bookhoard/internal/services/media_scanner_audio_multifile.go

package services

// Multi-file audiobook ingestion: folder-of-tracks books and zip-of-audio
// books. Both produce the SAME payload shape single-file books write into
// media_items.audio_metadata — the chapters array stays the format
// abstraction — plus a tracks array naming one playable file per track
// (see media_scanner_audio.go). The /audio/:trackIndex endpoint resolves
// those paths to bytes; the web player maps global seconds ⇄ track/local
// offset, so progress/bookmarks keep their existing coordinates.
//
// Grouping rule (folder books): the book root is the nearest ancestor of
// the track file that holds audio files DIRECTLY, merged upward through
// disc/part-style directories ("CD1", "Disc 2", "Part 3", "05") whose
// parent has no audio of its own. So
//
//	Author/Title/01.mp3, 02.mp3            → book at Author/Title
//	Author/Title/CD1/*.mp3, CD2/*.mp3      → one book at Author/Title
//	Author/BookA/*.mp3, Author/BookB/*.mp3 → two books (siblings never merge)
//
// A root holding fewer than two tracks falls back to the single-file path.

import (
	"archive/zip"
	"bufio"
	"context"
	"encoding/json"
	"errors"
	"fmt"
	"io"
	"os"
	"path/filepath"
	"regexp"
	"sort"
	"strconv"
	"strings"
	"time"

	"github.com/jackc/pgx/v5"
	"github.com/jackc/pgx/v5/pgtype"

	"bookhoard/internal/database"
	"bookhoard/internal/utils"
)

// discPartPattern matches directory names that group discs/parts of one
// book rather than naming a standalone book: "CD1", "disc 2", "Part3",
// "pt. 1", "Vol 2", or a bare two-digit ordering number ("01").
var discPartPattern = regexp.MustCompile(`(?i)^(disc|cd|part|pt|vol|volume)\s*\.?\s*\d{0,3}$|^\d{1,2}$`)

// naturalLess orders strings with embedded numbers numerically, so
// "02 Track" sorts before "10 Track". Digit runs compare by value; other
// bytes compare bytewise.
func naturalLess(a, b string) bool {
	i, j := 0, 0
	for i < len(a) && j < len(b) {
		ca, cb := a[i], b[j]
		da, db := ca >= '0' && ca <= '9', cb >= '0' && cb <= '9'
		switch {
		case da && db:
			si, sj := i, j
			for i < len(a) && a[i] >= '0' && a[i] <= '9' {
				i++
			}
			for j < len(b) && b[j] >= '0' && b[j] <= '9' {
				j++
			}
			na, nb := strings.TrimLeft(a[si:i], "0"), strings.TrimLeft(b[sj:j], "0")
			if len(na) != len(nb) {
				return len(na) < len(nb)
			}
			if c := strings.Compare(na, nb); c != 0 {
				return c < 0
			}
		default:
			if ca != cb {
				return ca < cb
			}
			i++
			j++
		}
	}
	return len(a)-i < len(b)-j
}

// sortTracksNatural orders full paths by natural comparison.
func sortTracksNatural(paths []string) {
	sort.Slice(paths, func(x, y int) bool { return naturalLess(paths[x], paths[y]) })
}

// subtreeAudioTracks collects playable audio tracks under dir
// recursively, natural-sorted by path. Zips don't count (containers,
// never tracks — see isAudioTrackFile).
func subtreeAudioTracks(dir string) []string {
	var tracks []string
	_ = filepath.WalkDir(dir, func(path string, d os.DirEntry, err error) error {
		if err != nil || d.IsDir() {
			return nil // unreadable subtree: skip silently, best effort
		}
		if isAudioTrackFile(path) {
			tracks = append(tracks, path)
		}
		return nil
	})
	sortTracksNatural(tracks)
	return tracks
}

// directAudioCount counts playable audio files directly inside dir (not
// recursive).
func directAudioCount(dir string) int {
	entries, err := os.ReadDir(dir)
	if err != nil {
		return 0
	}
	n := 0
	for _, e := range entries {
		if !e.IsDir() && isAudioTrackFile(e.Name()) {
			n++
		}
	}
	return n
}

// bookRootDirForTrackFile resolves the folder-book root DIRECTORY for an
// audio file without walking its subtree (cheap enough to call for every
// track visit). ok is false when no ancestor holds audio directly.
func bookRootDirForTrackFile(path, libraryRoot string) (root string, ok bool) {
	dir := filepath.Dir(path)

	// Climb to the nearest ancestor holding audio files directly.
	candidate := dir
	for directAudioCount(candidate) == 0 {
		parent := filepath.Dir(candidate)
		if parent == candidate || candidate == libraryRoot || !strings.HasPrefix(candidate, libraryRoot) {
			return "", false
		}
		candidate = parent
	}

	// Merge upward through disc/part groupings. Stop at the first
	// directory that doesn't look like a disc marker, or whose parent
	// holds audio of its own (that would swallow sibling books).
	for discPartPattern.MatchString(strings.ToLower(filepath.Base(candidate))) {
		parent := filepath.Dir(candidate)
		if parent == candidate || candidate == libraryRoot || !strings.HasPrefix(candidate, libraryRoot) {
			break
		}
		if directAudioCount(parent) > 0 {
			break
		}
		candidate = parent
	}

	return candidate, true
}

// bookRootForTrackFile resolves the folder-book root for an audio file.
// ok is false when the resolved root holds fewer than two tracks — the
// caller falls back to the single-file ingestion path.
func bookRootForTrackFile(path, libraryRoot string) (root string, tracks []string, ok bool) {
	candidate, ok := bookRootDirForTrackFile(path, libraryRoot)
	if !ok {
		return "", nil, false
	}
	tracks = subtreeAudioTracks(candidate)
	return candidate, tracks, len(tracks) >= 2
}

// maybeProcessAudioBook resolves the owning library + root folder and
// hands an audio file to processAudioBookFile. Called from the scan walk
// BEFORE the single-file path; handled=true consumes the file.
func (s *MediaScanner) maybeProcessAudioBook(ctx context.Context, path string) (bool, error) {
	if !isAudioFile(path) {
		return false, nil
	}
	// Extension allowlisting already scopes audio to audiobook libraries,
	// so any scannable audio file belongs to one.
	rootFolder := ""
	for _, folder := range s.folders {
		if strings.HasPrefix(path, folder) {
			rootFolder = folder
			break
		}
	}
	if rootFolder == "" {
		return false, nil
	}

	libraryID := s.defaultLibraryID
	if !libraryID.Valid {
		lib, err := s.db.GetLibraryByFolder(ctx, rootFolder)
		if err != nil {
			return false, fmt.Errorf("failed to find library for audio file %s: %w", path, err)
		}
		libraryID = lib.LibraryID
	}
	return s.processAudioBookFile(ctx, path, libraryID, rootFolder)
}

// processAudioBookFile dispatches one audio file: zips become zip books,
// other audio resolves its folder-book root (falls back to the caller's
// single-file path when the root holds fewer than two tracks).
func (s *MediaScanner) processAudioBookFile(ctx context.Context, path string, libraryID pgtype.UUID, rootFolder string) (bool, error) {
	if strings.EqualFold(filepath.Ext(path), ".zip") {
		if err := s.processZipAudioBook(ctx, path, libraryID, rootFolder); err != nil {
			return true, err
		}
		return true, nil
	}

	root, ok := bookRootDirForTrackFile(path, rootFolder)
	if !ok {
		return false, nil // no audio-bearing ancestor: shouldn't happen
	}
	if s.audioBookRoots[root] {
		return true, nil // already ingested this walk (≥2 verified then)
	}

	tracks := subtreeAudioTracks(root)
	if len(tracks) < 2 {
		return false, nil // lone file: single-file book, existing path
	}
	s.audioBookRoots[root] = true

	if err := s.processAudioFolderBook(ctx, root, tracks, libraryID, rootFolder); err != nil {
		return true, err
	}
	return true, nil
}

// audioTrackInfo is the probed fact set for one track file.
type audioTrackInfo struct {
	relPath  string // library-relative (folder books) or entry name (zip books)
	title    string
	duration float64
	size     int64
	disc     int // tag disc number (TPOS/disk); 0 = absent
	trackNum int // tag track number (TRCK/trkn); 0 = absent
}

// parseTagOrdinal reads "N" or "N/M" tag values (disc, track) into N.
func parseTagOrdinal(v string) int {
	v = strings.TrimSpace(v)
	if idx := strings.Index(v, "/"); idx != -1 {
		v = v[:idx]
	}
	n, err := strconv.Atoi(strings.TrimSpace(v))
	if err != nil || n < 0 {
		return 0
	}
	return n
}

// applyTrackProbe fills one track's probed facts from its ffprobe
// report: duration, disc/track ordering keys, and the display title per
// the precedence chain (movement_name → title → caller's fallback).
func applyTrackProbe(report *ffprobeReport, info *audioTrackInfo) {
	if report == nil {
		return
	}
	info.duration = float64(report.Format.Duration)
	info.disc = parseTagOrdinal(report.tag("disc"))
	info.trackNum = parseTagOrdinal(report.tag("track"))
	if t := report.tag("movement_name"); t != "" {
		info.title = t
	} else if t := report.tag("title"); t != "" {
		info.title = t
	}
}

// orderTracksByTags reorders probed tracks by (disc, track) when EVERY
// track carries a track number — tag order then outranks filename
// guesses for multi-disc rips whose files are flat `01.mp3` dumps. Any
// gap in coverage falls back to the natural-sorted discovery order.
func orderTracksByTags(probed []audioTrackInfo) {
	if len(probed) < 2 {
		return
	}
	for _, tr := range probed {
		if tr.trackNum <= 0 {
			return
		}
	}
	sort.SliceStable(probed, func(x, y int) bool {
		if probed[x].disc != probed[y].disc {
			return probed[x].disc < probed[y].disc
		}
		if probed[x].trackNum != probed[y].trackNum {
			return probed[x].trackNum < probed[y].trackNum
		}
		return naturalLess(probed[x].relPath, probed[y].relPath)
	})
}

// playlistHint is the ordering/title intent of an m3u/m3u8 playlist in a
// book root — a curator's explicit track order and names.
type playlistHint struct {
	refs   []string          // entry refs in playlist order (slash-normalized)
	titles map[string]string // normalized ref → #EXTINF title
}

// findPlaylistHint parses THE playlist of a book root (exactly one
// .m3u/.m3u8 among its direct children); zero or several → no hint.
func findPlaylistHint(rootDir string) *playlistHint {
	matches, err := filepath.Glob(filepath.Join(rootDir, "*.m3u"))
	if err != nil {
		matches = nil
	}
	m8, err := filepath.Glob(filepath.Join(rootDir, "*.m3u8"))
	if err == nil {
		matches = append(matches, m8...)
	}
	if len(matches) != 1 {
		return nil
	}

	f, err := os.Open(matches[0])
	if err != nil {
		return nil
	}
	defer func() { _ = f.Close() }()

	hint := &playlistHint{titles: map[string]string{}}
	pendingTitle := ""
	scanner := bufio.NewScanner(f)
	for scanner.Scan() {
		line := strings.TrimSpace(scanner.Text())
		if line == "" {
			continue
		}
		if strings.HasPrefix(line, "#") {
			if strings.HasPrefix(strings.ToUpper(line), "#EXTINF:") {
				if idx := strings.Index(line, ","); idx != -1 {
					pendingTitle = strings.TrimSpace(line[idx+1:])
				}
			}
			continue
		}
		ref := filepath.ToSlash(line)
		ref = strings.TrimPrefix(ref, "./")
		hint.refs = append(hint.refs, ref)
		if pendingTitle != "" {
			hint.titles[strings.ToLower(ref)] = pendingTitle
		}
		pendingTitle = ""
	}
	if len(hint.refs) == 0 {
		return nil
	}
	return hint
}

// reorder returns the absolute track paths in playlist order (refs
// matched by relative path first, then by basename), with any tracks
// the playlist didn't mention appended in their natural order.
func (h *playlistHint) reorder(absTracks []string) []string {
	byRel := make(map[string]string, len(absTracks))
	byBase := make(map[string][]string, len(absTracks))
	for _, abs := range absTracks {
		rel := filepath.ToSlash(abs)
		byRel[strings.ToLower(rel)] = abs
		byBase[strings.ToLower(filepath.Base(abs))] = append(byBase[strings.ToLower(filepath.Base(abs))], abs)
	}

	ordered := make([]string, 0, len(absTracks))
	used := make(map[string]bool, len(absTracks))
	for _, ref := range h.refs {
		abs, ok := byRel[strings.ToLower(ref)]
		if !ok {
			if candidates := byBase[strings.ToLower(filepath.Base(ref))]; len(candidates) > 0 {
				abs = candidates[0]
				ok = true
			}
		}
		if ok && !used[abs] {
			ordered = append(ordered, abs)
			used[abs] = true
		}
	}
	for _, abs := range absTracks {
		if !used[abs] {
			ordered = append(ordered, abs)
		}
	}
	return ordered
}

// titleFor returns the playlist's #EXTINF title for an absolute track
// path (matched by relative path, then basename); "" when none.
func (h *playlistHint) titleFor(abs string) string {
	if h == nil {
		return ""
	}
	if t, ok := h.titles[strings.ToLower(filepath.ToSlash(abs))]; ok {
		return t
	}
	return h.titles[strings.ToLower(filepath.Base(abs))]
}

// cleanTrackTitle derives a display title from a filename when the track
// carries no tag: strips extension, leading track numbers and separators.
func cleanTrackTitle(path string) string {
	base := filepath.Base(path)
	title := strings.TrimSpace(strings.TrimSuffix(base, filepath.Ext(base)))
	trimmed := strings.TrimLeft(title, "0123456789")
	if trimmed != title && trimmed != "" {
		title = strings.TrimLeft(trimmed, " -_.")
	}
	if title == "" {
		title = base
	}
	return title
}

// buildMultiFileAudioPayload assembles the audio_metadata payload for a
// multi-file book: layout, cumulative track start/end, chapters (sidecar
// chapters win; otherwise the track boundaries ARE the chapters) and the
// aggregate duration.
func buildMultiFileAudioPayload(layout string, payload *AudioMetadataPayload, tracks []audioTrackInfo) *AudioMetadataPayload {
	if payload == nil {
		payload = &AudioMetadataPayload{}
	}
	payload.Layout = layout
	payload.Tracks = nil
	clock := 0.0
	for i, tr := range tracks {
		end := clock + tr.duration
		payload.Tracks = append(payload.Tracks, AudioTrack{
			Index:           i,
			Path:            tr.relPath,
			Title:           tr.title,
			Start:           clock,
			End:             end,
			DurationSeconds: tr.duration,
			SizeBytes:       tr.size,
		})
		clock = end
	}
	payload.DurationSeconds = clock

	// Curated sidecar chapters describe the whole book in global seconds
	// and keep priority; otherwise tracks are the chapters (the
	// documented folder-as-book shape: same {index,title,start,end}).
	if len(payload.Chapters) == 0 && len(payload.Tracks) > 0 {
		for _, tr := range payload.Tracks {
			payload.Chapters = append(payload.Chapters, AudioChapter{
				Index: tr.Index,
				Title: tr.Title,
				Start: tr.Start,
				End:   tr.End,
			})
		}
	}
	return payload
}

// extractAudioBookMetadata gathers book-level metadata for a multi-file
// book: sidecars next to the representative path (a sentinel inside the
// book root for folders, the zip itself for archives — the sidecar checks
// all look at Dir(representative)), then folder-structure patterns, then
// the first track's tags (album/artist/composer/grouping/freeform atoms
// via applyAudioTagFallbacks), then fallbackTitle (the book directory
// name for folder books, the zip name for zip books — passed explicitly
// because Dir(representative) means different things per layout).
// Embedded cover art is the cover fallback of last resort: sidecar
// covers win. firstTrackAbs is the first track's absolute path for
// folder books ("" for zip books — their first entry is temp-extracted
// on demand).
func (s *MediaScanner) extractAudioBookMetadata(representative, rootFolder string, firstReport *ffprobeReport, firstTrackAbs, fallbackTitle string) *MediaMetadata {
	extracted, err := s.extractMetadata(representative)
	var metadata *MediaMetadata
	if err != nil || extracted == nil {
		metadata = &MediaMetadata{}
	} else {
		metadata = extracted
	}
	// The folder-book representative is a sentinel INSIDE the root (the
	// sidecar checks all look at Dir(representative)). The generic
	// extractor's filename-title fallback would name the book
	// ".audiobook-root" and every real source below would see a non-empty
	// title — strip it so sidecars, folder structure, tags and finally
	// the directory name get their turn.
	if metadata.Title != "" && strings.HasPrefix(filepath.Base(representative), ".") &&
		metadata.Title == strings.TrimSuffix(filepath.Base(representative), filepath.Ext(representative)) {
		metadata.Title = ""
	}
	if metadata.AudioMeta == nil {
		metadata.AudioMeta = readSidecarAudio(representative)
	}

	if folderMetadata := s.extractFolderStructureMetadata(representative, rootFolder); folderMetadata != nil {
		if metadata.Title == "" && folderMetadata.Title != "" {
			metadata.Title = folderMetadata.Title
		}
		if metadata.Author == "" && folderMetadata.Author != "" {
			metadata.Author = folderMetadata.Author
		}
		if metadata.Series == "" && folderMetadata.Series != "" {
			metadata.Series = folderMetadata.Series
		}
		if metadata.SeriesNumber == 0 && folderMetadata.SeriesNumber > 0 {
			metadata.SeriesNumber = folderMetadata.SeriesNumber
		}
	}

	// Tag fallbacks run after sidecars and folder structure — the shared
	// helper fills title/author/narrators/series/date/genre from the
	// first track's tags (album/artist/composer/grouping/date/genre plus
	// Audible freeform atoms), then embedded cover art as the cover
	// fallback of last resort.
	applyAudioTagFallbacks(metadata, firstReport)

	if metadata.Title == "" && fallbackTitle != "" {
		metadata.Title = strings.ReplaceAll(fallbackTitle, "_", " ")
	}
	if metadata.Author == "" {
		metadata.Author = "Unknown"
	}
	if metadata.CoverPath == "" {
		if cover := findSidecarCover(representative); cover != "" {
			metadata.CoverPath = s.getRelativePath(cover)
		}
	}
	if metadata.CoverPath == "" && firstTrackAbs != "" {
		if cover := extractEmbeddedAudioCover(firstTrackAbs, representative); cover != "" {
			metadata.CoverPath = s.getRelativePath(cover)
		}
	}
	return metadata
}

// createAudioBookItem inserts the media_items row shared by folder and
// zip books. The hash columns stay NULL — content dedup is path-based for
// multi-file books (there is no single byte stream to hash).
func (s *MediaScanner) createAudioBookItem(ctx context.Context, libraryID pgtype.UUID, metadata *MediaMetadata, relPath string, totalSize int64, mimeType string, modTime pgtype.Timestamptz) (database.MediaItems, error) {
	return s.db.CreateMediaItem(ctx, database.CreateMediaItemParams{
		LibraryID:       libraryID,
		Title:           metadata.Title,
		Author:          pgtype.Text{String: metadata.Author, Valid: metadata.Author != ""},
		Isbn:            pgtype.Text{String: utils.NormalizeISBNSafe(metadata.ISBN), Valid: metadata.ISBN != ""},
		Asin:            pgtype.Text{String: metadata.ASIN, Valid: metadata.ASIN != ""},
		Description:     pgtype.Text{String: metadata.Description, Valid: metadata.Description != ""},
		FilePath:        relPath,
		FileSize:        pgtype.Int8{Int64: totalSize, Valid: true},
		MimeType:        pgtype.Text{String: mimeType, Valid: mimeType != ""},
		CoverImagePath:  pgtype.Text{String: metadata.CoverPath, Valid: metadata.CoverPath != ""},
		Series:          pgtype.Text{String: metadata.Series, Valid: metadata.Series != ""},
		SeriesNumber:    pgtype.Int4{Int32: metadata.SeriesNumber, Valid: metadata.SeriesNumber > 0},
		Publisher:       pgtype.Text{String: metadata.Publisher, Valid: metadata.Publisher != ""},
		DatePublished:   pgtype.Date{Time: metadata.PublishDate, Valid: !metadata.PublishDate.IsZero()},
		Genre:           pgtype.Text{String: metadata.Genre, Valid: metadata.Genre != ""},
		Contributors:    utils.NormalizeContributors(metadata.Contributors),
		Narrators:       nilToEmpty(metadata.Narrators),
		NarratorsSearch: utils.NormalizeNarratorsSearch(metadata.Narrators),
		Abridged:        metadata.Abridged,
		Tags:            utils.NormalizeTags(metadata.Tags),
		AddedByAdminID:  s.adminID,
		CreatedAt:       modTime,
		ImportedAt:      pgtype.Timestamptz{Time: time.Now(), Valid: true},
		AudioMetadata:   marshalAudioMetadata(metadata.AudioMeta),
	})
}

// setAudioFormatGroup writes the audiobook format facts after create.
func (s *MediaScanner) setAudioFormatGroup(ctx context.Context, itemID pgtype.UUID, mimeType string, chapterCount int) error {
	return s.db.UpdateMediaItemFormatGroup(ctx, database.UpdateMediaItemFormatGroupParams{
		ID:             itemID,
		FormatGroup:    "audiobook",
		FormatMimetype: pgtype.Text{String: mimeType, Valid: mimeType != ""},
		IsReflowable:   pgtype.Bool{Bool: false, Valid: true},
		HasFixedLayout: pgtype.Bool{Bool: false, Valid: true},
		ChapterCount:   pgtype.Int4{Int32: int32(chapterCount), Valid: chapterCount > 0},
	})
}

// audioPayloadTracksEqual reports whether the stored audio_metadata
// already describes exactly the given track manifest (paths + sizes).
// Timestamp-free manifests make rescans idempotent.
func audioPayloadTracksEqual(stored []byte, payload *AudioMetadataPayload) bool {
	if len(stored) == 0 {
		return false
	}
	var storedPayload AudioMetadataPayload
	if err := json.Unmarshal(stored, &storedPayload); err != nil {
		return false
	}
	if storedPayload.Layout != payload.Layout || len(storedPayload.Tracks) != len(payload.Tracks) {
		return false
	}
	for i, tr := range storedPayload.Tracks {
		if tr.Path != payload.Tracks[i].Path || tr.SizeBytes != payload.Tracks[i].SizeBytes {
			return false
		}
	}
	return true
}

// absorbOneTrackItems removes the one-track audiobook items a previous
// scan ingested individually from the tracks now grouped under a folder
// book root. Annotations/progress cascade with the rows, and multi-file
// sources never carried listening progress — nothing real is lost. Only
// audio files under the root are deleted; real books and the folder book
// row itself are never touched.
func (s *MediaScanner) absorbOneTrackItems(ctx context.Context, libraryID pgtype.UUID, rootRel string) {
	items, err := s.db.ListMediaItemsByLibraryIncludingArchived(ctx, libraryID)
	if err != nil {
		fmt.Printf("Warning: absorb pass could not list library items for %s: %v\n", rootRel, err)
		return
	}
	prefix := rootRel + "/"
	removed := 0
	for _, item := range items {
		if !strings.HasPrefix(item.FilePath, prefix) || !isAudioTrackFile(item.FilePath) {
			continue
		}
		if err := s.db.DeleteMediaItem(ctx, item.ID); err != nil {
			fmt.Printf("Warning: failed to absorb one-track item %s into folder book: %v\n", item.FilePath, err)
			continue
		}
		fmt.Printf("[AUDIO-MERGE] Absorbed one-track item '%s' into folder book %s\n", item.Title, rootRel)
		s.logger.LogDelete(fmt.Sprintf("[AUDIO-MERGE] Absorbed one-track item '%s' into folder book '%s'", item.Title, rootRel))
		removed++
	}
	if removed > 0 {
		fmt.Printf("[AUDIO-MERGE] %d one-track item(s) absorbed into %s\n", removed, rootRel)
	}
}

// audioLayoutOf reads the layout key from a stored audio_metadata
// payload; "" when absent (single-file book or not audio).
func audioLayoutOf(stored []byte) string {
	if len(stored) == 0 {
		return ""
	}
	var probe struct {
		Layout string `json:"layout"`
	}
	if err := json.Unmarshal(stored, &probe); err != nil {
		return ""
	}
	return probe.Layout
}

// rescanMultiFileAudioBook is the per-item rescan for folder/zip books:
// rebuild the track manifest from disk and re-derive book metadata
// (sidecars, folder structure, first-track tags), then write through the
// shared editorial updater so override semantics match single-file
// books. The generic per-file updater is bypassed by the caller — see
// RescanMediaItem.
func (s *MediaScanner) rescanMultiFileAudioBook(ctx context.Context, item database.MediaItems, fullPath string) error {
	layout := audioLayoutOf(item.AudioMetadata)

	var representative string
	var probed []audioTrackInfo
	switch layout {
	case "folder":
		tracks := subtreeAudioTracks(fullPath)
		if len(tracks) == 0 {
			return fmt.Errorf("folder book %s has no tracks on disk", item.FilePath)
		}
		// Sentinel INSIDE the root: sidecar checks look at Dir().
		representative = filepath.Join(fullPath, ".audiobook-root")
		for _, abs := range tracks {
			info := audioTrackInfo{
				relPath: s.getRelativePath(abs),
				title:   cleanTrackTitle(abs),
			}
			if st, err := os.Stat(abs); err == nil {
				info.size = st.Size()
			}
			if report := runFFProbe(abs); report != nil {
				applyTrackProbe(report, &info)
			}
			probed = append(probed, info)
		}
	case "zip":
		representative = fullPath
		entries, err := zipAudioEntries(fullPath)
		if err != nil {
			return fmt.Errorf("failed to read audio zip %s: %w", fullPath, err)
		}
		for _, e := range entries {
			info := audioTrackInfo{
				relPath: e.name,
				title:   cleanTrackTitle(e.name),
				size:    e.size,
			}
			if report := probeZipEntry(fullPath, e.name); report != nil {
				applyTrackProbe(report, &info)
			}
			probed = append(probed, info)
		}
	default:
		return fmt.Errorf("unknown audio layout %q for %s", layout, item.FilePath)
	}
	if len(probed) == 0 {
		return fmt.Errorf("multi-file book %s has no tracks", item.FilePath)
	}
	orderTracksByTags(probed)

	// First track's tags name the book when sidecars/folders don't.
	var firstReport *ffprobeReport
	switch layout {
	case "folder":
		firstReport = runFFProbe(fullPath + string(filepath.Separator) + probed[0].relPath)
	case "zip":
		firstReport = probeZipEntry(fullPath, probed[0].relPath)
	}

	rootFolder := ""
	for _, folder := range s.folders {
		if strings.HasPrefix(fullPath, folder) {
			rootFolder = folder
			break
		}
	}

	metadata := s.extractAudioBookMetadata(representative, rootFolder, firstReport, firstTrackAbsFor(layout, fullPath, probed), layoutFallbackTitle(layout, fullPath))
	payload := buildMultiFileAudioPayload(layout, metadata.AudioMeta, probed)
	metadata.AudioMeta = payload
	if len(payload.Narrators) > 0 && len(metadata.Narrators) == 0 {
		metadata.Narrators = payload.Narrators
	}
	if payload.ASIN != "" && metadata.ASIN == "" {
		metadata.ASIN = payload.ASIN
	}
	metadata.Contributors = utils.NormalizeContributors(metadata.Contributors)
	metadata.Tags = utils.NormalizeTags(metadata.Tags)

	if err := s.updateMediaItemWithMetadata(ctx, item, representative, metadata); err != nil {
		return fmt.Errorf("failed to update multi-file book %s: %w", item.FilePath, err)
	}

	// file_size/chapter_count ride the focused payload query (the generic
	// update doesn't carry file_size).
	var totalSize int64
	for _, tr := range probed {
		totalSize += tr.size
	}
	firstMime := mimeForAudioEntry(probed[0].relPath)
	if layout == "folder" {
		firstMime = s.getMimeType(fullPath + string(filepath.Separator) + probed[0].relPath)
	}
	if _, err := s.db.UpdateMediaItemAudioBookPayload(ctx, database.UpdateMediaItemAudioBookPayloadParams{
		ID:            item.ID,
		FileSize:      pgtype.Int8{Int64: totalSize, Valid: true},
		MimeType:      pgtype.Text{String: firstMime, Valid: firstMime != ""},
		AudioMetadata: marshalAudioMetadata(payload),
		ChapterCount:  pgtype.Int4{Int32: int32(len(payload.Chapters)), Valid: len(payload.Chapters) > 0},
	}); err != nil {
		return fmt.Errorf("failed to refresh audio book payload %s: %w", item.FilePath, err)
	}

	fmt.Printf("[AUDIO] Rescanned %s book '%s' (%d tracks)\n", layout, item.Title, len(probed))
	return nil
}

// layoutFallbackTitle is the last-resort book title per layout: the book
// directory name for folder books, the zip's own name (extension
// stripped) for zip books.
func layoutFallbackTitle(layout, fullPath string) string {
	if layout == "zip" {
		return strings.TrimSuffix(filepath.Base(fullPath), filepath.Ext(fullPath))
	}
	return filepath.Base(fullPath)
}

// firstTrackAbsFor returns the first track's absolute path for folder
// books (zip books carry in-archive entries — no absolute path).
func firstTrackAbsFor(layout, rootPath string, probed []audioTrackInfo) string {
	if layout == "folder" && len(probed) > 0 {
		return filepath.Join(rootPath, probed[0].relPath)
	}
	return ""
}

// audioTrackCountOf reads the stored tracks-array length from an
// audio_metadata payload; -1 when absent/unparseable.
func audioTrackCountOf(stored []byte) int {
	if len(stored) == 0 {
		return -1
	}
	var probe struct {
		Tracks []json.RawMessage `json:"tracks"`
	}
	if err := json.Unmarshal(stored, &probe); err != nil {
		return -1
	}
	return len(probe.Tracks)
}

// multiFileBookUnchanged is the rescan fast path: track count and total
// size match the stored row AND the stored manifest carries real
// durations, so the (expensive) per-track ffprobe pass is skipped. A
// zero book duration means the last scan ran without ffprobe — treat it
// as changed so a ffprobe-equipped rescan can fill durations in (the
// probe pass is nearly free in ffprobe-less environments anyway).
// Book-level metadata is not re-derived on unchanged scans — a
// metadata.json edit needs a force rescan or a track change to pick up,
// same tradeoff the single-file path makes on size-unchanged skips.
func multiFileBookUnchanged(existing database.MediaItems, totalSize int64, trackCount int) bool {
	if existing.FileSize.Int64 != totalSize {
		return false
	}
	if audioTrackCountOf(existing.AudioMetadata) != trackCount {
		return false
	}
	var probe struct {
		DurationSeconds float64 `json:"duration_seconds"`
	}
	if len(existing.AudioMetadata) > 0 {
		_ = json.Unmarshal(existing.AudioMetadata, &probe)
	}
	return probe.DurationSeconds > 0
}

// processAudioFolderBook ingests or refreshes one folder-of-tracks book.
// The item's file_path is the book root directory (library-relative);
// file_size is the track-size sum; audio_metadata carries layout "folder"
// plus the track manifest.
func (s *MediaScanner) processAudioFolderBook(ctx context.Context, rootDir string, tracks []string, libraryID pgtype.UUID, rootFolder string) error {
	rootRel := s.getRelativePath(rootDir)
	fmt.Printf("[AUDIO] Folder book at %s (%d tracks)\n", rootRel, len(tracks))

	// Cheap pass first: track sizes only. The per-track ffprobe pass
	// below is expensive (a 1000-track book must not re-probe on every
	// auto-scan), so the existence check runs before it.
	var totalSize int64
	for _, abs := range tracks {
		if st, err := os.Stat(abs); err == nil {
			totalSize += st.Size()
		}
	}

	existing, existingErr := s.getMediaItemByFilePath(ctx, rootRel, libraryID)
	if existingErr == nil && !s.forceRescan && multiFileBookUnchanged(existing, totalSize, len(tracks)) {
		if existing.ArchivedAt.Valid || existing.MissingScanCount > 0 {
			if err := s.db.ClearMediaItemArchive(ctx, existing.ID); err != nil {
				fmt.Printf("Warning: failed to unarchive folder book %s: %v\n", rootRel, err)
			} else {
				fmt.Printf("[AUDIO] Restored folder book '%s' (root is back)\n", existing.Title)
			}
		}
		// One-track strays from pre-grouping scans go away either way.
		s.absorbOneTrackItems(ctx, libraryID, rootRel)
		return nil
	}

	// Playlist hint: a lone m3u/m3u8 in the root is a curator's explicit
	// order (and #EXTINF names) for the track list.
	hint := findPlaylistHint(rootDir)
	if hint != nil {
		tracks = hint.reorder(tracks)
		fmt.Printf("[AUDIO] Playlist ordering applied to %s (%d refs)\n", rootRel, len(hint.refs))
	}

	probed := make([]audioTrackInfo, 0, len(tracks))
	var firstReport *ffprobeReport
	for i, abs := range tracks {
		info := audioTrackInfo{
			relPath: s.getRelativePath(abs),
			title:   cleanTrackTitle(abs),
		}
		if t := hint.titleFor(abs); t != "" {
			info.title = t
		}
		if st, err := os.Stat(abs); err == nil {
			info.size = st.Size()
		}
		report := runFFProbe(abs)
		if report != nil {
			applyTrackProbe(report, &info)
			if i == 0 {
				firstReport = report
			}
		}
		probed = append(probed, info)
	}
	if len(probed) == 0 {
		return fmt.Errorf("folder book %s has no probeable tracks", rootRel)
	}
	orderTracksByTags(probed)

	// The representative path drives the sidecar/folder-structure checks;
	// they all look at Dir(representative), so a sentinel INSIDE the root
	// resolves covers and metadata.json living there (including disc-
	// layout roots where track zero sits deeper).
	sentinel := filepath.Join(rootDir, ".audiobook-root")
	metadata := s.extractAudioBookMetadata(sentinel, rootFolder, firstReport, tracks[0], filepath.Base(rootDir))
	payload := buildMultiFileAudioPayload("folder", metadata.AudioMeta, probed)
	metadata.AudioMeta = payload
	if len(payload.Narrators) > 0 && len(metadata.Narrators) == 0 {
		metadata.Narrators = payload.Narrators
	}
	if payload.ASIN != "" && metadata.ASIN == "" {
		metadata.ASIN = payload.ASIN
	}
	metadata.Contributors = utils.NormalizeContributors(metadata.Contributors)
	metadata.Tags = utils.NormalizeTags(metadata.Tags)

	firstMime := s.getMimeType(tracks[0])
	var modTime pgtype.Timestamptz
	if rootInfo, err := os.Stat(rootDir); err == nil {
		modTime = pgtype.Timestamptz{Time: rootInfo.ModTime(), Valid: true}
	}

	err := existingErr
	switch {
	case err == nil:
		changed := existing.FileSize.Int64 != totalSize || !audioPayloadTracksEqual(existing.AudioMetadata, payload)
		if changed || s.forceRescan {
			if err := s.updateAudioBookItem(ctx, existing, payload, firstMime, totalSize, len(payload.Chapters)); err != nil {
				return err
			}
		} else if existing.ArchivedAt.Valid || existing.MissingScanCount > 0 {
			// Unchanged content that was archived while the root was gone.
			if err := s.db.ClearMediaItemArchive(ctx, existing.ID); err != nil {
				fmt.Printf("Warning: failed to unarchive folder book %s: %v\n", rootRel, err)
			} else {
				fmt.Printf("[AUDIO] Restored folder book '%s' (root is back)\n", existing.Title)
			}
		}
	case errors.Is(err, pgx.ErrNoRows):
		created, err := s.createAudioBookItem(ctx, libraryID, metadata, rootRel, totalSize, firstMime, modTime)
		if err != nil {
			return fmt.Errorf("failed to create folder book %s: %w", rootRel, err)
		}
		if err := s.setAudioFormatGroup(ctx, created.ID, firstMime, len(payload.Chapters)); err != nil {
			fmt.Printf("Warning: failed to set format group for folder book %s: %v\n", rootRel, err)
		}
		fmt.Printf("[AUDIO] Created folder book '%s' (%d tracks) at %s\n", metadata.Title, len(tracks), rootRel)
		s.newItems++
	default:
		return fmt.Errorf("failed to check folder book existence %s: %w", rootRel, err)
	}

	// One-track strays from pre-grouping scans go away either way.
	s.absorbOneTrackItems(ctx, libraryID, rootRel)
	return nil
}

// updateAudioBookItem refreshes an existing folder/zip book row after a
// rescan saw the track set change (or the admin forced a rescan). Only
// scan-derived fields move — editorial columns keep admin edits.
func (s *MediaScanner) updateAudioBookItem(ctx context.Context, existing database.MediaItems, payload *AudioMetadataPayload, mimeType string, totalSize int64, chapterCount int) error {
	if _, err := s.db.UpdateMediaItemAudioBookPayload(ctx, database.UpdateMediaItemAudioBookPayloadParams{
		ID:            existing.ID,
		FileSize:      pgtype.Int8{Int64: totalSize, Valid: true},
		MimeType:      pgtype.Text{String: mimeType, Valid: mimeType != ""},
		AudioMetadata: marshalAudioMetadata(payload),
		ChapterCount:  pgtype.Int4{Int32: int32(chapterCount), Valid: chapterCount > 0},
	}); err != nil {
		return fmt.Errorf("failed to update audio book %s: %w", existing.FilePath, err)
	}
	fmt.Printf("[AUDIO] Refreshed multi-file book '%s' (%d tracks)\n", existing.Title, len(payload.Tracks))
	return nil
}

// processZipAudioBook ingests a zip archive of audio tracks as ONE book.
// The zip is the item's file (downloadable like any single file);
// audio_metadata carries layout "zip" and per-entry track paths the
// streaming endpoint extracts on demand.
func (s *MediaScanner) processZipAudioBook(ctx context.Context, zipPath string, libraryID pgtype.UUID, rootFolder string) error {
	entries, err := zipAudioEntries(zipPath)
	if err != nil {
		return fmt.Errorf("failed to read audio zip %s: %w", zipPath, err)
	}
	if len(entries) == 0 {
		fmt.Printf("[AUDIO] Skipping %s: zip contains no audio tracks\n", zipPath)
		return nil
	}

	zipRel := s.getRelativePath(zipPath)
	fmt.Printf("[AUDIO] Zip book at %s (%d tracks)\n", zipRel, len(entries))

	// Cheap pass first: central-directory sizes only. The per-entry
	// extract+ffprobe pass is expensive and skipped when the manifest is
	// unchanged (see multiFileBookUnchanged).
	var totalSize int64
	for _, entry := range entries {
		totalSize += entry.size
	}

	existing, existingErr := s.getMediaItemByFilePath(ctx, zipRel, libraryID)
	if existingErr == nil && !s.forceRescan && multiFileBookUnchanged(existing, totalSize, len(entries)) {
		return nil
	}

	probed := make([]audioTrackInfo, 0, len(entries))
	var firstReport *ffprobeReport
	for i, entry := range entries {
		info := audioTrackInfo{
			relPath: entry.name,
			title:   cleanTrackTitle(entry.name),
			size:    entry.size,
		}
		report := probeZipEntry(zipPath, entry.name)
		if report != nil {
			applyTrackProbe(report, &info)
			if i == 0 {
				firstReport = report
			}
		}
		probed = append(probed, info)
	}
	orderTracksByTags(probed)

	metadata := s.extractAudioBookMetadata(zipPath, rootFolder, firstReport, "", layoutFallbackTitle("zip", zipPath))
	if metadata.CoverPath == "" && len(entries) > 0 {
		// Zip books keep their art as the zip's sibling "<zip>.cover.jpg";
		// the first entry is temp-extracted for ffmpeg to read.
		if tmp, cleanup, err := zipEntryToTemp(zipPath, entries[0].name); err == nil {
			if cover := extractEmbeddedAudioCover(tmp, zipPath); cover != "" {
				metadata.CoverPath = s.getRelativePath(cover)
			}
			cleanup()
		}
	}
	payload := buildMultiFileAudioPayload("zip", metadata.AudioMeta, probed)
	metadata.AudioMeta = payload
	if len(payload.Narrators) > 0 && len(metadata.Narrators) == 0 {
		metadata.Narrators = payload.Narrators
	}
	if payload.ASIN != "" && metadata.ASIN == "" {
		metadata.ASIN = payload.ASIN
	}
	metadata.Contributors = utils.NormalizeContributors(metadata.Contributors)
	metadata.Tags = utils.NormalizeTags(metadata.Tags)

	firstMime := mimeForAudioEntry(entries[0].name)
	var modTime pgtype.Timestamptz
	if info, err := os.Stat(zipPath); err == nil {
		modTime = pgtype.Timestamptz{Time: info.ModTime(), Valid: true}
	}

	err = existingErr
	switch {
	case err == nil:
		changed := existing.FileSize.Int64 != totalSize || !audioPayloadTracksEqual(existing.AudioMetadata, payload)
		if changed || s.forceRescan {
			if err := s.updateAudioBookItem(ctx, existing, payload, firstMime, totalSize, len(payload.Chapters)); err != nil {
				return err
			}
		}
	case errors.Is(err, pgx.ErrNoRows):
		created, err := s.createAudioBookItem(ctx, libraryID, metadata, zipRel, totalSize, firstMime, modTime)
		if err != nil {
			return fmt.Errorf("failed to create zip book %s: %w", zipRel, err)
		}
		if err := s.setAudioFormatGroup(ctx, created.ID, firstMime, len(payload.Chapters)); err != nil {
			fmt.Printf("Warning: failed to set format group for zip book %s: %v\n", zipRel, err)
		}
		fmt.Printf("[AUDIO] Created zip book '%s' (%d tracks) at %s\n", metadata.Title, len(entries), zipRel)
		s.newItems++
	default:
		return fmt.Errorf("failed to check zip book existence %s: %w", zipRel, err)
	}
	return nil
}

// zipAudioEntry is one audio track inside a zip book.
type zipAudioEntry struct {
	name string
	size int64
}

// zipAudioEntries lists the audio entries of a zip, natural-sorted,
// skipping directories and macOS/hidden cruft.
func zipAudioEntries(zipPath string) ([]zipAudioEntry, error) {
	r, err := zip.OpenReader(zipPath)
	if err != nil {
		return nil, err
	}
	defer func() { _ = r.Close() }()

	var entries []zipAudioEntry
	for _, f := range r.File {
		if f.FileInfo().IsDir() {
			continue
		}
		name := filepath.ToSlash(f.Name)
		base := strings.ToLower(filepath.Base(name))
		if strings.HasPrefix(base, ".") || strings.Contains(name, "__MACOSX") {
			continue
		}
		if !isAudioFile(name) {
			continue
		}
		entries = append(entries, zipAudioEntry{name: name, size: int64(f.UncompressedSize64)})
	}
	sort.Slice(entries, func(x, y int) bool { return naturalLess(entries[x].name, entries[y].name) })
	return entries, nil
}

// zipEntryToTemp extracts one zip entry to a temp file (seekable input
// for ffprobe/ffmpeg). The caller MUST run the returned cleanup.
func zipEntryToTemp(zipPath, entryName string) (string, func(), error) {
	r, err := zip.OpenReader(zipPath)
	if err != nil {
		return "", nil, err
	}

	var entry *zip.File
	for _, f := range r.File {
		if filepath.ToSlash(f.Name) == filepath.ToSlash(entryName) {
			entry = f
			break
		}
	}
	if entry == nil {
		_ = r.Close()
		return "", nil, fmt.Errorf("entry %q not found", entryName)
	}

	src, err := entry.Open()
	if err != nil {
		_ = r.Close()
		return "", nil, err
	}

	tmp, err := os.CreateTemp("", "bookhoard-zipentry-*"+filepath.Ext(entryName))
	if err != nil {
		_ = src.Close()
		_ = r.Close()
		return "", nil, err
	}
	if _, err := io.Copy(tmp, src); err != nil {
		_ = tmp.Close()
		_ = src.Close()
		_ = r.Close()
		_ = os.Remove(tmp.Name())
		return "", nil, err
	}
	if err := tmp.Close(); err != nil {
		_ = src.Close()
		_ = r.Close()
		_ = os.Remove(tmp.Name())
		return "", nil, err
	}
	_ = src.Close()
	tmpName := tmp.Name()
	cleanup := func() {
		_ = r.Close()
		_ = os.Remove(tmpName)
	}
	return tmpName, cleanup, nil
}

// probeZipEntry ffprobes one zip entry by extracting it to a temp file
// (ffprobe wants a seekable path). The temp file is always removed.
func probeZipEntry(zipPath, entryName string) *ffprobeReport {
	tmp, cleanup, err := zipEntryToTemp(zipPath, entryName)
	if err != nil {
		fmt.Printf("Warning: zip entry probe failed for %s/%s: %v\n", zipPath, entryName, err)
		return nil
	}
	defer cleanup()
	return runFFProbe(tmp)
}

// mimeForAudioEntry maps an in-zip track name to its mime type.
func mimeForAudioEntry(name string) string {
	if m := MimeTypes[strings.ToLower(filepath.Ext(name))]; m != "" {
		return m
	}
	return "audio/mpeg"
}