Commit Graph

151 Commits

Author SHA1 Message Date
Jonathan 4907d0d5a4 feat: library scan — probe seam + scan_library (have + monitored + followed)
Adds the AudioProbe seam (probe.py, _probe.py MutagenProbe with a lazy
mutagen import, FakeAudioProbe) and scan_library, which walks /music,
resolves each album against MusicBrainz (falling back to embedded tags
when the folder name doesn't resolve), and idempotently records matched
albums as have (LibraryItem), monitored (MonitoredRelease, fulfilled),
and followed (WatchedArtist, autoMonitorFuture=false).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-11 17:00:33 +02:00
Jonathan af67d01aa2 feat: LibraryItem.requestId nullable (scanned haves need no request) + test cleanup
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-11 16:56:59 +02:00
Jonathan 446d54f819 feat: MBTarget exposes rgMbid + artistMbid from the resolver 2026-07-11 16:54:11 +02:00
Jonathan f421904c9b docs: add library-scan implementation plan
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-11 16:50:57 +02:00
Jonathan a4ae60a99e docs: add library-scan design (ingest on-disk albums as have+monitored)
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-11 16:45:58 +02:00
Jonathan f59f972c1b fix: guard test cleanup to a dedicated lyra_test DB (never wipe live data)
Test suites were running DELETE FROM.../deleteMany() against whatever
DATABASE_URL pointed to, which was the live app database, wiping user
credentials and library. Add a guard in both the worker conftest and
web test setup that raises before any cleanup unless the target DB
name ends in '_test'. Default DATABASE_URL to lyra_test when unset in
the worker, and make `npm test` target lyra_test explicitly.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-11 16:16:12 +02:00
Jonathan 88ef4149d9 fix: atomic import swap, disk/DB replace consistency, guarded pipeline loop
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-11 15:27:48 +02:00
Jonathan fe973343ad feat: sweep orphaned staging dirs on worker startup 2026-07-11 15:18:30 +02:00
Jonathan 29604db9cd feat: download to staging + atomically promote complete album into the library
Rewrite run_pipeline's download->import section to download into a per-job
staging dir (cleared before each candidate), verify completeness there,
promote only a complete album into the library via import_album, tag the
final dir, then DB-import. The staging dir is always removed in a
try/finally, so a partial/failed download never touches or corrupts the
library, and a retry can no longer leave orphan files behind.

Also point test_upgrade_e2e.py's run_pipeline call at dest_root="/tmp/lib"
(matching every other pipeline test) since import_album now unconditionally
creates the final dir, which requires a writable dest_root.
2026-07-11 15:13:31 +02:00
Jonathan 0eb36e8362 feat: staging-dir + clean import_album library helpers 2026-07-11 15:08:24 +02:00
Jonathan 1d847e2aeb docs: add staged-download + atomic-import implementation plan
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-11 15:07:01 +02:00
Jonathan 2614e32f55 docs: add staged-download + atomic-import design (pipeline fix)
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-11 14:59:55 +02:00
Jonathan 9f895cfc10 fix: search-now dedupes against an active job (match worker enqueue invariant)
The worker's enqueue_due refuses to enqueue when a non-terminal job
already exists for a release. The search-now route enqueued
unconditionally, so two clicks (or a click racing the worker sweep)
could create two concurrent pipelines for the same album, both
importing and colliding on LibraryItem's @@unique([artist,album]).

Mirror the worker's guard: no-op if a job in a non-terminal state
(anything other than imported/needs_attention) already exists for the
release.
2026-07-11 14:19:31 +02:00
Jonathan 7087b2780e feat: artist discography page + wanted page 2026-07-11 14:11:53 +02:00
Jonathan f7135b9046 feat: artists page with live-search follow picker 2026-07-11 14:08:10 +02:00
Jonathan cd652c7285 feat: wanted list API — list + standalone add 2026-07-11 14:04:35 +02:00
Jonathan 6615ecf3a7 feat: releases API — monitor toggle + search-now enqueue 2026-07-11 14:00:49 +02:00
Jonathan f5ceea8fde fix: narrow artist PATCH/DELETE 404 catch to Prisma P2025 (rethrow real errors) 2026-07-11 13:58:40 +02:00
Jonathan 1e39c74b16 feat: artist item API — detail, auto-monitor toggle, unfollow 2026-07-11 13:55:36 +02:00
Jonathan 7f0ac68657 feat: artists API — follow (persist discography) + list 2026-07-11 13:51:14 +02:00
Jonathan 8e97796bad feat: MusicBrainz proxy API routes (artist search + discography browse)
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-11 13:45:24 +02:00
Jonathan 341da71505 test: type the fetch mock so the MB test typechecks (tsc --noEmit clean) 2026-07-11 13:15:21 +02:00
Jonathan cdcda9e501 feat: read-only MusicBrainz module + monitoring test cleanup
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-11 13:11:11 +02:00
Jonathan eb2a2207a1 docs: add monitoring-web implementation plan (slice 2, plan 2)
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-11 13:08:19 +02:00
Jonathan 2273bf86ee test: pin that an at-cutoff upgrade job short-circuits intake (no candidates) 2026-07-11 12:56:54 +02:00
Jonathan 4481f1d998 fix: make intake dedupe cutoff-aware so monitor upgrade path works end-to-end
The monitor's quality-upgrade path was inert: run_pipeline's intake
short-circuit treated any existing LibraryItem as "done", so a
monitor-driven upgrade job for a below-cutoff copy jumped straight to
imported without downloading. Intake now checks whether the job is
monitor-driven (Request.monitoredReleaseId IS NOT NULL) and, if so,
only short-circuits when the existing copy already meets the quality
cutoff; plain manual requests keep exact slice-1 behavior. Also drops
an unused `field` import in browser.py.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-11 12:52:03 +02:00
Jonathan 3934a0c45d feat: import stage upserts to replace lower-quality library items 2026-07-11 12:40:11 +02:00
Jonathan 133f6219ae feat: monitor sweep, real MB browser, and worker-loop wiring 2026-07-11 12:35:07 +02:00
Jonathan 1f23aaa0e8 feat: monitor outcome reconcile 2026-07-11 12:29:26 +02:00
Jonathan 8f72568d87 feat: monitor retry/enqueue phase 2026-07-11 12:24:39 +02:00
Jonathan 7192752d8f test: genuinely exercise discovery ON CONFLICT dedup + config defaults 2026-07-11 12:22:29 +02:00
Jonathan 9a4ce3d5b5 feat: monitor discovery phase + MonitorConfig 2026-07-11 12:16:17 +02:00
Jonathan 2ca0f7a385 feat: MbBrowser seam and FakeMbBrowser test double
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-11 12:12:17 +02:00
Jonathan ec85016d88 feat: monitoring data model (WatchedArtist, MonitoredRelease, Request link) 2026-07-11 12:07:09 +02:00
Jonathan 915bd96cd3 docs: add monitoring-core implementation plan (slice 2, plan 1)
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-11 12:01:57 +02:00
Jonathan dcddcac86c docs: add library-manager design (slice 2)
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-11 11:46:41 +02:00
Jonathan c75b9d3562 fix: force IPv4 in worker, count real Qobuz files, flatten nested output
- worker had no IPv6 route -> Qobuz CDN (dual-stack) failed ~half the tracks
  with 'Network is unreachable'; disable IPv6 in the container (IPv4 only).
- StreamripClient counted metadata tracks (always full) -> a partial download
  falsely imported; now count the actual files written (short count -> needs_attention).
- streamrip writes into a nested 'Artist - Album [..]/' folder the tagger never saw;
  flatten audio files up into the album dir so tagging/organization runs.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-11 01:01:23 +02:00
Jonathan 00b3dddca6 feat: Qobuz auth-token (user_id + token) login as a reliable alternative to email/password
Qobuz's email/password API login returns 401 even with valid credentials
(a known Qobuz limitation, unaffected by streamrip version). Add settings
fields for a Qobuz user ID + auth token (token stored encrypted); when both
are present StreamripClient uses use_auth_token=True, else falls back to
email/password.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-11 00:46:17 +02:00
Jonathan caa0d542ea fix: MD5-hash Qobuz password and never log credential-bearing login errors
streamrip's email/password login expects the MD5 digest, not plaintext
(caused 'Invalid credentials'). Also wrap login() so streamrip's exception —
which embeds the email/password/app_id — can never reach the worker logs;
re-raise a clean credential-free RuntimeError instead.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-11 00:30:34 +02:00
Jonathan f91117092b refactor: use one album path for tagging and library record
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-11 00:09:42 +02:00
Jonathan 1596bab23e feat: wire the mutagen tagger into the worker loop 2026-07-10 23:59:28 +02:00
Jonathan d1fcfd544f fix: make album tagging best-effort per file
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-10 23:57:28 +02:00
Jonathan 32deac65a6 feat: real mutagen tagger and MusicBrainz tracklist 2026-07-10 23:52:19 +02:00
Jonathan 1c02a870f3 feat: Tagger seam and pipeline tag-stage integration
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-10 23:47:20 +02:00
Jonathan 2812695465 feat: organized album directory and persistent /music bind-mount
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-10 23:43:15 +02:00
Jonathan 6cce95cec6 docs: add tagging, import & persistent library plan (slice 1, plan 4b)
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-10 23:41:46 +02:00
Jonathan fef8cf0681 feat: wire MusicBrainz resolver into the worker loop 2026-07-10 23:27:28 +02:00
Jonathan 85c74ba109 feat: add real MusicBrainz resolver and opt-in live test 2026-07-10 23:21:50 +02:00
Jonathan c16e9f1035 feat: MusicBrainz-aware intake and completeness check against canonical track count 2026-07-10 23:17:06 +02:00
Jonathan cf5632aa5f fix: isolate per-adapter search failures in the pipeline match loop
Wrap each adapter's search() call in the match stage in try/except so
a down source (dead slskd, Qobuz auth error, etc.) contributes no
candidates instead of crashing run_pipeline and killing the worker.
2026-07-10 23:12:33 +02:00