Files
yellowjacket/backend/explore/dumppatch.go
T
yonluandClaude Opus 5 e7748f1fd5
CI / check (push) Successful in 3m7s
CI / e2e (push) Canceled after 1m45s
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
2026-08-16 13:58:15 -04:00

282 lines
6.5 KiB
Go

//go:build indexbuild
package explore
import (
"context"
"strings"
)
// Stage 4 of the dump import: small, idempotent API patch passes that
// fill in what the dumps can't provide. All calls go through the
// shared rate-limited ListenBrainz client and its HTTP cache, so
// re-running after an interruption is cheap.
const (
// Listener-count patch budgets: only the most popular rows get
// listener counts (a secondary ranking signal). 1000 MBIDs per
// batched POST call → ~350 API calls total.
listenerPatchArtists = 50_000
listenerPatchRGs = 100_000
listenerPatchRecordings = 200_000
// popularityBatchSize is the number of MBIDs per LB popularity
// or metadata request. LB accepts up to 1000 per call.
popularityBatchSize = 1000
)
// runPatchPasses fills listener counts, artist metadata, and the
// similar-artist map from the ListenBrainz API.
func (imp *dumpImporter) runPatchPasses(ctx context.Context) {
if imp.lb == nil {
return
}
imp.patchArtistMetadata(ctx)
if ctx.Err() != nil {
return
}
imp.patchSimilarArtists(ctx)
if ctx.Err() != nil {
return
}
imp.patchListenerCounts(ctx)
}
// patchArtistMetadata batch-fetches type/country/name for indexed
// artists that are missing them. Also creates rows for kept artists
// whose name wasn't derivable from the canonical dump (multi-artist
// credits only).
func (imp *dumpImporter) patchArtistMetadata(ctx context.Context) {
rows, err := imp.si.db.QueryContext(`
SELECT mbid FROM explore_index
WHERE entity_type = 1 /* artist */
AND (artist_type = '' OR country = '' OR title = '' OR title = mbid)
`)
if err != nil {
return
}
var mbids []string
for rows.Next() {
var m string
if err := rows.Scan(&m); err == nil {
mbids = append(mbids, m)
}
}
_ = rows.Close()
if len(imp.pendingArtists) > 0 {
mbids = append(mbids, imp.pendingArtists...)
}
if len(mbids) == 0 {
return
}
imp.logger.Info("dump import: patching artist metadata", "artists", len(mbids))
batches := chunkStrings(mbids, popularityBatchSize)
patched := 0
for i, batch := range batches {
if ctx.Err() != nil {
return
}
meta, err := imp.lb.BatchArtistMetadata(ctx, batch)
if err != nil || len(meta) == 0 {
continue
}
entries := make([]SearchIndexResult, 0, len(meta))
for mbid, m := range meta {
if m.Name == "" {
continue
}
entries = append(entries, SearchIndexResult{
EntityType: "artist",
MBID: mbid,
Title: m.Name,
ArtistName: m.Name,
ArtistMBID: mbid,
ArtistType: m.Type,
Country: m.Country,
})
// Pre-populate the MB rels cache so on-demand artist
// image resolution skips a MusicBrainz call.
if imp.si.artistImg != nil {
imp.si.artistImg.PreloadArtistRels(mbid, m)
}
}
imp.si.upsertBatch(entries)
patched += len(entries)
imp.setStageProgress(dumpStagePatch, i+1, len(batches))
}
imp.logger.Info("dump import: artist metadata patched", "artists", patched)
}
// patchSimilarArtists refreshes the similar-artist map for library
// artists (one API call per library artist, cached for a week).
func (imp *dumpImporter) patchSimilarArtists(ctx context.Context) {
libraryMBIDs := imp.si.getLibraryArtistMBIDs()
if len(libraryMBIDs) == 0 {
return
}
for i := 0; i < len(libraryMBIDs); i += similarArtistsBatchSize {
if ctx.Err() != nil {
return
}
end := min(i+similarArtistsBatchSize, len(libraryMBIDs))
grouped := imp.si.fetchSimilarArtistsBatch(ctx, imp.lb, libraryMBIDs[i:end])
for seed, similar := range grouped {
imp.si.storeSimilarArtists(seed, similar)
// Flag indexed similar artists for personalized ranking.
for _, s := range similar {
_, _ = imp.si.db.ExecContext(
"UPDATE explore_index SET is_similar = 1 WHERE artist_mbid = ?",
dbMBID(s.ArtistMBID),
)
}
}
}
imp.logger.Info("dump import: similar artists patched", "libraryArtists", len(libraryMBIDs))
}
// patchListenerCounts fills listener_count for the most popular rows
// of each entity type. Popularity (listen count) is NOT overwritten —
// the dump-derived counts stay authoritative so the ranking scale is
// consistent across the whole index.
func (imp *dumpImporter) patchListenerCounts(ctx context.Context) {
kinds := []struct {
entityType string
limit int
fetch func(context.Context, []string) (map[string]PopularityData, error)
}{
{"artist", listenerPatchArtists, imp.lb.ArtistPopularity},
{"release_group", listenerPatchRGs, imp.lb.ReleaseGroupPopularity},
{"recording", listenerPatchRecordings, imp.lb.RecordingPopularity},
}
for _, kind := range kinds {
if ctx.Err() != nil {
return
}
mbids := imp.topMBIDs(kind.entityType, kind.limit)
if len(mbids) == 0 {
continue
}
batches := chunkStrings(mbids, popularityBatchSize)
filled := 0
for i, batch := range batches {
if ctx.Err() != nil {
return
}
pops, err := kind.fetch(ctx, batch)
if err != nil {
continue
}
filled += imp.si.updateListenerCounts(pops)
imp.setStageProgress(dumpStageListeners, i+1, len(batches))
}
imp.logger.Info("dump import: listener counts patched",
"entityType", kind.entityType,
"rows", filled,
)
}
}
// topMBIDs returns the most popular index MBIDs for an entity type
// that don't have listener counts yet.
func (imp *dumpImporter) topMBIDs(entityType string, limit int) []string {
rows, err := imp.si.db.QueryContext(`
SELECT mbid FROM explore_index
WHERE entity_type = ? AND listener_count = 0
ORDER BY popularity DESC
LIMIT ?
`, dbEntityType(entityType), limit)
if err != nil {
return nil
}
defer func() { _ = rows.Close() }()
var mbids []string
for rows.Next() {
var m dbMBID
if err := rows.Scan(&m); err == nil {
mbids = append(mbids, string(m))
}
}
return mbids
}
// updateListenerCounts writes listener counts only (never popularity),
// keeping the dump-derived popularity scale consistent. Returns the
// number of rows updated.
func (si *SearchIndex) updateListenerCounts(updates map[string]PopularityData) int {
if len(updates) == 0 {
return 0
}
tx, err := si.db.BeginTx()
if err != nil {
return 0
}
defer func() { _ = tx.Rollback() }()
updated := 0
for mbid, data := range updates {
if data.ListenerCount <= 0 {
continue
}
res, err := tx.Exec(
`UPDATE explore_index
SET listener_count = ?
WHERE mbid = ? AND listener_count < ?`,
data.ListenerCount, dbMBID(strings.ToLower(mbid)), data.ListenerCount,
)
if err != nil {
continue
}
if n, err := res.RowsAffected(); err == nil {
updated += int(n)
}
}
_ = tx.Commit()
return updated
}