feat: add 'Ask' quality option, language auto-select, and m4a audio remux

- Add 'ask' option to audio/video quality settings (stored as '' in DB)
- Auto-select when both media type AND quality are set; show filtered
  format picker when quality is Ask
- Auto-skip language picker when user has a matching language preference
- Remux audio-only downloads to m4a instead of webm for Telegram compat
- Update README to document the new flow and m4a audio output

.gitignore

@@ -1,10 +1,11 @@
-.build/
-*.exe
-*.db
-*.db-wal
-*.db-shm
-.env
-.DS_Store
-tmp/
-opencode.json
 
+.build/
+cookies.txt
+*.db
+*.db-shm
+*.db-wal
+.DS_Store
+.env
+*.exe
+opencode.json
+tmp/

README.md

@@ -10,11 +10,12 @@ Telegram bot for downloading media from any site yt-dlp supports.
 - Thumbnail preview with metadata (title, uploader, duration) before download
 - Audio-only sites (Spotify, SoundCloud, YouTube Music, etc.) auto-detect and skip to audio quality selection
 - Multi-step format selection: type picker (Video/Audio/Ask), quality, secondary format (video+audio combine), language
-- Auto-select: skip the format picker entirely when defaults are set in user preferences
+- Auto-select: skip all pickers when both media type AND quality are set in preferences (quality "Ask" shows a filtered picker)
+- Audio remuxed to m4a (Telegram-friendly), video remuxed to mp4
 - Download progress updates (with speed and ETA)
 - Cancel downloads at any time (terminates yt-dlp process)
 - Per-user quota system (daily, weekly, monthly; 0 = unlimited)
-- User preferences (default media type, audio/video quality, language)
+- User preferences (default media type, audio/video quality with "Ask" option, language)
 - Download history with pagination
 - Configurable file size limit (default 50 MB)
 - Cookies support for age-restricted and authenticated content

docker-compose.yml

@@ -12,10 +12,11 @@ services:
       - HTTP_PROXY=${HTTP_PROXY}
       - HTTPS_PROXY=${HTTPS_PROXY}
       - DB_PATH=${DB_PATH}
+      - COOKIES_FILE=${COOKIES_FILE}
       - TZ=${TZ}
     volumes:
       - uptodown_data:/data
-      - ./cookies.txt:/cookies.txt:ro # Optional: cookies file for yt-dlp auth
+      - ./cookies.txt:/cookies.txt # Optional: cookies file for yt-dlp auth
 
 volumes:
   uptodown_data:

internal/dl/ytdlp.go

@@ -250,8 +250,10 @@ func (c *Client) StartDownload(job *domain.DownloadJob, downloadDir string) (<-c
 	outputPath := filepath.Join(downloadDir, job.ID+"_%(id)s.%(ext)s")
 	args = append(args, "--output", outputPath)
 
-	// Remux to mp4 for video output — Telegram-friendly container.
+	// Remux to Telegram-friendly containers: mp4 for video, m4a for audio-only.
-	if job.MediaType != domain.MediaTypeAudio {
+	if job.MediaType == domain.MediaTypeAudio {
+		args = append(args, "--merge-output-format", "m4a")
+	} else {
 		args = append(args, "--merge-output-format", "mp4")
 	}
 

internal/handler/download.go

@@ -161,9 +161,29 @@ func (h *Handler) handleURLInput(ctx context.Context, chatID int64, text string,
 
 	prefs, prefErr := h.Repo.GetOrCreatePreferences(userID)
 	if prefErr == nil && prefs.DefaultMediaType != domain.MediaTypeAsk {
+		// Media type is set — check quality preference
+		var qualityPref string
+		switch prefs.DefaultMediaType {
+		case domain.MediaTypeVideo:
+			qualityPref = prefs.DefaultVideoFormat
+		case domain.MediaTypeAudio:
+			qualityPref = prefs.DefaultAudioFormat
+		}
+		if qualityPref != "" {
 			h.autoSelectAndDownload(ctx, chatID, userID, url, info, prefs)
 			return
 		}
+		// Quality is Ask — show filtered format picker
+		ss := &stepState{
+			URL:       url,
+			MediaInfo: info,
+			Step:      0,
+			CreatedAt: time.Now(),
+		}
+		h.setUserStepState(chatID, ss)
+		h.showFormatsForType(ctx, chatID, userID, ss, prefs.DefaultMediaType.String())
+		return
+	}
 
 	ss := &stepState{
 		URL:       url,
@@ -239,6 +259,34 @@ func (h *Handler) autoSelectAndDownload(ctx context.Context, chatID int64, userI
 	h.advanceToLanguage(ctx, chatID, msgID, userID, ss)
 }
 
+// showFormatsForType sends the format picker filtered by the given media type.
+// Called when the user has a type preference but quality is Ask.
+func (h *Handler) showFormatsForType(ctx context.Context, chatID int64, userID int64, ss *stepState, mediaType string) {
+	ids, err := buildFilteredFormatList(ss.MediaInfo.Formats, mediaType)
+	if err != nil {
+		h.sendText(ctx, chatID, "No formats available.")
+		return
+	}
+	ss.FormatIDs = ids
+	h.setUserStepState(chatID, ss)
+
+	labels := make([]string, len(ids))
+	for i, id := range ids {
+		if id == "best" {
+			labels[i] = "Best"
+		} else {
+			f := findFormatByID(ss.MediaInfo.Formats, id)
+			if f != nil {
+				labels[i] = formatLabelForDisplay(*f)
+			} else {
+				labels[i] = id
+			}
+		}
+	}
+	kb := formatKeyboard("format_", labels, ids)
+	h.sendWithKB(ctx, chatID, "Select format:", kb)
+}
+
 // handleTypeSelection processes the initial type picker (Video/Audio/Ask).
 func (h *Handler) handleTypeSelection(ctx context.Context, chatID int64, msgID int, userID int64, mediaType string) {
 	ss := h.getUserStepState(chatID)
@@ -394,6 +442,18 @@ func (h *Handler) advanceToLanguage(ctx context.Context, chatID int64, msgID int
 	h.setUserStepState(chatID, ss)
 
 	if len(ss.MediaInfo.Languages) > 1 {
+		// Check if user has a default language and it's available.
+		prefs, prefErr := h.Repo.GetOrCreatePreferences(userID)
+		if prefErr == nil && prefs.DefaultLanguage != "" {
+			for _, lang := range ss.MediaInfo.Languages {
+				if lang == prefs.DefaultLanguage {
+					ss.Language = lang
+					h.startDownload(ctx, chatID, msgID, userID, ss)
+					return
+				}
+			}
+		}
+		// No match or no preference — show picker.
 		kb := languageKeyboard(ss.MediaInfo.Languages)
 		h.editText(ctx, chatID, msgID, "Select language:", &kb)
 		return

internal/handler/settings.go

@@ -14,11 +14,11 @@ import (
 func (h *Handler) buildSettingsKeyboard(userID int64, prefs *domain.UserPreferences) (string, models.InlineKeyboardMarkup) {
 	audioFmt := prefs.DefaultAudioFormat
 	if audioFmt == "" {
-		audioFmt = "best"
+		audioFmt = "ask"
 	}
 	videoFmt := prefs.DefaultVideoFormat
 	if videoFmt == "" {
-		videoFmt = "best"
+		videoFmt = "ask"
 	}
 	langLabel := prefs.DefaultLanguage
 	if langLabel == "" {
@@ -101,11 +101,13 @@ func (h *Handler) buildQualityAudioVideoPicker(prefs *domain.UserPreferences) (s
 }
 
 func (h *Handler) buildAudioQualityPicker(prefs *domain.UserPreferences) (string, models.InlineKeyboardMarkup) {
-	bitrates := []string{"best", "128k", "192k", "256k", "320k"}
+	bitrates := []string{"ask", "best", "128k", "192k", "256k", "320k"}
 	rows := [][]models.InlineKeyboardButton{}
 	for _, b := range bitrates {
 		label := b
-		if b == prefs.DefaultAudioFormat {
+		if b == "ask" && prefs.DefaultAudioFormat == "" {
+			label = "> " + label
+		} else if b != "ask" && b == prefs.DefaultAudioFormat {
 			label = "> " + label
 		}
 		rows = append(rows, []models.InlineKeyboardButton{
@@ -119,12 +121,14 @@ func (h *Handler) buildAudioQualityPicker(prefs *domain.UserPreferences) (string
 }
 
 func (h *Handler) buildVideoQualityPicker(prefs *domain.UserPreferences) (string, models.InlineKeyboardMarkup) {
-	resolutions := []string{"best", "2160p", "1440p", "1080p", "720p", "480p", "360p"}
+	resolutions := []string{"ask", "best", "2160p", "1440p", "1080p", "720p", "480p", "360p"}
 	rows := [][]models.InlineKeyboardButton{}
 	row := []models.InlineKeyboardButton{}
 	for i, r := range resolutions {
 		label := r
-		if r == prefs.DefaultVideoFormat {
+		if r == "ask" && prefs.DefaultVideoFormat == "" {
+			label = "> " + label
+		} else if r != "ask" && r == prefs.DefaultVideoFormat {
 			label = "> " + label
 		}
 		row = append(row, models.InlineKeyboardButton{
@@ -220,9 +224,15 @@ func (h *Handler) handleSettingsSet(ctx context.Context, chatID int64, msgID int
 		}
 		changed = true
 	case "audioquality":
+		if value == "ask" {
+			value = ""
+		}
 		prefs.DefaultAudioFormat = value
 		changed = true
 	case "videoquality":
+		if value == "ask" {
+			value = ""
+		}
 		prefs.DefaultVideoFormat = value
 		changed = true
 	case "language":

internal/repo/store.go

@@ -65,7 +65,7 @@ func (s *Store) GetOrCreateUser(chatID int64) (int64, error) {
 
 func (s *Store) GetOrCreatePreferences(userID int64) (*domain.UserPreferences, error) {
 	if _, err := s.db.Exec(
-		"INSERT OR IGNORE INTO user_preferences (user_id, default_media_type) VALUES (?, 'ask')",
+		"INSERT OR IGNORE INTO user_preferences (user_id, default_media_type, default_audio_format, default_video_format) VALUES (?, 'ask', '', '')",
 		userID,
 	); err != nil {
 		return nil, err