState(false);
@@ -121,6 +129,17 @@ export function VideoActions({ video, onHideVideo, hideSaving }: Props) {
         <EyeOff size={16} />
         <span>{hideSaving ? "处理中" : "不再显示"}</span>
       </button>
+
+      <button
+        type="button"
+        className="vd-actions__btn vd-actions__delete"
+        onClick={onDeleteVideo}
+        disabled={deleteSaving}
+        aria-label="删除这个视频"
+      >
+        <Trash2 size={16} />
+        <span>{deleteSaving ? "删除中" : "删除"}</span>
+      </button>
     </div>
   );
 }

src/data/videos.ts

@@ -52,6 +52,19 @@ export function hideVideo(id: string): Promise<{ ok: boolean }> {
   );
 }
 
+export function deleteVideo(
+  id: string,
+  options: { deleteSource?: boolean } = {}
+): Promise<{ ok: boolean; deletedSource: boolean }> {
+  return apiJSON<{ ok: boolean; deletedSource: boolean }>(
+    `/admin/api/videos/${encodeURIComponent(id)}`,
+    {
+      method: "DELETE",
+      body: JSON.stringify({ deleteSource: !!options.deleteSource }),
+    }
+  );
+}
+
 export function recordView(id: string): Promise<{ views: number }> {
   return apiJSON<{ views: number }>(
     `/api/video/${encodeURIComponent(id)}/view`,

src/pages/VideoDetailPage.tsx

@@ -7,6 +7,7 @@ import { VideoMetaHeader } from "@/components/VideoMetaHeader";
 import { VideoInfoPanel } from "@/components/VideoInfoPanel";
 import { RecommendedRail } from "@/components/RecommendedRail";
 import {
+  deleteVideo,
   fetchTags,
   fetchVideoDetail,
   hideVideo,
@@ -23,6 +24,10 @@ export default function VideoDetailPage() {
   const [loading, setLoading] = useState(true);
   const [tagSaving, setTagSaving] = useState(false);
   const [hideSaving, setHideSaving] = useState(false);
+  const [deleteOpen, setDeleteOpen] = useState(false);
+  const [deleteSource, setDeleteSource] = useState(false);
+  const [deleteSaving, setDeleteSaving] = useState(false);
+  const [deleteError, setDeleteError] = useState("");
   const detailTopRef = useRef<HTMLDivElement | null>(null);
 
   useEffect(() => {
@@ -76,6 +81,36 @@ export default function VideoDetailPage() {
     }
   }
 
+  function handleOpenDelete() {
+    if (!detail || deleteSaving) return;
+    setDeleteSource(false);
+    setDeleteError("");
+    setDeleteOpen(true);
+  }
+
+  function handleCloseDelete() {
+    if (deleteSaving) return;
+    setDeleteOpen(false);
+    setDeleteError("");
+  }
+
+  async function handleConfirmDelete() {
+    if (!detail || deleteSaving) return;
+    setDeleteSaving(true);
+    setDeleteError("");
+    try {
+      await deleteVideo(detail.id, { deleteSource });
+      navigate("/list", { replace: true });
+    } catch {
+      setDeleteError(
+        deleteSource
+          ? "删除失败。源文件未能删除时，管理库记录会保留。"
+          : "删除失败，请稍后重试。"
+      );
+      setDeleteSaving(false);
+    }
+  }
+
   function handleFirstPlay() {
     if (!detail) return;
     // 失败静默忽略，不打扰用户播放体验
@@ -199,7 +234,9 @@ export default function VideoDetailPage() {
                 <VideoActions
                   video={detail}
                   onHideVideo={handleHideVideo}
+                  onDeleteVideo={handleOpenDelete}
                   hideSaving={hideSaving}
+                  deleteSaving={deleteSaving}
                 />
               </section>
 
@@ -215,6 +252,59 @@ export default function VideoDetailPage() {
           </div>
         </div>
       </div>
+
+      {deleteOpen && (
+        <div className="vd-delete-modal" role="presentation">
+          <div
+            className="vd-delete-dialog"
+            role="dialog"
+            aria-modal="true"
+            aria-labelledby="vd-delete-title"
+          >
+            <div className="vd-delete-head">
+              <h2 id="vd-delete-title" className="vd-delete-title">
+                删除视频
+              </h2>
+              <p className="vd-delete-text">
+                确定删除「{detail.title}」吗？此操作会从管理库移除该视频。
+              </p>
+            </div>
+
+            <label className="vd-delete-option">
+              <input
+                type="checkbox"
+                checked={deleteSource}
+                disabled={deleteSaving}
+                onChange={(e) => setDeleteSource(e.target.checked)}
+              />
+              <span>
+                <strong>同时删除网盘中的源文件</strong>
+              </span>
+            </label>
+
+            {deleteError && <div className="vd-delete-error">{deleteError}</div>}
+
+            <div className="vd-delete-actions">
+              <button
+                type="button"
+                className="vd-delete-action vd-delete-cancel"
+                onClick={handleCloseDelete}
+                disabled={deleteSaving}
+              >
+                取消
+              </button>
+              <button
+                type="button"
+                className="vd-delete-action vd-delete-confirm"
+                onClick={handleConfirmDelete}
+                disabled={deleteSaving}
+              >
+                {deleteSaving ? "删除中..." : "删除"}
+              </button>
+            </div>
+          </div>
+        </div>
+      )}
     </AppShell>
   );
 }

src/styles/admin.css

@@ -1871,6 +1871,50 @@
   line-height: 1.7;
 }
 
+.admin-delete-source-option {
+  display: grid;
+  grid-template-columns: 18px minmax(0, 1fr);
+  gap: var(--space-2);
+  align-items: start;
+  margin-top: var(--space-3);
+  padding: var(--space-3);
+  border: 1px solid var(--border-subtle);
+  border-radius: var(--radius-sm);
+  background: var(--bg-sunken);
+  color: var(--text-default);
+  cursor: pointer;
+}
+
+.admin-delete-source-option input {
+  width: 16px;
+  height: 16px;
+  margin: 2px 0 0;
+  accent-color: var(--danger);
+}
+
+.admin-delete-source-option span {
+  min-width: 0;
+  display: grid;
+  gap: 2px;
+}
+
+.admin-delete-source-option strong {
+  color: var(--text-strong);
+  font-size: var(--font-sm);
+  font-weight: var(--weight-semibold);
+}
+
+.admin-delete-source-option small {
+  color: var(--text-muted);
+  font-size: var(--font-xs);
+  line-height: var(--line-relaxed);
+}
+
+.admin-delete-source-option:has(input:disabled) {
+  cursor: default;
+  opacity: 0.72;
+}
+
 /* =========================================================
  * Toast
  * ========================================================= */

src/styles/video-detail.css

@@ -751,7 +751,7 @@
   }
 }
 
-/* "不再显示"按钮：克制、hover 才露出 danger */
+/* 次要操作按钮：克制、hover 才露出 danger */
 .vd-actions__btn {
   display: inline-flex;
   align-items: center;
@@ -788,6 +788,182 @@
   margin-left: auto;
 }
 
+/* ---------- Delete confirm modal ---------- */
+.vd-delete-modal {
+  position: fixed;
+  inset: 0;
+  z-index: var(--z-modal);
+  display: grid;
+  place-items: center;
+  padding: var(--space-4);
+  background: var(--bg-overlay);
+  backdrop-filter: blur(8px);
+  animation: vd-delete-fade var(--duration-fast) var(--ease-out);
+}
+
+@keyframes vd-delete-fade {
+  from {
+    opacity: 0;
+  }
+
+  to {
+    opacity: 1;
+  }
+}
+
+.vd-delete-dialog {
+  width: min(460px, 100%);
+  max-height: 90vh;
+  overflow: auto;
+  padding: var(--space-5);
+  border: 1px solid var(--border-default);
+  border-radius: var(--radius-sm);
+  background: var(--bg-surface);
+  color: var(--text-default);
+  box-shadow: var(--shadow-xl);
+  animation: vd-delete-pop var(--duration-normal) var(--ease-out);
+}
+
+@keyframes vd-delete-pop {
+  from {
+    opacity: 0;
+    transform: translateY(8px) scale(0.98);
+  }
+
+  to {
+    opacity: 1;
+    transform: none;
+  }
+}
+
+.vd-delete-head {
+  display: grid;
+  gap: var(--space-2);
+  margin-bottom: var(--space-4);
+}
+
+.vd-delete-title {
+  margin: 0;
+  color: var(--text-strong);
+  font-size: var(--font-xl);
+  font-weight: var(--weight-bold);
+  line-height: var(--line-tight);
+  letter-spacing: 0;
+}
+
+.vd-delete-text {
+  margin: 0;
+  color: var(--text-default);
+  font-size: var(--font-md);
+  line-height: var(--line-relaxed);
+  overflow-wrap: anywhere;
+}
+
+.vd-delete-option {
+  display: grid;
+  grid-template-columns: 18px minmax(0, 1fr);
+  gap: var(--space-2);
+  align-items: start;
+  padding: var(--space-3);
+  border: 1px solid var(--border-subtle);
+  border-radius: var(--radius-sm);
+  background: var(--bg-sunken);
+  cursor: pointer;
+}
+
+.vd-delete-option input {
+  width: 16px;
+  height: 16px;
+  margin: 2px 0 0;
+  accent-color: var(--danger);
+}
+
+.vd-delete-option span {
+  min-width: 0;
+  display: grid;
+  gap: 2px;
+}
+
+.vd-delete-option strong {
+  color: var(--text-strong);
+  font-size: var(--font-sm);
+  font-weight: var(--weight-semibold);
+}
+
+.vd-delete-option small {
+  color: var(--text-muted);
+  font-size: var(--font-xs);
+  line-height: var(--line-relaxed);
+}
+
+.vd-delete-option:has(input:disabled) {
+  cursor: default;
+  opacity: 0.72;
+}
+
+.vd-delete-error {
+  margin-top: var(--space-3);
+  padding: var(--space-3);
+  border: 1px solid rgba(241, 85, 108, 0.3);
+  border-radius: var(--radius-sm);
+  background: var(--danger-soft);
+  color: var(--danger);
+  font-size: var(--font-sm);
+  line-height: var(--line-relaxed);
+}
+
+.vd-delete-actions {
+  display: flex;
+  justify-content: flex-end;
+  gap: var(--space-2);
+  margin-top: var(--space-5);
+}
+
+.vd-delete-action {
+  display: inline-flex;
+  align-items: center;
+  justify-content: center;
+  min-width: 86px;
+  height: 40px;
+  padding: 0 var(--space-4);
+  border: 1px solid var(--border-default);
+  border-radius: var(--radius-sm);
+  background: var(--bg-elevated);
+  color: var(--text-default);
+  font-size: var(--font-sm);
+  font-weight: var(--weight-semibold);
+  cursor: pointer;
+  transition: all var(--transition-fast);
+}
+
+.vd-delete-action:hover:not(:disabled) {
+  border-color: var(--border-strong);
+  color: var(--text-strong);
+}
+
+.vd-delete-action:focus-visible {
+  outline: 2px solid var(--accent);
+  outline-offset: 2px;
+}
+
+.vd-delete-action:disabled {
+  opacity: 0.55;
+  cursor: not-allowed;
+}
+
+.vd-delete-confirm {
+  border-color: var(--danger);
+  background: var(--danger);
+  color: #fff;
+}
+
+.vd-delete-confirm:hover:not(:disabled) {
+  border-color: var(--danger);
+  background: var(--danger);
+  color: #fff;
+  filter: brightness(1.04);
+}
+
 /* ---------- Info card (description + tags) ---------- */
 .vd-info {
   display: flex;
@@ -1743,6 +1919,36 @@
     display: none;
   }
 
+  .vd-actions__delete {
+    width: 44px;
+    padding: 0;
+    justify-content: center;
+    flex: 0 0 auto;
+  }
+
+  .vd-actions__delete span {
+    display: none;
+  }
+
+  .vd-delete-modal {
+    align-items: end;
+    padding: var(--space-3);
+  }
+
+  .vd-delete-dialog {
+    width: 100%;
+    padding: var(--space-4);
+  }
+
+  .vd-delete-actions {
+    display: grid;
+    grid-template-columns: 1fr 1fr;
+  }
+
+  .vd-delete-action {
+    width: 100%;
+  }
+
   .vd-info__desc {
     padding: var(--space-3);
     font-size: var(--font-base);