feat(database): shape the library like files, and shrink the catalog

Plans 013 and 014, the album page that prompted them, and the smaller
fixes they turned up. Changelog, largest first.

## The local library is shaped like files, not like MusicBrainz

`audio_files` carries its own tags and points at `albums` and
`artists`; `file_genres` is the one real many-to-many. `recordings`,
`release_group_recordings`, `artist_credit`, `artist_credit_artist`,
`recording_genres`, `release_groups` and `release_to_rg` are gone from
the local side, and with them a six-way join in every read, a
`MIN(release_group_id)` subquery in eleven queries and a
first-credited-artist subquery in nine. Measured on a real 25,966-file
library, every many-to-many that model expressed was 1:1 in the data.

- Ownership is a file. `GetFilePathsByRecordingMBIDs`,
  `LibraryMBIDIndex.CheckMBIDs`, `collectLibraryEntities` and
  `pruneStaleLocalCrossReferences` all join `audio_files`, so the 812
  orphaned recordings, 216 release groups and 260 artists that library
  carried are now structurally impossible.
- One projection: every track query selects from the `track_metadata`
  view, one row type, one mapper. Nine hand-rolled copies had drifted
  far enough to report different years on different screens.
- `library_id = 0` means every library, so each list query exists once
  instead of scoped and unscoped with a branch at every call site.
- No migration chain. `sql/schemas/` is the one description of the
  shape; `sql/migrations/`, `applyMigrations` and `schema_migrations`
  are squashed away, along with the drift between them that had sqlc
  generating against a stale schema.
- `database.InsertTestTrack` is the one test seeder; twenty test files
  had been assembling the old FK chain each in its own order.

## The catalog stores its ids as bytes

`explore_index`'s three 36-char MBID columns and its entity-type text
are 16 raw bytes and a small integer. The table and its six indexes go
780 MB to 405 MB on a real 2,052,200-row catalog, which is why a fresh
install is ~0.6 GB rather than ~1.0 GB.

- `backend/explore/mbid.go` is the only place the encoding is known;
  everything above it speaks dashed strings.
- `CHECK(length(mbid) = 16)` makes a stringly write fail at the insert
  rather than silently returning no rows, since SQLite does not coerce
  between TEXT and BLOB.
- The importer asks the artifact what encoding it carries and converts
  on the way in, so the artifact already published keeps working and no
  format bump is needed.
- `indexRowColumns`/`scanIndexRow` replace four copies of a 22-column
  list, and `TestStoredEncodingRoundTrips` sweeps every read path.

## An album page that says how much of the album is yours

- One question, asked once: is there a file. `filePaths` is filled by a
  single batched lookup when the tracklist settles, and the badge, the
  Play count, the dimmed rows and every menu item read it — replacing
  four claims of decreasing confidence that could show a green tick on
  an album whose every action did nothing.
- Play, Play 7 of 12, or no play button at all.
- `total_tracks` on `explore_index` (~2 bytes over 400,677 release
  groups) and on `audio_files` from tags that have always carried it:
  a complete MBID-matched album now makes no catalog call at all, where
  it used to spend the most expensive request the app makes.
- A merged cluster shows the running order the most releases agree on,
  and the version list marks the release you own rather than standing a
  synthetic entry in for it.
- `AlbumReleasesFailed`: a slow fetch is no longer reported as a failed
  one by a 12-second timer.
- Rows not in the library are dimmed in place (with `aria-disabled`)
  instead of the owned ones wearing a green tick and a legend.

## Caches and cover art get ceilings

- Only the three tiers of a cover are stored; the full-resolution copy
  nothing rendered was 1,134 MB of a 1.4 GB covers directory.
- One artist portrait is downloaded and the rest are remembered as
  URLs — 4.1 GB of a 5.3 GB cache was candidates no code path reads.
- `browsedArtBudget` and `httpCacheBudget` bound what an age cannot:
  the same install held art for 5,770 artists in a 1,301-artist
  library.
- `OrphanedArtistImagesJob` joined a bare MBID onto a sharded
  directory, so it deleted the rows that were the only record of the
  files it left behind. `explore.ArtistImageDir` is that layout's one
  definition now.

## The autotag queue asks whether there is work

`tagging_items` was a row per album folder, not a queue, and no query
read the `tag_status` column that held the answer. The four queue
queries ask the files, which matters most where it is least visible:
`startPrefetch` was scoring every album in a tagged library against
MusicBrainz.

## Phantom playlist tracks resolve in place

An M3U8 imported before its files leaves phantom rows; they now match
by path and fall back to position, keep their place in the playlist
when resolved, and pair best-first so two phantoms cannot claim the
same file.

## Playing a track plays the list it is in

Double-click, and Play on a single row's menu, queue the list as
displayed with `startIndex` on that row — the album page and the track
list used to queue one track and discard the album around it. A
multi-row selection still plays exactly itself.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AfVYUVExXsx1nSWrXN8mAh

e2e/specs/explore-shelves.spec.ts

@@ -139,14 +139,25 @@ test.describe('Explore before anyone has typed', () => {
 async function stageCatalogIfEmpty(app: Page): Promise<void> {
   if ((await catalogRows(app)) > 0) return;
 
+  // The catalog stores an MBID as its 16 raw bytes and an entity type
+  // as a small integer, so a staged row has to be spelled the way the
+  // app spells one: a real UUID, converted at the boundary, and a code
+  // rather than the word.  `'e2e-ar-a'` is 8 characters and fails
+  // `CHECK(length(mbid) = 16)` -- which `INSERT OR IGNORE` then
+  // swallows, so the staging step looked exactly like a staging step
+  // and the page had nothing to draw.  That is the same fault this
+  // helper's own comment below describes, one layer down.
+  const ARTIST = 1;
+  const RELEASE_GROUP = 2;
+
   const rows = [
-    ['artist', 'e2e-ar-a', 'Staged Alpha', 'Staged Alpha', 'e2e-ar-a', 9000],
-    ['artist', 'e2e-ar-b', 'Staged Beta', 'Staged Beta', 'e2e-ar-b', 500],
-    ['artist', 'e2e-ar-c', 'Staged Gamma', 'Staged Gamma', 'e2e-ar-c', 300],
-    ['release_group', 'e2e-rg-a1', 'Alpha One', 'Staged Alpha', 'e2e-ar-a', 8000],
-    ['release_group', 'e2e-rg-a2', 'Alpha Two', 'Staged Alpha', 'e2e-ar-a', 7000],
-    ['release_group', 'e2e-rg-a3', 'Alpha Three', 'Staged Alpha', 'e2e-ar-a', 6000],
-    ['release_group', 'e2e-rg-b1', 'Beta One', 'Staged Beta', 'e2e-ar-b', 400],
+    [ARTIST, uuid('ar-a'), 'Staged Alpha', 'Staged Alpha', uuid('ar-a'), 9000],
+    [ARTIST, uuid('ar-b'), 'Staged Beta', 'Staged Beta', uuid('ar-b'), 500],
+    [ARTIST, uuid('ar-c'), 'Staged Gamma', 'Staged Gamma', uuid('ar-c'), 300],
+    [RELEASE_GROUP, uuid('rg-a1'), 'Alpha One', 'Staged Alpha', uuid('ar-a'), 8000],
+    [RELEASE_GROUP, uuid('rg-a2'), 'Alpha Two', 'Staged Alpha', uuid('ar-a'), 7000],
+    [RELEASE_GROUP, uuid('rg-a3'), 'Alpha Three', 'Staged Alpha', uuid('ar-a'), 6000],
+    [RELEASE_GROUP, uuid('rg-b1'), 'Beta One', 'Staged Beta', uuid('ar-b'), 400],
   ];
 
   for (const row of rows) {
@@ -158,7 +169,8 @@ async function stageCatalogIfEmpty(app: Page): Promise<void> {
           sql: `INSERT OR IGNORE INTO explore_index
                   (entity_type, mbid, title, artist_name, artist_mbid,
                    popularity, listener_count, primary_type)
-                VALUES (?, ?, ?, ?, ?, ?, ?, 'Album')`,
+                VALUES (?, unhex(replace(?, '-', '')), ?, ?,
+                        unhex(replace(?, '-', '')), ?, ?, 'Album')`,
           args: [...args, 10],
         }),
       });
@@ -171,12 +183,52 @@ async function stageCatalogIfEmpty(app: Page): Promise<void> {
     // and the staging step looked exactly like a staging step. A setup
     // whose failure is not checked is not setup.
     expect(result.status, `staging failed: ${result.body}`).toBe(200);
+
+    // …and `OR IGNORE` means a 200 is not a write. A CHECK the row
+    // violates is *ignored*, not reported, so the count below is the
+    // only thing that can tell staging from silence.
+    expect(
+      (JSON.parse(result.body) as { rowsAffected?: number }).rowsAffected,
+      `staged nothing: ${result.body}`,
+    ).toBe(1);
   }
 
   expect(await catalogRows(app)).toBeGreaterThan(0);
 }
 
-/** How many catalog rows this environment has. CI has none. */
+/**
+ * A stable, valid MBID from a short label.
+ *
+ * The column is `CHECK(length(mbid) = 16)` after `unhex`, so a fixture
+ * id has to be a real UUID rather than a readable string -- the same
+ * trade the Go fixtures make with `testMBID()`, and for the same
+ * reason: a readable id that cannot be stored is not readable, it is
+ * absent.
+ */
+function uuid(label: string): string {
+  const hex = [...label]
+    .map((c) => c.charCodeAt(0).toString(16).padStart(2, '0'))
+    .join('')
+    .padEnd(32, '0')
+    .slice(0, 32);
+
+  return [
+    hex.slice(0, 8),
+    hex.slice(8, 12),
+    hex.slice(12, 16),
+    hex.slice(16, 20),
+    hex.slice(20),
+  ].join('-');
+}
+
+/**
+ * Whether this environment has a catalog at all. CI has none.
+ *
+ * `exploreIndex` is deliberately 0-or-1 rather than a row count: a real
+ * catalog is ~1.1M rows, and a cold `COUNT(*)` over it took 65 seconds
+ * on the first call after a seed was extracted -- which timed out
+ * whichever spec ran first and looked like flake.
+ */
 async function catalogRows(app: Page): Promise<number> {
   const health = await app.evaluate(async () => {
     const res = await fetch('/__test/health');

e2e/specs/queue-reorder.spec.ts

@@ -31,9 +31,12 @@ async function order(app: Page): Promise<string[]> {
 
 async function queueFourAndOpen(app: Page): Promise<string[]> {
   const paths: string[] = await app.evaluate(async () => {
+    // One argument, and 0 means every library: the scoped and
+    // unscoped list queries collapsed into one when the schema did
+    // (plan 013 R3), so `GetTracks()` no longer exists to call.
     const tracks = await window.__yjEvents.call(
-      'library.Library.GetAllTracks',
-      [],
+      'library.Library.GetTracks',
+      [0],
       10_000,
     );
 

e2e/specs/reduced-motion.spec.ts

@@ -57,9 +57,12 @@ async function playTheLongOne(app: Page): Promise<void> {
   });
 
   const paths: string[] = await app.evaluate(async (needle) => {
+    // One argument, and 0 means every library: the scoped and
+    // unscoped list queries collapsed into one when the schema did
+    // (plan 013 R3), so `GetTracks()` no longer exists to call.
     const tracks = await window.__yjEvents.call(
-      'library.Library.GetAllTracks',
-      [],
+      'library.Library.GetTracks',
+      [0],
       10_000,
     );
 

e2e/specs/requested-badge.spec.ts

@@ -23,8 +23,16 @@ import { test, expect, callBinding } from '../support/fixtures.js';
  */
 
 /** A release group that exists whether or not this environment has a
- *  catalog — CI's `YJ_CORE_INDEX_URL` is deliberately dead. */
-const MBID = 'e2e-rg-badge-0001';
+ *  catalog — CI's `YJ_CORE_INDEX_URL` is deliberately dead.
+ *
+ *  It is a real UUID because the catalog stores an MBID as its 16 raw
+ *  bytes under `CHECK(length(mbid) = 16)`. A readable id fails that
+ *  check, `INSERT OR IGNORE` swallows the failure, and the staging step
+ *  below reports 200 having written nothing. */
+const MBID = 'e2ebad9e-0001-4000-8000-000000000001';
+
+/** The staged album's artist, for the same reason. */
+const ARTIST_MBID = 'e2ebad9e-0002-4000-8000-000000000002';
 const TITLE = 'Requested Album';
 const ARTIST = 'Badge Artist';
 
@@ -45,18 +53,29 @@ test.describe('the requested badge', () => {
             sql: `INSERT OR IGNORE INTO explore_index
                     (entity_type, mbid, title, artist_name, artist_mbid,
                      popularity, listener_count, primary_type)
-                  VALUES ('release_group', ?, ?, ?, 'e2e-ar-badge', 10, 10, 'Album')`,
-            args: [row.mbid, row.title, row.artist],
+                  VALUES (2 /* release_group */,
+                          unhex(replace(?, '-', '')), ?, ?,
+                          unhex(replace(?, '-', '')),
+                          10, 10, 'Album')`,
+            args: [row.mbid, row.title, row.artist, row.artistMbid],
           }),
         });
 
         return { status: r.status, body: await r.text() };
       },
-      { mbid: MBID, title: TITLE, artist: ARTIST },
+      { mbid: MBID, title: TITLE, artist: ARTIST, artistMbid: ARTIST_MBID },
     );
 
-    // A setup step whose failure is not checked is not setup.
+    // A setup step whose failure is not checked is not setup — and with
+    // `OR IGNORE` a 200 is not a write either: a CHECK the row violates
+    // is ignored rather than reported. `rowsAffected` is 0 on a second
+    // run against the same backend, so what is asserted is that the
+    // statement did not error.
     expect(res.status, `staging failed: ${res.body}`).toBe(200);
+    expect(
+      (JSON.parse(res.body) as { error?: string }).error,
+      `staging failed: ${res.body}`,
+    ).toBeUndefined();
 
     // A previous run that died between adding and removing would leave
     // this album requested, and the first assertion here is that it is