Adds nullable storage fields + availability column on tracks, remote
source/source_id identity on albums/artists, TrackRepository.materialize()
and get_or_create_remote() repos — groundwork for on-demand YTM library
(placeholders saved without audio, materialized in-place on first play).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Add an optional `source` filter to `GET /api/v1/tracks` (and the
`TrackRepository.list`/`count` port + SQLAlchemy adapter). Lets clients
query, e.g., only uploaded tracks (`?source=upload`) newest-first — the
backing for the webui's persistent "Recently uploaded" view.
- test: upload then list with `?source=upload` (hit) / `?source=youtube`
(miss)
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Adds inline AcoustID match-finding (multiple ranked candidates via
lookup_all) and PUT /tracks/{id}/metadata for manual edits, resolving
artist/album and setting metadata_status=manual. Extends TrackOut with
genre/year/track_number.
Two related gaps surfaced from "uploaded a track, nothing changed / no status":
- A track could stay stuck on `pending` forever (an unexpected worker error
rolled back the run without recording anything), and `failed` carried no
reason. Add `tracks.metadata_error` + `tracks.enriched_at` (migration), stamp
the outcome in apply_enrichment, add TrackRepository.mark_enrichment_failed,
wrap enrich_task to persist crashes as `failed` in a fresh session, and emit a
human-readable no-match reason. Expose metadata_error/enriched_at in TrackOut.
- The tag-first merge let junk embedded tags (e.g. "Music Track"/"Sound_13958")
override even a 0.99-confidence AcoustID match. Add acoustid_trust_score
(default 0.85): above it the acoustic identity wins for title/artist/album/
year, tags are fallback; below it, tag-first as before.
Add a license-free real-file fixture (Scarlet Fire / Otis McDonald) whose junk
tags AcoustID overrides, with an always-on tag-reader test plus fpcalc/AcoustID/
network-gated identity + full-pipeline tests (skip on host, run in the container).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Resolve, store and serve album cover art.
Sources (tag-first, mirroring enrichment): embedded artwork extracted
offline via mutagen (ID3 APIC / FLAC+OGG Picture / MP4 covr), then Cover
Art Archive by release-group MBID as a network fallback. Resolution runs
inside MetadataEnrichmentService after album resolution, only when the
album has no cover yet (idempotent, never overwrites), and is best-effort
so a cover failure never affects enrichment status.
- CoverArt value object + CoverArtExtractor/CoverArtProvider ports
- MutagenCoverExtractor + CoverArtArchiveClient adapters
- AcoustID parser now captures release_group_mbid
- Covers stored via FileStorage at covers/{album_id}.{ext} (local + S3)
- AlbumRepository.set_cover_path
- Serve real covers: GET /api/v1/albums|tracks/{id}/cover (StreamUser,
?token=), Subsonic getCoverArt (placeholder fallback)
- has_cover flag on AlbumOut/TrackOut
- coverart_enabled / coverart_base_url settings
- tests: cover resolution units + release_group parse + DB-backed
test_cover_api.py (139 green via make test-api)
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Implements the §6.2 enrichment pipeline: embedded tags → Chromaprint
fingerprint → AcoustID lookup. Well-tagged files get correct
artist/album/title offline; the rest are identified via AcoustID
(which also yields a MusicBrainz recording id in one call).
- domain: AudioTags/Fingerprint/RecordingMatch value objects; ports
AudioTagReader, AudioFingerprinter, AcoustIdClient; TrackRepository
.apply_enrichment (gap-fill, never erases) + AlbumRepository.get_or_create
- infrastructure/metadata: MutagenTagReader, FpcalcFingerprinter,
AcoustIdHttpClient (rich meta=recordings+releasegroups, throttled)
- application: MetadataEnrichmentService — tags preferred, AcoustID fills
gaps; resolves artist/album; status enriched/failed; skips manual;
every external step wrapped (graceful degradation)
- workers: enrich_task registered; enqueue_enrich is best-effort and
deferred so the caller's txn commits before the worker reads the row
- wiring: upload enqueues after add; import returns imported_ids and
enqueues post-commit (mid-scan would race the worker); manual
POST /tracks/{id}/metadata/enrich endpoint
- deps: add mutagen (fpcalc/ffmpeg already in the image)
Tests: metadata service orchestration, AcoustID parser, tag helpers.
125 passed; mypy strict + ruff clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add S3FileStorage adapter (any S3-compatible backend: AWS, MinIO, Garage)
alongside the local adapter, selected via STORAGE_BACKEND. Proxied range
streaming via get_object+Range; as_local_path downloads to a tempfile for
ffmpeg/fpcalc. Rename track.file_path -> storage_uri across domain entity,
ORM, repositories, port, and services, with an Alembic migration. Adds
mocked S3 unit tests.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Senko-san
olly