Discord (DXP4800 NAS, 7148 tracks): library updates kept dying with "Update appears stuck — no
progress for 300s (last phase: Fetching all tracks in bulk...)". not actually hung.
root cause: the bulk track/album fetch used a single 10000-item page, so a whole library came
back in ONE request that emitted NO progress while in flight. the watchdog (database_update_health)
kills a job with no progress for 300s — so on a slow server that one silent request tripped it even
though it was alive, not stuck. raising the timeout cap only buys the silent request more rope; a
bigger library or slower disk just needs a higher number. the per-batch progress line also only ran
when there was a NEXT page, so a sub-page-size library reported nothing at all.
fix: extract a pure paginate_all_items seam (core/library/bulk_paginate.py) that pages in 1000s and
reports progress after EVERY page — so the watchdog is fed on a cadence set by page size, not library
size, and can't starve mid-fetch no matter how big the library. both Jellyfin bulk loops (tracks +
albums, same defect) now route through it. preserves the failure-shrink resilience (halve to a floor,
then give up). does NOT change what's fetched — same query, fields, items.
note: changes nothing about WHICH tracks come back; only how they're paged + that every page reports.
keep the raised cap on dev as a margin — this is the actual fix. Plex/Navidrome don't share the
pattern (checked). 9 seam tests incl. the watchdog-feed invariant (progress count scales with
N/page_size, never one call for the whole library) + the sub-page regression + failure-shrink.
467 jellyfin/library tests green, ruff clean.
#942's normalize_spotify_oauth_config trimmed whitespace/quotes (good — those can't be part of a
real credential) but ALSO rstrip("/")'d the redirect_uri. that's unsafe: Spotify matches the
redirect URI EXACTLY against the app's dashboard registration, and a trailing slash is a
legitimate part of a URI. stripping it would silently break anyone who registered '…/callback/'
(we'd send '…/callback' → INVALID_CLIENT: Invalid redirect URI) — trading one failure mode for a
sneakier one the user can't diagnose (SoulSync no longer sends what they typed).
drop the rstrip; keep the whitespace/quote trim. the value is now preserved verbatim apart from
unambiguous paste garbage. flipped the test that asserted the strip to assert the slash is kept
(and that whitespace/quotes around it are still trimmed), + a dedicated regression guard.
the #942 integration test mocks normalize, so it's unaffected. 262 spotify/oauth tests green.
credit: builds on HellRa1SeR's #942.
reopened by diegocade1: pasting a Qobuz track link still showed unrelated results. the earlier
fix (b1f061a) only BUBBLED the linked track to the top — but a pasted link is resolved to an
"artist title" text query and searched, and for an obscure track ("foreign lavennew" by colacola)
that text search returns broad lookalikes ("Foreign Bird", "Foreign Spies", …) and never the
actual track. nothing to bubble → user sees junk.
fix: since the link is already resolved via get_track(id), fetch that exact track AS a downloadable
result and inject it at the top (Qobuz downloads by id, so the result is fully usable). the text
search still runs for alternatives.
- QobuzClient.get_track_result(id): get_track + _qobuz_to_track_result; None on any failure.
- _qobuz_to_track_result gains require_streamable (default True for bulk search). the link fetch
passes False: track/get may OMIT the streamable flag, which would default-False and wrongly drop
the exact track the user explicitly asked for. (this closes the one shape assumption that
couldn't be verified against a live Qobuz API — the track is no longer gated on it.)
- track_link.inject_linked_track_first(tracks, linked_result, id): pure seam — prepend the fetched
result + drop any search duplicate; falls back to the bubble when no result was fetched.
- manual-search endpoint fetches linked_result defensively (getattr 'get_track_result') and calls
the seam. Tidal/HiFi (get_track returns a dict but the converter wants an object — shape
mismatch) have no get_track_result, so they keep the existing bubble path: NO regression.
14 tests: inject puts the fetched track first when search missed it / dedups a search copy / falls
back to bubble / str-safe id / noop; get_track_result convert/none/exception; and the REAL
converter builds a valid downloadable result from a track/get dict that OMITS streamable (search
path still rejects it). 85 track-link/qobuz tests green, ruff clean.
the frontend keeps its own copy of the lossless set (settings.js RT_LOSSLESS_FORMATS + the
index.html quality-profile dropdown) — runtime-fetching a yearly-changing list from the backend
isn't worth the coupling. but that duplication IS the exact root cause of #941 (a format added
to one place, not another). so instead of unifying, pin it: two tests parse the frontend lists
and assert they match the backend LOSSLESS_FORMATS. adding a new lossless format now fails CI
until it's added everywhere, instead of silently shipping a half-wired feature.
verified the guard catches drift (not a tautology): a simulated backend-only 'ape' addition
makes the equality fail. 18 lossless tests green, ruff clean.
radoslav-orlov: "create lossy copies of lossless tracks" only recognized FLAC, even though ALAC/
WAV/AIFF/DSD are now quality-profile formats. the FLAC knowledge was hardcoded in 3 separate
places (the import path, the Lossy Converter scan, and the fix executor) — exactly how a format
gets added in one spot but not another.
kettui-style fix — one canonical seam both sites route through, instead of 3 more string edits:
- new core/quality/lossless.py: is_lossless_format / is_lossless_audio_path (pure; injects a
codec probe for the ambiguous .m4a/.mp4 — ALAC vs AAC — so the decision stays testable with no
I/O), LOSSLESS_FORMATS (single source of truth, derived-consistent with model.tier_score), and
the lossy_output_would_overwrite_source safety invariant.
- create_lossy_copy + the Lossy Converter scan + repair_worker._fix_missing_lossy_copy all route
through it. SQL pre-filters by candidate extensions, then each file is confirmed (probing .m4a).
- SAFETY: a lossy copy must never be written over its own source — an .m4a ALAC source + AAC
target lands on the same .m4a path, and ffmpeg runs with -y. all three sites now bail on the
overwrite case BEFORE ffmpeg (the existing delete-original guard was too late — the source was
already clobbered). dropped a vestigial mutagen FLAC import; updated FLAC-only UI strings.
19 tests: full seam coverage (formats, the .m4a ALAC/AAC probe branch, candidate extensions, the
overwrite guard), a tier-model consistency test that fails if the lossless set drifts, and import-
site wiring tests — WAV now converts (was rejected), and the .m4a-ALAC+AAC overwrite case proves
ffmpeg NEVER runs. 286 quality/import/repair tests green, ruff clean.
diegocade1: DSD files (.dsf, ~500MB DSD64) were labeled "Low Quality" and nagged to upgrade.
two independent causes, both fixed (additive — no existing format/behaviour changed):
1) DSF was an unrecognized format -> bottom 'unknown' tier -> "Low Quality":
- source_map: map .dsf/.dff -> 'dsf' (also lights it up in AUDIO_EXTENSIONS, so Soulseek can
match a DSF if one exists)
- model.tier_score: 'dsf' base 102 (just above FLAC) — lands in the lossless range
- probe_audio_quality: add a DSD branch returning format='dsf' (mutagen.dsf for .dsf detail;
.dff classifies lossless without measured detail) instead of None
- settings UI: DSD in RT_LOSSLESS_FORMATS + a "DSD (DSF / DFF)" option in the profile dropdown
2) the actual cause of the screenshot's findings — the truncation guard falsely called DSF
"broken (only ~12% decodes)": ffmpeg decodes DSD to PCM at a different rate than the DSD
container's 2.8 MHz, so astats samples ÷ container-rate massively under-counts. now
detect_broken_audio skips the truncation check for DSD (silence detection still applies).
8 seam tests: dsf/dff -> 'dsf'; dsf tier in lossless range (with + without measured bitrate);
is_dsd_path; and a contrast pair proving the same 12%-decode numbers flag a .flac but skip a
.dsf. 230 quality/import/silence tests green, ruff + JS integrity clean.
ramonskie's thread finding: opening the Import page fires staging files + groups + hints
together, and each one independently os.walk'd the whole staging folder AND mutagen-read every
file's tags — 3x the directory walk and 3x the per-file tag I/O on every page open (the import
scan storm + memory spike on large staging folders).
they all need the same per-file tag data, so scan ONCE: _scan_staging_records walks staging and
reads each file's metadata a single time, returning per-file records that files/groups/hints all
derive from in-memory. a short TTL (6s) + a lock means the three near-simultaneous page-open
requests share one scan instead of each kicking off a full re-read; the lock also prevents
concurrent full scans. hints now derives from the same read_staging_file_metadata the other two
use (same underlying tags) instead of a separate read_tags pass. album/singles process drop the
cache on completion so the list updates immediately after files leave staging.
net: 1 walk + 1 tag-read-per-file on page open instead of 3. 2 tests (shared-scan: 3 endpoints =
1 read per file, not 3; hints updated to the shared reader) + autouse cache-clear fixture; 695
import/staging tests green.
reported (radoslav-orlov): the Stats page hangs ~20s on a 16GB/HDD box, GET /api/stats/cached
?range=7d -> 200 in ~20000ms, while it's instant on Boulder's SSD. the endpoint is documented
"instant" — it reads 3 small precomputed metadata blobs — but it then ran every top
artist/album/track image through normalize_image_url ON THE READ. that fixer calls
cache_url_for, which does a SQLite INSERT/UPDATE + commit under a global lock PER image. on an
HDD each commit fsyncs and contends with the background image fetcher -> ~20s for ~50 images.
(this is exactly the "image caching when we open a page" ramonskie flagged in the thread.)
fix: do the normalization in the background ListeningStatsWorker when it builds the cache
(_enrich_stats_items), so the cache stores browser-ready /api/image-cache/... URLs and the read
does ZERO per-image work. the read-path fixer stays as a cheap no-op (normalize_image_url early-
returns on already-proxied URLs), so an old raw-URL cache self-heals on the next rebuild with no
broken art in between. the registration writes now happen once per rebuild, off the hot path.
2 tests (existing enrich test relaxed to truthy since the value is normalized now; new
deterministic test stubs the fixer to prove every section's url is fixed at build time).
136 stats/listening/image tests green.
Docker report (HellRa1SeR, 2.7.8): pasting YouTube cookies threw
`ERROR: unsupported browser: "custom"`. the 'Paste cookies.txt' dropdown value is the sentinel
'custom', and youtube_client built `cookiesfrombrowser=('custom',)` from it — yt-dlp rejects
that since 'custom' isn't a browser. (server/Docker users have no local browser, so pasted
cookies.txt is the ONLY way to authenticate yt-dlp — e.g. for private 'Liked Music'.)
the correct precedence already existed in core.youtube_cookies.build_youtube_cookie_opts
('custom' -> cookiefile, browser name -> cookiesfrombrowser, falsy -> anonymous) and the web
layer used it — but youtube_client never got migrated and kept 5 inline cookiesfrombrowser
sites. route all of them through a shared _resolve_cookie_opts() that delegates to the tested
helper (reads youtube.cookies_browser + youtube.cookies_file, checks the file exists). also:
the no-cookies download retry now drops cookiefile too (not just cookiesfrombrowser), and
removed 3 now-dead config_manager imports.
3 regression tests (custom->cookiefile, browser unchanged, custom+missing-file->anonymous);
97 youtube tests green.
since 9a0e3b40 persisted completed downloads in the Downloads view, the Clear Completed button
was hidden for those rows and clear-completed only pruned live session tasks. after a restart
the page filled with persisted completed downloads with no way to clear them.
now Clear Completed clears BOTH:
- live session completed/failed tasks (clear_completed_local, unchanged), AND
- the persisted download-history tail: new clear_completed_download_history() deletes every
library_history event_type='download' row, so the list actually empties and stays empty.
this includes unverified rows (the verification review queue) by design: on a library where
verification never confirmed the imports, ALL completed downloads are 'unverified', so preserving
them made the button a no-op. it only removes HISTORY rows — the actual files and their tracks
entries are untouched, so nothing in the library is lost, only the 'needs verification' flags.
the action confirms first (showConfirmDialog, destructive) and the button now shows whenever any
completed/failed row is present.
3 seam tests (clears all incl unverified; leaves non-download history; empty=0); reconcile +
orphan + JS integrity suites green.
four fixes from the review (and a self-correction):
1) close the connection. reconcile_unverified_history_from_tracks opened a connection with no
finally/close. runs once per boot so GC reclaimed it, but now it's consistent + robust.
2) scope the tracks scan to the review queue. it built lookup dicts from EVERY verified/
human_verified track (~350k on a large library) on every boot while anything is unverified
(the normal state). now it loads the stuck rows first and skips verified tracks whose path
AND basename can't match any queued row, so dicts stay proportional to the queue, not the
library. behaviour identical (all 13 PR reconcile tests still pass).
3) close the title-less basename collision. a title-less history row fell back to filename-only
matching with no ambiguity check, so a generic name like "01 - Intro.flac" could heal a
DIFFERENT song to verified. now a title-less basename heal only fires when that basename is
unique among verified tracks; unique-basename rows still heal (recall preserved).
4) "Clean orphaned" protects force_imported rows (deliberate user decision, keep for human
approval) without weakening the mount-down safety gate. CRUCIAL self-correction: filtering
them out BEFORE the orphan check (my first cut) shrank the checked count below the threshold
and would have let a few unverified orphans be deleted during a mount outage. instead,
find_orphan_history_ids now takes a deletable predicate: protected rows still count toward
checked / all-missing (gate stays strong) but never enter the orphan_ids delete set.
3 new regression tests (title-less collision; deletable protects from delete; protected rows
still count toward the gate). 936 verification/acoustid/history/downloads tests green. builds
on nick2000713's #938.
clicking Download Discography → Add all to wishlist added ~1 track every 15-30s. trace: the
endpoint's per-track library-ownership check (track_already_owned → check_track_exists) ran the
LEGACY path — firing search_tracks for every title-variation × artist-variation, per track. on
a large library and an artist you own NOTHING of, STRATEGY-1 (indexed LIKE) always missed and
fell through to the fuzzy fallback (full-table scan), ~10-15 scans/track = the 15-30s. metadata
fetch was never the bottleneck (deezer returns each album in ~1s).
fix: pre-fetch the artist's owned tracks ONCE (get_candidate_albums_for_artist →
get_candidate_tracks_for_albums) and pass candidate_tracks to check_track_exists's batched
in-memory path — the same path the discography backfill job + completion-stream already use.
pass an EMPTY list (not None) when nothing is owned so the owns-nothing case still takes the
fast path → instant. per-track cost drops from ~20s to ~1ms.
safe: track_already_owned's only real caller is this endpoint; the new param is optional
(None = unchanged legacy behaviour for any other caller). normal ownership still detected (same
artist-variation breadth); the one divergence is a track owned ONLY via a compilation → a
harmless redundant wishlist add, which is the endpoint's explicitly-accepted failure mode and
already how the backfill job behaves. 4 new tests; 1134 discog/metadata/wishlist tests green.
#936 added fragmented-row grouping. two follow-ups found in review:
1) BUG: an excluded canonical sibling (a row pinned to the same canonical edition whose tracks
fail the strict fragment match) was emitted with canonical_items but no owned set, so
_build_missing_tracks flagged the ENTIRE tracklist as missing — including tracks the row
already owns (e.g. a 3-track fragment of a 12-track edition reported 12 missing, not 9).
now compute that sibling's own owned slots from its local tracks (shared
_owned_reference_for_tracks helper, same logic as the anchor) so it reports only what it's
actually missing and the count stays internally consistent.
2) PERF: _build_candidate_groups rescanned all albums for every canonical group — O(G*N), which
degrades badly once many editions are pinned (fine today at G=1, latent at scale). invert to
a single O(N) pass that assigns each row to its unique group; identical 'exactly one match'
semantics. also added a Stop/Pause check in _prepare_work_items, which now front-loads the
canonical lookups + matching.
3 new tests (excluded-sibling reports only its missing tracks; ambiguous candidate stays
independent; unambiguous candidate joins) — 17 completeness + 123 repair tests green.
root cause: the library stores album/artist art as media-server RELATIVE paths (Plex
/library/metadata/.., Jellyfin /Items/.., Navidrome /rest/..), which don't render in a browser
<img>. normal wishlist items carry Spotify CDN urls so they show fine, but LIBRARY-sourced
items — dead-file re-downloads and preview-clip re-fetches — carry the raw relative path, so
their album art came up blank. and the nebula only had artist photos for WATCHLISTED artists,
so non-watchlist orbs showed initials.
fix on READ in the wishlist tracks endpoint (so it also repairs items already in the wishlist,
no re-run needed), using the library data we already have:
- normalize each track's album.images url that needs it — relative/internal only, via the
canonical normalize_image_url; CDN urls are left untouched so already-rendering items can't
regress.
- build an artist-name -> normalized library-photo map and return it; the nebula seeds its
artist-image map from it (every wishlist artist), with curated watchlist photos overriding.
8 tests (predicate: relative/internal fixed, CDN untouched; album normalize in-place; artist
map build/skip-empty/idempotent/graceful). 237 wishlist+repair+JS tests green, ruff clean.
two field reports:
1) FIX-ALL skipped these findings — bulk_fix_findings() has a hardcoded fixable_types
allowlist that didn't include 'short_preview_track'. added it, so select-all/fix-all and
the per-page bulk-fix now cover this tool.
2) RE-WISHLISTED ITEM WAS ART-LESS — the payload pulled album art from the library thumb,
which is empty for un-enriched HiFi previews, so the wishlist orb showed initials (no album
OR artist image, since the orb falls back to album art). now the duration lookup also
captures the metadata source's CDN art from raw_data (spotify album.images / itunes
artworkUrl, upscaled) and stores it on the finding; the fix prefers that over the empty thumb.
3 new tests (art capture from spotify raw_data, itunes artwork upscale, fix uses finding art);
8 job tests + repair suite green.
The reconcile heals rows whose file is still in the library; it deliberately
leaves ORPHANS — history rows whose file is gone (deleted / replaced /
re-downloaded elsewhere). Those can never be healed (no file left to confirm)
and linger in the Unverified list forever. This adds an explicit, user-initiated
cleanup for them.
- core/downloads/orphan_history.py: pure, tested rule. A row is an orphan when
its file resolves nowhere; flags `suspicious` when EVERY reviewed file is
unreachable (the mount-down signature) so the caller refuses rather than
mass-delete a healthy log during an outage.
- POST /api/verification/clean-orphans (admin-only): runs it against
_resolve_history_audio_path (raw path -> prefix-swap resolver -> tracks-table
title fallback), refuses on the suspicious signature, and deletes only history
ROWS — never a file (the files are already gone).
- UI: "🧹 Clean orphaned" button in the Unverified bulk-actions row, with a
confirm dialog spelling out that it removes log rows only and refuses if the
library looks offline.
NEVER automatic / never at boot — a filesystem check during a mount outage would
otherwise wipe good history. 5 pure-rule tests + safety-gate coverage.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01LWJk7EuM7YktQeNyqQwTZY
Second-pass audit of the startup reconcile found a real correctness bug in it:
the basename fallback had NO title guard, so a shared track-number filename
("01 - Intro.flac" in different albums) would heal the WRONG song — marking an
actually-unverified file 'verified' and silently dropping it from the review
queue. This is the exact collision the scan-time matcher (history_match.py)
guards against; the reconcile now mirrors it.
- basename match now requires the history row's title and the candidate track's
title to agree (alphanumeric-lowercase), only when BOTH are present (legacy
rows without a title still fall back to filename-only, like the matcher).
- exact-path matches stay unguarded (same path = same file, unambiguous).
- cheap early-out: skip the tracks scan entirely when no 'unverified' rows exist
(keeps the every-boot cost ~nil on healthy libraries).
3 new tests (collision must-not-heal, titles-agree heals, missing-title falls
back). 8 reconcile tests green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01LWJk7EuM7YktQeNyqQwTZY
two field-reported issues: clicking the job sat at 'Starting…' with Scanned: 0 forever and
never showed the current track.
1) HANG — spotify get_track_details() defaults to allow_fallback=True, which scrapes the
configured metadata source when the official API isn't authed (HiFi users). that scrape is
slow and blocked the scan loop on the first track. now pass allow_fallback=False (official
only — fast, returns None cleanly) and fall through to iTunes/MusicBrainz.
2) NO LIVE UPDATE — progress was only pushed every 5 tracks via update_progress, never the
current item. now report_progress() every track (phase + 'artist — title' + scanned/total)
plus a start phase, so the UI moves and shows what it's checking.
also made the test track ids INTEGER to match production (tracks.id is INTEGER PRIMARY KEY),
exercising the real str(id) finding -> WHERE id=? round-trip. 5 tests green.
HiFi (and occasionally other) downloads sometimes deliver a ~30s preview clip instead of the
full song; it lands in the library looking real. new repair job scans short tracks (duration
<= 30s, configurable), looks up the EXPECTED length from the track's metadata source
(spotify/itunes/mb get_track_details), and flags any whose real length is much longer than the
file (default: >= 30s longer) as a preview clip.
approving the finding (repair_worker._fix_short_preview_track) deletes the preview file (path
resolved via _resolve_file_path like the other delete tools), drops the DB row so the track
goes missing, and re-adds it to the wishlist with the full payload (mirrors _fix_dead_file)
so the real version downloads. scan ONLY creates findings — nothing destructive without user
approval, like every other tool.
conservative: genuine short tracks (source agrees they're short) and tracks whose length can't
be verified are skipped, never flagged. registered the job + finding-type label/fix-button in
the UI. 5 tests (scan flag/skip/scope + fix delete+remove+wishlist); 89 repair tests green.
the duration-agreement check used abs() drift, so a file running LONGER than the metadata
(a remaster with a longer outro, an extended cut) was rejected the same as a truncated one —
e.g. A-Ha 'Take on Me' remaster at 228.5s vs 225.0s expected, quarantined for +3.5s.
but a longer file is the OPPOSITE of truncated. make the auto tolerance asymmetric: keep the
tight 3s/5s bound for SHORTER files (the truncation case the check exists for), allow up to
15s in the LONGER direction for version/master differences. a wrong song still trips it (off
by far more than 15s), and a user-pinned tolerance is honoured symmetrically. direction-aware
rejection message too. 4 new tests; 274 integrity/import tests green.
Complements the AcoustID-scan-time heal: re-links library_history rows still
showing 'unverified' to the verified/human_verified status their file already
carries in the tracks table — matching exact path AND basename, so a file that
moved (media-server import / reorganize) heals even though the stored history
path is frozen. Upgrade-only and non-destructive (no deletes, no bulk
migration).
Why this is needed on top of the scan-time fix:
- It clears the EXISTING backlog (e.g. 5551 rows) on the next restart with NO
re-fingerprinting and no AcoustID API calls — the file's status is already in
tracks from the prior scan; this just propagates it to the frozen history row.
- It covers human_verified files, which the AcoustID scan skips entirely
(file_verif_status == 'human_verified' returns early), so their stale history
rows would otherwise never heal.
Runs once on DB init (cheap, idempotent). 5 real-sqlite tests.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01LWJk7EuM7YktQeNyqQwTZY
the AcoustID scanner matched library_history rows by EXACT file_path, but that path is
frozen at import time while the file moves afterward (media-server import / reorganize) —
so tracks.file_path (what the scan reads) no longer equals it. two failures resulted, both
introduced in 37ea6604: verified status never reached the history row (verified tracks kept
showing 'unverified'), and a fresh acoustid_scan row was INSERTed every run (5551 rows for
3675 songs).
- new pure, tested matcher (core/downloads/history_match.py): exact path → filename guarded
by title; prefers a real download row over a synthetic scan row.
- _persist_status now HEALS the matched row's path + status (so future scans match cleanly),
DELETES synthetic acoustid_scan duplicates by exact path (collision-free, never a real row),
and inserts only when the file genuinely has no row.
- a full AcoustID job now self-cleans existing duplicates — no destructive bulk migration.
8 matcher + 4 real-DB heal/dedup/insert tests; existing scanner tests updated to the new
seam (heal vs insert). 1076 acoustid/verification/download tests green.
save_watchlist_scan_run had a single caller — the manual scan endpoint. the automatic/
scheduled path (process_watchlist_scan_automatically) ran the full scan but never wrote a
history row, so nightly scans never showed up in the History modal — only manual ones.
- new shared helper persist_scan_run(database, state, ...) extracts the run from the
finished watchlist_scan_state and writes one history row
- the automatic path now stamps scan_run_id/scan_track_events and calls it
- the manual path is refactored onto the same helper so the two can't drift apart again
- history is global (no profile filter), so the all-profiles nightly scan records one
aggregate row (profile_id None → 1, never NULL)
tests: 4 new persist_scan_run seam tests (real DB) + 2 new auto-scan integration tests
proving the auto path actually records (completed + cancelled, exactly once). 420
watchlist/automation tests green.
a pasted track link IS resolved + searched, but the 'bubble the exact track to the top'
step read getattr(t,'id') — and TrackResult has no top-level id (the source id lives in
_source_metadata['track_id']). so the bubble was a silent no-op: the linked track sat buried
among fuzzy text-search lookalikes and the user saw unrelated tracks. qobuz made it worse —
_qobuz_to_track_result never stamped _source_metadata at all, so the track had no id to match.
- stamp _source_metadata={'source':'qobuz','track_id':...} on qobuz TrackResults (mirrors tidal)
- extract the bubble into pure, tested helpers (linked_track_id / bubble_linked_track_first)
that read _source_metadata['track_id'] — fixes it for tidal too, str/int-safe, stable no-op
19 track-link tests (+6 new) + 87 qobuz/download tests green.
_normalize_album_for_match stripped ANY trailing '- clause', so a real distinguishing
subtitle ('Clair Obscur: Expedition 33 - Nos vies en Lumière (Bonus Edition)') collapsed to
the same name as the OST → _albums_likely_match treated them as one album → the watchlist
marked unowned tracks of one edition as owned via the other and under-wishlisted.
- strip a trailing '- ...' clause ONLY when every token in it is an edition/format
qualifier (+ connectors / year-ordinal): '- Single', '- Acoustic Version', '- 2011
Remaster' still collapse, but real subtitles ('- Nos vies en Lumière', '- Volume 2',
'- Live in Berlin') are kept. Avoids the inverse regression (a same-album pair splitting
into a redownload loop), which a naive narrow strip would have caused.
- drop the loose substring shortcut + raise the fuzzy floor 0.6→0.85; genuine drift already
collapses to an EXACT match, so the looseness only ever produced false fuses.
blast radius: _albums_likely_match has exactly one caller (the allow-duplicates skip).
48 album-match tests pass (qualifier-suffix merges + edition-subtitle splits) + 219 watchlist.
the file faked spotipy + config.settings in sys.modules at import time with no teardown.
the fake config.settings had no ConfigManager, so depending on collection order it leaked
into tests/test_config_save_retry and intermittently failed the full suite. the real
modules import fine in the test env (spotipy is installed, config.settings has both
ConfigManager + config_manager), so the stubs were pure liability — removed them. album
tests still pass (10), the album+config combo that errored now passes (17), 573 repair/
config/canonical tests green.
#929 added 'musicbrainz' to library_reorganize._ALBUM_ID_COLUMNS but not to the
canonical layer, breaking the equality invariant test (canonical reads exactly what
reorganize reads). add musicbrainz to CANONICAL_ALBUM_SOURCES, and move it from the
'can't pin' param group to the 'pins' group in the manual-lock tests — now consistent,
and forward-compatible with pinning a deliberately-matched MB edition. inert at runtime
today (mb isn't in the manual source selector, so should_pin is never called with it).
540 canonical/reorganize tests green.
test_personalized_playlists_id_gate asserted the OLD (wrong) deezer thresholds (>=100000, raw-rank assumption) — the same bug fixed in c033656f. The discovery pool synthesizes deezer popularity to 0-100, so the test now asserts (60, 50). This is the CI failure from running the full suite (my -k subset missed it).
The discovery pool synthesizes deezer popularity onto a 0-100 score (base 45 + bonuses, capped at 100), but _get_popularity_thresholds had deezer on the raw-rank scale (500000/100000). So Popular Picks' 'popularity >= 500000' matched nothing — empty for every deezer-primary user — while Hidden Gems' '< 100000' caught the whole pool. Deezer thresholds now sit on the 0-100 scale (60/50, like Spotify's 60/40). Tested.
Every library track was stored with disc_number=1 because the Jellyfin/Plex/Navidrome scan parsed the track number but never the disc field. Multi-disc albums collapsed onto disc 1, so disc-2+ tracks were mis-filed (shown under disc 1) and flagged 'missing' — the frontend title-fallback band-aid couldn't recover it (breaks on iTunes title mismatches).
Now the shared insert_or_update_media_track reads the disc number (Jellyfin .discNumber=ParentIndexNumber, Navidrome .discNumber, Plex .parentIndex), floors to >=1, and stores it in the INSERT + UPDATE. The disc_number column is ensured on init (it was only added by a migration that doesn't run on fresh installs, so the new INSERT would have hard-failed for new users). The enhanced album view already carries disc_number through (SELECT * -> dict), so the display fixes itself once the column is populated — a re-scan backfills existing libraries. Seam-tested across Jellyfin/Navidrome/Plex shapes + the floor-to-1 + re-scan-update cases.
Opens to the same category-card landing the Discovery Pool uses, with two cards: 'guesses to review' (unverified wing-it) and 'resolved manually' (ones you've Fixed) — click to drill in, Back to return. Previously it jumped straight to a single list.
To populate the resolved list, the /fix endpoint now stamps was_wing_it on the rewritten extra_data (the wing_it_fallback flag is otherwise lost on fix), and get_wing_it_pool gained a resolved flag + the stats return both counts. Fixing/re-matching from either card refreshes in place. Seam test updated for both states.
Wing It auto-matches tracks to the server library on a best-effort guess; those tracks are flagged wing_it_fallback in extra_data and count as 'discovered', so the Discovery Pool hides them — there was no way to see or audit the guesses. New 'Wing It Pool' button (next to Discovery Pool on the Mirrored Playlists tab) opens a modal listing them with a per-playlist filter + search; 'Fix Match' reuses the Discovery Pool's fix flow (/api/discovery-pool/fix), and a manual match drops the track from the pool on refresh.
No new table or provider hooks needed — the wing-it flag is already persisted, so this is a pure query (get_wing_it_pool / get_wing_it_pool_stats, cloning the failed-pool LIKE pattern) + a /api/wing-it-pool endpoint + a cloned modal. Found 81 wing-it tracks on a real library. Seam-tested (include unverified / exclude manual-matched / scope by playlist+profile).
Adds an opt-in `rank_candidates_by_quality` profile flag. When on, the
priority-mode download walk orders candidates by the ranked-target quality
(confidence/speed only break ties) instead of confidence-first. Default off
keeps the byte-for-byte old behaviour, so existing installs are unaffected.
Best-quality search mode is always quality-first regardless of the flag; the
toggle only affects priority mode. Search-time source selection is unchanged —
nothing is skipped, so a track can never go missing, only the order in which
copies are tried changes.
The version-mismatch force-import follows automatically: it accepts the
first-tried (= best-ordered) quarantined candidate, which is the highest-quality
one once the walk is quality-first. No change to its selection logic needed.
- core/quality/selection.py: load_rank_candidates_by_quality() (fail-closed).
- core/downloads/task_worker.py: _best_quality_ordering -> _candidate_ordering;
quality-first when best_quality mode OR the toggle is on.
- database/music_database.py: default profile carries the flag (False).
- web_server.py: flag is preserved globally across preset apply/reset, like
search_mode.
- core/imports/version_mismatch_fallback.py: comment clarified (no behaviour
change).
Tests (TDD): load_rank_candidates_by_quality default/enabled/disabled/error;
_candidate_ordering across all mode+toggle combinations + fail-closed.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Two subsystems post-process the same completed transfer: the browser-poll
status endpoint (web_server) and the background download monitor. Both watch
the same slskd/streaming transfers and each launches the verification
pipeline. When one path quarantines + requeues the next-best candidate
(clearing username/filename, status -> 'searching'), the monitor's
already-submitted run_post_processing_worker then runs, finds no source info,
and falsely marks the task 'failed' ("missing file or source information") —
clobbering the in-flight retry while a parallel attempt imports the song.
Fix: a single atomic claim (downloading/queued -> post_processing under
tasks_lock) so exactly one path processes each download.
- runtime_state: new claim_for_post_processing() helper
- post_processing: race guard — worker bails (no fail/notify) if the task is
no longer 'post_processing' when it runs
- web_server: both poll paths (Soulseek + streaming) claim before launching;
claim is released on thread-launch failure
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The tab reads the v2 personalized framework (personalized_playlists), but the Discover page generates through the legacy path and nothing seeded those v2 rows -> the tab was empty. Fixes:
- New 'listening_mix' v2 generator: hands the scan's stored 'listening_recs_tracks_full' tracks to the personalized manager so the Listening Mix can mirror + Auto-Sync like every other kind (no pool hydration; can't shrink on rotation). Registered + tested.
- Sync tab now lists every registered SINGLETON kind (Listening Mix, Fresh Tape, Archives, Hidden Gems, Discovery Shuffle, Popular Picks) as a card, not just already-generated rows. Clicking 'Refresh & Mirror' runs the generator + mirrors. Variant kinds (decade/genre/daily) need a picker, so they're not auto-listed; existing variant rows still show.
Additive: new generator + frontend merge, no backend endpoint changes. End-to-end verified (refresh -> generate -> persist -> syncable tracks).
The mix is (artist, title) pairs acquired via Soulseek, so the recommendation fetch needn't match the user's active source. When the active source can't fetch top tracks (iTunes/Discogs/MusicBrainz — or Spotify when unauthed), fall back to Deezer's public artist/{id}/top (no auth, available to everyone). All five active sources now build a full mix without switching; the name-search + names_match guard still prevents wrong-artist results. New pure helper choose_mix_fetch_source + tests.
#913 was silently producing 0 recs: similar_artists.source_artist_id is a SOURCE id (Spotify/etc.), but the scan keyed id->name by internal artists.id (resolved nothing), and the consensus ranker was fed the name-collapsed get_top_similar_artists (consensus could never fire). Fixed + elevated:
- id->name keyed by source-id columns; raw per-seed edges (real consensus); similarity_rank threaded into the score; recency-weighted seeds (recent plays boost lifetime favs)
- new 'Based On Your Listening' artist row (/api/discover/listening-recommendations) with 'because you listen to X' explanations
- new 'Your Listening Mix' track row: each rec's top tracks via a guarded, name-resolved Spotify/Deezer fetch (falls back to the discovery pool), stored as full render dicts so the row can't shrink on pool rotation
- pure tested core: similarity_from_rank, build_recency_weighted_seeds, to_mix_track, names_match (+ rank-aware grouping)
Fresh Tape (5-10 tracks): future-dated albums sorted to the top of get_discovery_recent_albums and ate the 50-album budget before the is_future_release skip ran. Add exclude_future_years + fetch a generous budget; downstream caps unchanged. Regression tested.
Also drop the per-track block 'X' from the compact playlist rows (wrong spot). Plan/audit in DISCOVER_BEST_IN_CLASS_PLAN.md.
