fix: same-titled albums of two release groups are two albums, and a missing track with no disc is on disc 1

The scanner found an album by artist and title alone, so Weezer's six
self-titled albums (1994–2026) were one 118-track row and their merges,
wants and upgrades counted them as one. A same-named row whose release group
is known and differs from the file's is now another album; tagReadVersion 7
re-reads the library so existing tracks move to their own rows.

WantHoldings compared discs with IS NOT DISTINCT FROM, so a missing track 6
with no disc was never covered by a present disc 1 track 6. Air's Love 2
stayed wanted and was downloaded again for nothing.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

internal/db/dbq/albums.sql.go

@@ -72,17 +72,27 @@ func (q *Queries) DeleteAlbumIfEmpty(ctx context.Context, id pgtype.UUID) (Delet
 }
 
 const getAlbumByArtistAndTitle = `-- name: GetAlbumByArtistAndTitle :one
-SELECT id, title, sort_title, artist_id, release_date, mbid, cover_art_path, created_at, updated_at, cover_art_source, cover_art_sources_version, release_group_mbid FROM albums WHERE artist_id = $1 AND title = $2 LIMIT 1
+SELECT id, title, sort_title, artist_id, release_date, mbid, cover_art_path, created_at, updated_at, cover_art_source, cover_art_sources_version, release_group_mbid FROM albums
+ WHERE artist_id = $1 AND title = $2
+   AND ($3::text IS NULL OR release_group_mbid IS NULL
+        OR release_group_mbid = $3)
+ ORDER BY (release_group_mbid IS NOT DISTINCT FROM $3) DESC, id
+ LIMIT 1
 `
 
 type GetAlbumByArtistAndTitleParams struct {
-	ArtistID pgtype.UUID
+	ArtistID         pgtype.UUID
-	Title    string
+	Title            string
+	ReleaseGroupMbid *string
 }
 
-// Scanner uses this for the no-mbid dedupe path: resolve-or-create.
+// Scanner uses this for the no-mbid dedupe path: resolve-or-create. An album
+// of the same name in another release group is a different album (Weezer's
+// six self-titled albums shared one row until 2026-10-20), so a row whose
+// group is known and differs is not returned. A row of the file's own group
+// comes first, then one with no group yet.
 func (q *Queries) GetAlbumByArtistAndTitle(ctx context.Context, arg GetAlbumByArtistAndTitleParams) (Album, error) {
-	row := q.db.QueryRow(ctx, getAlbumByArtistAndTitle, arg.ArtistID, arg.Title)
+	row := q.db.QueryRow(ctx, getAlbumByArtistAndTitle, arg.ArtistID, arg.Title, arg.ReleaseGroupMbid)
 	var i Album
 	err := row.Scan(
 		&i.ID,

internal/db/dbq/wanted.sql.go

@@ -1572,11 +1572,13 @@ SELECT a.id AS album_id,
        COALESCE(bool_or(t.missing_since IS NOT NULL AND t.missing_since < $1::timestamptz), false)::bool AS missing_before,
        -- Missing tracks no track on disk stands in for, by position or
        -- title. A missing row whose song is back at a new path (#5245) is
-       -- covered; a lost bonus track is not.
+       -- covered; a lost bonus track is not. A track with no disc number is
+       -- on disc 1: Air's Love 2 stayed wanted, and was downloaded again for
+       -- nothing, over a missing track 6 with no disc beside a present 1/6.
        COUNT(*) FILTER (WHERE t.missing_since IS NOT NULL AND NOT EXISTS (
            SELECT 1 FROM tracks p
             WHERE p.album_id = a.id AND p.missing_since IS NULL
-              AND ((p.track_number = t.track_number AND p.disc_number IS NOT DISTINCT FROM t.disc_number)
+              AND ((p.track_number = t.track_number AND COALESCE(p.disc_number, 1) = COALESCE(t.disc_number, 1))
                    OR lower(p.title) = lower(t.title))))::int AS uncovered_missing
   FROM albums a
   JOIN tracks t ON t.album_id = a.id

internal/db/queries/albums.sql

@@ -36,8 +36,17 @@ WHERE albums.artist_id = $1
 ORDER BY release_date NULLS LAST, sort_title;
 
 -- name: GetAlbumByArtistAndTitle :one
--- Scanner uses this for the no-mbid dedupe path: resolve-or-create.
+-- Scanner uses this for the no-mbid dedupe path: resolve-or-create. An album
-SELECT * FROM albums WHERE artist_id = $1 AND title = $2 LIMIT 1;
+-- of the same name in another release group is a different album (Weezer's
+-- six self-titled albums shared one row until 2026-10-20), so a row whose
+-- group is known and differs is not returned. A row of the file's own group
+-- comes first, then one with no group yet.
+SELECT * FROM albums
+ WHERE artist_id = sqlc.arg(artist_id) AND title = sqlc.arg(title)
+   AND (sqlc.narg(release_group_mbid)::text IS NULL OR release_group_mbid IS NULL
+        OR release_group_mbid = sqlc.narg(release_group_mbid))
+ ORDER BY (release_group_mbid IS NOT DISTINCT FROM sqlc.narg(release_group_mbid)) DESC, id
+ LIMIT 1;
 
 -- name: ListAlbumsByArtist :many
 SELECT * FROM albums WHERE artist_id = $1 ORDER BY release_date NULLS LAST, sort_title;

internal/db/queries/wanted.sql

@@ -146,11 +146,13 @@ SELECT a.id AS album_id,
        COALESCE(bool_or(t.missing_since IS NOT NULL AND t.missing_since < sqlc.arg(since)::timestamptz), false)::bool AS missing_before,
        -- Missing tracks no track on disk stands in for, by position or
        -- title. A missing row whose song is back at a new path (#5245) is
-       -- covered; a lost bonus track is not.
+       -- covered; a lost bonus track is not. A track with no disc number is
+       -- on disc 1: Air's Love 2 stayed wanted, and was downloaded again for
+       -- nothing, over a missing track 6 with no disc beside a present 1/6.
        COUNT(*) FILTER (WHERE t.missing_since IS NOT NULL AND NOT EXISTS (
            SELECT 1 FROM tracks p
             WHERE p.album_id = a.id AND p.missing_since IS NULL
-              AND ((p.track_number = t.track_number AND p.disc_number IS NOT DISTINCT FROM t.disc_number)
+              AND ((p.track_number = t.track_number AND COALESCE(p.disc_number, 1) = COALESCE(t.disc_number, 1))
                    OR lower(p.title) = lower(t.title))))::int AS uncovered_missing
   FROM albums a
   JOIN tracks t ON t.album_id = a.id

internal/library/scanner.go

@@ -74,7 +74,11 @@ var audioExtensions = map[string]bool{
 //	   dhowden/tag reads it as 0, and Minstrel's own importer wrote it that
 //	   way, so all 245 FLAC tracks it imported came in with no track or disc
 //	   number and their albums lost their order.
-const tagReadVersion int16 = 6
+//	7: an album of the same name in another release group gets its own row
+//	   (resolveAlbum). Weezer's six self-titled albums were one 118-track row,
+//	   so their merges, wants and upgrades all counted as one album; a re-read
+//	   files each existing track under its own album.
+const tagReadVersion int16 = 7
 
 type Stats struct {
 	Scanned int `json:"scanned"`
@@ -562,7 +566,12 @@ func (s *Scanner) resolveArtist(ctx context.Context, q *dbq.Queries, name, mbid
 }
 
 func (s *Scanner) resolveAlbum(ctx context.Context, q *dbq.Queries, artistID pgtype.UUID, title string, year int, mbid, releaseGroupMBID string) (dbq.Album, error) {
-	existing, err := q.GetAlbumByArtistAndTitle(ctx, dbq.GetAlbumByArtistAndTitleParams{ArtistID: artistID, Title: title})
+	lookup := dbq.GetAlbumByArtistAndTitleParams{ArtistID: artistID, Title: title}
+	if releaseGroupMBID != "" {
+		rg := releaseGroupMBID
+		lookup.ReleaseGroupMbid = &rg
+	}
+	existing, err := q.GetAlbumByArtistAndTitle(ctx, lookup)
 	if err == nil {
 		// Heal the release group the same way. Not unique, so no conflict to
 		// handle: every release of a group may carry it.

internal/library/scanner_test.go

@@ -385,3 +385,94 @@ func TestScanner_ReleaseGroupMBID_Integration(t *testing.T) {
 		t.Errorf("album A release group changed to %v", rg)
 	}
 }
+
+// Two albums of one name by one artist, in different release groups, are two
+// albums: Weezer's six self-titled albums shared one row until 2026-10-20. A
+// track of the same group joins its album, and a track filed under the other
+// album's row by the old lookup moves to its own on the tagReadVersion re-read.
+func TestScanner_SameTitledAlbumsOfTwoGroupsAreTwoAlbums_Integration(t *testing.T) {
+	if testing.Short() {
+		t.Skip("skipping scanner integration in -short mode")
+	}
+	dsn := os.Getenv("MINSTREL_TEST_DATABASE_URL")
+	if dsn == "" {
+		t.Skip("MINSTREL_TEST_DATABASE_URL not set")
+	}
+	ctx := context.Background()
+	logger := slog.New(slog.NewTextHandler(io.Discard, nil))
+	if err := db.Migrate(dsn, logger); err != nil {
+		t.Fatalf("migrate: %v", err)
+	}
+	pool, err := pgxpool.New(ctx, dsn)
+	if err != nil {
+		t.Fatalf("pool: %v", err)
+	}
+	t.Cleanup(pool.Close)
+	if _, err := pool.Exec(ctx, "TRUNCATE tracks, albums, artists RESTART IDENTITY CASCADE"); err != nil {
+		t.Fatalf("truncate: %v", err)
+	}
+
+	const (
+		blue  = "aaaaaaaa-1994-2222-3333-444444444444"
+		green = "bbbbbbbb-2001-2222-3333-444444444444"
+	)
+	root := t.TempDir()
+	tagged := func(rel, title, group string) string {
+		p := filepath.Join(root, rel)
+		writeTestMP3(t, p, map[string]string{
+			"TIT2": title, "TPE1": "Weezer", "TALB": "Weezer",
+			"TXXX": "MusicBrainz Release Group Id\x00" + group,
+		})
+		return p
+	}
+	tagged("Weezer (1994)/01.mp3", "My Name Is Jonas", blue)
+	tagged("Weezer (1994)/02.mp3", "No One Else", blue)
+	greenPath := tagged("Weezer (2001)/01.mp3", "Don't Let Go", green)
+
+	scanner := New(pool, logger, []string{root}, nil)
+	if _, err := scanner.Scan(ctx, nil); err != nil {
+		t.Fatalf("scan: %v", err)
+	}
+	albumsByGroup := func() map[string]int {
+		t.Helper()
+		rows, err := pool.Query(ctx, `SELECT a.release_group_mbid, count(t.id) FROM albums a
+			JOIN tracks t ON t.album_id = a.id WHERE a.title = 'Weezer' GROUP BY a.id, a.release_group_mbid`)
+		if err != nil {
+			t.Fatal(err)
+		}
+		defer rows.Close()
+		out := map[string]int{}
+		for rows.Next() {
+			var g *string
+			var n int
+			if err := rows.Scan(&g, &n); err != nil {
+				t.Fatal(err)
+			}
+			if g == nil {
+				t.Fatal("an album with no release group")
+			}
+			out[*g] = n
+		}
+		return out
+	}
+	if got := albumsByGroup(); len(got) != 2 || got[blue] != 2 || got[green] != 1 {
+		t.Fatalf("albums by group = %v, want blue with 2 tracks and green with 1", got)
+	}
+
+	// The green track as the old lookup left it: on the blue album's row,
+	// under an older tag-read version, its own album row gone.
+	if _, err := pool.Exec(ctx, `
+		WITH b AS (SELECT id FROM albums WHERE release_group_mbid = $1)
+		UPDATE tracks SET album_id = (SELECT id FROM b), tag_read_version = 6 WHERE file_path = $2`, blue, greenPath); err != nil {
+		t.Fatal(err)
+	}
+	if _, err := pool.Exec(ctx, `DELETE FROM albums WHERE release_group_mbid = $1`, green); err != nil {
+		t.Fatal(err)
+	}
+	if _, err := scanner.Scan(ctx, nil); err != nil {
+		t.Fatalf("rescan: %v", err)
+	}
+	if got := albumsByGroup(); len(got) != 2 || got[blue] != 2 || got[green] != 1 {
+		t.Errorf("after the re-read, albums by group = %v, want the green track on its own album", got)
+	}
+}

internal/wanted/upgrade_test.go

@@ -6,6 +6,7 @@ import (
 	"testing"
 	"time"
 
+	"github.com/jackc/pgx/v5/pgtype"
 	"github.com/jackc/pgx/v5/pgxpool"
 
 	"git.fabledsword.com/bvandeusen/minstrel/internal/db/dbq"
@@ -135,3 +136,39 @@ func TestCheckUpgrades_ARipIsReplacedWhateverTheSetting_Integration(t *testing.T
 		t.Errorf("a rip: upgrade %q", w.Upgrade)
 	}
 }
+
+// A missing track with no disc number is on disc 1, so a present 1/2 stands in
+// for a missing "2". Air's Love 2 stayed wanted over one such row and was
+// downloaded again for nothing (2026-10-20). A missing track no present track
+// covers still counts.
+func TestWantHoldings_AMissingTrackWithNoDiscIsOnDiscOne_Integration(t *testing.T) {
+	pool := newPool(t)
+	ctx := context.Background()
+	album := seedMP3s(t, pool, 320, false)
+	if _, err := pool.Exec(ctx, `UPDATE tracks SET disc_number = 1 WHERE album_id = $1`, album.ID); err != nil {
+		t.Fatal(err)
+	}
+	q := dbq.New(pool)
+	for _, n := range []int32{2, 3} {
+		num := n
+		tr, err := q.UpsertTrack(ctx, dbq.UpsertTrackParams{
+			Title: fmt.Sprintf("Stale %d", n), AlbumID: album.ID, ArtistID: album.ArtistID, TrackNumber: &num,
+			DurationMs: 200000, FilePath: fmt.Sprintf("/music/ballad/stale-%d.mp3", n), FileSize: 1, FileFormat: "mp3",
+		})
+		if err != nil {
+			t.Fatal(err)
+		}
+		if _, err := pool.Exec(ctx, `UPDATE tracks SET missing_since = now() - interval '2 days' WHERE id = $1`, tr.ID); err != nil {
+			t.Fatal(err)
+		}
+	}
+	rows, err := q.WantHoldings(ctx, dbq.WantHoldingsParams{
+		Since: pgtype.Timestamptz{Time: time.Now(), Valid: true}, GroupMbid: theBallad, AlbumID: album.ID,
+	})
+	if err != nil {
+		t.Fatal(err)
+	}
+	if len(rows) != 1 || rows[0].UncoveredMissing != 1 {
+		t.Fatalf("holdings = %+v, want one album with 1 uncovered missing track (track 3 only)", rows)
+	}
+}