Compare commits

...
Author SHA1 Message Date
Renovate Bot 0500f6aac5 chore(deps): update dependency @sveltejs/kit to v3
renovate/stability-days Updates have met minimum release age requirement
renovate/artifacts Artifact file update failure
2026-10-08 14:45:01 +00:00
bvandeusenandClaude Opus 5.5 308045d056 chore(deps): vite 8, vite-plugin-svelte 7 and vitest 5 together (#5021)
release / govulncheck (push) Successful in 33s
release / web (push) Successful in 1m41s
release / go (push) Successful in 1m50s
release / integration (push) Successful in 4m57s
release / android (push) Successful in 6m25s
release / Build signed APK (releases and dev) (push) Successful in 6m37s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 2m13s
release / Verify release artifacts (tag releases only) (push) Skipped
Merges Renovate's three branches (PRs #145, #146, #147), which fail
alone: vite-plugin-svelte 7 requires vite 8, and vitest 5 is the vitest
for vite 8. Renovate changed only package.json, so npm ci failed on each.

The lockfile is regenerated for just these packages: the stale entries
for vite, vitest, @vitest/* and vite-plugin-svelte (with its old
inspector) were dropped and re-resolved, leaving everything else locked.
SvelteKit stays on 2.70.3, which accepts vite 8 and plugin 7. Vite 8
builds with Rolldown, so esbuild moves to 0.28 and rollup leaves the tree.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-08 10:24:51 -04:00
bvandeusen 5d5af359b6 Merge remote-tracking branch 'origin/renovate/vite-8.x' into dev 2026-10-08 10:23:28 -04:00
Renovate Bot 67f1975f53 chore(deps): update dependency vitest to v5
renovate/artifacts Artifact file update failure
renovate/stability-days Updates have met minimum release age requirement
2026-10-08 14:19:20 +00:00
Renovate Bot 5dfbe1361e chore(deps): update dependency vite to v8
renovate/stability-days Updates have met minimum release age requirement
renovate/artifacts Artifact file update failure
2026-10-08 14:19:18 +00:00
Renovate Bot 39c33e4306 chore(deps): update dependency @sveltejs/vite-plugin-svelte to v7
renovate/stability-days Updates have met minimum release age requirement
renovate/artifacts Artifact file update failure
2026-10-08 14:19:16 +00:00
bvandeusenandClaude Opus 5.5 9be5bbae3c fix(discover): cap the taste-matched arm per album and artist before its LIMIT (#5356)
release / web (push) Successful in 1m37s
release / govulncheck (push) Successful in 53s
release / go (push) Successful in 2m7s
release / integration (push) Successful in 5m2s
release / android (push) Successful in 6m34s
release / Build signed APK (releases and dev) (push) Successful in 6m7s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 1m36s
release / Verify release artifacts (tag releases only) (push) Skipped
Summed tag weight rewards a track for carrying many of the user's tags, so
on the deploy two artists whose every track carries the whole lo-fi profile
took all 120 rows of the taste-unheard query. capByAlbumAndArtist ran after
the LIMIT and left 6, and the arm with the lowest skip rate (12% against
~23%) handed its slots to dormant and random.

The query now ranks within album, then within artist over what the album
cap kept, before the LIMIT: the same walk the Go cap makes, so the bucket
fills from as many artists as match. The caps come from the Go constants.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-08 09:24:20 -04:00
bvandeusenandClaude Opus 5.5 9940bc375d feat(android): notifications when the app is closed, a specialUse delivery service (#5347)
release / govulncheck (push) Successful in 37s
release / web (push) Successful in 1m8s
release / go (push) Successful in 1m31s
release / integration (push) Successful in 5m6s
release / android (push) Successful in 6m17s
release / Build signed APK (releases and dev) (push) Successful in 6m13s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 1m13s
release / Verify release artifacts (tag releases only) (push) Skipped
Roundtable's shape: no FCM. A specialUse foreground service keeps the
process alive; EventsStream stays connected; DeliveryLauncher runs the
catch-up on every notification.created nudge and every reconnect, and is
the single owner of the service's lifetime (signed in AND the device's
"Notifications when the app is closed", on by default).

- NotificationSync pulls the newest page and announces unread notices past
  a high-water mark (auth_session.notifiedUpTo, read and written through
  the DAO), honouring the per-kind phone pref. A first look sets the mark
  without announcing; more than three collapse to one line; nothing is
  posted while the app is on screen.
- Two channels: "Your requests" and "Library health"; the ongoing notice
  sits on a MIN "Background connection" channel.
- EventsStream: a connected flow, 2s→5min backoff with ±25% jitter, and an
  immediate reconnect when a network comes up (a hint, not VALIDATED) or
  the app comes to the foreground.
- BootReceiver restarts delivery after a reboot or a self-update.
- A tap opens what the notice links to (routeForLink), a pile the inbox.
- POST_NOTIFICATIONS is asked for on Android 13+ once delivery is wanted,
  and again when the toggle is turned on.
- Room v12: auth_session.backgroundDelivery and notifiedUpTo.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-08 07:59:11 -04:00
bvandeusenandClaude Opus 5.5 63709a433d feat(notifications): grouped email digest, new music as a daily summary (#5346)
release / govulncheck (push) Successful in 21s
release / web (push) Successful in 1m19s
release / go (push) Successful in 1m39s
release / integration (push) Successful in 5m27s
release / android (push) Successful in 5m47s
release / Build signed APK (releases and dev) (push) Successful in 5m34s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 26s
release / Verify release artifacts (tag releases only) (push) Skipped
Nothing is emailed per event. New music (request_completed) goes out at
most once a day, at the summary hour in each user's own timezone, grouped
by artist. Everything else is batched: one email a window after the first
un-emailed item, holding whatever accumulated.

- Migration 0074: notification_email_settings (summary hour, batch window,
  admin-configurable) and user_notification_email_state (batch start, last
  sent, failures and retry_after per user and group). Existing rows are
  stamped emailed so the upgrade sends no backlog.
- The Notifier stamps emailed_at at write time when the recipient's email
  channel is off, so turning email on later doesn't send old items.
- Read rows are never selected. A row is stamped only after the mailer
  accepts, in one transaction with the state, against the read's clock, so
  a coalesced row updated mid-send stays pending.
- A failed send backs off 5m doubling to 6h; SMTP not configured just waits.
- Links come from the public address; without one the email has none.
- The mailer now RFC 2047-encodes subjects and strips line breaks from them.
- Admin → Integrations gains a Notification emails card.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-08 07:48:22 -04:00
bvandeusenandClaude Opus 5.5 62f76290fb fix(android): split notification kinds out of the replayer's dispatch (detekt)
release / govulncheck (push) Successful in 16s
release / web (push) Successful in 1m16s
release / go (push) Successful in 1m37s
release / integration (push) Successful in 4m39s
release / android (push) Successful in 4m59s
release / Build signed APK (releases and dev) (push) Successful in 5m9s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 14s
release / Verify release artifacts (tag releases only) (push) Skipped
dispatch reached cyclomatic complexity 16 with the three M489 kinds; they
now share one entry that hands off to dispatchNotification.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-08 07:38:43 -04:00
bvandeusenandClaude Opus 5.5 c416157a1b feat(android): notifications bell, inbox screen and settings, offline-first (#5343, #5345)
release / govulncheck (push) Successful in 16s
release / web (push) Successful in 1m11s
release / go (push) Successful in 1m29s
release / android (push) Failing after 1m41s
release / Attach APK to the Release (tag releases only) (push) Canceled after 0s
release / Build + push container image (push) Canceled after 0s
release / Verify release artifacts (tag releases only) (push) Canceled after 0s
release / integration (push) Canceled after 2m49s
release / Build signed APK (releases and dev) (push) Canceled after 2m50s
- The top bar carries a bell with a badge. The badge is inverse, not
  error red, and counts to 9, then shows 9+. It opens the Notifications
  screen: one line per notice, how long ago, "Mark all read", pull to
  refresh. A tap marks the notice read and opens its screen (albums,
  artists, requests, admin requests and quarantine; other admin pages
  land on Admin).
- The newest 50 notices live in Room (cached_notifications, v11 with
  MIGRATION_10_11), so the badge and the list work offline. A refresh
  keeps a read made here that the server hasn't seen yet.
- Reads, read-all and setting toggles go through the MutationQueue
  (rule 100): NOTIFICATION_READ, NOTIFICATIONS_READ_ALL and
  NOTIFICATION_SETTING_SET.
  - Read-all sends the newest notice shown, rounded up a millisecond
    (Room keeps ms, the server µs), so a late replay leaves newer notices
    unread.
  - Settings collapse per kind and channel.
- The `notification.created` live event, a return to the foreground and
  reconnecting all refresh the inbox.
- Settings → Notifications: a row per kind with Inbox, Phone and Email.
  Admin kinds sit under "Library health". Phone and email ride on the
  inbox. One line says why email is off. If the system blocks
  notifications, a row opens Minstrel's notification settings; it is
  re-checked on resume.
- Signing out clears the cached inbox.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-08 07:35:56 -04:00
bvandeusenandClaude Opus 5.5 f11dba2336 feat(api): read-all takes an optional up_to cutoff (M489)
release / govulncheck (push) Successful in 16s
release / web (push) Successful in 1m22s
release / go (push) Successful in 1m40s
release / integration (push) Successful in 4m40s
release / android (push) Successful in 5m1s
release / Build signed APK (releases and dev) (push) Successful in 5m12s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 25s
release / Verify release artifacts (tag releases only) (push) Skipped
POST /api/me/notifications/read-all accepts {"up_to": RFC3339}. Android
queues "mark all read" for replay when offline, and a replay landing later
must not mark notices that arrived in between, which the user never saw.
A coalesced notice updated since then has a newer created_at, so it stays
unread. An empty body still marks everything.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-08 07:27:29 -04:00
bvandeusenandClaude Opus 5.5 87f8ed6147 feat(web): notification settings per kind and channel (#5345, web half)
- A Notifications section on Settings has a row per kind and a toggle
  each for Inbox, Phone and Email. Labels are short, menu-style.
- Admin kinds sit under "Library health", for admins only.
- Toggles are optimistic and send only the kind and channel touched. A
  failed save reverts unless something newer has happened (snippet
  #5106's generation counter).
- With the inbox off, phone and email are disabled: they ride on it.
- When email isn't usable, one line says why. With no address it links
  to the profile. With no SMTP an admin gets a link to Integrations and a
  listener is simply told. Saving the profile refreshes the settings so
  the line clears.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-08 07:25:31 -04:00
bvandeusenandClaude Opus 5.5 956058d4a0 feat(web): notifications bell, unread badge and inbox panel in the header (#5342)
- The bell sits between search and the user menu. Its badge is
  parchment on obsidian, not the accent, which the house style keeps
  off general chrome. It counts up to 9, then shows 9+.
- The panel lists the server-rendered title, body and relative time,
  newest first, with unread rows marked. Clicking a row marks it read
  and opens its link. "Mark all read" appears while anything is unread,
  and an empty inbox says "Nothing waiting for you."
- createNotificationsQuery and createUnreadCountQuery poll every 60s
  while the tab is visible. The `notification.created` live event
  invalidates ['notifications'] so the badge and list refresh promptly.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-08 07:23:32 -04:00
bvandeusenandClaude Opus 5.5 1e9408c835 feat(notifications): library health reaches admins, coalesced (#5341)
release / govulncheck (push) Successful in 18s
release / web (push) Successful in 1m22s
release / go (push) Successful in 1m43s
release / integration (push) Successful in 4m56s
release / android (push) Successful in 5m16s
release / Build signed APK (releases and dev) (push) Successful in 5m27s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 15s
release / Verify release artifacts (tag releases only) (push) Skipped
- A failed scan run sends scan_failed. Each failure adds to the count and
  the notice shows the latest error. A scan cut short by shutdown says
  nothing.
- Marking tracks missing sends tracks_missing with a running count.
- A duplicate sweep that proposes a group it had not proposed before
  sends duplicates_found, counting everything awaiting review. A sweep
  that only re-finds known groups stays quiet, so a read notice isn't
  repeated every sweep (CountDuplicateGroupsDetectedSince).
- A playback-error report sends playback_errors, counting the unresolved
  errors (CountUnresolvedPlaybackErrors).

The library package gets its notifier as a package-level SetNotifier
beside SetEventBus, for the same reason the bus is package-level.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-08 07:20:33 -04:00
bvandeusenandClaude Opus 5.5 069baeb14d feat(notifications): requests and flags reach the people who act on them (#5340)
- A new request still pending after any auto-approval notifies the
  admins (request_pending), but not the requester if they are an admin.
  A request that dedups into one already in flight is not announced again.
- Approving or rejecting a request notifies the requester, and a
  rejection carries the admin's notes as the reason. An admin deciding
  their own request gets nothing.
- The reconciler notifies the requester when their request arrives
  (request_completed), linking the matched album or artist.
- A request the re-acquisition sweeper files and cannot approve itself
  notifies the admins.
- A quarantine flag notifies every admin except the flagger, naming the
  track, the flagger and the reason.

lidarrrequests.Service.CreateTracked reports whether a request was
inserted or deduped; Create wraps it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-08 07:17:11 -04:00
bvandeusenandClaude Opus 5.5 94ef8c1887 feat(api): notifications inbox and per-user settings endpoints (#5339)
release / go (push) Successful in 1m35s
release / govulncheck (push) Successful in 17s
release / web (push) Successful in 1m18s
release / integration (push) Successful in 4m35s
release / android (push) Successful in 4m58s
release / Build signed APK (releases and dev) (push) Successful in 5m14s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 14s
release / Verify release artifacts (tag releases only) (push) Skipped
M489 step 2.

- GET /api/me/notifications?limit&before: newest first, keyset-paged on
  (created_at, id) with an opaque cursor, plus the unread count.
- GET /api/me/notifications/unread-count: the badge's cheap call.
- POST /api/me/notifications/{id}/read and /read-all. Mark-read is
  idempotent; another user's id is a 404, the same as a malformed one.
- GET/PUT /api/me/notification-settings: every kind the caller can receive
  (admin kinds only for admins) with inbox/phone/email. PUT is partial, so an
  offline replay sends only what was touched, and a batch with any invalid
  change applies nothing. The response says whether email can be delivered
  at all: no address on file, or SMTP not configured. A failed SMTP config
  read is a 500, not "not configured".

notifications.Render turns kind + payload into title, body and link on the
server, so the web inbox, the Android inbox, the phone's shade and the email
digest all say the same thing. mailer.Configured lifts Send's readiness
check out so settings can report it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-08 07:10:05 -04:00
bvandeusenandClaude Opus 5.5 8bf333e748 feat(notifications): the inbox store, one writer, coalescing and retention (#5338)
release / govulncheck (push) Successful in 42s
release / web (push) Successful in 1m34s
release / go (push) Successful in 1m51s
release / integration (push) Successful in 4m59s
release / android (push) Successful in 5m24s
release / Build signed APK (releases and dev) (push) Successful in 5m32s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 1m20s
release / Verify release artifacts (tag releases only) (push) Skipped
M489 step 1. The event bus is fire-and-forget, so a client that isn't
connected never hears that a request completed or that tracks went missing.
user_notifications is the durable record; the bus only nudges.

- Migration 0073: user_notifications (kind CHECK-gated, payload jsonb,
  read_at, coalesce_key, emailed_at) and user_notification_prefs (per user,
  per kind: inbox, phone, email). A missing pref row means the kind's
  defaults, so nothing is seeded.
- internal/notifications.Notifier is the only writer. It resolves recipients
  (admin kinds reach admins only, and never the excepted user), honours the
  inbox pref (phone and email ride on it), writes, and publishes a
  contentless notification.created nudge per recipient.
- Burst-prone admin kinds coalesce into one unread row: tracks_missing and
  scan_failed add up their counts, duplicates_found and playback_errors take
  the latest total. Once read, the next event is a new row.
- Retention: read rows go after 90 days, anything after a year, on the
  library_changes compactor's daily shape.

Tests: unit (channel rules, kind table) and integration (recipients, nudge,
coalescing both ways, prefs, owner-scoped idempotent mark-read, every kind
against both schema CHECKs, retention cut-offs).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-08 07:06:39 -04:00
bvandeusenandClaude Opus 5.5 5dffe51b95 feat(web): duplicates report flags identical audio under different titles (#3885)
release / govulncheck (push) Successful in 15s
release / Build + push container image (push) Successful in 26s
release / Verify release artifacts (tag releases only) (push) Skipped
release / web (push) Successful in 1m10s
release / go (push) Successful in 1m28s
release / integration (push) Successful in 4m21s
release / android (push) Successful in 4m43s
release / Build signed APK (releases and dev) (push) Successful in 4m57s
release / Attach APK to the Release (tag releases only) (push) Skipped
An exact-tier group whose copies carry different titles means at least one
file's tags are wrong, and the recording the other title names may be missing
from the library. WWW (2020) was this: "WWW" was a second copy of the
instrumental, the vocal was absent, and nothing said so. The report now names
the titles and says what it implies, so the absence surfaces at the moment of
choosing which copy to keep.

Titles compare case- and whitespace-insensitively. Acoustic-tier groups are
left alone: across encodings a "Remastered" suffix is routine, not a mislabel.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-08 06:33:37 -04:00
bvandeusenandClaude Opus 5.5 43c369f082 test(android): queue drag math is a pure function, held to web's cases (#2436)
release / govulncheck (push) Successful in 15s
release / web (push) Successful in 1m11s
release / go (push) Successful in 1m29s
release / android (push) Successful in 4m46s
release / Build signed APK (releases and dev) (push) Successful in 4m58s
release / integration (push) Successful in 15m34s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 14s
release / Verify release artifacts (tag releases only) (push) Skipped
The drag-offset-to-row arithmetic lived inline in queueReorderDrag's
onDragEnd lambda, which no JVM test can reach. Web's copy, offsetToDelta,
has been extracted and tested since the start, and the Android comment
says it mirrors web, but only one side could be held to that.

- QueueDragMath.kt adds queueDragDelta (web's offsetToDelta) and
  queueDragTarget (delta plus the clamp to the queue). Kotlin's roundToInt
  breaks ties toward positive infinity, the same as JS Math.round, so the
  web cases carry over exactly, including half a row up staying put.
- queueDragTarget returns the start index for an empty queue instead of
  letting coerceIn(0, -1) throw.
- QueueDragMathTest mirrors queue-row-math.test.ts one case at a time,
  plus clamping past either end, a sub-half-row drag, an unmeasured row,
  and the empty queue.

No behaviour change for a non-empty queue.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-08 00:18:43 -04:00
bvandeusenandClaude Opus 5.5 b46c080d19 fix(web): all accent text and icons use accent-fg (#5318)
release / govulncheck (push) Successful in 17s
release / web (push) Successful in 1m11s
release / go (push) Successful in 1m26s
release / integration (push) Successful in 4m18s
release / android (push) Successful in 4m44s
release / Build signed APK (releases and dev) (push) Successful in 4m53s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 26s
release / Verify release artifacts (tag releases only) (push) Skipped
The raw accent fails AA as text on every dark surface, not only on its
own tint: 3.04:1 on the page, 2.70 on iron, 2.21 on slate, against 4.5.
accent-fg (the house formula, 45% toward parchment) measures 5.62 at
worst across both modes. The operator chose the readable colour over the
signature teal for text, on 2026-10-08.

- 36 sites swap. They are 35 Tailwind uses: links, "Now playing", the
  ingest progress line, active shuffle/repeat, the liked heart, the app
  download icon and its hover. The last is the alphabet rail's pending
  spinner in CSS. Icons follow the text: as graphics they need only
  3:1, and the raw accent misses even that on iron.
- check-tint-contrast adds accent to TEXT_NEVER_RAW, so a new raw
  text-accent or color: var(--fs-accent) fails the web lane. Run against
  the files before the swap, it finds all 36. Borders, rings and fills
  keep the raw accent.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-08 00:17:19 -04:00
bvandeusenandClaude Opus 5.5 37d3a5bcd3 fix(library): read every genre of FLAC, Ogg, Opus and MP4 files (#2500)
release / govulncheck (push) Successful in 50s
release / web (push) Successful in 2m3s
release / go (push) Successful in 2m18s
release / integration (push) Successful in 5m42s
release / android (push) Successful in 6m30s
release / Build signed APK (releases and dev) (push) Successful in 6m0s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 1m20s
release / Verify release artifacts (tag releases only) (push) Skipped
Vorbis comments repeat a field to give it several values (GENRE=Boom Bap,
GENRE=Downtempo, ...). dhowden/tag keeps comments in a map keyed by field
name, so each repeat overwrote the previous one and only the last genre
was stored. MP4 has the same gap: several data atoms in one ©gen atom, or
repeated ©gen atoms, collapse to one value.

On the operator's library 3,658 of 4,092 FLACs declare more than one
genre. Kupla's Life Forms carries eight and was stored as "Instrumental
Hip Hop" alone, so browse and the taste profile never saw the other seven.

- vorbisgenre.go reads the comment block directly: FLAC's metadata block
  (including FLACs behind an ID3v2 tag), and the comment packet of Ogg
  Vorbis and Opus, reassembled across pages when cover art makes it span
  several.
- mp4genre.go walks moov > udta > meta > ilst and returns every ©gen text
  value. It handles ISO and QuickTime meta layouts and a moov placed after
  mdat. A file with only the numeric gnre atom still falls back to
  dhowden, which resolves it.
- extractGenres routes VORBIS and MP4 through them, as #2499 did for
  ID3v2.
- tagReadVersion 3 -> 4, so the next scan re-reads the tags of files
  already indexed. Unchanged files keep their duration and fingerprint,
  so the pass costs tag reads only.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-07 23:40:56 -04:00
bvandeusenandClaude Opus 5.5 efa3bf54ce fix(playlists): a missing track never seeds For You or Songs-like (#2701)
release / govulncheck (push) Successful in 14s
release / web (push) Successful in 1m11s
release / go (push) Successful in 1m29s
release / integration (push) Successful in 4m27s
release / android (push) Successful in 5m2s
release / Build signed APK (releases and dev) (push) Successful in 5m17s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 1m15s
release / Verify release artifacts (tag releases only) (push) Skipped
PickTopPlayedTracksForUser's liked tier read general_likes without
joining tracks. With no plays to seed from, For You could pick a liked
track whose file is gone. The play tiers were already safe: their
play_events join filters missing_since.

The same gap was in PickTopPlayedTrackForArtistByUser's fallback, which
seeds Songs-like from the artist's newest album when there are no recent
plays. It could pick a missing track, and since #5296 a missing track is
never fetched for similarity, so that seed has no edges either. The
caller already skips an empty seed, so an artist whose tracks are all
missing gets no Songs-like mix instead of one aimed at nothing.

Integration tests cover both cases: a liked-but-missing track is not a
seed, and the Songs-like fallback moves to the next album once the
newest one's track goes missing.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-07 23:10:17 -04:00
bvandeusenandClaude Opus 5.5 743b6f5eac fix(web): error text uses error-fg on every surface, not only on tints (#3150)
release / govulncheck (push) Successful in 17s
release / web (push) Successful in 1m6s
release / go (push) Successful in 1m29s
release / integration (push) Successful in 4m30s
release / android (push) Successful in 5m18s
release / Build signed APK (releases and dev) (push) Successful in 5m32s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 25s
release / Verify release artifacts (tag releases only) (push) Skipped
Raw error red fails AA as text even with no tint behind it: in dark mode
it measures 3.63:1 on obsidian, 3.23 on iron and 2.64 on slate, against
4.5. error-fg (the house formula, 50% toward parchment) measures 5.30 at
worst across both modes.

- All 19 text-error uses become text-error-fg: the "Couldn't load"
  messages on the admin pages, the integrations form errors, the flag
  popover, and the error toast's text. The toast keeps its error border,
  since a border is a graphic with a 3:1 floor.
- check-tint-contrast flags raw error text anywhere (text-error,
  class:text-error, color: var(--fs-error)) and leaves borders and
  outlines alone. Run against the files before the swap, it finds all 19.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-07 23:00:28 -04:00
bvandeusenandClaude Opus 5.5 fdee77eaec fix(web): text on a tint of its own hue uses the house -fg tokens (#3150)
release / govulncheck (push) Successful in 38s
release / integration (push) Successful in 5m1s
release / Build + push container image (push) Successful in 1m17s
release / Verify release artifacts (tag releases only) (push) Skipped
release / web (push) Successful in 1m22s
release / go (push) Successful in 1m48s
release / android (push) Successful in 5m41s
release / Build signed APK (releases and dev) (push) Successful in 5m11s
release / Attach APK to the Release (tag releases only) (push) Skipped
A hue painted as text on a color-mix tint of itself sits close to the
surface under it. On Minstrel's surfaces the raw accent on its 15% tint
measures 1.97:1 at worst (dark mode, hover surface), against AA's 4.5.

- tokens.json gains colors.fg: the five FabledSword -fg formulas (accent
  45%, success 45%, warning, error and info 50%), each mixed toward
  parchment so one declaration serves both modes. Success is Minstrel's
  moss. tokens-to-css emits them in :root.
- Tailwind exposes them as text-accent-fg, text-warning-fg, text-error-fg
  and text-info-fg.
- 23 sites swapped: 14 Tailwind class strings (PlayerBar and the admin
  count pills) and 9 CSS rules (StatusPill's four tones and five accent
  chips). Worst case after: accent-fg 5.03, error-fg 4.75, warning-fg
  4.92, success-fg 4.85.
- scripts/check-tint-contrast.js finds the pair in either spelling. Its
  test scans src in the web Vitest lane and fails on any new site, with
  fixture cases showing it can fail.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-07 22:51:52 -04:00
bvandeusenandClaude Opus 5.5 c17f4273c6 test(web): stub SvelteKit app modules suite-wide so no test loads the client runtime (#3943)
release / govulncheck (push) Successful in 32s
release / web (push) Successful in 1m25s
release / go (push) Successful in 1m41s
release / integration (push) Successful in 4m55s
release / android (push) Successful in 6m15s
release / Build signed APK (releases and dev) (push) Successful in 6m42s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Verify release artifacts (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 2m4s
The home page reaches the real $app/navigation through AlbumCard and
AlbumMenu. That loads SvelteKit's client runtime, whose $app/paths reads
__SVELTEKIT_PAYLOAD__ at module load. The global is only there when the
kit plugin's define reaches the module, and under vitest that is not
reliable: page.test.ts failed to load on CI run 6576 and passed on its
re-run. #374 was the same class of failure.

vitest.setup.ts now mocks $app/navigation, $app/state and $app/paths for
every test. Per-file mocks still win. A small guard test fails if the
suite-wide mocks are removed.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-07 21:26:21 -04:00
bvandeusenandClaude Opus 5.5 e5dac9ddf0 fix(similarity): keep ListenBrainz's whole answer and resolve it locally (#5296)
release / govulncheck (push) Successful in 45s
release / web (push) Successful in 1m27s
release / go (push) Successful in 1m51s
release / integration (push) Successful in 5m36s
release / android (push) Successful in 7m47s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build signed APK (releases and dev) (push) Successful in 8m29s
release / Build + push container image (push) Successful in 1m47s
release / Verify release artifacts (tag releases only) (push) Skipped
The worker kept only the similar recordings already in the library, at most
20 of ListenBrainz's 50, and judged freshness by the edges it had written.
Two failures followed, both measured on the operator's library (#3879):

- A seed whose answer matched nothing wrote nothing, so it was never fresh.
  With the queue ordered by id, 25 such seeds held its head and were
  re-asked every hour; 17 of 2,466 played seeds had any edges.
- A recording that reached the library after its seed was fetched (a
  Lidarr import, an MBID from the AcoustID lookup) was never linked until
  a refetch, which for the stuck seeds never came.

Now every answer is cached whole in listenbrainz_similar_recordings and
every answer, an empty one or a permanent 4xx included, is recorded in
track_similarity_fetches. The queue reads the fetch record: never-fetched
first, then the oldest, refreshed after 30 days. The listenbrainz edges are
derived in SQL from the cache, one present track per recording and no cap,
for the seed just fetched and for every seed once per tick, so new arrivals
link within the hour without asking ListenBrainz again.

Artists get the same queue fix via artist_similarity_fetches; their answer
was already kept in artist_similarity and artist_similarity_unmatched.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-07 19:56:22 -04:00
bvandeusenandClaude Opus 5.5 aa390c711a fix(recommendation): artist-level arms take one track per related artist in turn (#5297)
The similar_artists and coplay_artists arms ordered by artist score, so the
closest related artist's catalogue filled the whole LIMIT. On the operator's
library the 30-row similar_artists arm held exactly one artist for all 17
seeds measured (#3879), though each seed had 7-33 similar artists in the
library. Both arms now rank tracks within each artist and take every
artist's first track, best artist first, before anyone's second.

Both also skip missing tracks: the outer select already dropped them, but
only after they had taken places in the arm's LIMIT.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-07 19:56:21 -04:00
bvandeusenandClaude Opus 5.5 865a3176c9 feat(library): fold a missing track into its on-disk replacement (M485 #5286 #5287)
release / web (push) Successful in 2m21s
release / go (push) Successful in 2m34s
release / govulncheck (push) Successful in 40s
release / integration (push) Successful in 6m17s
release / android (push) Successful in 6m36s
release / Build signed APK (releases and dev) (push) Successful in 6m6s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 1m19s
release / Verify release artifacts (tag releases only) (push) Skipped
A track marked missing whose replacement is already on disk under another
row — on the operator's library 355 of 451 missing tracks, nearly all Lidarr
mp3 -> flac upgrades — is folded into the replacement: likes, plays, playlist
entries and tags move across and the missing row goes. Move adoption could
not catch these: the replacements were re-encodes (no shared audio hash) with
no recording MBID at import.

Pairs (ListMissingTrackPairs): the same recording MBID within the album
group, or the same album row and title ignoring case. Each side must have
exactly one candidate; conflicting MBIDs refuse a pair. No duration or track
position gate: on the 142 pairs known to be one recording, 18% differed by
over 2s and the poorly tagged set is where numbering is broken (spike #5274).

Each pair folds in its own transaction after locking both rows and checking
the pair still holds. Runs automatically (operator, 2026-10-07) after a full
scan, after a watcher batch that added or updated tracks, and after an
AcoustID pass that matched any track. The first scan after deploy repairs the
existing rows.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-07 17:23:32 -04:00
bvandeusenandClaude Opus 5.5 2df28345b4 refactor(library): the per-copy fold is a helper the duplicate merge calls (M485 #5285)
MergeDuplicateGroup's loop body — repoint and copy every FK from a removed
copy onto the survivor, inherit its MBID, delete its row, tidy an emptied
album — moves into foldTrackInto, so the missing-pair pass can fold a stale
missing row into its replacement with the same mechanics. The group lock,
file removal and group bookkeeping stay in the merge. No behaviour change.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-07 17:23:32 -04:00
bvandeusenandClaude Opus 5.5 0766349397 feat(brand): the browser tab icon is the full logo (#5267)
release / web (push) Successful in 1m15s
release / govulncheck (push) Successful in 47s
release / go (push) Successful in 2m19s
release / integration (push) Successful in 5m34s
release / Build signed APK (releases and dev) (push) Successful in 6m23s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / android (push) Successful in 6m10s
release / Build + push container image (push) Successful in 1m30s
release / Verify release artifacts (tag releases only) (push) Skipped
The tab icon was a hand-drawn reduced hat, made because the traced art loses
detail at 16px. A redrawing reads as a different logo. The tab icon now uses
the same traced mark as the header: a high-resolution screen draws a tab icon
from 32px, where it holds, and at 16px it keeps the logo's shape.

The drawn reduced mark had no other consumer, so it leaves the generator,
and mark-small.svg (referenced nowhere) is removed. Regenerating changed only
favicon.svg and favicon.png; every other brand asset is byte-identical.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-07 15:19:10 -04:00
bvandeusenandClaude Opus 5.5 fcbad3ad40 fix(deps): raise golang.org/x/text to v0.41.0 for GO-2026-6629
release / govulncheck (push) Successful in 12s
release / web (push) Successful in 1m18s
release / go (push) Successful in 1m35s
release / integration (push) Successful in 4m34s
release / Attach APK to the Release (tag releases only) (push) Canceled after 0s
release / Build + push container image (push) Canceled after 0s
release / Verify release artifacts (tag releases only) (push) Canceled after 0s
release / Build signed APK (releases and dev) (push) Canceled after 6m5s
release / android (push) Canceled after 6m8s
govulncheck began failing on a new advisory: a panic parsing crafted input in
x/text/secure/precis, reachable from db.Open via pgxpool. x/text is indirect
(through pgx), so the Dependency Dashboard has no update queued for it.
x/text v0.41.0 requires x/sync v0.22.0, which go mod tidy raised with it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-07 15:13:09 -04:00
bvandeusenandClaude Opus 5.5 4e3ce4065c fix(lidarr): complete a request only when the album actually came back (#5263)
release / govulncheck (push) Failing after 33s
release / web (push) Successful in 1m21s
release / go (push) Successful in 1m44s
release / integration (push) Successful in 4m37s
release / android (push) Successful in 5m29s
release / Build signed APK (releases and dev) (push) Successful in 5m34s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Skipped
release / Verify release artifacts (tag releases only) (push) Skipped
Re-acquisition targets albums with ANY track missing, and completion only
asked for a track on disk, so the tracks that never left completed every
re-acquisition request the moment Lidarr accepted the add (52 on the deploy,
each ~150ms after its add).

An album or track request now completes when an album named by its release
or group has a track on disk AND either a track arrived after the request
(a new album, or Lidarr fetching another release into its own row) or no
track that was missing at the request is still missing. added_at is the
arrival clock; updated_at moves on every tag re-read.

Migration 0071 reopens completed album/track requests whose matched album
fails that test, as approved with the match cleared; the Lidarr add stays
confirmed, so nothing is re-sent.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-07 14:55:32 -04:00
bvandeusenandClaude Opus 5.5 e509d7d5a9 feat(lidarr): ask Lidarr for release groups; repair stored release-id requests (M483 #5244)
release / web (push) Successful in 1m41s
release / go (push) Successful in 2m33s
release / govulncheck (push) Successful in 22s
release / integration (push) Successful in 6m0s
release / android (push) Successful in 6m7s
release / Build signed APK (releases and dev) (push) Successful in 6m8s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 1m43s
release / Verify release artifacts (tag releases only) (push) Skipped
Lidarr's metadata is keyed by MusicBrainz release group, but re-acquisition
requested albums by their release id, so every add came back "not found".

- Sweeper requests an album by its tag-supplied release group, else the one
  MusicBrainz names (cached onto the album). An album MusicBrainz cannot name
  is skipped and counted, with no attempt spent.
- Reconciler: an add refused as not found re-reads the request's album id as
  a release (library first, then MusicBrainz), rewrites the request to the
  group and adds again. This repairs the requests already stored.
- Completion matches an album by release id or release group, and only once a
  track of it is on disk, so a re-acquisition request no longer completes
  against the row of the album it is trying to bring back.

Closes #5241.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-07 11:38:45 -04:00
bvandeusenandClaude Opus 5.5 2b4e274b12 feat(tags): resolve a MusicBrainz release to its release group (M483 #5243)
ReleaseGroupForRelease asks /ws/2/release/<id>?inc=release-groups through
the registered MusicBrainz provider, so it shares that provider's client
and 1 req/s limiter with tag enrichment and respects its on/off switch.
ErrNotFound when switched off or MusicBrainz has no such release (an id
that is already a release group included); ErrTransient to retry.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-07 11:34:04 -04:00
bvandeusenandClaude Opus 5.5 5c19a916ba feat(library): store each album's MusicBrainz release-group id (M483 #5242)
release / govulncheck (push) Successful in 22s
release / web (push) Successful in 1m14s
release / go (push) Successful in 1m33s
release / integration (push) Successful in 4m36s
release / Attach APK to the Release (tag releases only) (push) Canceled after 0s
release / Build + push container image (push) Canceled after 0s
release / Verify release artifacts (tag releases only) (push) Canceled after 0s
release / android (push) Canceled after 5m24s
release / Build signed APK (releases and dev) (push) Canceled after 5m29s
albums.mbid is the release id (Picard's musicbrainz_albumid, one edition).
Lidarr names albums by release group, so re-acquisition and request
completion need that id too (#5241).

Migration 0070 adds albums.release_group_mbid (nullable, non-unique index:
several releases share a group). The scanner reads musicbrainz_releasegroupid
through extractReleaseGroupMBID, writes it on insert and heals it onto
existing rows when NULL. tagReadVersion goes to 3 so the next scan fills
it for the library already indexed, bound by tag reads (no ffprobe, no
decode).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-07 11:33:26 -04:00
bvandeusenandClaude Opus 5.5 bf6364b709 fix(lidarr): send the artist add's monitor choice in addOptions (#5239)
release / govulncheck (push) Successful in 29s
release / web (push) Successful in 1m58s
release / go (push) Successful in 2m14s
release / integration (push) Successful in 5m17s
release / android (push) Successful in 6m21s
release / Build signed APK (releases and dev) (push) Successful in 6m38s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 2m1s
release / Verify release artifacts (tag releases only) (push) Skipped
ArtistResource has no top-level `monitor`. Lidarr reads the choice from
AddOptions (AddArtistOptions, a MonitoringOptions), so the "all"/"future"
we sent there was dropped on deserialisation. AddOptions.Monitor stayed
Unknown, and AlbumMonitoredService.SetAlbumMonitoredStatus returns early
on Unknown. The request's monitoring was never applied.

Send monitor and monitored inside addOptions with searchForMissingAlbums,
the shape Lidarr's getNewArtist.js posts, and set monitorNewItems "all"
explicitly: both choices mean new releases are watched.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-07 10:58:55 -04:00
bvandeusenandClaude Opus 5.5 670b30c954 fix(lidarr): add an album as the looked-up resource with its artist nested (#5234)
release / web (push) Successful in 1m42s
release / go (push) Successful in 2m7s
release / govulncheck (push) Successful in 37s
release / integration (push) Successful in 5m31s
release / Attach APK to the Release (tag releases only) (push) Canceled after 0s
release / Build + push container image (push) Canceled after 0s
release / Verify release artifacts (tag releases only) (push) Canceled after 0s
release / android (push) Canceled after 5m14s
release / Build signed APK (releases and dev) (push) Canceled after 4m42s
Lidarr's POST /api/v1/album validates `artist` as a nested resource
(AlbumController: RuleFor(s => s.Artist).NotNull()), so the flat payload
we sent was refused with "'Artist' must not be empty" every time. The
album add has never worked against a real Lidarr; approved album and
track requests sat in the reconciler retrying every 5 minutes.

AddAlbum now does what Lidarr's own add-album UI does (getNewAlbum /
getNewArtist): look the album up by MBID (album/lookup?term=lidarr:<mbid>),
then POST that resource back with monitored + searchForNewAlbum. When
Lidarr doesn't have the artist yet, the nested artist gets the request's
quality/metadata profile and root folder, monitors this album only
(monitor "none" + albumsToMonitor, which AlbumMonitoredService prefers)
and no future releases. An artist Lidarr already has is left as it is.
An MBID Lidarr's metadata doesn't know is ErrNotFound with no POST.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-07 10:55:41 -04:00
bvandeusenandClaude Opus 5.5 e8eee55325 fix(web): AcoustID card's loading line names itself (M401 #3922)
release / integration (push) Successful in 6m4s
release / android (push) Successful in 8m1s
release / Build signed APK (releases and dev) (push) Successful in 8m22s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 1m31s
release / Verify release artifacts (tag releases only) (push) Skipped
release / govulncheck (push) Successful in 28s
release / go (push) Successful in 1m41s
release / web (push) Successful in 1m21s
Its bare "Loading…" made the Integrations page's cover-providers test find
two matches for /loading…/i (Vitest, run 8487). "Loading AcoustID
settings…" also says which card is loading.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 23:47:14 -04:00
bvandeusenandClaude Opus 5.5 03e07e7c66 fix(web): send the empty body api.post requires for AcoustID run-now (M401 #3922)
release / govulncheck (push) Successful in 25s
release / web (push) Failing after 1m18s
release / go (push) Successful in 1m43s
release / Build + push container image (push) Canceled after 0s
release / Attach APK to the Release (tag releases only) (push) Canceled after 0s
release / Verify release artifacts (tag releases only) (push) Canceled after 0s
release / integration (push) Canceled after 4m51s
release / android (push) Canceled after 4m51s
release / Build signed APK (releases and dev) (push) Canceled after 4m6s
svelte-check on 0a7f7883: api.post takes a body. Every other body-less POST
in the client passes {}.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 23:42:25 -04:00
bvandeusenandClaude Opus 5.5 0a7f788390 feat(web): AcoustID card on Integrations — key, threshold, coverage by source (M401 #3922)
release / go (push) Successful in 2m25s
release / web (push) Failing after 26s
release / govulncheck (push) Successful in 21s
release / Attach APK to the Release (tag releases only) (push) Canceled after 0s
release / Build + push container image (push) Canceled after 0s
release / Verify release artifacts (tag releases only) (push) Canceled after 0s
release / integration (push) Canceled after 3m34s
release / android (push) Canceled after 1m3s
release / Build signed APK (releases and dev) (push) Canceled after 1m3s
The card takes the slot of the unimplemented "MusicBrainz overrides"
placeholder. Rows:
- the on switch
- a write-only key field (the stored key is never sent back), with a
  link to register an application
- the minimum score (0.5 to 1)

Below them, recording-id coverage reads as a column: from tags, looked
up, none, and of the none how many are waiting, no match, ambiguous or
failed. There is a "Look up now" button and a folded list of the tracks
the lookup could not settle.

Off, keyless and stopped-short passes are each a visible state with the
reason (rule 164).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 23:29:34 -04:00
bvandeusenandClaude Opus 5.5 3c575b137c feat(library): AcoustID lookup worker fills the MBIDs tags leave empty (M401 #3920 #3921)
release / web (push) Successful in 1m44s
release / go (push) Successful in 2m1s
release / govulncheck (push) Successful in 17s
release / integration (push) Successful in 5m22s
release / android (push) Successful in 5m48s
release / Build signed APK (releases and dev) (push) Successful in 5m53s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 2m6s
release / Verify release artifacts (tag releases only) (push) Skipped
Migration 0069 adds tracks.mbid_source (tag | acoustid), a lookup state per
track (matched | ambiguous | no_match | failed) and the acoustid_settings
row (off, no key, min score 0.85).

The file's tag outranks a lookup (D4). UpsertTrack keeps a looked-up id
through a re-read that finds no tag id and replaces it as soon as one
appears. SetTrackMbidFromAcoustID refuses to write over a tag id.

The worker fingerprints each untagged track with fpcalc's compressed
print, looks it up and writes an id only when D5 settles it: one
recording at or above the threshold, or one left after matching title and
length. Ambiguous and no-match results write nothing. A key AcoustID
refuses, or the service being unreachable, stops the pass and is reported
in the worker's status. It never counts as a verdict on a track.

A changed file drops its lookup in the scan. The re-lookup takes back an
id that no longer matches.

Admin API: GET /api/admin/library/acoustid (settings, status, coverage by
source), PUT …/acoustid-settings (write-only key), POST …/acoustid/run,
GET …/acoustid/unsettled.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 23:26:03 -04:00
bvandeusenandClaude Opus 5.5 f13da62797 feat(library): AcoustID lookup client and the compressed fpcalc print (M401 #3919)
internal/acoustid posts AcoustID's v2 lookup as a gzip form with
meta=recordings, one request per 400ms (their limit is 3/s) and a 20s
deadline. It returns every linked recording with its best score; choosing
among them is the worker's job (D5). The server's error codes map to an
invalid key (stop and say so), a rejected fingerprint (a verdict on the
track) or unavailable (try again later). A cancelled caller stays a
cancellation.

fpcalcLookupArgs and parseFpcalcCompressed read fpcalc's default output,
the compressed string the lookup takes (D1), always over the first 120s.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 23:16:52 -04:00
bvandeusenandClaude Opus 5.5 b28cbe0600 feat(android): refuse plain http:// to a public server address (#5111)
release / go (push) Successful in 1m47s
release / govulncheck (push) Successful in 18s
release / web (push) Successful in 1m15s
release / integration (push) Successful in 4m44s
release / android (push) Successful in 5m37s
release / Build signed APK (releases and dev) (push) Successful in 5m43s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 1m13s
release / Verify release artifacts (tag releases only) (push) Skipped
Cleartext stays permitted app-wide for LAN servers and UPnP (#2439),
but a password or session cookie sent over plain HTTP to a public
address can be read by anyone on the path. A network interceptor now
refuses a cleartext request to the Minstrel server when the connection
lands on a public address, before any request byte is written.

Checked per connection, on the address actually reached, rather than
when the URL is typed: a name that resolved to the home network at
entry resolves to a public address once the phone leaves home.
Allowed: loopback, 10/8, 172.16/12, 192.168/16, link-local, 100.64/10
(Tailscale and other overlay VPNs) and fc00::/7. Only requests
BaseUrlInterceptor tagged as server-bound are checked; external
fetches and UPnP are untouched. The refusal has its own message.

Family baseline #5105, practice 13.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 23:02:07 -04:00
bvandeusenandClaude Opus 5.5 40dd5bb52c ci(android): pin the release certificate in the signer check (#5116)
release / govulncheck (push) Successful in 17s
release / web (push) Successful in 1m20s
release / go (push) Successful in 1m36s
release / integration (push) Successful in 5m25s
release / android (push) Successful in 6m25s
release / Build signed APK (releases and dev) (push) Successful in 7m0s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Verify release artifacts (tag releases only) (push) Canceled after 0s
release / Build + push container image (push) Canceled after 1m3s
The check failed only on a debug signer, so an APK signed by any other
wrong key (a regenerated keystore, a swapped secret) would publish and
then reach no installed phone: Android updates in place only when the
signer matches. The step now requires exactly one signer whose
SHA-256 digest is the release certificate's (CN=Minstrel,
O=FabledSword, read from run 8446), and names a debug key or the
digest it got when it fails. Rotating the key on purpose changes the
digest in the same commit.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 22:51:33 -04:00
bvandeusenandClaude Opus 5.5 a1e9de2c84 ci(android): never ship a debug-signed APK; check the signer (#5116)
release / integration (push) Successful in 5m19s
release / android (push) Successful in 5m57s
release / Build signed APK (releases and dev) (push) Successful in 5m54s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 1m26s
release / Verify release artifacts (tag releases only) (push) Skipped
release / govulncheck (push) Successful in 26s
release / go (push) Successful in 2m5s
release / web (push) Successful in 1m24s
Adopts the rest of family idea #5103 (distributing your own APK):

- Practice 2: build.gradle.kts no longer falls back to the debug key
  when ANDROID_KEYSTORE_PATH is unset; the release build is signed with
  the release key or left unsigned. Main no longer builds and uploads a
  debug-signed app-debug.apk, which no install could ever update.
- Practice 3: android-release runs apksigner on the built APK, prints
  the signer's DN and SHA-256 digest, and fails on a debug signer.
  An unsigned build fails the same step, since there is no
  app-release.apk to verify.
- Practice 9: debug builds offer no server update. The banner does not
  poll and the About card says updates come from Android Studio, since
  the release-signed APK cannot install over a debug-signed app.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 21:33:22 -04:00
bvandeusenandClaude Opus 5.5 13a7a3629a fix(library): MBID backfills skip tracks whose files are missing (#5139)
release / govulncheck (push) Successful in 37s
release / web (push) Successful in 1m13s
release / go (push) Successful in 1m34s
release / integration (push) Successful in 4m55s
release / Build signed APK (releases and dev) (push) Successful in 6m39s
release / android (push) Successful in 6m22s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 1m53s
release / Verify release artifacts (tag releases only) (push) Skipped
The track backfill listed every track with a NULL mbid, missing or
not, so each scan tried to open every missing file and logged an
"open failed" warning per track. Nothing ever healed. The album
backfill could pick a missing track as the one to read, and since
that pass is capped per scan, albums stuck that way were retried
ahead of the rest every time.

Both now read only tracks still on disk; an album with none left is
skipped until a scan finds its files again.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 21:22:05 -04:00
bvandeusenandClaude Opus 5.5 2e36e70268 feat(android): Sonos/UPnP queue plays leveled URLs, rendered a track ahead (M464 #5002)
release / go (push) Successful in 2m49s
release / web (push) Successful in 2m21s
release / govulncheck (push) Successful in 25s
release / integration (push) Successful in 5m54s
release / android (push) Successful in 8m10s
release / Build signed APK (releases and dev) (push) Successful in 8m35s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 1m42s
release / Verify release artifacts (tag releases only) (push) Skipped
Every URL the Sonos queue loader sends is minted with level=true and
the track's album-play verdict from its neighbours in the queue; the
server returns the plain stream when leveling is off or changes
nothing. The playing track and the one after it are rendered ahead,
and each time the renderer moves on, the next is.

Server: a mint no longer prerenders on its own. A queue load mints
every track, which would have started an ffmpeg render per track at
once. The request now carries prerender, and at most two prerenders
run at a time; past that they are dropped, since a fetch renders on
demand anyway.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 19:38:21 -04:00
bvandeusenandClaude Opus 5.5 f34423a0e0 feat: leveled FLAC stream for Sonos/UPnP speakers (M464 #5001)
release / go (push) Successful in 2m17s
release / govulncheck (push) Successful in 26s
release / web (push) Successful in 2m4s
release / Attach APK to the Release (tag releases only) (push) Canceled after 0s
release / Build + push container image (push) Canceled after 0s
release / Verify release artifacts (tag releases only) (push) Canceled after 0s
release / integration (push) Canceled after 4m39s
release / android (push) Canceled after 2m53s
release / Build signed APK (releases and dev) (push) Canceled after 2m24s
Speakers fetch their own audio, so the phone cannot level it. A cast
token minted with level=true (and the client's asAlbum, which only the
queue holder knows) now returns GET /api/tracks/{id}/leveled.flac: the
track rendered by ffmpeg at the user's gain (volume=XdB, plus
alimiter at -1 dBFS for a limiter-mode boost), metadata stripped, FLAC
at 16 or 24 bits and at most 48 kHz. The gain is computed server-side
from the user's preference and the stored loudness, carried as
?g=<centi-dB>&lim=0|1 and signed into the token, so an edited URL does
not verify. Unity gains get the plain stream.

Renders are written beside the cache file and renamed in, keyed by the
source's size and mtime, coalesced per file (singleflight, detached
from the requesting speaker so a retry finds the render running),
started at mint time so the fetch finds them ready, and evicted least
recently used past leveled_cache_mb, a new admin setting (migration
0068, Loudness analysis card).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 19:31:07 -04:00
bvandeusenandClaude Opus 5.5 92c3f9bdb8 fix(android): look gains up by key, not value (M464 #5000)
release / govulncheck (push) Successful in 16s
release / web (push) Successful in 1m28s
release / go (push) Successful in 1m43s
release / integration (push) Successful in 4m38s
release / android (push) Successful in 5m19s
release / Build signed APK (releases and dev) (push) Successful in 5m30s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 1m13s
release / Verify release artifacts (tag releases only) (push) Skipped
`id in map` on a ConcurrentHashMap resolves to its legacy contains(),
which tests values (KT-18053); the compiler refuses it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 19:18:15 -04:00
bvandeusenandClaude Opus 5.5 1013c283da feat(android): level playback with a gain processor in the audio sink (M464 #5000)
release / go (push) Successful in 1m43s
release / web (push) Successful in 1m27s
release / govulncheck (push) Successful in 35s
release / integration (push) Successful in 5m16s
release / android (push) Failing after 3m52s
release / Build signed APK (releases and dev) (push) Failing after 3m20s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Skipped
release / Verify release artifacts (tag releases only) (push) Skipped
Media3 1.10.1 -> 1.11.0, the version Renovate proposes; 1.11 flushes the
sink's audio processors at every item boundary with the playlist timeline
and the new item's period. GainAudioProcessor uses that to find the track
and its play-order neighbours (auto mode's album rule) and applies the
gain from the first sample, gapless transitions included, with a -1 dBFS
peak limiter in limiter mode and a full-scale clamp otherwise.

Gains come from the library cache first (sync now carries track and album
ReplayGain values; Room v10 adds the columns and rewinds the sync cursor
so an existing cache re-pulls them), then GET /api/tracks/replay-gain,
then none. The player service refreshes the leveling preference at start.

Web: a same-album neighbour without a track number no longer counts as
in-order album play, matching Android.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 19:10:35 -04:00
bvandeusenandClaude Opus 5.5 38290bf8f9 feat(web): level playback by the user's normalization preference (M464 #4999)
release / integration (push) Successful in 4m26s
release / android (push) Successful in 5m20s
release / Build signed APK (releases and dev) (push) Successful in 5m31s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 32s
release / Verify release artifacts (tag releases only) (push) Skipped
release / govulncheck (push) Successful in 18s
release / web (push) Successful in 1m13s
release / go (push) Successful in 1m29s
Cuts go through element.volume. Boosts route the element through a Web
Audio GainNode and a DynamicsCompressor (a -1 dBFS limiter in limiter
mode, a pass-through otherwise), built only when a track wants a boost
and only once an AudioContext is confirmed running; iOS never gets the
graph. Auto mode takes album gain when a queue neighbour is from the
same album in track order. Gains are fetched for the next 50 tracks as
the queue moves, with a 10s deadline. The prefetch element is untouched.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 18:21:41 -04:00
bvandeusenandClaude Opus 5.5 2a7eb3dd19 ci(integration): give the suite a 20m package timeout
release / govulncheck (push) Successful in 37s
release / go (push) Successful in 1m32s
release / web (push) Successful in 1m12s
release / android (push) Successful in 6m26s
release / Build signed APK (releases and dev) (push) Successful in 6m49s
release / integration (push) Successful in 19m9s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 1m18s
release / Verify release artifacts (tag releases only) (push) Skipped
internal/api takes ~6.5 min under -race on an idle runner; with a second
run on the same runner it crossed go test's default 10m (run 8368: FAIL
at 600.016s with the running test 2s old, so load, not a hang).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 17:52:01 -04:00
bvandeusenandClaude Opus 5.5 d5dfcf5b7c fix: boost control is a switch; MutationQueue keeps one enqueue per kind (M464 #4998)
release / go (push) Successful in 2m15s
release / govulncheck (push) Successful in 29s
release / web (push) Successful in 1m42s
release / android (push) Successful in 5m57s
release / Build signed APK (releases and dev) (push) Successful in 5m49s
release / integration (push) Failing after 20m33s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Skipped
release / Verify release artifacts (tag releases only) (push) Skipped
The ListenBrainz settings test finds the page's one checkbox, and the
boost control is a toggle anyway. detekt counts MutationQueue's enqueue
functions; suppressed as the replayer's dispatchers already are.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 14:34:28 -04:00
bvandeusenandClaude Opus 5.5 af36b2f24a feat: per-user volume leveling preference, synced across devices (M464 #4998)
release / web (push) Failing after 1m5s
release / govulncheck (push) Successful in 22s
release / go (push) Successful in 1m17s
release / android (push) Failing after 1m51s
release / integration (push) Canceled after 4m12s
release / Attach APK to the Release (tag releases only) (push) Canceled after 0s
release / Build + push container image (push) Canceled after 0s
release / Verify release artifacts (tag releases only) (push) Canceled after 0s
release / Build signed APK (releases and dev) (push) Canceled after 3m21s
Mode (off, auto, track, album), target (-18, -16, -14 LUFS) and boost
(within headroom, or fully with a limiter), stored per user on the server
so the web player, the Android app and casts apply the same one.

- Server: user_normalization_prefs (migration 0067), GET/PUT
  /api/me/normalization; a whole-body PUT, validated, last write wins.
- Web: Settings > Playback > Volume leveling. Saves at once, restores the
  old choice if the save fails, and caches the value for the player.
- Android: Settings card. The device keeps a copy for offline playback
  (Room v9 with an explicit migration, so the upgrade wipes nothing).
  Writes are offline-first: shown at once, PUT best effort, queued on
  failure (NORMALIZATION_SET, collapsed to the newest). A refresh never
  overwrites a change still queued.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 14:29:40 -04:00
bvandeusenandClaude Opus 5.5 e3aa8629d3 feat(api): deliver loudness gains to every client (M464 #4997)
release / web (push) Successful in 1m44s
release / go (push) Successful in 2m12s
release / govulncheck (push) Successful in 40s
release / android (push) Successful in 5m28s
release / Build signed APK (releases and dev) (push) Successful in 4m42s
release / integration (push) Successful in 15m33s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 1m21s
release / Verify release artifacts (tag releases only) (push) Skipped
ReplayGain 2.0 values (gain to -18 LUFS, linear peak) derived from the
stored track and album loudness:

- Web: GET /api/tracks/replay-gain?ids=... (up to 200), a lookup the
  player calls for its queue, rather than a field on every TrackRef
  surface.
- Android: track_gain/track_peak and album_gain/album_peak on the sync
  views, so cached tracks level offline. Storing a measurement logs a
  track change, and an album's values moving logs an album change, both
  before the write (#2704), so caches pick the gains up.
- OpenSubsonic: replayGain on every song (album, getSong, search3,
  starred), as a JSON object and an XML element.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 13:54:05 -04:00
bvandeusenandClaude Opus 5.5 c2f81bf8df feat(library): album loudness from the tracks' summed block histograms (M464 #4996)
release / go (push) Successful in 1m47s
release / govulncheck (push) Successful in 27s
release / web (push) Successful in 1m27s
release / integration (push) Successful in 4m51s
release / android (push) Successful in 6m23s
release / Build signed APK (releases and dev) (push) Successful in 6m25s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 1m30s
release / Verify release artifacts (tag releases only) (push) Skipped
Album-mode normalization plays a whole album at one gain. That gain comes
from album_loudness (migration 0066): BS.1770's gated loudness over every
block on the album, computed by summing the tracks' stored histograms and
gating the sum. No audio is decoded again. Album true peak is the loudest
track's.

- Recomputed by the loudness worker each tick, after the track pass and
  whether or not analysis is switched on. ListAlbumsNeedingLoudness lists
  albums whose md5 over (present track id, measurement version and time) no
  longer matches the stored digest. One comparison covers every way
  membership changes (scan retag, duplicate merge, delete, missing and
  restored) without hooking each.
- No album value until every present track has a settled measurement, so
  an album's gain doesn't shift mid-listen as the rest is measured. Silent
  and unreadable tracks count as settled.
- Rows for albums with no present track left are dropped.
- The parser and the merge share trimBins.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 13:27:02 -04:00
bvandeusenandClaude Opus 5.5 aee4b50bd4 test(api): pass the loudness settings to Mount in the library test router (M464 #4995)
release / go (push) Successful in 2m21s
release / web (push) Successful in 2m7s
release / govulncheck (push) Successful in 19s
release / integration (push) Successful in 5m15s
release / android (push) Successful in 6m24s
release / Build signed APK (releases and dev) (push) Successful in 6m37s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 2m24s
release / Verify release artifacts (tag releases only) (push) Skipped
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 13:06:44 -04:00
bvandeusenandClaude Opus 5.5 f3196b3443 feat(library): measure every track's loudness in the background (M464 #4995)
release / go (push) Failing after 1m18s
release / govulncheck (push) Successful in 35s
release / web (push) Successful in 1m27s
release / android (push) Canceled after 5m47s
release / Build signed APK (releases and dev) (push) Canceled after 4m21s
release / integration (push) Failing after 4m13s
release / Attach APK to the Release (tag releases only) (push) Canceled after 0s
release / Build + push container image (push) Canceled after 0s
release / Verify release artifacts (tag releases only) (push) Canceled after 0s
The first step of loudness normalization: the server measures each track
with ffmpeg's EBU R128 filter (true peak, mono as dual mono) and stores the
integrated loudness, true peak and loudness range in track_loudness
(migration 0065).

It also keeps a histogram of the 400 ms gating blocks at 0.1 LU, so album
loudness can be computed exactly later with no second decode (#4996). The
histogram reproduces ffmpeg's own figure (-10.68 against -10.7 on the
captured fixture), and the analyzer logs a warning if the two ever drift.

- A background worker, cloned from the fingerprint backfill, measures every
  track, new ones included. Measuring inline in the scan was dropped: the
  analysis decodes the whole file, and a large import could pass the scan's
  one-hour stuck threshold. The scan only deletes a changed file's
  measurement; the worker ticks every 10 minutes.
- Timeouts, the cancel/missing-binary split and settled verdicts follow the
  fingerprint runner. Silence and undecodable files are stored as verdicts;
  stalls are retried. The deadline scales with track length.
- loudness_settings (enabled, files at once) and an admin card with the
  coverage gauge, under GET/PUT /api/admin/library/loudness-settings and
  GET /api/admin/library/loudness.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 13:01:01 -04:00
bvandeusenandClaude Opus 5.5 edd9a3a6db fix(auth): the Subsonic password is generated, never the login password (M462 #5026)
release / govulncheck (push) Successful in 39s
release / web (push) Successful in 1m8s
release / go (push) Successful in 1m30s
release / integration (push) Successful in 4m37s
release / android (push) Successful in 5m56s
release / Build signed APK (releases and dev) (push) Successful in 5m55s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 1m23s
release / Verify release artifacts (tag releases only) (push) Skipped
`minstrel admin reset-password` copied the new login password into
subsonic_password, which is stored in plain text because Subsonic t/s
sign-in needs it. Every account recovered through the CLI had its login
password readable in the database, and changing the password later left
the copy behind.

- reset-password now changes only password_hash.
- Migration 0064 clears every subsonic_password, removing the copies.
- Settings gets a Subsonic password card: the server generates a random
  password, shows it once, and it can be regenerated or turned off
  (GET/POST/DELETE /api/me/subsonic-password, audited). Generated rather
  than user-chosen so it can never be a reused password.
- docs/security.md describes the separate password instead of the known
  issue.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 10:54:30 -04:00
bvandeusenandClaude Opus 5.5 522503e011 docs: hosting guide and security notes; README setup and HTTPS guidance (M462 #4986)
- docs/hosting.md: LAN vs internet; binding 4533 to 127.0.0.1 behind an
  HTTPS proxy (Caddy example, no buffering, long read timeouts for SSE and
  streams); the Client IP detection hop count (default 1, so 0 with no
  proxy or clients can forge X-Forwarded-For); the public address that
  password-reset links need; finding the setup token.
- docs/security.md: sessions, API keys, rate limits, headers and CSP; why
  CSRF rests on SameSite=Strict plus JSON-only cookie writes; the Subsonic
  password column, including the known issue that admin reset-password
  writes the login password there (#5026); why Android allows plain HTTP;
  the CI publish gate.
- README: keeps the LAN-first port mapping with a pointer for internet
  hosts, scopes "plain http:// is fine" to trusted networks, explains the
  setup token in first-run step 1, and links both docs.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 10:37:32 -04:00
bvandeusenandClaude Opus 5.5 6de8d4136d feat(android): keep the session cookie in Keystore-encrypted storage (M462 #4985)
release / govulncheck (push) Successful in 18s
release / web (push) Successful in 1m18s
release / go (push) Successful in 1m40s
release / integration (push) Successful in 4m29s
release / android (push) Successful in 5m17s
release / Build signed APK (releases and dev) (push) Successful in 5m20s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 16s
release / Verify release artifacts (tag releases only) (push) Skipped
The session cookie is a bearer credential, and it sat in plain text in the
Room auth_session row. It now lives in a SessionVault: AES-256-GCM under a
key held in the Android Keystore, with only the ciphertext in a private
prefs file. A copy of the app's files no longer yields a usable session.
Platform APIs only, no new dependency (androidx.security-crypto is
deprecated).

Nobody is signed out by the upgrade. On first launch AuthStore moves a
cookie still in the row into the vault and clears the column. If the
Keystore can't be used on a device, the cookie stays in the row as before
rather than being lost. A sign-in or 401 that lands during the move wins
over the value it read, and the move never throws. The auth gate now
waits for this before choosing Login or Home, with a 10s deadline so a
wedged Keystore can't leave the start screen spinning.

Tests: AuthStoreSessionVaultTest (upgrade move, vault-only load, Keystore
fallback, sign-in/out, hydration race) and SealedBoxTest (round trip,
fresh IV, tamper and wrong-key rejection). The real Keystore path needs
a device; the first launch after updating is that check.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 10:29:32 -04:00
bvandeusenandClaude Opus 5.5 60c87da38e ci(govulncheck): check out with plain git; the golang image has no node (M462 #4984)
release / go (push) Successful in 2m13s
release / web (push) Successful in 1m40s
release / govulncheck (push) Successful in 41s
release / integration (push) Successful in 4m54s
release / android (push) Successful in 5m32s
release / Build signed APK (releases and dev) (push) Successful in 5m11s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Successful in 1m15s
release / Verify release artifacts (tag releases only) (push) Skipped
actions/checkout runs on node, which golang:1.26-bookworm does not carry,
so the lane died at checkout (run 8272, exit 127) before scanning
anything. That run was also the gate's first red: every other lane
passed, and image-release and release-assets both skipped, leaving :dev
on the previous build.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 10:00:24 -04:00
bvandeusenandClaude Opus 5.5 b36125fa67 ci: one workflow graph, so nothing publishes on red; add govulncheck and npm audit (M462 #4984)
release / govulncheck (push) Failing after 2s
release / go (push) Successful in 1m49s
release / web (push) Successful in 1m8s
release / integration (push) Successful in 4m39s
release / android (push) Successful in 5m45s
release / Build signed APK (releases and dev) (push) Successful in 5m58s
release / Attach APK to the Release (tag releases only) (push) Skipped
release / Build + push container image (push) Skipped
release / Verify release artifacts (tag releases only) (push) Skipped
test-go, test-web and android were separate workflows on the same push as
release.yml, so the image build could not see their verdict: :dev meant
"it built", never "it passed". All lanes now live in release.yml, and
both publishing jobs (image-release and the new release-assets) need
every lane and require `result == 'success'` from each by name, so a
skipped lane blocks the publish just as a failed one does (rule 177).

- New lanes: govulncheck (in golang:1.26-bookworm, the builder's image,
  so it checks the stdlib that ships) and `npm audit --omit=dev` in web.
- Attaching the APK to a Release moved out of android-release into
  release-assets, behind the gate; the APK still builds in parallel.
- `docker buildx build --pull`, so floating base tags can't serve a
  stale Go patch release from the runner's cache.
- Integration wait uses `pg_isready` via docker exec: the old /dev/tcp
  probe never connects under dash (rule 81) and burned two minutes a run.
- workflow_dispatch input force_red fails the go lane on purpose, to
  watch the gate refuse.
- release_gate_test.go pins the gate: every job must be classified, and
  every publisher must need and require success from every lane.

Lanes have no path filters any more; a web-only push runs the Go suite
too, because "not run" must never read as "passed".

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 09:51:37 -04:00
bvandeusenandClaude Opus 5.5 3217e10168 fix(deps): clear known vulnerabilities in what ships; build on the Go line CI tests (M462 #4984)
- golang.org/x/text v0.37.0 -> v0.39.0 (GO-2026-5970, infinite loop on
  invalid input, reachable from pgxpool). x/sync follows to v0.21.0.
- web lockfile: in-range updates from `npm audit fix` for devalue (high)
  and svelte (moderate), both of which ship in the browser bundle.
  package.json is unchanged.
- Dockerfile builder golang:1.25 -> golang:1.26. CI has tested on 1.26
  since the ci-go migration while the image was still compiled with 1.25,
  left over from the April skeleton; the shipped binary now uses the
  toolchain the tests ran on. govulncheck under golang:1.26-bookworm
  (go1.26.8) reports 0 vulnerabilities reachable from our code.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 09:51:37 -04:00
bvandeusenandClaude Opus 5.5 327d49428f feat(auth): store Subsonic API keys hashed; a new key is shown once (M462 #4983)
test-web / test (push) Successful in 2m4s
test-go / test (push) Successful in 2m23s
test-go / integration (push) Successful in 5m28s
release / Build signed APK (releases and dev) (push) Successful in 6m29s
release / Build + push container image (push) Successful in 29s
release / Verify release artifacts (tag releases only) (push) Skipped
users.api_token held each user's apiKey in plaintext and was looked up by
equality, so a leaked row or backup handed out working keys. Migration
0063 replaces it with api_token_hash (sha256, hex), computed in place
from the existing keys so every Subsonic client keeps working.

The key can no longer be read back: GET /api/me/api-token is gone, and
POST returns the new key once. Settings shows it right after Regenerate
with a copy button and a "won't be shown again" note.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 09:42:35 -04:00
bvandeusenandClaude Opus 5.5 2f3fbccab6 feat(auth): first account on a new server needs the setup token from the server log (M462 #4982)
test-go / test (push) Successful in 2m3s
test-web / test (push) Successful in 1m14s
test-go / integration (push) Successful in 4m38s
release / Build signed APK (releases and dev) (push) Successful in 5m36s
release / Build + push container image (push) Successful in 1m24s
release / Verify release artifacts (tag releases only) (push) Skipped
While no accounts exist, the server mints a random setup token at boot and
logs it. Registering the first account (which becomes admin) must carry it,
so whoever reaches a freshly exposed instance first cannot claim it. The
register page asks GET /api/auth/setup-status and shows a "Setup token"
field in place of the invite field while setup is pending.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 09:35:52 -04:00
bvandeusenandClaude Opus 5.5 46194a609d fix(auth): build password-reset links from an operator-set public address, never the Host header (M462 #4981)
test-go / test (push) Successful in 1m29s
test-web / test (push) Successful in 1m37s
release / Build + push container image (push) Canceled after 0s
release / Verify release artifacts (tag releases only) (push) Canceled after 0s
test-go / integration (push) Canceled after 2m45s
release / Build signed APK (releases and dev) (push) Canceled after 3m40s
buildResetURL used r.Host and r.TLS, so a forgot-password request with a
forged Host emailed the victim a real reset token on a link to the
attacker's server. Links now come only from network_settings.public_url
(migration 0062), and no reset email is sent while it is empty; the response
stays the same opaque 200 and the log says why.

The address is set on a new "Public address" card under Admin → Integrations,
which offers the page's own origin and warns while unset. PUT
/api/admin/network-settings takes either field alone, so the proxy card and
this one can't overwrite each other.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 09:32:10 -04:00
bvandeusenandClaude Opus 5.5 d411693bb2 feat(server): security headers and a hash-based CSP for the web app (M462 #4980)
test-web / test (push) Successful in 55s
test-go / test (push) Successful in 1m14s
release / Build + push container image (push) Canceled after 0s
release / Verify release artifacts (tag releases only) (push) Canceled after 0s
release / Build signed APK (releases and dev) (push) Canceled after 2m50s
test-go / integration (push) Canceled after 2m50s
There were no security headers at all. Now:
- every response: nosniff, Referrer-Policy strict-origin-when-cross-origin,
  Permissions-Policy (no camera/mic/geolocation), X-Frame-Options DENY;
  each set only if the handler hasn't.
- HSTS only when the trusted proxy reports HTTPS (rule 94); never a redirect.
- index.html carries a Content-Security-Policy whose script-src is 'self'
  plus the sha256 of each inline script in the page as served, computed
  after the branding template runs. No 'unsafe-inline' or 'unsafe-eval'
  for scripts. img-src admits remote https/http because Lidarr suggestion
  art is a remote poster URL.

Hashing in Go rather than via SvelteKit's kit.csp covers the inline scripts
SvelteKit doesn't know about (app.html's theme bootstrap and the branding
global injected at build) and stays correct whatever the app name is.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 09:29:26 -04:00
bvandeusenandClaude Opus 5.5 24b767c23b feat(server): cap request bodies, bound body reads, private cache headers, JSON-only cookie writes (M462 #4979)
test-go / test (push) Successful in 1m48s
test-web / test (push) Successful in 2m0s
android / Build + lint + test (push) Successful in 6m27s
release / Build signed APK (releases and dev) (push) Successful in 6m42s
test-go / integration (push) Successful in 4m55s
release / Build + push container image (push) Successful in 1m20s
release / Verify release artifacts (tag releases only) (push) Skipped
- Every request body is capped at 4 MiB and must arrive within 30s. The
  deadline is set per request and cleared at end of body rather than via
  http.Server.ReadTimeout, which would cancel audio streams and the SSE
  stream once the background read hit it.
- IdleTimeout 120s closes idle keep-alive connections. Still no global
  WriteTimeout, for the same streaming reason.
- Streams, album covers and playlist covers are Cache-Control: private, so
  a shared cache never keeps an authenticated response for others.
- A cookie-authenticated write to /api must be application/json (415
  otherwise). SameSite=Strict can't see a sibling app on the same
  registrable domain; forms and no-preflight fetches can't send JSON.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 08:34:26 -04:00
bvandeusenandClaude Opus 5.5 755f997b0d feat(auth): sessions expire server-side; password change and reset end other sessions (M462 #4978)
Sessions had no server-side expiry: only the web cookie's 30-day Max-Age
limited them, and a bearer token (Android) lived until revoked by hand.
GetSessionByTokenHash and ListSessionsForUser now ignore sessions idle for
30 days or older than a year, and the GC worker deletes them hourly.

A password change was a plain UPDATE, so a session opened with the old
password survived it. Now:
- self-service change signs out every other device and keeps this one;
- reset by email ends every session the account has;
- an admin reset ends the target's sessions (keeping the admin's own when
  they reset themselves).

The success copy on web and Android says the other devices were signed out.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 08:32:39 -04:00
bvandeusenandClaude Opus 5.5 719dc62b0d fix(auth): set the session cookie's Secure flag behind a TLS-terminating proxy (M462 #4977)
The cookie set Secure from r.TLS, which is nil whenever TLS terminates at
Traefik or Cloudflare, so a public deployment's session cookie went out
without Secure. auth.IsHTTPS now decides it from X-Forwarded-Proto, believed
only through the trusted-hop count (rule 94) and read positionally like
X-Forwarded-For. Plain-HTTP and LAN logins are unchanged; nothing redirects.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 08:30:46 -04:00
bvandeusenandClaude Opus 5.5 3bfddd0862 feat(auth): throttle login, register, password reset and Subsonic auth failures (M462 #4976)
test-go / test (push) Successful in 1m54s
test-web / test (push) Successful in 1m34s
test-go / integration (push) Successful in 4m56s
android / Build + lint + test (push) Successful in 5m41s
release / Build signed APK (releases and dev) (push) Successful in 5m52s
release / Build + push container image (push) Canceled after 0s
release / Verify release artifacts (tag releases only) (push) Canceled after 0s
Every password-shaped check was mounted bare, so guessing was limited only
by bcrypt cost. A shared in-memory AttemptLimiter now sits in front of them:

- login: 10 failures per account and 50 per address per 15 min, checked
  before the user lookup and bcrypt; 429 with Retry-After. A success clears
  the account's count but not the address's.
- unknown usernames run a dummy bcrypt compare, so timing no longer says
  which accounts exist.
- register: 10 per address per hour; forgot-password: 5 per address and 3
  per email per hour (applied whether or not the email matches); reset: 20
  failed tokens per address per 15 min.
- Subsonic /rest: same limits as login, counting only wrong credentials,
  since clients authenticate on every request.

Web login, register, reset and forgot-password screens say how long to
wait; web and Android carry copy for the rate_limited code.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-06 08:28:29 -04:00
bvandeusenandClaude Opus 5 516413f4ca fix(admin): re-acquisition settings take effect without a restart, and say why a save was refused (#3936, #3937)
test-web / test (push) Successful in 1m9s
test-go / test (push) Successful in 1m28s
test-go / integration (push) Successful in 3m57s
release / Build signed APK (releases and dev) (push) Successful in 5m20s
release / Build + push container image (push) Successful in 1m23s
release / Verify release artifacts (tag releases only) (push) Skipped
#3936: Router() built a reacquisition.SettingsService of its own, so a save
from the admin card refreshed that instance's cache while the sweeper in
main.go kept serving what it loaded at boot. The card showed the new
policy, the feature ran the old one, and only a restart reconciled them.
main.go now hands its instance to the server (srv.ReacqSettings), as it
already did for RecSettings, TagSettings and FingerprintSettings, and
Router() constructs one only when that field is nil. The regression test
saves through the router and reads the sweeper's instance.

#3937: the card's catch tested `e instanceof Error`, but api.put throws a
plain {code, message, status} object, so every reason the server gave was
discarded in favour of "Couldn't save settings." It now uses errMessage,
which appends the server's message for invalid_setting. Its test rejected
with an Error no code path produces, so it passed throughout; it now
rejects with what the client actually throws.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-11 20:15:35 -04:00
bvandeusenandClaude Opus 5 37d4906033 test(api): pass fingerprint settings to Mount in the route-registration test (M400 #3913)
test-go / test (push) Successful in 1m0s
test-go / integration (push) Successful in 3m25s
release / Build signed APK (releases and dev) (push) Successful in 4m35s
release / Build + push container image (push) Successful in 25s
release / Verify release artifacts (tag releases only) (push) Skipped
The unprefixed Mount call in library_test.go was missed when #3913 added
the parameter, failing go vet. Also pins the fingerprint coverage,
fingerprint settings and duplicates routes as admin-gated.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-11 17:58:05 -04:00
bvandeusenandClaude Opus 5 077ae61235 feat(admin): fingerprinting settings — on/off, length, match threshold, concurrency, sweep interval (M400 #3913)
test-go / test (push) Failing after 44s
test-web / test (push) Successful in 49s
test-go / integration (push) Failing after 2m42s
release / Build + push container image (push) Canceled after 0s
release / Verify release artifacts (tag releases only) (push) Canceled after 0s
release / Build signed APK (releases and dev) (push) Canceled after 4m8s
Rule 25: the fingerprinting knobs move out of source into a DB-backed
singleton (migration 0061), edited from a card on the Duplicates page and
shared live with the scanner, the backfill and the duplicate sweep through
one service instance, so a save needs no restart.

The length is the knob that can silently break the library: prints taken
at two lengths never match. Each track_fingerprints row now records the
length it was taken at, and every reader filters on the current one — the
backfill treats another length as stale, the gauge counts it pending, the
sweep never streams it. Equivalent to a version bump, except that setting
the length back makes rows not yet redone current again. The card warns
before a length change re-fingerprints the library.

Off stops every decode: the scan takes only the stream hash (a demux, and
what recognises a moved file) and stores nothing, dropping a changed file's
stale row; the backfill idles. A save also makes a sweep due, since a new
threshold or length changes what the same prints group into, and the sweep
interval gains slack so an hourly interval on an hourly tick doesn't skip
every other tick.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-11 17:53:55 -04:00
bvandeusenandClaude Opus 5 c8bf9dc929 refactor(library): move detection matches on the audio hash, not size and duration (M400 #3914)
test-go / test (push) Successful in 1m9s
test-go / integration (push) Successful in 3m45s
release / Build signed APK (releases and dev) (push) Successful in 4m52s
release / Build + push container image (push) Successful in 14s
release / Verify release artifacts (tag releases only) (push) Skipped
A file that comes back renamed or moved keeps its track row, and with
it its likes and play history, by being matched to the missing row it
replaces (#2528). Untagged files were matched on (file_size,
duration_ms), which was never a fingerprint. It could pair two
unrelated files that happened to share a byte count and a duration,
and it missed a file retagged in place, whose size changes. The only
defence was requiring a unique match and otherwise giving up.

Now there is a real identity. FindMissingTrackByAudioHash matches a
missing track by the SHA-256 of its encoded audio (track_fingerprints,
#3906). That survives a rename, a move and a retag, and only an
identical recording can match it. adoptMovedTrack takes the new file's
hash, which the scan already computes before adoption. The size and
duration query and fallback are removed outright, with no second path
(rule 22).

Unchanged:
- MBID first: it identifies the recording and survives a re-encode
  that even the hash does not
- a unique match is still required
- an absent hash is never looked up, so unhashable files cannot pair
  with each other

The test fake answers the hash lookup only for the hash it holds, so
the tests can tell adoption by identity apart from adoption by
coincidence. That includes the case the old pair got wrong: different
audio of equal size and duration is not adopted.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-11 17:31:42 -04:00
bvandeusenandClaude Opus 5 11ef044ef6 feat(library): merge duplicates without losing history (M400 #3911)
test-web / test (push) Successful in 57s
test-go / test (push) Successful in 1m16s
test-go / integration (push) Successful in 3m39s
release / Build signed APK (releases and dev) (push) Successful in 4m46s
release / Build + push container image (push) Successful in 26s
release / Verify release artifacts (tag releases only) (push) Skipped
Merge keeps one copy of a duplicate group and removes the rest. Every
table that references tracks does so ON DELETE CASCADE, so deleting a
duplicate's row outright would silently destroy its likes, plays,
playlist entries and tags. The merge moves all of that onto the kept
copy first, then deletes the empty row.

In one transaction, holding a lock on the group:
- repoints play_events, skip_events, contextual_likes, playback_errors,
  lidarr_requests.matched_track_id and playlist_tracks. The last is
  keyed by position, so every entry stays where it was.
- merges general_likes one per user, dated to the earlier like
- takes the union of track_tags, keeping the kept copy's own weight on
  a shared tag
- rewrites track_similarity onto the kept copy, dropping edges that
  would point a track at itself and keeping the kept copy's existing
  edge on a collision
- lets the kept copy take a recording MBID only the removed copy had
- deletes the removed copies' rows, tidies emptied albums and artists,
  marks the group merged
- logs sync changes: track deletes, and like and playlist-track
  delete/upsert pairs

The removed copies' files are deleted first, before any row changes,
through the same helper as DeleteTrackFile (now shared, along with the
album tidy-up). A merge that left the file behind would be undone by
the next scan re-importing it. An unwritable library answers 409
library_not_writable and nothing changes.

tracks.Service.MergeDuplicates wraps it with the opt-in Lidarr unmonitor
from RemoveTrack, skipped when the removed copy is a second file of the
kept copy's own album track: unmonitoring that would stop Lidarr
managing the kept file. It writes a duplicate_merge audit row after
commit, per the audit package's best-effort contract, naming both
paths.

POST /api/admin/library/duplicates/{id}/merge takes an optional
survivor_track_id (the report's proposal otherwise) and unmonitor.

On the report page:
- each copy gets a Keep choice, defaulting to the proposed one
- Merge needs a second click, on a button that says how many files it
  removes, with the consequence stated beside an opt-in Lidarr checkbox

Integration tests cover:
- every piece of history landing on the kept copy exactly: likes
  deduped at the earlier time, plays and skips counted, playlist
  position unchanged, tags unioned, similarity rewritten with no
  duplicate or self-edge, MBID inherited
- the removed file gone, and a second merge refused
- an unwritable file leaving likes, plays, row and group untouched
- a survivor outside the group refused

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-11 17:25:06 -04:00
bvandeusenandClaude Opus 5 ff493a8c7d feat(admin): the duplicates report — review proposed duplicate groups (M400 #3912)
test-web / test (push) Successful in 52s
test-go / test (push) Successful in 1m9s
test-go / integration (push) Successful in 3m31s
release / Build signed APK (releases and dev) (push) Successful in 4m32s
release / Build + push container image (push) Successful in 24s
release / Verify release artifacts (tag releases only) (push) Skipped
A new admin tab, Duplicates, beside Missing files: the proposals from
the duplicate sweep, with a Sweep now trigger and a Not duplicates
dismissal. Nothing on it merges or deletes; the merge is #3911.

Each group shows:
- whether it is identical audio or the same recording, with a match
  percentage from the weakest link between members
- every copy's format, size, duration, path, and the likes and plays
  it carries (every user's; this is admin-only, and it is what decides
  which copy to keep)
- the copy proposed to keep, and the rule that chose it

The survivor rule is library.ProposeSurvivor, a pure function the
merge will reuse: lossless over lossy, then the larger file, then the
copy in the library longest, then lowest id. Bitrate is not in it
because the scanner never fills tracks.bitrate, and for one recording
at one duration a larger file is the higher bitrate. m4a is not counted
as lossless: it may be AAC. The reason names the rule that separated
first place from second, not every rule the winner passed.

An empty report has three causes, and the page says which: still
fingerprinting, the sweep has never run, or it ran and found nothing.
The sweep's state and the backfill's progress come back with the groups
for that reason. Groups left with fewer than two members since the
sweep are not shown.

GET /api/admin/library/duplicates, POST .../sweep (202, or 409
sweep_in_progress), POST .../{id}/dismiss (404
duplicate_group_not_pending when already resolved).

Migration 0060 indexes play_events by track_id. Its only indexes led
with user_id, so each copy's play count, and the merge's repointing of
play history, would scan the whole table.

Web only, like Missing files: Android has no library-health admin
screens.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-11 17:11:11 -04:00
bvandeusenandClaude Opus 5 6379b6c31d feat(library): the duplicate sweep — propose duplicate groups from fingerprints (M400 #3910)
test-go / test (push) Successful in 1m0s
test-go / integration (push) Successful in 3m17s
release / Build signed APK (releases and dev) (push) Successful in 4m51s
release / Build + push container image (push) Successful in 14s
release / Verify release artifacts (tag releases only) (push) Skipped
Reads fingerprints, runs them through the matcher, and records
proposals in duplicate_groups (migration 0059). Nothing is merged or
deleted: a group is a proposal for the admin report (#3912).

Streaming. The whole library's fingerprints are hundreds of megabytes,
but tracks are only compared within 3s of each other in duration. So
candidates stream in (duration_ms, id) order, keyset-paged on a new
tracks(duration_ms, id) index. The grouper holds only the tracks within
3s of the oldest one not yet settled. A seed is settled once a track
arrives beyond its window, which gives the same result as grouping the
whole sorted list. groupDuplicates is rebuilt on the same streamGrouper,
so there is one grouping rule and the #3909 tests still cover it. Each
fingerprint's alignment index and variety check are computed once
instead of for every pair.

Exact duplicates are grouped library-wide in SQL. The first member the
stream meets stands in for the whole group in the acoustic pass. An
exact group caught in an oversize acoustic cluster is still proposed:
the acoustic evidence is discarded, identical bytes are not.

Re-sweeping:
- a group is identified by its sorted member ids, so finding it again
  refreshes the row in place
- a proposal whose members all sat in one dismissed group is not
  proposed again (a subset repeats the verdict; a superset is new
  evidence)
- a pending proposal no sweep has found again is retired, but only
  after a complete sweep, and only if an earlier sweep last confirmed
  it, so two overlapping sweeps cannot delete each other's findings
- dismissals are kept

DuplicateSweepWorker checks hourly and sweeps only when a fingerprint
was written after the last sweep started. TryStartDuplicateSweep guards
against two sweeps at once and reaps one stuck in flight for 2h. The
sweep row is closed on a detached context with a deadline, so a sweep
cancelled at shutdown still records that it ended.

The integration test pages one row at a time and checks:
- an acoustic pair and an exact pair are found
- a track with no fingerprint, a missing track and a near-duration
  unrelated song are left out
- a dismissed group is suppressed while the pending one refreshes
  without duplicating
- a proposal that stops holding is retired and the dismissal survives

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-11 17:00:07 -04:00
bvandeusenandClaude Opus 5 c06af48cd6 feat(library): the duplicate matcher — a pure comparison over fingerprints (M400 #3909)
test-go / test (push) Successful in 1m2s
test-go / integration (push) Successful in 3m24s
release / Build signed APK (releases and dev) (push) Successful in 5m8s
release / Build + push container image (push) Successful in 1m15s
release / Verify release artifacts (tag releases only) (push) Skipped
Decides whether tracks are proposed as one recording. No database, no
files, so every rule is falsifiable in a unit test.

Two tiers:
- exact: equal audio_stream_sha256 (identical encoded audio bytes). No
  threshold and no false positives.
- acoustic: chromaprint fingerprints that agree once aligned. Two
  fingerprints can start at slightly different points in the audio
  (padding trimmed differently), so offsets within ±120 items (~15s)
  are voted on using items that share their high 14 bits. Bit-error
  rate is then measured over the overlap at the winning offset. The
  approach and both constants follow AcoustID's pg_acoustid; it was
  reimplemented from that description and no code was copied.

No verdict below ~10s of overlap, or for low-information fingerprints
(silence, a sustained tone). Two such tracks agree without being one
recording.

Grouping uses complete linkage: a track joins a group only if it
matches every member. Otherwise A close to B and B close to C would
merge A and C, which are not close, and it means any member can be the
survivor. Other rules:
- durations must be within 3s
- acoustic groups are capped at 8, and larger clusters are reported
  and discarded as a likely shared jingle
- an exact group absorbed into an acoustic one takes the acoustic tier
- output does not depend on input order

The acoustic threshold is 0.15 bit-error rate: deliberately
conservative, since the operator's concern is different recordings of
one song being merged, and an instrumental shares its vocal's harmony.
It is unmeasured, and needs calibrating against real pairs once the
backfill has populated fingerprints (#3913 exposes it).

Nothing calls this yet; the sweep (#3910) does.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-11 16:48:06 -04:00
bvandeusenandClaude Opus 5 21c698a616 test(web): give the admin page mock the fingerprint coverage query
test-web / test (push) Successful in 35s
release / Build signed APK (releases and dev) (push) Successful in 4m25s
release / Build + push container image (push) Successful in 34s
release / Verify release artifacts (tag releases only) (push) Skipped
b8855b48 made the admin overview page create a fingerprint coverage
query, but admin.test.ts mocks $lib/api/admin with an explicit factory
that only returned the cover coverage query. Every test that renders
the page threw on the missing export (run 6512, 11 failures). The mock
now returns it in the same empty-store shape, so the gauge stays hidden
in these tests the way the cover gauge does.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-11 15:00:38 -04:00
bvandeusenandClaude Opus 5 b8855b480f feat(library): backfill fingerprints for the existing library — M400 #3908
test-web / test (push) Failing after 50s
test-go / test (push) Successful in 1m7s
test-go / integration (push) Successful in 3m27s
release / Build + push container image (push) Canceled after 0s
release / Verify release artifacts (tag releases only) (push) Canceled after 0s
release / Build signed APK (releases and dev) (push) Canceled after 4m23s
The scan fingerprints only bytes it has not seen, so everything imported
before fingerprinting existed, and any row derived by an older
fingerprintVersion, needs a pass of its own.

That pass is a background worker, not a stage in RunScan. RunScan runs
at boot and then every 12h, and an in-flight scan older than an hour is
reaped and a second started beside it. A stage would have to stop inside
the hour: a few hundred decodes a run, so about a month for a 50k-track
library. It would also hold the run in flight and answer manual
rescans with 409 while it worked.

FingerprintBackfillWorker runs once at start, then hourly. Nothing a
pass does (error or panic) can stop the next tick. A pass walks tracks
with no fingerprint or a stale version, skipping missing tracks,
keyset-paged on id. The cursor is what lets a pass end: an inconclusive
attempt writes no row, so a file that keeps timing out would otherwise
be re-listed and retried forever. Two decodes at a time, deliberately:
they compete with transcoding for CPU and with streaming for the mount.

storeFingerprint is now one package function shared by the scan and
the worker, and reports whether the attempt was fingerprinted,
rejected, inconclusive or failed to store.

Progress is a live gauge on the Admin scan card, served by
GET /api/admin/library/fingerprints: fingerprinted / rejected / pending
of total, with missing tracks excluded so it can reach the end.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-11 14:56:18 -04:00
bvandeusenandClaude Opus 5 71d4335584 docs(readme): the music mount is writable — Minstrel deletes when asked
test-go / test (push) Successful in 1m8s
test-go / integration (push) Successful in 4m10s
release / Build signed APK (releases and dev) (push) Successful in 5m23s
release / Build + push container image (push) Successful in 1m16s
release / Verify release artifacts (tag releases only) (push) Skipped
The quickstart mounted the library :ro and promised "Minstrel never
writes to your library". That stopped being true long before #3918:
quarantine's Delete file removes files, and under :ro it failed. The
operator has accepted delete ownership (Scribe note #3926).

The quickstart now mounts it writable and says exactly what Minstrel
writes: it deletes a file when an admin asks, and never moves, renames
or retags. It notes that uid 1000 needs write access, and that :ro
still works, with deletes refusing and explaining why.

Reorganising and tag writes stay out, pending whether Minstrel absorbs
Lidarr's role.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-11 14:34:28 -04:00
bvandeusenandClaude Opus 5 702b48ce36 fix(lidarrquarantine): pass dataDir at the four stub-client test constructors
d7a8e5f3 added a dataDir parameter to NewService and updated the 13
call sites spelled NewService(pool, lidarrconfig.New(pool), nil). Four
more build their client from a Lidarr stub, NewService(pool, cfg,
clientFn), and were missed, so the package's tests did not compile and
run 6495 failed both go vet and the integration build.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-11 14:34:28 -04:00
bvandeusenandClaude Opus 5 d7a8e5f300 fix(library): a track delete that cannot remove its file deletes nothing — #3918
test-go / test (push) Failing after 55s
test-web / test (push) Successful in 56s
test-go / integration (push) Failing after 4m50s
android / Build + lint + test (push) Successful in 5m52s
release / Build signed APK (releases and dev) (push) Successful in 6m5s
release / Build + push container image (push) Successful in 1m14s
release / Verify release artifacts (tag releases only) (push) Skipped
Two delete paths had opposite failure policies. tracks.RemoveTrack
logged a failed os.Remove and deleted the row anyway, which CASCADEs
likes, plays, playlist memberships and tags, while the file survived
for the next scan to re-import as a stranger. library.DeleteTrackFile
stopped correctly but reported it as a bare 500 nobody could read.

One path now: library.DeleteTrackFile removes the file first and, on
anything but ErrNotExist, returns *FileRemoveError with nothing
deleted. Only then does it delete the row and tidy an emptied album
and artist in one transaction, log the sync change and clear orphaned
artist art. RemoveTrack calls it, which also fixes RemoveTrack never
logging a sync change. Quarantine Delete file now tidies emptied
albums and artists too.

Both endpoints answer an unwritable library (EROFS, EACCES, EPERM) with
409 library_not_writable. The message names the directory (removal
writes to the parent), the uid:gid the server runs as, and that
nothing was deleted. Other remove errors are 500 file_delete_failed
with the path.

The reachable surface is quarantine Delete file, which failed
silently: no copy for the code on either client, and Android swallowed
the exception so the row just reappeared. Web and Android now have
copy for both codes and append the server message for exactly those
two. Android's quarantine screen shows it in a snackbar.

DELETE /api/admin/tracks/{id} has had no client since f7278f24, which
kept it on purpose for a safer admin surface, so its history loss was
latent. Fixed rather than removed.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-11 14:23:01 -04:00
bvandeusenandClaude Opus 5 cba77a5187 feat(library): fingerprint every new or changed file — M400 #3905-#3907
test-go / test (push) Successful in 1m9s
test-go / integration (push) Successful in 3m28s
release / Build signed APK (releases and dev) (push) Successful in 4m38s
release / Build + push container image (push) Successful in 1m26s
release / Verify release artifacts (tag releases only) (push) Skipped
Two identities per track, because they answer different questions:

- audio_stream_sha256: SHA-256 of the ENCODED audio packets
  (ffmpeg -map 0:a -c:a copy -f hash). Equal means identical audio
  whatever the tags say. Measured against the #3885 pair: the two WWW
  files hash identically here and differently as whole files. Packets
  rather than decoded samples, so an ffmpeg upgrade cannot silently
  change every stored hash, and nothing is decoded.
- chromaprint: fpcalc -raw -signed. The same recording at another
  bitrate or codec, for the acoustic tier.

fpcalc ships in the image (libchromaprint-tools); shelled out because
CGO_ENABLED=0 rules out bindings.

Stored in a track_fingerprints table rather than on tracks: eight
queries read tracks with SELECT *, including album pages, search and
the Subsonic surface, and a ~4 KB array there would be de-TOASTed on
every one of them.

The scan fingerprints only bytes it has not seen (a new path, or mtime
past the row's). A tag-repair pass leaves fingerprints alone, and
unchanged files with no fingerprint are the backfill's job (#3908).
Folding that into the skip check would re-decode the whole library on
the first scan after upgrade and push a sync change per track.

A failure that says nothing about the file (timeout, cancelled scan,
tool not installed) is never stored, and on changed bytes it removes
the old row. A tool that rejects the file stores NULL at the current
version, so the backfill does not retry it every boot.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-11 13:21:15 -04:00
bvandeusenandClaude Opus 5 eff3d88931 fix(recommendation): make the candidate draw reproducible, not accidentally so
test-go / test (push) Successful in 1m18s
test-go / integration (push) Successful in 4m52s
release / Build signed APK (releases and dev) (push) Successful in 6m9s
release / Build + push container image (push) Successful in 2m5s
release / Verify release artifacts (tag releases only) (push) Skipped
Four arms of the candidate query ended in a bare `ORDER BY random()` with no
seed: similar_artists, likes_overlap, coplay_artists and random_fill.

Such an arm returns a STABLE set only while its LIMIT exceeds the rows
eligible for it — at that point it returns all of them and the order stops
mattering, because scoreAndSortCandidates sorts by track id before drawing
jitter. Below that threshold it returns a random SUBSET, and two builds on
the same day draw different ones.

So daily determinism held BY ACCIDENT, and only for libraries smaller than
the limits. Any real library is larger, which means same-day rebuilds have
been producing different mixes since those arms were written — invisible,
because a mix that changes after a refresh looks like a feature rather than
a broken promise.

Found by breaking it: cutting RandomFill to 10 while tuning Songs-like
turned TestBuildSystemPlaylists_DailyNonceDeterminism red. That test seeds
~20 tracks against a default RandomFill of 30, so its determinism came from
the limit exceeding the library, not from the code being right. It is now a
real guard.

The arms order by md5(id || $12) instead. The CALLER decides what that
means, which is the point: system mixes pass a per-(user, day) seed and get
the determinism they promise, radio passes a fresh value per request and
keeps varying, which is what a radio should do. Same shape the browse
queries in this file already use (`md5(id::text || current_date::text)`) —
existing idiom, not a new one.

This also unblocks the trim that #3881 wanted and could not have. Shrinking
a randomly-ordered arm was what broke membership; a seeded one takes a
smaller but REPRODUCIBLE slice. Songs-like's seed-independent share drops
from 29% to 12%, which was the original intent before determinism forced it
back to 20%.

TestSongsLikeLimits_DoNotShrinkTheUnseededRandomArms is DELETED rather than
kept passing. It existed to stop anyone trimming those arms while the
ordering was broken; the ordering is fixed, so the constraint is gone and a
guard enforcing it would now forbid correct code.

Was filed as blocked on tooling. It was not: `make generate-go` runs sqlc as
a pinned Go tool and is the same path CI takes.

One thing worth knowing for next time: three files in internal/db/dbq are
owned by root, left by `make generate` running sqlc in Docker. sqlc errored
on the first it could not write. They are untouched by this change and the
regeneration of recommendation.sql.go completed, but `make generate` will
keep failing until they are chowned.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-11 08:44:33 -04:00
bvandeusenandClaude Opus 5 4ce47397a9 fix(recommendation): a nil LibrarySize must degrade, not panic
test-go / test (push) Successful in 1m4s
test-go / integration (push) Successful in 3m41s
release / Build signed APK (releases and dev) (push) Successful in 4m37s
release / Build + push container image (push) Successful in 1m58s
release / Verify release artifacts (tag releases only) (push) Skipped
Fixes the integration failure from 72115484: a SIGSEGV inside handleRadio
took down TestHandleRadio_ColdStart_OnlySeedReturned.

    recommendation.(*LibrarySize).Get(0x0, ...)
      library_scale.go:146
    api.(*handlers).handleRadio(...)
      radio.go:95

internal/api builds its handlers struct directly in a dozen tests, none of
which know about every field, so librarySize arrives nil there. Get took
l.mu.Lock() straight off the nil receiver.

The shape of the bug is what matters more than the nil check. This value's
entire contract is that it degrades — an errored count keeps the last known
number, a never-counted cache returns 0, and 0 scales to the base limits,
i.e. today's behaviour. A pool-sizing HINT then turned a request into a
crash, which is the precise opposite of that.

A nil receiver is now VALID and means "no cache": the count still runs, it
is just not memoised. Correct-but-uncached rather than zero, so a wiring
miss in production would cost a query per request, not silently unscale
every pool — a performance bug is findable, a quietly-wrong pool is not.

Patching the test constructors was the alternative and is worse: a dozen
call sites, and the next test to build a handlers literal reintroduces it.

Guarded with the nil path exercised directly, including that it counts
again rather than memoising, and still returns 0 on a failed count. The old
shape fails it by panicking.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-10 23:29:35 -04:00
bvandeusenandClaude Opus 5 721154847e fix(recommendation): size the candidate pool to the library
test-go / test (push) Successful in 1m5s
test-go / integration (push) Failing after 3m39s
release / Build signed APK (releases and dev) (push) Successful in 4m55s
release / Verify release artifacts (tag releases only) (push) Canceled after 0s
release / Build + push container image (push) Canceled after 1m38s
Operator, 2026-09-10: "is the pool that we draw from somehow scaled to the
amount of music in the library... my earlier understanding of the tuning and
work may have been skewed by what was in my library."

It was not. DefaultCandidateSourceLimits returns what its own comment calls
"the v1 hardcoded constants per spec" — ~170 candidates for a 500-track
library and a 100,000-track one alike. The pool therefore samples a
shrinking FRACTION of a growing collection: 17% of 1,000 tracks, 1.7% of
10,000, 0.17% of 100,000. RandomFill, whose whole job is exploration,
becomes a thinner and noisier slice at exactly the moment a library gets
more diverse — which is the "starting to feel weird" being reported.

    1,000 tracks -> pool 170     (unchanged)
    5,000        -> pool 170     (unchanged)
   20,000        -> pool 280
   80,000        -> pool 500     (ceiling)

THE SCALING IS PER-ARM, and that is the substance rather than a refinement.
A limit only matters if there are rows for it to cut off, so what an arm is
BOUNDED BY decides whether library size can help it. LBSimilar,
SimilarArtist, TagOverlap and RandomFill grow: they are bounded by
similarity/tag data and by the library itself. LikesOverlap, UserCoplay and
TasteOverlap do not: they are bounded by the user's likes, the instance's
co-play graph and the taste profile, none of which grow when the library
does. Raising those would sample more of a set that did not change — churn,
not reach. It also keeps this from inflating the sim_score-0 share, since
TasteOverlap is one of the two zero-similarity arms.

sqrt, not linear: linear would put a 100,000-track library at a
3,400-candidate pool, long past where more candidates improve the answer.
A 4x ceiling bounds it at ~500.

Never shrinks an arm. The base limits are a floor, and #3889 makes that
load-bearing rather than tidy — shrinking an arm ordered by unseeded
random() changes pool membership between same-day rebuilds.

Library size comes from a TTL-cached count reusing CountTracksMatching with
an empty pattern (rule 28 — a new query would need sqlc regeneration, which
is blocked). The ILIKE defeats every index, so it is a full scan and must
not run per request. It degrades rather than fails: an error keeps the last
known value, a never-counted cache returns 0, and 0 scales to the base
limits — today's behaviour exactly. Nothing about sizing a pool justifies
failing the request it is sizing. Bounded by a 3s deadline (rule 156), and
a failed refresh does not stamp the clock, so a blip cannot pin a stale
value for the whole TTL.

THE REFERENCE IS ASSUMED, NOT MEASURED. libraryScaleReference = 5000 is
where growth starts, and the size the v1 constants were really tuned against
is unrecorded. #3879 should replace it; until then that constant is the one
thing to change. Deliberately conservative: below it nothing scales at all,
so no existing install changes behaviour.

Falsification caught a weak guard: the sqrt-vs-linear assertion was written
at SIXTEEN times the reference, where linear has already been clamped by the
ceiling and both curves land on 4x. It proved nothing. Moved to four times
the reference, below the ceiling for both, where sqrt gives 2x and linear
would give 4x.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-10 23:23:04 -04:00
bvandeusenandClaude Opus 5 633d4f591f fix(radio): cap any one artist's share of a radio session
test-go / test (push) Successful in 1m13s
test-go / integration (push) Successful in 3m53s
release / Build signed APK (releases and dev) (push) Successful in 5m9s
release / Build + push container image (push) Successful in 1m52s
release / Verify release artifacts (tag releases only) (push) Skipped
Operator, 2026-09-10: started radio from a song and "literally all of the
songs in the playlist after that were from a single artist which was not
expected."

There was no per-artist cap anywhere in the radio path. radio.go built the
pool and handed it straight to Shuffle, which scores, sorts and takes the
top N — nothing between those steps bounded any artist's share, so a pool
dominated by one artist produced an output dominated by it. The asymmetry
was the tell: discover.go, you_might_like.go and home.go all cap; radio
never got one.

With the fixture that reproduces it — 20 liked tracks by one artist plus 10
by ten others — the old path returns 10 tracks from 1 artist. It now returns
10 from 8.

TWO PASSES, and that is the whole design. A hard cap was the easy mistake:
radio asks for 50 tracks by default and 200 at most, so capping at three per
artist over a concentrated pool would hand back a six-track "radio". Pass
one takes candidates that fit under the caps; pass two fills any remaining
slots from those it skipped, still in score order. The result always holds
min(limit, len(candidates)) — the caps change WHICH tracks are picked, never
HOW MANY. Rule 131's principle past the system mixes it was written for.

The caps SCALE with the requested length rather than being a constant.
Three-per-artist is a sensible 12% of a 25-track mix and an absurd 1.5% of a
200-track radio, where every selection would sit in the relaxation path and
the cap would be decorative. RadioDiversityCaps holds the system mixes'
proportion at any length: 3/2 at 25, 6/4 at 50, 24/16 at 200, with floors so
a very short radio is not capped down to one track per artist.

A BOUND, NOT AN EXCLUSION — the operator asked for the opposite of removal:
"again it should be able to add songs from the same artist." The dominant
artist still appears, just not exclusively. Guarded, because the tempting
wrong fix is the filter songs-like used to carry.

Shuffle grew the parameter rather than gaining a capped twin: radio is its
only production caller, so a second function would have left the original
dead (rule 22).

Falsified against each named regression: uncapped gives 10/10 to one artist;
a hard cap returns 3 of 10 on a single-artist pool; a cap-as-exclusion drops
the artist entirely; a fixed cap stays 3 where the scaled one reaches 24.

Caught while writing the guards: the artist-key constant was hand-written
hex and wrong — the fixture's artist UUID carries 0001 in its fourth group,
so the lookup missed and the assertion measured nothing. Derived from the
same construction the fixture uses now.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-10 22:14:42 -04:00
bvandeusenandClaude Opus 5 f5dd4462de test(playlists): the same-artist guard needed a fixture that has same artists
test-go / test (push) Successful in 1m25s
test-go / integration (push) Successful in 4m46s
release / Build signed APK (releases and dev) (push) Successful in 5m45s
release / Build + push container image (push) Successful in 17s
release / Verify release artifacts (tag releases only) (push) Skipped
Fixes the integration failure from 31190657. The test was wrong, not the
code: it could not have passed whatever produceSeedMixes did.

seedActiveLibrary builds its tracks through seedTrack, whose own comment
says "artist and album are not deduplicated across calls (mbid-less
upsert)". So every track gets a fresh artist row despite sharing a name —
4 artists x 5 tracks is really 20 artists with one track each. A seed
artist's only track IS the seed, which is excluded from its own mix, so
"does this mix contain a track by its seed artist" was structurally
answerable only as no.

That is the failure mode worth naming: the assertion was measuring the
fixture, not the behaviour, and it reported the behaviour as broken.

seedSharedArtistLibrary upserts each artist ONCE and reuses the id across
its tracks, so a seed artist genuinely owns five others. Albums are still
not deduplicated, which suits this test — the per-album cap never binds, so
the per-artist cap (3) is unambiguously what is under test. Noted in the
fixture, because "tidying" the album titles into something shared would
silently change which cap the assertion measures.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-10 21:25:49 -04:00
bvandeusenandClaude Opus 5 31190657d8 feat(recommendation): Songs-like can include the seed artist's own music
test-go / test (push) Successful in 1m16s
test-go / integration (push) Failing after 3m55s
release / Build signed APK (releases and dev) (push) Successful in 5m11s
release / Build + push container image (push) Successful in 16s
release / Verify release artifacts (tag releases only) (push) Skipped
Operator, 2026-09-10: "it should also be able to include music from the same
artist." Completes #3881 — the weights and pool landed in f367eeaa; this is
the eligibility half.

produceSeedMixes filtered the seed artist out entirely:

    // "Songs like X" excludes X's own songs.
    if !pgtypeUUIDEqual(c.Track.ArtistID, artistID) { ... }

That reads as obviously right and is not. The seed is a TRACK — the artist's
top-played one — and the tracks most likely to sound like it are usually the
rest of that artist's catalogue. The filter threw away the seed's nearest
neighbours, then reached FURTHER OUT to replace them. On the one surface
whose job is staying in a neighbourhood, that is backwards, and it worked
against the coherence tuning rather than with it.

Domination is bounded by the cap instead of by exclusion, which is the
distinction that makes this safe rather than a new problem:
capCandidatesByAlbumAndArtist already allows at most 3 tracks per artist in
a 25-track mix, so the seed artist gets 12% at most — a presence, not a
takeover. Without that bound this would just be the radio failure (#3882)
arriving on a different surface. The seed track itself still cannot appear;
it is passed to LoadCandidatesFromSimilarity as an exclusion.

Guarded end-to-end rather than by reading the source, for two reasons: the
check has to survive the filter returning in a different shape, and an
absence check would now match the comment that explains why the filter is
gone — rule 167's prose trap exactly. The test asserts both directions, that
at least one mix contains its seed artist and that none exceeds the cap.

Its falsification is by construction rather than by execution: under the
previous code every mix's own-artist count was necessarily zero, so the
assertion could not have passed. Running it needs Postgres, which is the
integration lane's job.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-10 21:13:32 -04:00
bvandeusenandClaude Opus 5 ecfa056d4d fix(recommendation): don't shrink a candidate arm ordered by unseeded random()
test-go / test (push) Successful in 1m10s
test-go / integration (push) Successful in 3m35s
release / Build signed APK (releases and dev) (push) Successful in 5m6s
release / Build + push container image (push) Successful in 16s
release / Verify release artifacts (tag releases only) (push) Skipped
Fixes the integration failure from f367eeaa:
"same-day rebuild produced different track lists".

Cutting RandomFill 30→10 for Songs-like broke
TestBuildSystemPlaylists_DailyNonceDeterminism, and the reason is worth
stating because the number is not the bug.

`likes_overlap` and `random_fill` end in a bare `ORDER BY random()` with no
daily seed (recommendation.sql:118, :161). Such an arm returns a STABLE set
only while its LIMIT exceeds the rows eligible for it — then it returns all
of them, and the random order stops mattering because scoreAndSortCandidates
sorts by track id before drawing jitter. Below that threshold the arm
returns a random SUBSET, and two builds on the same day draw different ones.

So the test was green by accident. It seeds ~20 tracks against a default
RandomFill of 30; the limit exceeded the library, so the arm returned
everything. Determinism held for a reason unrelated to the code being right.

Which means it does NOT hold in production. Any real library is larger than
30, so same-day rebuilds have been drawing different mixes since that arm
was written — invisible, because a mix changing after a refresh looks like a
feature. Filed as #3889; the fix is a seeded ordering per (user, day), which
needs a .sql change and sqlc regeneration and so cannot land from here.

The correction: grow an arm freely, never shrink one whose ordering is
unseeded random. LikesOverlap and RandomFill go back to the defaults;
TasteOverlap stays halved because it sorts by `tpa.weight DESC, t.id` and is
genuinely deterministic. Guarded by a test that names the reasoning, so the
next person to trim these has to read why first — and it should be DELETED
once #3889 lands rather than worked around.

The cost is honest: the seed-independent share of the Songs-like pool falls
from 29% to 20% instead of the intended cut. That matters less than it
sounds. The pool only biases the draw; the songs_like WEIGHTS are what
actually demote sim_score-0 candidates, and they are untouched here — a
perfect match still scores 5.00 against an unrelated favourite's 2.00.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-10 20:59:39 -04:00
bvandeusenandClaude Opus 5 f367eeaa9d fix(recommendation): Songs-like gets its own profile so it stops wandering
test-web / test (push) Successful in 1m7s
test-go / test (push) Successful in 1m31s
test-go / integration (push) Failing after 4m21s
release / Build signed APK (releases and dev) (push) Successful in 5m11s
release / Build + push container image (push) Successful in 1m52s
release / Verify release artifacts (tag releases only) (push) Skipped
Operator, 2026-09-10: "when I play it I'm expecting to get a consistent
sound and style from the experience... I was getting a seeming wide variety
of music from each one when I was hoping to stay in a certain neighborhood."

Songs-like shared the `daily_mix` weight profile with For-You, and that
sharing WAS the bug. The two surfaces want opposite things: For-You answers
"what will they enjoy today" and is supposed to roam; Songs-like answers
"what sounds like THIS". Under one profile the broad answer wins.

The arithmetic, from the shared weights:

    unrelated track, liked, not played recently → 1.0 + 2.0 + 1.0 = 4.0
    PERFECT similarity match, not liked         → 1.0 + 1.5       = 2.5

Liking something outranked sounding like the seed, because LikeBoost (2.0)
exceeded SimilarityWeight's whole range (1.5) and TasteWeight (1.5, and
seed-INDEPENDENT) matched it outright. Under the new profile the same pair
scores 5.00 vs 2.00.

Two levers, because either alone leaves the other's failure intact:

POOL. Songs-like now takes its own CandidateSourceLimits. The default gave
~29% of candidates a sim_score of literally zero — `taste_overlap` and
`random_fill` are both `0.0::float8` in recommendation.sql, seed-independent
by construction. Same total pool size; composition shifts to arms that
measure distance from the seed, LBSimilar doubled.

WEIGHTS. A third profile beside radio and daily_mix, DB-backed and live per
rule 25, with the property that similarity's range exceeds the combined
range of every seed-independent differentiator — so a closer match cannot
be beaten on likes, freshness and taste alone, while tracks within ~0.39
similarity of each other still get ordered by what the user likes.

Rule 131 changed the pool design mid-way and for the better. Zeroing the
two seed-independent arms was the first instinct and is exactly the
vanish-or-nothing shape that rule forbids: a seed with thin ListenBrainz
coverage would yield a short mix or none. They are the tier-3 FLOOR — cut
hard, never removed — and the weights keep them at the bottom of the
ranking rather than out of the pool. "A few tracks further from the seed
than we'd like" beats "no playlist".

Caught while wiring it: switching only pickTopN's final Score would have
been nearly INERT. scoreAndSortCandidates does the selection sort, and the
caller caps and truncates in that order — so the playlist would still have
been chosen by daily_mix and merely relabelled with songs_like numbers. It
now takes the profile as a parameter, and each surface passes its own.

Also corrects the daily_mix card's blurb, which claimed Songs-like as one
of its surfaces and no longer is.

Guards pin behaviour rather than the numbers, since numbers get retuned:
that similarity beats an unrelated liked track, that daily_mix still
DOESN'T (or the split buys nothing), that the tier-3 floor is non-zero,
and that the UI card shows its own values rather than falling back. Each
falsified against its named regression first.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-10 20:51:15 -04:00
bvandeusenandClaude Opus 5 270ad7a71b fix(ci): tests do not ship, so they must not re-version an artifact
test-go / test (push) Successful in 1m1s
test-go / integration (push) Successful in 3m17s
release / Build signed APK (releases and dev) (push) Successful in 4m41s
release / Build + push container image (push) Successful in 16s
release / Verify release artifacts (tag releases only) (push) Skipped
Completes the pathspec. 17212e9e excluded CI, docs and tooling but left
tests in the shipped set, so its own commit re-versioned the image on the
strength of a _test.go file. `go build` drops *_test.go outright and the
Vite build never imports a .test.ts — neither reaches an image or an APK.

Globs over files rather than a directory exclusion, because this repo has
no tests/ tree to exclude: Go tests sit inline beside the code they cover
(158 files) and the web suite beside its modules (115). Patterns match what
exists and nothing speculative — there are no .spec.* files, no __tests__/
directories and no androidTest/ tree. If any appear they re-version until
named, which is the harmless direction and the point of a denylist.

The guard that matters is not "a test-only commit is inert" — it is that a
commit touching a test AND its source still moves the version. `':!internal'`
would satisfy every inertness assertion while silently excluding the entire
server, which is the stale-version-on-changed-artifact failure this whole
derivation exists to prevent.

Falsified: drop the Go exclusion and a _test.go commit moves the version;
drop the web one and a .test.ts does; replace the globs with `':!internal'`
and the source-alongside-test case breaks.

That last check failed first time, on a bug in the FIXTURE rather than the
derivation, and it is worth recording because it makes a test pass for the
wrong reason. Both commit helpers wrote the constant "x\n", so re-writing a
file with identical bytes recorded NOTHING — the "source and test together"
commit actually contained only the test, and the assertion was quietly
checking the case it was meant to contrast against. Content is now derived
from the commit's epoch, and the test asserts HEAD really contains both
paths before drawing any conclusion from it.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-10 18:21:12 -04:00
bvandeusenandClaude Opus 5 17212e9eb4 fix(ci): version derives from the shipped set; untrack an 18MB binary
test-go / test (push) Successful in 1m5s
test-go / integration (push) Successful in 3m30s
release / Build signed APK (releases and dev) (push) Successful in 5m10s
release / Build + push container image (push) Successful in 1m31s
release / Verify release artifacts (tag releases only) (push) Skipped
Three build-hygiene fixes that turned up while explaining the pathspec.

**version.sh derives from what SHIPPED.** It read bare HEAD, so any commit
moved the version — including one touching only CI or a README. Rules 148
and 149 both specify the pathspec form. Now a denylist, and the direction
is the point: as an allowlist the list must be updated by whoever adds a
directory and nothing fails if they don't, so the failure mode is a changed
artifact keeping its old version silently on a green run. Inverted, new
content counts by default.

android/ is deliberately NOT excluded, and that is the subtle part. This
repo ships TWO artifacts from ONE derivation: android/ is in no server
image, but it is the APK's entire source, and excluding it would stop an
Android-only commit from moving the APK's own version — the silent
downgrade the versioning rework exists to prevent. So the list is the
union: exclude only what ships in neither, and accept that an Android
commit also nudges the server's reported version. Over-inclusion across the
two, which is the harmless direction. roundtable/roundtable-android each
keep tighter lists because they are one-artifact repos; don't copy theirs.

**.dockerignore excluded the wrong CI directory.** It named .forgejo/ and
.github/, neither of which this repo has. Gitea Actions reads .gitea/, so
the one directory that exists was the one not excluded. The "Flutter mobile
client" block had also lost its PATTERN when flutter_client/ was deleted,
leaving a comment describing an exclusion that was not happening — android/
never took its place, so 4.1MB of Gradle project entered the context and
busted the `COPY . .` layer on every Android-only change. bin/ excluded too.

**bin/minstrel was tracked** — an 18MB binary last refreshed by a commit
about web test mocks, and re-dirtied by every `make build` since. Untracked
and ignored; the file stays on disk.

Guards are behavioural rather than textual: they build throwaway repos with
pinned commit timestamps and run version.sh against them, so they break when
the derivation changes rather than when the wording does. Falsified — drop
the .gitea exclusion and the CI-only commit moves the version; add an
android exclusion and an Android commit stops moving it; exclude everything
and a source commit refuses.

One honest note on the refusal test: the script already refused an empty
result via the downstream date check, so the new explicit check improves the
diagnostic ("no commit touches the shipped file set — shallow clone?") and
not the safety. The test pins the property, which is defended in depth.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-10 17:55:36 -04:00
bvandeusenandClaude Opus 5 8f4b76a638 fix(ci): artifacts move to stock upload-artifact@v7 / download-artifact@v8
test-go / test (push) Successful in 1m4s
test-go / integration (push) Successful in 3m54s
android / Build + lint + test (push) Successful in 4m58s
release / Build signed APK (releases and dev) (push) Successful in 5m11s
release / Build + push container image (push) Successful in 1m52s
release / Verify release artifacts (tag releases only) (push) Skipped
android.yml's debug upload and release.yml's minstrel-apk pair went
through the bvandeusen fork mirrors, with comments saying stock actions
refuse this hostname, that the pair had to be matched on the bundled
@actions/artifact major, and that download v7 was off-limits for node24.
None of that holds on gitea/runner 3.x: the runner edits the GHES refusal
out of the action bundles, every download major v4-v8 reads every upload
major v4-v7 (Scribe spike #3843, CI-runner run 6312), and every CI image
carries Node 24. The mirror pair itself was last verified at tag run 6286.

Same artifact names, paths and if-no-files-found. ci-requirements.md
drops the pairing table and keeps what is still true: @v3 is invisible.

Scribe snippet #2271, milestone 395.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DwoKYuw3qJmUUYsJeNherB
2026-09-10 17:25:10 -04:00
bvandeusenandClaude Opus 5 aeb8781c4e fix(release): drop version image tags, mint the rollback unit on main
test-go / test (push) Successful in 1m43s
test-web / test (push) Successful in 1m13s
test-go / integration (push) Successful in 4m12s
release / Build signed APK (releases and dev) (push) Successful in 5m11s
release / Build + push container image (push) Successful in 38s
release / Verify release artifacts (tag releases only) (push) Skipped
The image tag map was the inverse of family rules 145 and 147 on every
count: it published :vYYYY.MM.DD.HHMM that nobody pinned, published :main
that rule 147 says should not exist, and published no commit-addressable
image at all — so the rollback unit the rule names did not exist in this
repo. A bad main push had nothing to roll back to but the previous
release tag, which may be many commits back.

The whole map is now:

  dev  → :dev
  main → :latest + :<sha>
  tag  → :latest

A release refreshes the channel and mints nothing else. The tag build
rebuilds the SAME SOURCE as main's build minutes earlier, differing only
in which APK is baked in, so rule 145's immutability clause applies
directly: move the channel tag, never re-push a commit-addressable one.
:latest has to move here rather than waiting for the next main push, or
the channel would carry the previous release's APK indefinitely — a
channel that cannot refresh itself (rule 146).

Two consequences that are not optional:

The verify job asserted the :<version> image existed. With version tags
gone that would fail every release for a tag nothing mints. Re-pointed at
the :<sha> image rather than deleted — deleting it is the tempting way to
make a failing guard go green, and it earns its keep twice now: it still
catches an image push that silently did not happen, and it additionally
proves the ordering, since a tag cut on a commit whose main build never
completed has no rollback target.

The server's self-reported version was the literal string "main" or
"dev". That was survivable while :vYYYY.MM.DD.HHMM existed to identify a
build; with version tags gone it is the ONLY thing that says which build
is running, and two dev images months apart were indistinguishable. It
now carries the derived name from ci/version.sh on every lane, with the
channel as a sibling field (rule 149) rather than folded into the string.
Surfaced at /healthz and beside the version in Settings.

Guards added for each arm of the policy, and every one was falsified
against the specific regression it names before committing. That caught
two real bugs in the guards themselves: stepBody cut at the next
`- name:`, which returns an EMPTY body for the last step in a job and
made the assertions pass vacuously, and its replacement cut at any blank
line followed by indentation, which truncated a step mid-run-block. The
helper now refuses an empty body outright.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-10 15:10:15 -04:00
bvandeusenandClaude Opus 5 88508b536b fix(release): a non-matching grep must not kill the rebundle step
test-go / test (push) Successful in 1m39s
release / Build signed APK (releases and dev) (push) Successful in 6m11s
release / Build + push container image (push) Successful in 14s
release / Verify release artifacts (tag releases only) (push) Skipped
test-go / integration (push) Successful in 5m59s
The first `main` build after the version rework failed, and the bug was
mine. :latest was never moved — "Build and push" was skipped — so nothing
reached production, but every subsequent main push would have failed the
same way.

The runner invokes `shell: bash` as `bash -e -o pipefail`. Under pipefail a
command substitution reports the FIRST non-zero status in its pipeline, not
the last, so

  VAR="$(printf ... | grep -oP ... | grep -E '\.apk\.version$' | head -1)"

exits non-zero when that grep matches nothing, even though `head` succeeded.
With -e the step dies AT THE ASSIGNMENT — before reaching the `if` written
to handle exactly the empty case.

Which is what happened: v2026.09.09 predates sidecar assets, so its
`.apk.version` grep matched nothing and the step aborted instead of falling
through to the name-only branch I added in 9f3e0b8c for precisely that
release. The transition case was described correctly in that commit message
and then not handled in code.

The other two assignments carried the same latent hazard and had simply
never fired, because a release always has a tag_name and an .apk asset. So
the step's documented promise — "degrades to an empty client/ (404 update
channel) — never a wrong version — if no release or APK asset can be
resolved" — was never actually reachable under pipefail. All three now
carry `|| true`.

Reproduced under the runner's exact shell before fixing: without `|| true`
the script exits 1 with no output at all, proving it never reaches the
branch; with it, the fallback runs and emits the name-only sidecar.

Guarded, since the graceful degradation depends on this and the failure is
invisible until the one release that triggers it: the new test asserts every
command-substitution grep in that step ends with `|| true`, and was
falsified by removing it from the sidecar line.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-10 08:37:23 -04:00
bvandeusenandClaude Opus 5 90bb3538c6 feat(release): build a dev channel so testing stops requiring a release
test-go / test (push) Successful in 1m11s
release / Build signed APK (releases and dev) (push) Successful in 5m0s
test-go / integration (push) Successful in 5m55s
release / Build + push container image (push) Successful in 1m52s
release / Verify release artifacts (tag releases only) (push) Skipped
There was no test channel at all. release.yml ran only on main and tags, so
no :dev image existed and no APK was produced outside a release — the only
way to get a build onto a phone was to ship one, which made `main` the
staging area by default.

A push to dev now builds a signed APK, bundles it, and publishes :dev.

Signed with the SAME key as release builds, deliberately. A differently
signed APK cannot install over the stable app, so anyone moving between
channels would have to uninstall and lose their local data. Same key means
both directions work.

:dev is published ALONE, with no per-commit tag. A rolling channel is
rolling by definition; a commit-addressable image for it would be a rollback
target nobody ever pulls, kept forever. Recovery on dev is to fix forward,
and that is a deliberate trade rather than an omission.

The channel is derived from the REF, not the commit, which is why it is
computed in the workflow and not in ci/version.sh. The same commit built on
dev and on main reports the same version NAME and differs only in the
channel field — that separation is the entire point of keeping the three
values apart.

What this repo deliberately does NOT get: a cross-repo dispatch to refresh
the channel when its bundled APK is rebuilt. That mechanism exists elsewhere
in the family because the app and server live in separate repos, and a
channel that can only be refreshed by an unrelated commit is not a channel.
Minstrel is a monorepo — one push builds the APK and the image in the same
run from the same commit, so the channel cannot go stale against its own
artifact. The requirement is met structurally; copying the mechanism would
add a moving part to fix a problem that does not exist here.

Two guards, for the two ways this wiring can fail quietly:

A dev push must never move :latest. That would ship untested code to every
stable operator on their next pull, with the build green and the image
perfectly valid — just the wrong audience. Nothing else in the suite would
notice.

The two bundling paths must stay mutually exclusive. The rebundle step is
now gated to main specifically, not to "not a tag": under the looser
condition a dev push would run BOTH steps, staging its fresh APK and then
overwriting it with the previous release's. The image still builds, the
sidecar still parses, and the channel whose whole job is being current
quietly serves stale art.

Both falsified against the regressions they name before committing.

Scribe task #3819, milestone #390.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-09 22:47:42 -04:00
bvandeusenandClaude Opus 5 a687ef439c fix(release): refuse an ordering key that overflows versionCode
test-go / test (push) Successful in 1m31s
test-go / integration (push) Successful in 5m48s
The script asserted the key was positive but never that it fits. Android's
versionCode is a signed 32-bit int and the platform rejects an APK above it,
so a build machine with a badly wrong clock would emit a code the script
happily hands on and the install then refuses.

Worse than a rejected build: an over-ceiling code is also unreachably high,
so every correct build afterwards would fail to outrank it and the update
channel would be permanently stuck. Cheaper to refuse at the source than to
diagnose it from a phone that will not update.

The Go guard already asserted this, but only against a pinned value. The
script is what actually runs at build time, so the check belongs here too.

Falsified at the boundary rather than by eye — exactly at the ceiling exits
0, one minute past exits 1. My first probe used a year-6000 clock and did
NOT fire, which turned out to be the probe being wrong rather than the
check: that epoch still lands under the ceiling. The ceiling is reached in
6103, roughly 4079 years out, so this only ever catches a misconfigured
clock.

This commit deliberately touches ci/version.sh alone, to verify the path
filters added in eaf4654c actually fire the Go lane for release-machinery
changes. That run proved nothing about them, because it also touched
internal/** and would have run regardless.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-09 22:17:07 -04:00
bvandeusenandClaude Opus 5 eaf4654c0a test(release): make the version derivation executable, and guard it on dev
test-go / test (push) Successful in 1m10s
test-go / integration (push) Canceled after 5m10s
Steps 1 and 2 of this milestone shipped with no CI coverage at all, and the
reason generalises: release.yml triggers only on main and tags, so nothing
inside it is exercised until a release is already running. That is the worst
place in the repo to be unguarded, because the failure mode is silence — a
version nobody can compare looks exactly like being up to date, and nobody
reports an update they were never offered.

The fix is not a test that reads YAML. The derivation moved into
ci/version.sh, so it can be RUN, and internal/server/release_version_test.go
runs it on every push. release.yml now calls the same script, so the thing
that ships and the thing under test are one artifact rather than two copies
that agree until they don't.

test-go.yml gains 'ci/**' and '.gitea/workflows/release.yml' in its paths.
Without that the guard exists but never fires on the changes it protects,
which is the same nothing it replaces.

What is pinned, and why each one:

  - HHMM is zero-padded. A build at 00:42 must emit "0042"; a stripped
    leading zero shifts the segment two orders of magnitude and reverses
    comparisons against every other build that day. It only bites for a
    tenth of the day, so it will not be found by chance.
  - The name derives from the COMMIT and the code from the BUILD. Asserted
    by holding one clock and moving the other: the name must not move, the
    code must.
  - The code clears 1895, the highest versionCode the retired commit-count
    scheme shipped. Below that Android refuses the upgrade as a downgrade
    and the channel becomes a one-way door.
  - The tag is the name with a `v`, never chosen.
  - release.yml still calls the script, and does not derive a commit count
    again. This pins the WIRING: without it every other assertion keeps
    passing while the shipped path silently drifts out of coverage.

The script rejects unusable clocks rather than emitting something plausible,
and those rejections are tested — a guard that cannot fail is worse than
none, because it reads as coverage.

Falsified before committing rather than after: ran the script against good
and broken inputs and watched all three failure paths fire; verified every
asserted value by executing it rather than by reading it; and checked the
two workflow predicates catch their regressions while staying immune to a
comment that merely names the old formula.

Step 5 of 5 — Scribe task #3812, milestone #390.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-09 22:11:51 -04:00
bvandeusenandClaude Opus 5 ca1c18bbbb fix(android): miniplayer content sat at the top of its bar, not centred
android / Build + lint + test (push) Successful in 4m9s
Reported with a screenshot: the bar drew at full height but the cover,
title and transport row hugged its top edge, leaving an empty strip of
surface above the gesture area.

The Surface is a fixed 80dp. Inside it a plain Column stacked a 4dp
progress fill and then MiniRow at its INTRINSIC height — 48dp, set by the
cover and the icon buttons. A Column stacks from the top and nothing
claimed the remainder, so 80 - 4 - 48 = 28dp collected at the bottom.

Measured off the screenshot rather than eyeballed, and the bands agree
exactly: progress fill 14px (4dp at 3.5x), surface 280px (80dp), cover
167px (48dp), empty below 99px (28.3dp). That the arithmetic lands on the
measurement is what makes this the whole cause rather than one contributor.

MiniRow was already centring its content correctly — inside a box that was
only ever 48dp tall. Giving it weight(1f) lets it take what the progress
fill leaves, so it measures 76dp and centres 48dp of content: 14dp above
and below. The fill stays pinned to the top edge, which is where a
progress indicator belongs.

Not the same bug as issue #2681. That was a dead strip ABOVE the
miniplayer from an unclaimed navigation-bar inset, fixed in v2026.08.18.
This is inside the bar, pure layout, no insets — the surface already
stopped correctly above the gesture area.

CI cannot see this one. There are no Compose UI tests in the repo; the
Android lane is ktlint, detekt and JVM unit tests, and a layout bug needs
an instrumented test to catch. Compilation and lint are all this commit
gets from CI — the visual check is on a device, and the APK only builds on
a tagged release.

Scribe issue #3826.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-09 22:07:23 -04:00
bvandeusenandClaude Opus 5 68136c64c0 fix(android): decide updates on the ordering key, not the version name
android / Build + lint + test (push) Successful in 3m55s
The app compared NAMES while Android installs by versionCode, with nothing
keeping the two orderings consistent. So it could offer a build the platform
then refused as a downgrade, or stay silent about one it would have
accepted. The offer and the install were asking different questions.

Both consumers — the shell banner and the About card — now route through
one isUpdateAvailable(): decide on the ordering key whenever the server
reports one, since that is the same value the package installer compares,
so an offer implies an install that will actually be accepted. Name
comparison survives only as the fallback for a server predating the field.

isVersionNewer is deliberately untouched. It already degrades per segment
and is not what was broken; rewriting it while nearby would have put the
fallback path at risk for no gain.

code is nullable on the wire, and that is load-bearing rather than
stylistic. The app's Json sets coerceInputValues = true, which replaces a
JSON null with the declared default on a NON-nullable property — so
`val code: Long = 0` would have turned "this server reports no ordering
key" into "its key is 0" silently, ranking every such server as infinitely
behind and offering its build to everyone forever. Reading the field
declaration alone would never show that; it lives in AppModule.

A third caller turned up during the sweep and was deliberately left alone.
NetworkStatusController compares the /healthz minClientVersion, which is a
server-declared compatibility floor rather than the bundled APK — there is
no ordering key on that wire at all, so names remain the only thing it can
compare. Different question, correctly still using the old helper.

The update channel had no tests whatsoever before this, which is worth
stating: the thing deciding whether anyone is ever offered an update fails
silently in both directions. The new suite pins that the key wins when it
disagrees with the name, that a null key falls back rather than reading as
zero, the recorded migration constraint (a new-scheme name outranks an
old-scheme one across a day boundary but NOT within the same day), and the
degradation cases — including that an unparseable DECIDING segment reads as
zero and loses, which is why the channel must never live inside the name.

Every assertion was checked against the real comparison by mirroring it,
rather than from reading it: two of my first-draft comments described the
wrong mechanism and were corrected on the evidence.

Step 4 of 5 — Scribe task #3811, milestone #390.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-09 22:01:26 -04:00
bvandeusenandClaude Opus 5 9f3e0b8cd3 feat(version): sidecar and /api/client/version carry name, code and channel
test-go / test (push) Successful in 1m25s
test-go / integration (push) Successful in 5m52s
The client compares names while Android installs by versionCode, and the
wire had no way to close that gap: the sidecar was one positional line and
the endpoint returned a name only. This is the plumbing that makes the
ordering key decidable by the client at all.

The sidecar is now JSON rather than a grown positional string. That shape
was chosen against a specific failure: the obvious growth path was
"<name> <code>", which a first-space split silently mangles the moment a
third field appears — the code stops parsing as an integer and the reader
falls back to name comparison WITHOUT erroring. JSON cannot mistake a new
field for an old one.

code is a POINTER on both sides, and omitempty on the wire. Absent has to
stay distinguishable from zero: a build published before ordering keys were
recorded genuinely has no code, and zero would claim it is infinitely old
rather than unknown.

A malformed sidecar now fails loudly instead of serving a blank version.
If an unreadable file produced an empty name, every client would compare
against nothing, conclude it was current, and go quiet — "I cannot read
this" and "there is nothing newer" would return the same answer, which is
the failure mode nobody reports because nobody is offered anything to
report.

The non-tag :latest path no longer RECONSTRUCTS the bundled APK's version.
android-release now publishes the sidecar as a release asset beside the
APK, and the image build downloads it. The old reconstruction duplicated a
derivation formula across two files, and could only ever recover the name —
the ordering key is build-time minutes and exists nowhere once that build
ends. Releases predating the sidecar report their name with a null code,
which is the honest answer rather than a guessed one.

image-release also drops to a shallow checkout: it needed full history and
tags only to re-derive versions from the tagged commit, and now touches git
for nothing. MINSTREL_VERSION comes from GITHUB_REF.

Two things checked rather than assumed. The Android Json sets
ignoreUnknownKeys, so the added fields cannot break already-installed apps.
It also sets coerceInputValues, which will silently turn a null code into 0
if step 4 declares the field non-nullable — recorded on task #3811, because
reading the field declaration alone would never reveal it.

Also fixes a stale comment block describing "the Flutter client", deleted
in v2026.08.18.

Step 3 of 5 — Scribe task #3810, milestone #390.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-09 21:51:41 -04:00
bvandeusenandClaude Opus 5 e46c6bcccf docs(release): tags become vYYYY.MM.DD.HHMM, and stop telling people to move them
The tag is now the artifact's own version name with a `v` in front, so
`v2026.09.10.1432` and `2026.09.10.1432` are one string. Nothing has to
reconcile what the tag claims against what the APK reports, and minting one
is arithmetic on the tagged commit's timestamp rather than a lookup.

The substantive change is the prose. release.yml's header instructed the
reader to `git push -f origin vYYYY.MM.DD` on a same-day re-cut. That is
the operation the family rulebook forbids outright, and it has incidents
behind it — moving a same-day tag forward once took a published release
down with it. Anyone who had installed from that tag was holding something
it no longer pointed at.

With HHMM there is nothing left for mutability to buy: every tag is unique
by construction, so a second release the same day is not a collision to
resolve, just another tag.

The old instruction is recorded as retired rather than deleted. Someone who
remembers it should learn it was withdrawn and why, not find it silently
absent and assume they misremembered.

README contradicted itself inside one sentence — "immutable per-day release
tags ... a same-day re-cut moves the tag forward" — and now says which it
is, plus a note that pre-2026-09-10 tags keep the old shape and still work.

Transition wrinkle, deliberately left for step 3: the non-tag :latest path
reconstructs the bundled APK's name from the latest release's commit
timestamp, which for the one existing old-shape release yields
2026.09.09.1828 while that APK actually declares 2026.09.09.1895. It fails
SAFE — 1828 compares lower, so no false update is offered — and it
self-corrects at the first new-scheme release. Step 3 removes the
reconstruction entirely by having the sidecar carry recorded values instead
of derived ones.

Step 2 of 5 — Scribe task #3809, milestone #390.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-09 21:43:33 -04:00
bvandeusenandClaude Opus 5 bfdaed9365 fix(release): derive versionCode from build time, versionName from commit time
android / Build + lint + test (push) Successful in 4m19s
versionCode was `git rev-list --count HEAD`, and build.gradle.kts called it
"monotonic forever". It is not, and that claim was sitting directly above
the bug it denied.

A commit count runs ahead on `dev`. So a dev build carried a HIGHER code
than the `main` release meant to supersede it, and Android refuses that
install as a downgrade — a channel you can enter and cannot leave without
uninstalling and losing local data.

Two clocks now, and the split is deliberate even though it reads like an
inconsistency:

The NAME answers "is this the same code?", so it derives from COMMIT time
and reads identically on every lane building this source. A dev build and
a main build of one commit must report the same string. Build time cannot
do that — it prints two numbers for one thing.

The ORDERING KEY answers "may this be installed over that?", so it must be
monotonic BY CONSTRUCTION: minutes since 2020-01-01. Commit time fails
here for the mirror-image reason — rebuild an older commit and it goes
DOWN, which on a phone is a refused install rather than a confusing label.

The non-tag :latest path reconstructed the bundled APK's name with the old
formula, so it is moved to the same commit-timestamp derivation. That
duplication is temporary: once the tag becomes `v<version-name>` it
collapses to `${TAG#v}` with nothing left to keep in step.

Verified locally by running the derivations rather than reasoning about
them: HEAD yields 2026.09.09.1828; the key yields 3519456 against ~1895
from the old scheme, inside int32 with ~4000 years of headroom; a commit
at 00:42 UTC yields "0042", not "42". The workflow now asserts the emitted
shape too — a malformed name builds, signs and publishes happily and only
surfaces as an update nobody is offered, which nobody reports.

That local check is the only verification this commit gets. release.yml
triggers on main and tags only, so nothing on `dev` executes the new
derivation; CI here proves the Gradle file still parses and nothing else.

Also confirms the migration constraint recorded in milestone #390: this
commit would name a release 2026.09.09.1828, which is LOWER than the
installed 2026.09.09.1895 under name comparison. The first new-scheme
release must be cut on a later calendar day, or existing installs will
never be offered it.

Step 1 of 5 — Scribe task #3808, milestone #390.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SQ31KQpYbStyK5y58UmPLH
2026-09-09 21:37:30 -04:00
510 changed files with 39515 additions and 3346 deletions
+19 -5
View File
@@ -6,9 +6,20 @@
**/build
web/build
# Flutter mobile client — built separately on developer machines / Flutter CI.
# Including it in the Go build context wastes ~70 files and invalidates the
# `COPY . .` layer cache on every Flutter-only change.
# The Android client — built by its own job, never from this context. The APK
# reaches the image through client/, downloaded as a CI artifact, so nothing
# here reads android/ sources.
#
# This block named `flutter_client/` until 2026-09-10 and lost its PATTERN when
# that tree was deleted, leaving a comment describing an exclusion that was no
# longer happening. android/ never took its place, so 4.1 MB of Gradle project
# has been entering the context and busting the `COPY . .` layer on every
# Android-only change.
android/
# Local `make build` output — an 18 MB binary the image never uses, since the
# builder stage compiles its own.
bin/
# Docs and IDE noise
docs/
@@ -26,5 +37,8 @@ docs/
!.env.example
# CI workflow files don't need to ship in the image.
.forgejo/
.github/
#
# This said `.forgejo/` and `.github/` — neither of which this repo has. Gitea
# Actions reads `.gitea/`, so the one directory that actually exists was the
# one not excluded, and every workflow edit invalidated the context.
.gitea/
-95
View File
@@ -1,95 +0,0 @@
name: android
# Native Android (Kotlin/Compose/Media3) — M8 rewrite, now the only client.
# This workflow is testing only — lint + detekt + unit tests on every push
# to dev/main, plus a debug APK artifact for main. The signed-release
# build + asset attach + image-bundling lives in release.yml under a
# `needs:` chain so the docker image cannot ship without the APK.
on:
push:
branches: [main, dev]
paths:
- 'android/**'
- '.gitea/workflows/android.yml'
# pull_request trigger intentionally omitted — see test-web.yml for
# the rationale (single-author repo, push covers PR-merge equivalent).
concurrency:
group: ${{ github.workflow }}-${{ github.ref }}
cancel-in-progress: true
env:
# Silences the JDK 22+ "restricted method in java.lang.System has been
# called" warning that Gradle 9.1's bundled native-platform jar trips
# at launch (System.load for native primitives). Affects the LAUNCHER
# JVM, not the daemon — that's why org.gradle.jvmargs in
# gradle.properties isn't enough. Future-compat: required opt-in once
# JDK 25 promotes the warning to an error.
JAVA_TOOL_OPTIONS: "--enable-native-access=ALL-UNNAMED"
jobs:
build:
name: Build + lint + test
# Using flutter-ci runner label because it's the only proven-working
# label with docker that can pull our container.image. Switch to
# android-ci once the operator registers that runner label.
runs-on: flutter-ci
container:
image: git.fabledsword.com/bvandeusen/ci-android:36
defaults:
run:
working-directory: android
steps:
- name: Checkout
uses: actions/checkout@v4
- name: Cache Gradle dirs
# Resolved deps + Gradle distribution + Kotlin daemon caches.
# Saves ~3 min per CI run after the first warm-up.
uses: actions/cache@v4
with:
path: |
~/.gradle/caches
~/.gradle/wrapper
~/.kotlin
key: gradle-${{ runner.os }}-${{ hashFiles('android/gradle/wrapper/gradle-wrapper.properties', 'android/gradle/libs.versions.toml', 'android/**/*.gradle.kts') }}
restore-keys: |
gradle-${{ runner.os }}-
- name: Make gradlew executable
run: chmod +x ./gradlew
- name: Gradle wrapper validation
run: ./gradlew --version
- name: ktlint
run: ./gradlew ktlintCheck
- name: detekt
run: ./gradlew detekt
- name: Unit tests
run: ./gradlew testDebugUnitTest
- name: Assemble debug
if: github.event_name == 'push' && github.ref == 'refs/heads/main'
run: ./gradlew assembleDebug
- name: Upload debug APK
if: github.event_name == 'push' && github.ref == 'refs/heads/main'
# Mirrored action, never actions/upload-artifact. @v4+ throws
# GHESNotSupportedError client-side on the hostname (no server setting
# reaches that check), and @v3 is worse — it reports success while Gitea
# serves artifacts back only through the v4 API, so the upload is stored
# and invisible to every retrieval path. @v3 is what left 72 unreachable
# artifacts on this repo. Pinned by SHA because the mirror auto-syncs;
# full URL because DEFAULT_ACTIONS_URL sends bare owner/repo to github.com.
# See Scribe issues 2255 / 2270.
uses: https://git.fabledsword.com/bvandeusen/upload-artifact@cb8afe72b42edc798abfb8fcb556cf660d894245
with:
name: minstrel-android-debug-${{ github.sha }}
path: android/app/build/outputs/apk/debug/app-debug.apk
if-no-files-found: error
+673 -108
View File
@@ -2,55 +2,407 @@ name: release
# Builds and pushes the minstrel container image to the Gitea registry.
#
# push to main → :main and :latest (latest-release APK bundled)
# push tag vYYYY.MM.DD → :vYYYY.MM.DD and :latest (freshly-built APK bundled)
# workflow_dispatch → manual trigger (same rules based on the ref)
# push to dev → :dev (freshly-built dev APK bundled)
# push to main → :latest + :<sha> (latest-release APK bundled)
# push tag vYYYY.MM.DD.HHMM → :latest (fresh APK bundled)
# workflow_dispatch → manual trigger (same rules based on the ref)
#
# Release model: per-day CalVer tags (no trailing patch digit). The day's
# tag is intentionally mutable — if a second release happens the same day,
# move the tag with `git push -f origin vYYYY.MM.DD` and the image tag of
# the same name gets overwritten. :latest is updated by every main push
# AND every tag push, so it always reflects the newest blessed image.
# That is the whole tag map, and it is family rule 145 + 147 as written.
#
# :<sha> on main is the ROLLBACK UNIT — every production commit addressable
# without a release ceremony. It is minted only on main, where rollback is
# actually worth having: merges are gated (rule 2) so they number in the dozens
# per year, while on dev they would be one per push, forever, for a channel
# whose entire contract is that it moves.
#
# There are NO :<version> image tags. This repo published :vYYYY.MM.DD.HHMM
# until 2026-09-10 and it was the inverse of the rule on both counts — minting
# a version tag nobody pinned while the rollback unit the rule names did not
# exist here at all. Git and the build's own self-reported version answer
# "which build is this"; a third name for the same thing is upkeep for a model
# we do not run. Operator, 2026-09-10: "only things like the APK need that kind
# of versioning for their update process."
#
# There is no :main either. :latest tracks main's tip with no gate between them
# (rule 147), so a second name for the same image sends readers looking for a
# distinction that does not exist.
#
# The dev channel exists so testing a build does not require shipping one.
# Before it, the only way to get an APK onto a phone was to cut a release,
# which made `main` the staging area by default. `:dev` carries its own
# freshly-built APK, signed with the SAME key as release builds — a different
# key cannot install over the stable app, so anyone crossing channels would
# have to uninstall and lose their data.
#
# :dev is published ALONE, with no per-commit tag. A rolling channel is
# rolling by definition; a commit-addressable image for it would be a
# rollback target nobody ever pulls, kept forever. Recovery on dev is to fix
# forward.
#
# Note what this repo does NOT need: a cross-repo dispatch to refresh the
# channel when its bundled APK is rebuilt. That mechanism exists elsewhere in
# the family because the app and the server live in separate repos. Minstrel
# is a monorepo — one push builds the APK and the image in the same run from
# the same commit, so the channel cannot go stale against its own artifact.
# The requirement is satisfied structurally; copying the mechanism would add
# a moving part to fix a problem that does not exist here.
#
# Release model: the tag IS the artifact's version name with a `v` in front.
# `v2026.09.10.1432` and `2026.09.10.1432` are the same string, derived from
# the tagged commit's UTC timestamp — so there is no mismatch to reconcile
# between what the tag says and what the APK reports, and nothing to look up
# when minting one.
#
# TAGS ARE IMMUTABLE. Never move, retarget or delete a published tag. A
# same-day second release is not a collision — HHMM makes every tag unique
# by construction, so the answer is simply another tag.
#
# This block used to say the opposite: that the per-day tag was
# "intentionally mutable" and that a same-day re-cut should
# `git push -f origin vYYYY.MM.DD`. That instruction is what the family
# rulebook now forbids outright, and it has incidents behind it — moving a
# same-day tag forward once took a published release down with it. Anyone
# installing from a tag is holding something the tag no longer points at,
# which is a worse failure than an extra row in the tag list.
#
# :latest is updated by every main push AND every tag push, so it always
# reflects the newest blessed image.
#
# APK pipeline: on tag pushes the android-release job builds + signs the
# Android APK and uploads it as a workflow artifact. The image-release
# job declares `needs: android-release`, so the docker image cannot
# start building until the APK is guaranteed-ready — no polling, no
# race, no silent-failure mode. Asset attachment to the gitea Release
# happens in the same android-release job, so the Release-page download
# link and the in-image bundled APK are both populated atomically.
# race, no silent-failure mode. Attaching the APK to the gitea Release is
# its own job (release-assets), behind the test gate below.
#
# :latest always carries an APK. Because every main push also moves
# :latest (not just tags), a main build with no APK would silently strip
# the in-app update channel off :latest until the next release. So on
# non-tag builds image-release pulls the MOST RECENT release's signed APK
# and reconstructs its exact versionName (tag + commit-count, the same
# formula android-release bakes in) for the version sidecar — no rebuild,
# just rebundle. Tag builds keep bundling their own freshly-built APK.
# AND the version sidecar published beside it — the recorded values, not
# recomputed ones — so no rebuild is needed, just a rebundle. Tag builds
# keep bundling their own freshly-built APK.
#
# Android testing (lint + detekt + unit tests, debug APK upload on main)
# lives in android.yml and runs independently on every push.
# THE GATE (rule 177, M462 #4984). Every verifying lane lives in this file —
# Go (vet, lint, short tests), the Postgres integration suite, the web app
# (npm audit, svelte-check, vitest), Android (ktlint, detekt, unit tests) and
# govulncheck — and every job that publishes something names each of them in
# `needs:` and requires `success` from each, by name. Nothing publishes on red.
#
# They used to be three separate workflows (test-go, test-web, android) on the
# same push trigger as this one. Separate workflows cannot see each other's
# verdict, so :dev meant "it built", never "it passed": a red test run and a
# fresh :dev could carry the same timestamp. One graph is the only place the
# edge can be written.
#
# A skipped lane is NOT a pass. The publishing conditions check
# `result == 'success'` per lane rather than `!failure()`, so a lane that
# never started blocks the publish exactly as a red one does. Lanes carry no
# path filters for the same reason: a web-only push still runs the Go suite,
# because "not run" must never read as "passed".
#
# What publishes, and is therefore gated: the image tags (image-release) and
# the APK + version sidecar attached to a tag's Release (release-assets).
# android-release only BUILDS the signed APK into a workflow artifact, which
# nobody outside this run can pull, so it runs in parallel with the lanes
# instead of after them; attaching it to the Release is the publishing half,
# and that half waits for the gate.
#
# To watch the gate refuse: dispatch this workflow with force_red=true. The go
# lane fails on purpose, and both publishing jobs must report skipped.
on:
push:
branches: [main]
branches: [main, dev]
tags: ['v*']
paths-ignore:
- 'docs/**'
- '**/*.md'
workflow_dispatch:
inputs:
force_red:
description: Fail the go lane on purpose, to check that nothing publishes on red
type: boolean
default: false
# Force-moving the per-day tag (or rapidly re-pushing to main) should
# supersede the in-flight build — the operator explicitly wants the
# later commit to win.
# A rapid re-push to main should supersede the in-flight build — the
# operator explicitly wants the later commit to win. Tags no longer enter
# into this: they are immutable and unique, so no tag build can ever be
# superseded by another run on the same ref.
concurrency:
group: ${{ github.workflow }}-${{ github.ref }}
cancel-in-progress: true
jobs:
# ---------------------------------------------------------------- lanes --
# Verifying jobs. Each one is named in the `needs:` of every publishing job
# below; add a lane here and it must be added there in the same commit.
go:
runs-on: go-ci
container:
image: git.fabledsword.com/bvandeusen/ci-go:1.26
steps:
- name: Checkout
uses: actions/checkout@v4
- name: Forced failure (gate check)
if: github.event.inputs.force_red == 'true'
run: |
echo "::error::force_red dispatch: failing on purpose so the publishing jobs must skip"
exit 1
- name: Toolchain versions
run: |
go version
golangci-lint --version
- name: Generated code matches queries (sqlc)
run: make verify-generate
- name: go vet
run: go vet ./...
- name: golangci-lint
run: golangci-lint run ./...
- name: go test (short, race)
run: go test -short -race ./...
# Full `go test -race` against an ephemeral Postgres.
#
# DB wiring follows the act_runner shared-daemon pattern: the runner's Docker
# daemon also runs the operator's dev compose stack, so service containers
# get NO published ports (collision) and no service-name DNS. We discover the
# service container through the mounted docker socket and reach it by bridge
# IP. The exactly-one assertion is a hard guard — pointing tests at the dev
# Postgres would truncate it (the disaster Fable #339 exists to prevent).
#
# The key stays `integration` with no `name:` (rule 80): act_runner derives
# the service container's name from the job's display name.
#
# `web/build/` has a committed placeholder index.html so go:embed succeeds
# without the SPA being built first.
integration:
runs-on: go-ci
container:
image: git.fabledsword.com/bvandeusen/ci-go:1.26
services:
postgres:
image: postgres:16-alpine
env:
POSTGRES_USER: minstrel
POSTGRES_PASSWORD: minstrel
POSTGRES_DB: minstrel_test
# No `ports:` — the runner shares the operator's dev compose
# Docker daemon; publishing a fixed host port collides.
steps:
- name: Checkout
uses: actions/checkout@v4
- name: Integration suite (discover service by bridge IP, migrate, test)
run: |
set -eux
# Discover THIS job's Postgres service container via the
# mounted docker socket. act_runner attaches the job
# container and its service container(s) to a shared per-job
# network, so scope discovery to a postgres that sits on a
# network THIS job container is also on. The old
# `--filter name=integration` matched EVERY concurrent
# integration run's postgres (a dev push + the main-merge run
# overlap → 2 candidates → false "expected exactly 1" abort).
# The operator's dev compose `minstrel-postgres-*` is never on
# this job's network; skip it explicitly as belt-and-suspenders
# (a wrong target would truncate real data).
SELF=$(cat /etc/hostname)
SELF_NETS=$(docker inspect -f '{{range $k,$v := .NetworkSettings.Networks}}{{$k}} {{end}}' "$SELF")
test -n "$SELF_NETS"
echo "self ($SELF) networks: $SELF_NETS"
PG_ID=""
PG_NAME=""
for cid in $(docker ps --filter "ancestor=postgres:16-alpine" -q); do
nm=$(docker inspect -f '{{.Name}}' "$cid" | sed 's#^/##')
case "$nm" in *minstrel-postgres*|*_postgres_*) continue ;; esac
for net in $(docker inspect -f '{{range $k,$v := .NetworkSettings.Networks}}{{$k}} {{end}}' "$cid"); do
case " $SELF_NETS " in *" $net "*) PG_ID="$cid"; PG_NAME="$nm"; break 2 ;; esac
done
done
test -n "$PG_ID" || { echo "FATAL: no postgres service container on this job's network (self nets: $SELF_NETS)"; exit 1; }
echo "selected postgres: $PG_ID $PG_NAME"
PG_IP=$(docker inspect -f '{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}' "$PG_ID")
test -n "$PG_IP"
export MINSTREL_TEST_DATABASE_URL="postgres://minstrel:minstrel@${PG_IP}:5432/minstrel_test?sslmode=disable"
# Wait for Postgres to accept connections. Asked of the service
# container itself: the run: shell is dash (rule 81), where the
# old `/dev/tcp` probe never connects and the loop silently
# burned its full two minutes on every run.
ready=""
for i in $(seq 1 60); do
if docker exec "$PG_ID" pg_isready -U minstrel -d minstrel_test -q; then ready=1; break; fi
sleep 2
done
test -n "$ready" || { echo "FATAL: postgres never became ready"; exit 1; }
# Relax durability on the throwaway CI Postgres. Our test pattern
# is dbtest.ResetDB → TRUNCATE … RESTART IDENTITY CASCADE before
# every test, and the per-TRUNCATE commit fsync is the dominant
# cost of the integration suite. The CI DB is rebuilt every run so
# fsync / full_page_writes / synchronous_commit buy nothing. Apply
# via docker exec because:
# - The act_runner `services:` block can't override the container
# command, so `postgres -c fsync=off` at boot isn't an option.
# - ALTER SYSTEM cannot run inside a transaction; psql -c
# auto-commits each statement, which is what we need.
# - fsync / full_page_writes are sighup GUCs and
# synchronous_commit is user-context, so pg_reload_conf() picks
# all three up with no restart.
# Non-fatal: a perms surprise degrades to "slower", never red CI.
docker exec "$PG_ID" psql -U minstrel -d minstrel_test \
-c "ALTER SYSTEM SET fsync = off" \
-c "ALTER SYSTEM SET synchronous_commit = off" \
-c "ALTER SYSTEM SET full_page_writes = off" \
-c "SELECT pg_reload_conf()" \
|| echo "WARN: durability relax failed; continuing"
# Apply embedded migrations to the fresh test DB, then run the
# full suite (no -short → integration tests execute). -p 1:
# every integration package TRUNCATEs the one shared test DB;
# concurrent package binaries → TRUNCATE deadlocks. Serialize
# package execution (the documented local invocation too).
# -timeout 20m: internal/api alone takes ~6.5 min under -race on
# an idle runner, and a dev and a main run sharing the runner
# pushed it past go test's default 10m (run 8368, 600.016s, the
# running test 2s old — nothing hung).
MINSTREL_DATABASE_URL="$MINSTREL_TEST_DATABASE_URL" go run ./cmd/minstrel migrate
go test -p 1 -race -timeout 20m ./...
web:
runs-on: go-ci
container:
image: git.fabledsword.com/bvandeusen/ci-go:1.26
defaults:
run:
working-directory: web
steps:
- name: Checkout
uses: actions/checkout@v4
- name: Install deps
run: npm ci
# What ships to browsers: `dependencies` and the runtime they pull in
# (svelte, devalue). Build and test tooling (vite, vitest, tailwind,
# kit's dev server) is left out because none of it reaches a user, and
# its open advisories need major-version upgrades tracked separately.
- name: npm audit (shipped dependencies)
run: npm audit --omit=dev --audit-level=moderate
- name: Type-check + svelte-check
run: npm run check
- name: Vitest
run: npm test
android:
# Using flutter-ci runner label because it's the only proven-working
# label with docker that can pull our container.image.
runs-on: flutter-ci
container:
image: git.fabledsword.com/bvandeusen/ci-android:36
defaults:
run:
working-directory: android
env:
# Silences the JDK 22+ "restricted method in java.lang.System has been
# called" warning that Gradle's bundled native-platform jar trips at
# launch. Affects the LAUNCHER JVM, not the daemon — that's why
# org.gradle.jvmargs in gradle.properties isn't enough.
JAVA_TOOL_OPTIONS: "--enable-native-access=ALL-UNNAMED"
steps:
- name: Checkout
uses: actions/checkout@v4
- name: Cache Gradle dirs
uses: actions/cache@v4
with:
path: |
~/.gradle/caches
~/.gradle/wrapper
~/.kotlin
key: gradle-${{ runner.os }}-${{ hashFiles('android/gradle/wrapper/gradle-wrapper.properties', 'android/gradle/libs.versions.toml', 'android/**/*.gradle.kts') }}
restore-keys: |
gradle-${{ runner.os }}-
- name: Make gradlew executable
run: chmod +x ./gradlew
- name: Gradle wrapper validation
run: ./gradlew --version
- name: ktlint
run: ./gradlew ktlintCheck
- name: detekt
run: ./gradlew detekt
- name: Unit tests
run: ./gradlew testDebugUnitTest
# No debug APK is built or uploaded here. Main used to upload a
# debug-signed app-debug.apk: a build signed by a key regenerated in
# every container, which no install can update (family idea #5103,
# practice 2). Phones get builds from android-release, signed with
# the one release key, on dev and on tags.
# Known vulnerabilities in the Go code and the standard library it is built
# with. Runs in the SAME image the Dockerfile's builder stage uses, so the
# standard library it checks is the one that ends up in the shipped binary;
# the ci-go image carries its own Go and would be checking a different
# toolchain. Keep this image and the Dockerfile's builder in step.
#
# govulncheck is fetched at CI time, unpinned (rule 154): the vulnerability
# database and the tool that reads it should both be current.
govulncheck:
runs-on: go-ci
container:
image: golang:1.26-bookworm
steps:
# Plain git, not actions/checkout: that action runs on node, which the
# golang image does not carry. This step is dash (rule 81).
- name: Checkout
env:
TOKEN: ${{ github.token }}
run: |
set -eu
auth=$(printf 'x-access-token:%s' "$TOKEN" | base64 -w0)
git init -q .
git remote add origin "${{ github.server_url }}/${{ github.repository }}.git"
git -c http.extraHeader="Authorization: Basic ${auth}" fetch -q --depth 1 origin "${{ github.sha }}"
git checkout -q FETCH_HEAD
git log -1 --format='%H %s'
- name: govulncheck
run: |
go version
go run golang.org/x/vuln/cmd/govulncheck@latest ./...
# ------------------------------------------------------------ artifacts --
android-release:
name: Build signed APK (tag releases only)
if: startsWith(github.ref, 'refs/tags/v')
name: Build signed APK (releases and dev)
# Also builds on `dev`, which is what makes a test channel possible at
# all. Without it the only way to get a build onto a phone was to cut a
# release, which quietly turns `main` into the staging area.
if: startsWith(github.ref, 'refs/tags/v') || github.ref == 'refs/heads/dev'
runs-on: flutter-ci
container:
image: git.fabledsword.com/bvandeusen/ci-android:36
@@ -75,14 +427,18 @@ jobs:
outputs:
version_name: ${{ steps.ver.outputs.name }}
version_code: ${{ steps.ver.outputs.code }}
channel: ${{ steps.ver.outputs.channel }}
steps:
- name: Checkout
uses: actions/checkout@v4
with:
# fetch-depth: 0 retrieves full history; default shallow clone
# would return 1 for `git rev-list --count HEAD`, breaking the
# iteration suffix.
# Full history. The version name now reads only the tip commit's
# timestamp, so a shallow clone would technically serve — but this
# job derives a value that ships to devices, and a shallow checkout
# changes what git-derived values resolve to WITHOUT failing. The
# whole failure class here is a green build carrying a wrong
# version, so the cheap guarantee is worth keeping.
fetch-depth: 0
- name: Compute release version
@@ -91,12 +447,23 @@ jobs:
working-directory: ${{ github.workspace }}
run: |
set -euo pipefail
TAG="${GITHUB_REF#refs/tags/v}"
COMMIT_COUNT=$(git rev-list --count HEAD)
VERSION_NAME="${TAG}.${COMMIT_COUNT}"
echo "name=${VERSION_NAME}" >> "$GITHUB_OUTPUT"
echo "code=${COMMIT_COUNT}" >> "$GITHUB_OUTPUT"
echo "::notice::APK version: ${VERSION_NAME} (code=${COMMIT_COUNT})"
# The derivation lives in ci/version.sh, not here, so it can be
# executed by a test on every push. Anything inline in this file is
# unverifiable until a release is already running.
out="$(ci/version.sh HEAD)"
printf '%s\n' "${out}" >> "$GITHUB_OUTPUT"
# The channel is a property of the LANE, not of the commit, which is
# why it is derived here rather than in version.sh. Same commit built
# on dev and on main reports the same NAME and differs only here —
# that is the whole point of separating the two values.
if [ "${GITHUB_REF}" = "refs/heads/dev" ]; then
channel=dev
else
channel=stable
fi
echo "channel=${channel}" >> "$GITHUB_OUTPUT"
echo "::notice::APK $(printf '%s' "${out}" | tr '\n' ' ') channel=${channel}"
# Checked BEFORE the expensive work, not after it. "Attach APK to gitea
# Release" below resolves the release by tag and fails if it is absent —
@@ -108,6 +475,7 @@ jobs:
# the release together, so this passes). A bare `git push origin vX` is the
# case this catches.
- name: Release must exist for this tag
if: startsWith(github.ref, 'refs/tags/v')
shell: bash
working-directory: ${{ github.workspace }}
env:
@@ -155,14 +523,49 @@ jobs:
-PMINSTREL_VERSION_NAME=${{ steps.ver.outputs.name }} \
-PMINSTREL_VERSION_CODE=${{ steps.ver.outputs.code }}
# The APK every phone updates from must carry THE release key: Android
# updates an app in place only when the signer matches, so an APK
# signed by any other key (debug, a regenerated keystore, a swapped
# secret) reaches no installed phone. The certificate's digest is
# pinned below; it is public, not a secret. Gradle signs with the
# release key or leaves the APK unsigned, and an unsigned build fails
# here too, as there is no app-release.apk to verify (family idea
# #5103, practice 3). apksigner, not keytool: keytool prints nothing
# for a v2-only APK.
#
# Rotating the key on purpose means every install must be removed and
# reinstalled; change the digest here in the same commit.
- name: The APK carries the release key
shell: bash
env:
# CN=Minstrel, O=FabledSword. Read from run 8446 (#5116).
RELEASE_CERT_SHA256: 43d183307bc46b821789d90444a960b137f78f2166ff431efb2406d0fceaf612
run: |
set -euo pipefail
sdk="${ANDROID_HOME:-${ANDROID_SDK_ROOT:-}}"
signer="$(ls "$sdk"/build-tools/*/apksigner 2>/dev/null | sort -V | tail -1 || true)"
test -n "$signer" || { echo "::error::no apksigner under '$sdk/build-tools'"; exit 1; }
certs="$("$signer" verify --print-certs app/build/outputs/apk/release/app-release.apk)"
printf '%s\n' "$certs" | grep -E '^Signer #[0-9]+ certificate (DN|SHA-256 digest)'
# One signer, and it is ours. A second signer would be a lineage or
# a mistake; either way not something to ship unexamined.
digests="$(printf '%s\n' "$certs" | sed -n 's/^Signer #[0-9]* certificate SHA-256 digest: //p')"
if [ "$digests" != "$RELEASE_CERT_SHA256" ]; then
if printf '%s' "$certs" | grep -q 'CN=Android Debug'; then
echo "::error::the release APK is signed with a debug key"
else
echo "::error::the release APK is not signed by the release key: got '${digests//$'\n'/ }', want ${RELEASE_CERT_SHA256}"
fi
exit 1
fi
- name: Upload APK as workflow artifact
# Mirrored action, never actions/upload-artifact — @v4+ refuses on the
# hostname, @v3 uploads something Gitea will never serve back. This is
# the producing half of a pair: image-release downloads `minstrel-apk`
# below with the matching download-artifact mirror. Both must stay on
# the v4 protocol — mixing a v3 upload with a v4 download (or the
# reverse) yields an empty listing, not an error. See Scribe 2255 / 2270.
uses: https://git.fabledsword.com/bvandeusen/upload-artifact@cb8afe72b42edc798abfb8fcb556cf660d894245
# Stock action (snippet #2271) — never @v3, which uploads something Gitea
# will never serve back. This is the producing half of a pair:
# image-release downloads `minstrel-apk` below. Any upload v4+ pairs with
# any download v4+ on this forge (every combination tested 2026-09-10,
# Scribe spike #3843), so the two pins need not move together.
uses: actions/upload-artifact@v7
with:
name: minstrel-apk
path: android/app/build/outputs/apk/release/app-release.apk
@@ -170,17 +573,63 @@ jobs:
# artifact existing, so an empty upload must fail here, not there.
if-no-files-found: error
# Publishes the signed APK and its version sidecar on the tag's Release.
# Split out of android-release so the APK can be BUILT in parallel with the
# lanes while being PUBLISHED only once they have all passed.
release-assets:
name: Attach APK to the Release (tag releases only)
needs: [go, integration, web, android, govulncheck, android-release]
if: >-
${{
!cancelled()
&& needs.go.result == 'success'
&& needs.integration.result == 'success'
&& needs.web.result == 'success'
&& needs.android.result == 'success'
&& needs.govulncheck.result == 'success'
&& needs.android-release.result == 'success'
&& startsWith(github.ref, 'refs/tags/v') }}
runs-on: go-ci
container:
image: git.fabledsword.com/bvandeusen/ci-go:1.26
steps:
- name: Download signed APK artifact
uses: actions/download-artifact@v8
with:
name: minstrel-apk
path: release-apk/
- name: Attach APK to gitea Release
# Tag releases only. A dev build has no Release to hang assets on and
# does not need one — the :dev image bundles the APK, and the server
# serves it from /api/client/apk like any other.
shell: bash
env:
CI_TOKEN: ${{ secrets.CI_TOKEN }}
VERSION_NAME: ${{ needs.android-release.outputs.version_name }}
VERSION_CODE: ${{ needs.android-release.outputs.version_code }}
run: |
set -euxo pipefail
TAG="${GITHUB_REF#refs/tags/}"
REPO="${GITHUB_REPOSITORY}"
APK_PATH="app/build/outputs/apk/release/app-release.apk"
APK_PATH="release-apk/app-release.apk"
ls -lh "${APK_PATH}"
# Publish the version sidecar as a release asset next to the APK.
#
# This is what lets a later :latest build stop RECONSTRUCTING the
# bundled APK's version and simply read what was recorded. The
# ordering key in particular cannot be re-derived after the fact —
# it is build-time minutes, so once this job ends the value exists
# nowhere else. Reconstruction could only ever recover the name,
# and only by duplicating a formula that then has to be kept in
# step across two files.
SIDECAR_PATH="/tmp/minstrel.apk.version"
printf '{"name":"%s","code":%s,"channel":"stable"}\n' \
"${VERSION_NAME}" "${VERSION_CODE}" > "${SIDECAR_PATH}"
cat "${SIDECAR_PATH}"
RELEASE_JSON="$(curl -fsSL \
-H "Authorization: token ${CI_TOKEN}" \
"https://git.fabledsword.com/api/v1/repos/${REPO}/releases/tags/${TAG}")"
@@ -202,15 +651,37 @@ jobs:
exit 1
fi
# Same treatment for the sidecar. Named `.apk.version` so the
# downloader's `\.apk$` match cannot pick it up by mistake.
SIDECAR_HTTP=$(curl -sS -L -o /tmp/upload-sidecar.out -w '%{http_code}' \
-H "Authorization: token ${CI_TOKEN}" \
-F "attachment=@${SIDECAR_PATH}" \
"https://git.fabledsword.com/api/v1/repos/${REPO}/releases/${RELEASE_ID}/assets?name=minstrel-${TAG}.apk.version")
echo "sidecar_upload_http=${SIDECAR_HTTP}"
cat /tmp/upload-sidecar.out || true
echo
if [ "${SIDECAR_HTTP}" -lt 200 ] || [ "${SIDECAR_HTTP}" -ge 300 ]; then
echo "::error::version sidecar upload returned HTTP ${SIDECAR_HTTP}"
exit 1
fi
image-release:
name: Build + push container image
# `needs:` waits for android-release. For tag pushes android-release
# runs and must succeed before this job starts — guaranteeing the
# APK artifact is present. For main pushes android-release is
# skipped; the `if: ...` below lets this job run anyway and the
# download/copy steps gate themselves on the tag context.
needs: [android-release]
if: ${{ !failure() && !cancelled() }}
# Every lane must have SUCCEEDED, each named here (rule 177). Then the
# APK: tag and dev pushes build one and it must have succeeded; main
# pushes skip android-release and bundle the latest release's APK
# instead, so for main alone a skipped android-release is expected.
needs: [go, integration, web, android, govulncheck, android-release]
if: >-
${{
!cancelled()
&& needs.go.result == 'success'
&& needs.integration.result == 'success'
&& needs.web.result == 'success'
&& needs.android.result == 'success'
&& needs.govulncheck.result == 'success'
&& (needs.android-release.result == 'success'
|| (needs.android-release.result == 'skipped' && github.ref == 'refs/heads/main')) }}
runs-on: go-ci
container:
image: git.fabledsword.com/bvandeusen/ci-go:1.26
@@ -222,11 +693,16 @@ jobs:
- name: Checkout
uses: actions/checkout@v4
with:
# Full history + tags so non-tag :latest builds can resolve the
# latest release tag's commit count and reconstruct the bundled
# APK's exact versionName (see "Bundle latest release APK" below).
# Full history, and rule 149 names this specifically: any job that
# DERIVES the version name needs it, because a shallow clone changes
# what git-derived values resolve to WITHOUT failing — a too-low
# value, silently, with every lane green.
#
# This job was depth-1 while it took the version from GITHUB_REF. It
# now runs ci/version.sh itself, because with :<version> image tags
# gone the server's self-reported version is the only thing that says
# which build an image is.
fetch-depth: 0
fetch-tags: true
- name: Detect buildable project
id: guard
@@ -244,21 +720,68 @@ jobs:
if: steps.guard.outputs.ready == 'true'
shell: bash
run: |
if [[ "${GITHUB_REF}" == refs/tags/v* ]]; then
VERSION="${GITHUB_REF#refs/tags/}"
echo "args=-t ${IMAGE}:${VERSION} -t ${IMAGE}:latest" >> "$GITHUB_OUTPUT"
echo "version=${VERSION}" >> "$GITHUB_OUTPUT"
echo "::notice::Release build: ${VERSION} + latest"
else
# Main is the protected, post-PR-merge branch. Treat it as the
# rolling stable channel — every main push moves :latest.
# Pinned consumers can target :vYYYY.MM.DD; everyone else
# gets the newest main.
echo "args=-t ${IMAGE}:main -t ${IMAGE}:latest" >> "$GITHUB_OUTPUT"
echo "version=main" >> "$GITHUB_OUTPUT"
echo "::notice::Main-branch build: :main + :latest"
set -euo pipefail
# THE VERSION, and it is derived the same way on every ref — the
# branch decides the CHANNEL, never the version (family rule 149).
#
# This used to be three different things: the literal string "main"
# on main, "dev" on dev, and the tag name on a tag. None of them
# ordered, and the first two were the same string forever — two dev
# images eight weeks apart were indistinguishable in the UI. That
# mattered little while :vYYYY.MM.DD.HHMM existed to identify a
# build; with version image tags gone, this IS how an operator tells
# which build a container is running.
#
# `sed -n s///p` rather than `grep`: it exits 0 when nothing matches,
# so the empty check below is actually reachable. A grep here would
# kill the step at the assignment under the runner's pipefail — the
# exact bug that took down the first main build after the version
# rework.
VERSION="$(ci/version.sh HEAD | sed -n 's/^name=//p')"
if [ -z "${VERSION}" ]; then
echo "::error::could not derive a build version from ci/version.sh"
exit 1
fi
if [[ "${GITHUB_REF}" == refs/tags/v* ]]; then
# A release refreshes the CHANNEL and mints nothing else.
#
# The tag build exists to produce the signed APK and attach it to
# the release; the image it rebuilds is the SAME SOURCE as the main
# build minutes earlier, differing only in which APK is baked in.
# Rule 145 is explicit about that case: when the same source is
# rebuilt with different contents, publish the moving channel tag
# and never a commit-addressable one.
#
# :latest must move here rather than waiting for the next main
# push, or the channel would carry the PREVIOUS release's APK
# indefinitely — a channel that cannot refresh itself (rule 146).
CHANNEL=stable
echo "args=-t ${IMAGE}:latest" >> "$GITHUB_OUTPUT"
echo "::notice::Release build ${VERSION}: refreshing :latest around the new APK"
elif [[ "${GITHUB_REF}" == "refs/heads/dev" ]]; then
# The rolling test channel, and :dev ALONE — deliberately no
# per-commit tag. A rolling channel is rolling by definition, so a
# commit-addressable image here would be a rollback target nobody
# has ever pulled, accumulating in the registry forever. Recovery
# on dev is to fix forward.
CHANNEL=dev
echo "args=-t ${IMAGE}:dev" >> "$GITHUB_OUTPUT"
echo "::notice::Dev-branch build ${VERSION}: :dev"
else
# The production line: :latest tracks main's tip (rule 147) and
# :<sha> is the rollback unit (rule 145). Full 40-char SHA, matching
# the family's other repos, so a rollback target is addressable
# straight from the commit anyone is reading.
CHANNEL=stable
echo "args=-t ${IMAGE}:latest -t ${IMAGE}:${GITHUB_SHA}" >> "$GITHUB_OUTPUT"
echo "::notice::Main-branch build ${VERSION}: :latest + :${GITHUB_SHA}"
fi
echo "version=${VERSION}" >> "$GITHUB_OUTPUT"
echo "channel=${CHANNEL}" >> "$GITHUB_OUTPUT"
- name: Registry login
if: steps.guard.outputs.ready == 'true'
shell: bash
@@ -267,54 +790,57 @@ jobs:
| docker login git.fabledsword.com -u "${{ github.actor }}" --password-stdin
- name: Download signed APK artifact
# Tag pushes only — android-release just produced this. Non-tag
# builds take the "Bundle latest release APK" path below instead.
if: steps.guard.outputs.ready == 'true' && startsWith(github.ref, 'refs/tags/v')
# Consuming half of the pair — never actions/download-artifact. Same fork,
# same reason: upstream's client-side GHES check rejects this hostname
# before it connects. bvandeusen/download-artifact mirrors
# code.forgejo.org/forgejo/download-artifact.
#
# SHA below is that fork's `v6` tag. Match on @actions/artifact, NOT on
# the action's own version number — the two actions release on unrelated
# cadences, and download v5 would pair a ^2.3.2 client with this file's
# ^4.0.0 uploader. v6 is the tag whose bundled library major (^4.0.0) is
# the same one proven against this instance by the upload side.
# Deliberately NOT v7: it moves to node24 and upstream requires runner
# >= 2.327.1 for it, which act_runner does not claim to satisfy.
# Pinned, not tagged — the mirror auto-syncs every 8h.
uses: https://git.fabledsword.com/bvandeusen/download-artifact@8d4e9521a5f7e5f8b6351f341f719f9f45a92a3a
# Tag and dev pushes — android-release just produced this. Only `main`
# takes the "Bundle latest release APK" path below, because it is the
# one ref that moves a channel without building an APK of its own.
if: >-
steps.guard.outputs.ready == 'true' &&
(startsWith(github.ref, 'refs/tags/v') || github.ref == 'refs/heads/dev')
# Consuming half of the pair: stock download-artifact, which works here for
# the same reason as the upload (gitea/runner 3.x edits the GHES refusal
# out of the bundle; snippet #2271). v8 runs on node24, which every
# CI-runner image carries — the runner uses the image's own node.
uses: actions/download-artifact@v8
with:
name: minstrel-apk
path: client/
- name: Stage bundled APK + version sidecar
if: steps.guard.outputs.ready == 'true' && startsWith(github.ref, 'refs/tags/v')
if: >-
steps.guard.outputs.ready == 'true' &&
(startsWith(github.ref, 'refs/tags/v') || github.ref == 'refs/heads/dev')
shell: bash
env:
# Pulled from android-release.outputs.version_name so the
# sidecar string the server hands clients matches the
# versionName baked into the APK they're comparing against.
# All three pulled from android-release's outputs so the sidecar the
# server hands clients matches exactly what is baked into the APK
# they are comparing against.
APK_VERSION_NAME: ${{ needs.android-release.outputs.version_name }}
APK_VERSION_CODE: ${{ needs.android-release.outputs.version_code }}
APK_CHANNEL: ${{ needs.android-release.outputs.channel }}
run: |
set -euxo pipefail
# The artifact lands as `app-release.apk` (the original Gradle
# output name). The Dockerfile COPYs client/* into /app/client/
# and the server reads minstrel.apk + minstrel.apk.version.
mv client/app-release.apk client/minstrel.apk
echo "${APK_VERSION_NAME}" > client/minstrel.apk.version
printf '{"name":"%s","code":%s,"channel":"%s"}\n' \
"${APK_VERSION_NAME}" "${APK_VERSION_CODE}" "${APK_CHANNEL}" \
> client/minstrel.apk.version
cat client/minstrel.apk.version
ls -lh client/
- name: Bundle latest release APK (non-tag :latest builds)
# Main pushes don't build an APK, but they DO move :latest — so
# without this the in-app update channel would vanish from :latest
# until the next tag. Pull the most-recent release's signed APK and
# reconstruct its exact versionName (${TAG#v}.$(git rev-list --count
# TAG) — identical to android-release's formula) so the version
# sidecar the server hands clients matches the installed build.
# the sidecar published beside it, so what the server reports is what
# that build actually recorded rather than something re-derived here.
# Degrades to an empty client/ (404 update channel) — never a wrong
# version — if no release / APK asset / tag-count can be resolved.
if: steps.guard.outputs.ready == 'true' && !startsWith(github.ref, 'refs/tags/v')
# version — if no release or APK asset can be resolved. That
# degradation only actually works because the greps below carry
# `|| true`; under the runner's default pipefail a non-matching grep
# kills the step instead of falling through to the empty-case branch.
if: steps.guard.outputs.ready == 'true' && github.ref == 'refs/heads/main'
shell: bash
env:
CI_TOKEN: ${{ secrets.CI_TOKEN }}
@@ -326,26 +852,53 @@ jobs:
if [ -z "${REL_JSON}" ]; then
echo "::notice::no published release — image ships without bundled APK"; exit 0
fi
TAG="$(printf '%s' "${REL_JSON}" | grep -oP '"tag_name":\s*"\K[^"]+' | head -1)"
APK_URL="$(printf '%s' "${REL_JSON}" | grep -oP '"browser_download_url":\s*"\K[^"]+' | grep -E '\.apk$' | head -1)"
# `|| true` on every one of these, and it is load-bearing rather
# than defensive habit. The runner already invokes this shell as
# `bash -e -o pipefail`, so a pipeline whose grep matches NOTHING
# exits non-zero even though `head` succeeded — and the step dies at
# the assignment, before ever reaching the `if` written to handle the
# empty case. Every "degrades gracefully" branch below is unreachable
# without this.
TAG="$(printf '%s' "${REL_JSON}" | grep -oP '"tag_name":\s*"\K[^"]+' | head -1)" || true
APK_URL="$(printf '%s' "${REL_JSON}" | grep -oP '"browser_download_url":\s*"\K[^"]+' | grep -E '\.apk$' | head -1)" || true
if [ -z "${TAG}" ] || [ -z "${APK_URL}" ]; then
echo "::notice::latest release '${TAG:-?}' has no APK asset — image ships without bundled APK"; exit 0
fi
COUNT="$(git rev-list --count "${TAG}" 2>/dev/null || true)"
if [ -z "${COUNT}" ]; then
echo "::notice::could not resolve commit count for ${TAG} (tag not fetched?) — skipping APK bundle"; exit 0
fi
VERSION_NAME="${TAG#v}.${COUNT}"
curl -fsSL -H "Authorization: token ${CI_TOKEN}" -o client/minstrel.apk "${APK_URL}"
echo "${VERSION_NAME}" > client/minstrel.apk.version
echo "::notice::bundled release APK ${TAG} as version ${VERSION_NAME}"
# Take the version the release RECORDED rather than recomputing it.
# This used to re-derive the name from the tagged commit, which meant
# the formula lived in two files that had to be kept in step, and it
# could only ever recover the name — the ordering key is build-time
# minutes and does not exist anywhere after that build ends.
SIDECAR_URL="$(printf '%s' "${REL_JSON}" | grep -oP '"browser_download_url":\s*"\K[^"]+' | grep -E '\.apk\.version$' | head -1)" || true
if [ -n "${SIDECAR_URL}" ]; then
curl -fsSL -H "Authorization: token ${CI_TOKEN}" -o client/minstrel.apk.version "${SIDECAR_URL}"
cat client/minstrel.apk.version
else
# Releases published before sidecars were attached. Their name is
# still recoverable from the tag, but their ordering key genuinely
# is not — so it is reported ABSENT rather than guessed. A wrong
# key is an install the platform refuses; an absent one just tells
# the client to fall back to comparing names, which is exactly
# what those builds already do.
echo "::notice::release ${TAG} predates the version sidecar — bundling with name only, no ordering key"
printf '{"name":"%s","code":null,"channel":"stable"}\n' "${TAG#v}" > client/minstrel.apk.version
fi
echo "::notice::bundled release APK from ${TAG}"
ls -lh client/
- name: Build and push
if: steps.guard.outputs.ready == 'true'
# --pull: the Dockerfile's base images are floating tags (golang:1.26,
# debian:bookworm-slim). Without it the runner's daemon reuses
# whatever it cached, and the shipped binary can sit on a Go patch
# release govulncheck already flagged while the lane, which pulls
# fresh, reports clean.
run: |
docker buildx build \
docker buildx build --pull \
--build-arg MINSTREL_VERSION="${{ steps.tags.outputs.version }}" \
--build-arg MINSTREL_CHANNEL="${{ steps.tags.outputs.channel }}" \
--push ${{ steps.tags.outputs.args }} .
# Verifies a tag release actually ended up complete, and names the specific
@@ -356,8 +909,8 @@ jobs:
# `failure` with none executed and image-release showed `skipped`. The run was
# red, but the *release page rendered fine*, and `main`'s own push build had
# already moved `:latest`, so the code was deployable and nothing looked
# obviously wrong. The release was simply missing its APK and its immutable
# `:vYYYY.MM.DD` image, which is easy to skim past.
# obviously wrong. The release was simply missing its APK and its image,
# which is easy to skim past.
#
# This job cannot prevent that (the cause was a runner failing to launch, not
# anything in this file). What it does is turn an incomplete release into an
@@ -368,7 +921,7 @@ jobs:
# above did NOT succeed.
verify-release:
name: Verify release artifacts (tag releases only)
needs: [android-release, image-release]
needs: [android-release, release-assets, image-release]
if: ${{ always() && startsWith(github.ref, 'refs/tags/v') }}
runs-on: go-ci
container:
@@ -408,18 +961,30 @@ jobs:
# missing when v2026.08.07 had to be re-cut. `always()` on this job means
# it runs even when image-release failed, so without this the guard would
# cheerfully verify an incomplete release.
- name: Immutable image tag must exist
#
# This asserted `:${TAG}` — the :vYYYY.MM.DD.HHMM image — until
# 2026-09-10. Version image tags are no longer published (rule 145), so
# that assertion would now fail every release for a tag nothing mints.
# The rollback target it was really protecting is the :<sha> image, which
# main's own build published for this same commit before the tag was cut.
#
# Checking it here earns its keep twice over: it still catches an image
# push that silently did not happen, and it additionally proves the
# ORDERING — a tag cut on a commit whose main build never completed has
# no rollback target, and that is worth failing on rather than
# discovering during an incident.
- name: Rollback image must exist for the tagged commit
shell: bash
run: |
set -euo pipefail
TAG="${GITHUB_REF#refs/tags/}"
IMAGE="git.fabledsword.com/bvandeusen/minstrel"
echo "${{ secrets.CI_TOKEN }}" \
| docker login git.fabledsword.com -u "${{ github.actor }}" --password-stdin
if ! docker manifest inspect "${IMAGE}:${TAG}" > /dev/null 2>&1; then
echo "::error::image ${IMAGE}:${TAG} was never pushed — the release tag has no immutable image, so there is nothing to pin or roll back to. Re-run this workflow run."
if ! docker manifest inspect "${IMAGE}:${GITHUB_SHA}" > /dev/null 2>&1; then
echo "::error::image ${IMAGE}:${GITHUB_SHA} does not exist — this commit has no rollback target."
echo "::error::That image is published by the MAIN build of this commit, not by the tag build. If main's build never ran or failed, fix that first; a release whose commit cannot be rolled back to is the thing this check exists to refuse."
exit 1
fi
echo "::notice::image verified: ${IMAGE}:${TAG}"
echo "::notice::rollback target verified: ${IMAGE}:${GITHUB_SHA}"
-150
View File
@@ -1,150 +0,0 @@
name: test-go
# Go server: vet + golangci-lint + short race tests. Runs on push to
# dev/main and PRs to main, scoped to Go-side files only — web-only or
# Flutter-only diffs don't trigger this workflow.
#
# Two jobs: `test` (fast — vet + lint + `go test -short -race`, no DB) and
# `integration` (full `go test -race` against an ephemeral Postgres).
#
# Integration-job DB wiring follows the act_runner shared-daemon pattern:
# the runner's Docker daemon also runs the operator's dev compose stack,
# so service containers get NO published ports (collision) and no
# service-name DNS. We discover the service container by the job-scoped
# name filter via the mounted docker socket and reach it by bridge IP.
# The exactly-one assertion is a hard guard — pointing tests at the dev
# Postgres would truncate it (the disaster Fable #339 exists to prevent).
#
# `web/build/` has a committed placeholder index.html so go:embed succeeds
# without needing the SPA to be freshly built. Real builds happen in
# release.yml (container) and locally during dev.
on:
push:
branches: [dev, main]
paths:
- '**/*.go'
- 'go.mod'
- 'go.sum'
- 'sqlc.yaml'
- 'Makefile'
- 'internal/**'
- 'cmd/**'
- '.golangci.yml'
- '.gitea/workflows/test-go.yml'
# pull_request trigger intentionally omitted — see test-web.yml for
# the rationale (single-author repo, push covers PR-merge equivalent).
concurrency:
group: ${{ github.workflow }}-${{ github.ref }}
cancel-in-progress: true
jobs:
test:
runs-on: go-ci
container:
image: git.fabledsword.com/bvandeusen/ci-go:1.26
steps:
- name: Checkout
uses: actions/checkout@v4
- name: Toolchain versions
run: |
go version
golangci-lint --version
- name: Generated code matches queries (sqlc)
run: make verify-generate
- name: go vet
run: go vet ./...
- name: golangci-lint
run: golangci-lint run ./...
- name: go test (short, race)
run: go test -short -race ./...
integration:
runs-on: go-ci
container:
image: git.fabledsword.com/bvandeusen/ci-go:1.26
services:
postgres:
image: postgres:16-alpine
env:
POSTGRES_USER: minstrel
POSTGRES_PASSWORD: minstrel
POSTGRES_DB: minstrel_test
# No `ports:` — the runner shares the operator's dev compose
# Docker daemon; publishing a fixed host port collides.
steps:
- name: Checkout
uses: actions/checkout@v4
- name: Integration suite (discover service by bridge IP, migrate, test)
run: |
set -eux
# Discover THIS job's Postgres service container via the
# mounted docker socket. act_runner attaches the job
# container and its service container(s) to a shared per-job
# network, so scope discovery to a postgres that sits on a
# network THIS job container is also on. The old
# `--filter name=integration` matched EVERY concurrent
# integration run's postgres (a dev push + the main-merge run
# overlap → 2 candidates → false "expected exactly 1" abort).
# The operator's dev compose `minstrel-postgres-*` is never on
# this job's network; skip it explicitly as belt-and-suspenders
# (a wrong target would truncate real data).
SELF=$(cat /etc/hostname)
SELF_NETS=$(docker inspect -f '{{range $k,$v := .NetworkSettings.Networks}}{{$k}} {{end}}' "$SELF")
test -n "$SELF_NETS"
echo "self ($SELF) networks: $SELF_NETS"
PG_ID=""
PG_NAME=""
for cid in $(docker ps --filter "ancestor=postgres:16-alpine" -q); do
nm=$(docker inspect -f '{{.Name}}' "$cid" | sed 's#^/##')
case "$nm" in *minstrel-postgres*|*_postgres_*) continue ;; esac
for net in $(docker inspect -f '{{range $k,$v := .NetworkSettings.Networks}}{{$k}} {{end}}' "$cid"); do
case " $SELF_NETS " in *" $net "*) PG_ID="$cid"; PG_NAME="$nm"; break 2 ;; esac
done
done
test -n "$PG_ID" || { echo "FATAL: no postgres service container on this job's network (self nets: $SELF_NETS)"; exit 1; }
echo "selected postgres: $PG_ID $PG_NAME"
PG_IP=$(docker inspect -f '{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}' "$PG_ID")
test -n "$PG_IP"
export MINSTREL_TEST_DATABASE_URL="postgres://minstrel:minstrel@${PG_IP}:5432/minstrel_test?sslmode=disable"
# Wait for Postgres to accept TCP (no health-check dependency).
for i in $(seq 1 60); do (echo > "/dev/tcp/${PG_IP}/5432") 2>/dev/null && break; sleep 2; done
# Relax durability on the throwaway CI Postgres. Our test pattern
# is dbtest.ResetDB → TRUNCATE … RESTART IDENTITY CASCADE before
# every test, and the per-TRUNCATE commit fsync is the dominant
# cost of the integration suite. The CI DB is rebuilt every run so
# fsync / full_page_writes / synchronous_commit buy nothing. Apply
# via docker exec because:
# - The act_runner `services:` block can't override the container
# command, so `postgres -c fsync=off` at boot isn't an option.
# - ALTER SYSTEM cannot run inside a transaction; psql -c
# auto-commits each statement, which is what we need.
# - fsync / full_page_writes are sighup GUCs and
# synchronous_commit is user-context, so pg_reload_conf() picks
# all three up with no restart.
# Non-fatal: a perms surprise degrades to "slower", never red CI.
docker exec "$PG_ID" psql -U minstrel -d minstrel_test \
-c "ALTER SYSTEM SET fsync = off" \
-c "ALTER SYSTEM SET synchronous_commit = off" \
-c "ALTER SYSTEM SET full_page_writes = off" \
-c "SELECT pg_reload_conf()" \
|| echo "WARN: durability relax failed; continuing"
# Apply embedded migrations to the fresh test DB, then run the
# full suite (no -short → integration tests execute). -p 1:
# every integration package TRUNCATEs the one shared test DB;
# concurrent package binaries → TRUNCATE deadlocks. Serialize
# package execution (the documented local invocation too).
MINSTREL_DATABASE_URL="$MINSTREL_TEST_DATABASE_URL" go run ./cmd/minstrel migrate
go test -p 1 -race ./...
-44
View File
@@ -1,44 +0,0 @@
name: test-web
# Web SPA: vitest + svelte-check. Runs on push to dev/main only —
# the `pull_request` trigger is intentionally omitted because every
# branch on this repo is local-only (no fork PRs), so the dev push
# fully covers what a PR run would re-execute. Keeping both events
# doubled CI cost on every commit.
on:
push:
branches: [dev, main]
paths:
- 'web/**'
- '.gitea/workflows/test-web.yml'
# Cancel an earlier in-flight run for the same ref when a newer
# commit arrives. With cancel-in-progress, rapid re-pushes don't
# pile up zombie runs.
concurrency:
group: ${{ github.workflow }}-${{ github.ref }}
cancel-in-progress: true
jobs:
test:
runs-on: go-ci
container:
image: git.fabledsword.com/bvandeusen/ci-go:1.26
defaults:
run:
working-directory: web
steps:
- name: Checkout
uses: actions/checkout@v4
- name: Install deps
run: npm ci
- name: Type-check + svelte-check
run: npm run check
- name: Vitest
run: npm test
+5
View File
@@ -12,6 +12,11 @@
# Test binary, built with `go test -c`
*.test
# `make build` output. bin/minstrel was tracked until 2026-09-10 — an 18 MB
# binary committed by accident, last refreshed by a commit about web test
# mocks, and re-dirtied by every local build since.
bin/
# Bundled Android APK + version sidecar (#397). Populated by CI for
# tag releases; never committed. README in client/ explains the flow.
client/minstrel.apk
+21 -6
View File
@@ -7,7 +7,7 @@ RUN npm ci
COPY web/ ./
RUN npm run build
FROM golang:1.25-bookworm AS builder
FROM golang:1.26-bookworm AS builder
WORKDIR /src
COPY go.mod go.sum ./
RUN go mod download
@@ -15,17 +15,32 @@ COPY . .
# Overwrite the committed placeholder with the freshly-built SPA assets.
COPY --from=web /web/build ./web/build
ENV CGO_ENABLED=0
# Version stamping: release.yml passes the git tag via MINSTREL_VERSION
# build-arg; local `docker build` falls back to "dev". Surfaced at
# /healthz for operator-side image-version verification.
# Version stamping. release.yml passes the DERIVED version name
# (YYYY.MM.DD.HHMM) and the lane's channel; a local `docker build` falls back
# to "dev"/"local". Both are surfaced at /healthz.
#
# These are two values on purpose (family rule 149): the same commit built on
# dev and on main reports the same NAME and differs only in CHANNEL. Folding
# the channel into the version string is what the rule forbids — the version
# used to BE the channel word here ("main"/"dev"), which meant two dev images
# eight weeks apart were indistinguishable.
ARG MINSTREL_VERSION=dev
ARG MINSTREL_CHANNEL=local
RUN go build -trimpath \
-ldflags="-s -w -X 'git.fabledsword.com/bvandeusen/minstrel/internal/server.ServerVersion=${MINSTREL_VERSION}'" \
-ldflags="-s -w \
-X 'git.fabledsword.com/bvandeusen/minstrel/internal/server.ServerVersion=${MINSTREL_VERSION}' \
-X 'git.fabledsword.com/bvandeusen/minstrel/internal/server.ServerChannel=${MINSTREL_CHANNEL}'" \
-o /out/minstrel ./cmd/minstrel
FROM debian:bookworm-slim
# ffmpeg: duration probes and the exact-tier audio hash (a SHA-256 of the
# encoded audio packets, so no decode). libchromaprint-tools: fpcalc, the
# acoustic fingerprint that tells the same recording at two bitrates apart
# from two different recordings (M400). Both are baked in at build time so a
# deployed instance never fetches either (rule 164); fpcalc is shelled out
# rather than bound because CGO_ENABLED=0 above rules out cgo.
RUN apt-get update \
&& apt-get install -y --no-install-recommends ca-certificates ffmpeg \
&& apt-get install -y --no-install-recommends ca-certificates ffmpeg libchromaprint-tools \
&& rm -rf /var/lib/apt/lists/*
RUN groupadd --system --gid 1000 minstrel \
+31 -11
View File
@@ -34,11 +34,18 @@ Minstrel is not affiliated with or endorsed by Lidarr, ListenBrainz, MusicBrainz
services:
minstrel:
image: git.fabledsword.com/bvandeusen/minstrel:latest
# Reachable from your LAN at http://<host>:4533. If this host faces the
# internet, bind it to 127.0.0.1 and put an HTTPS proxy in front instead:
# see docs/hosting.md.
ports: ['4533:4533']
volumes:
# Your music library. Point ./music at wherever your audio files
# live. Mounted read-only — Minstrel never writes to your library.
- ./music:/music:ro
# live. Writable, because Minstrel deletes a file when an admin asks
# it to (for example, quarantine's "Delete file"). It never moves,
# renames or retags anything. The container runs as uid 1000, so that
# user needs write access to the folders. Mount it :ro to forbid even
# deletes: those actions then refuse, say why, and delete nothing.
- ./music:/music
# Generated data: playlist cover collages, artist art, caches.
# The path must match MINSTREL_STORAGE_DATA_DIR, which the image
# sets to /app/data — keep this mount on /app/data or your cache
@@ -47,7 +54,7 @@ services:
environment:
MINSTREL_DATABASE_URL: postgres://minstrel:minstrel@db:5432/minstrel?sslmode=disable
# Colon-separated library roots to scan; must match the container
# path of the read-only music mount above (/music here).
# path of the music mount above (/music here).
MINSTREL_LIBRARY_SCAN_PATHS: /music
depends_on: [db]
@@ -72,9 +79,9 @@ docker compose up -d
## First run
With the stack up, a handful of in-app steps get you to a working library. Use your own host in place of `localhost` if you're reaching the server over a LAN/VPN address (plain `http://` is fine — no TLS required).
With the stack up, a handful of in-app steps get you to a working library. Use your own host in place of `localhost` if you're reaching the server over a LAN/VPN address. Plain `http://` is fine on a network you trust; a server reachable from the internet belongs behind HTTPS, which [docs/hosting.md](docs/hosting.md) walks through.
**1. Create your admin account.** Visit `http://localhost:4533/register`. The first account on a fresh instance is automatically the administrator; later users join through the same form or an invite token (step 5).
**1. Create your admin account.** Visit `http://localhost:4533/register`. The first account on a fresh instance becomes the administrator, and creating it asks for the **setup token** the server prints in its log (`docker compose logs minstrel | grep setup_token`), so nobody else can claim a newly exposed server first. Later users join through the same form or an invite token (step 5).
<a href="docs/screenshots/register.png"><img src="docs/screenshots/register.png" width="320" alt="Creating the first (admin) account on a fresh instance"></a>
@@ -96,6 +103,8 @@ With the stack up, a handful of in-app steps get you to a working library. Use y
For the full configuration surface, see [`config.example.yaml`](./config.example.yaml).
Hosting Minstrel on the internet: see [docs/hosting.md](docs/hosting.md). What Minstrel does to protect accounts, and why: [docs/security.md](docs/security.md).
## Configuration
Most operators only need the env vars in the quickstart above. A few extras worth knowing:
@@ -112,11 +121,21 @@ Most operational keys have a `MINSTREL_<SECTION>_<FIELD>` env override. Recommen
Image tags (`git.fabledsword.com/bvandeusen/minstrel:<tag>`):
- `:latest` — the newest blessed image. Moves on every `main` push **and** every release. Recommended for most operators.
- `:vYYYY.MM.DD` — immutable per-day release tags. Pin one of these for a deployment you don't want moving under you. (Per-day CalVer — no trailing patch digit; a same-day re-cut moves the tag forward.)
- `:main` — the rolling post-merge tip. Same image as `:latest` at push time; choose it if you want to track `main` explicitly rather than the release line.
- `:latest` — production. Tracks `main`'s tip and moves on every `main` push and every release. What most operators should run.
- `:<commit-sha>` — the rollback unit. Every `main` push publishes one, so any production commit is addressable without a release ceremony. Immutable: a given SHA tag is never re-pushed. Pin one if you need a deployment that cannot change under you, and use it to roll back.
- `:dev` — the rolling test channel, rebuilt on every push to `dev` and carrying its own freshly-built Android APK. Run this to try something before it ships. It moves constantly, has no per-commit tag, and its only recovery path is forward — if a `:dev` image is broken, the fix is the next push, not a rollback.
Every `:latest` and every `:vYYYY.MM.DD` bundles the current signed Android APK, so the in-app update channel is always live. Database migrations run automatically at startup; rollbacks require restoring a Postgres dump.
That is the whole tag map. **There are no version-numbered image tags**, and no `:main`. Git and the build's own self-reported version answer "which build is this" — the Settings page shows it, and so does `/healthz`. Release *tags* in git are still `vYYYY.MM.DD.HHMM`; they name a changelog entry and the APK attached to it, not an image.
Rolling back to `:<commit-sha>` pins the **server code** at that commit — not the server-and-app pair. The Android APK is baked in at image build time, so a SHA image carries whichever app was current when that commit was built, which may be older than what `:latest` bundles now. If both halves matter, check what the image bundles rather than trusting the tag's name.
Every `:latest`, `:<commit-sha>` and `:dev` bundles a signed Android APK, so the in-app update channel is always live. All are signed with the same key, so a phone can move between the stable and dev channels without uninstalling — point it at a `:dev` server and the in-app updater offers that channel's build.
The app reports which channel it is on alongside its version, and decides whether an update is available using the build's ordering key rather than its displayed name — the same value Android installs by, so an offer it makes is one the platform will accept.
Database migrations run automatically at startup; rollbacks require restoring a Postgres dump.
Releases up to 2026-09-10 also published a `:vYYYY.MM.DD[.HHMM]` image tag. Those images still exist and still work — they are simply not extended.
## Specs
@@ -140,7 +159,8 @@ Two concurrent dev processes:
truncates your dev `minstrel` data (admin user, library, likes). It
brings up the compose Postgres and creates the test DB if missing.
- CI runs both: a fast `go test -short -race` gate plus an integration
job with its own ephemeral Postgres (`.gitea/workflows/test-go.yml`).
job with its own ephemeral Postgres (the `integration` lane in
`.gitea/workflows/release.yml`, which also gates every image publish).
### Production build
@@ -150,7 +170,7 @@ Two concurrent dev processes:
- Day-to-day work happens on `dev` (or feature branches merged into `dev`).
- `main` is **protected** — changes land via PR from `dev`.
- Releases are cut by tagging `v*` off `main`; the release workflow builds and pushes the container image to the Gitea registry.
- Releases are cut by tagging `v*` off `main`; the release workflow builds the signed APK, attaches it to the release, and refreshes `:latest` around it.
Task and milestone tracking: Fable (`Minstrel` project, id 12).
+23 -8
View File
@@ -21,13 +21,24 @@ android {
applicationId = "com.fabledsword.minstrel"
minSdk = 26
targetSdk = 36
// versionName / versionCode are released-build values injected by
// CI from the git tag + commit count. Local / debug builds fall
// back to "dev" so the About card reads honestly. Releases ship
// versionName="YYYY.MM.DD.<commits>" (e.g. "2026.06.02.142") and
// versionCode=<commits>, which is monotonic forever and lets the
// shared isVersionNewer comparator distinguish two same-day
// re-cuts (the iteration suffix differs).
// versionName / versionCode are released-build values injected by CI.
// Local / debug builds fall back to "dev" so the About card reads
// honestly.
//
// versionName is "YYYY.MM.DD.HHMM" from the COMMIT's timestamp, so
// every lane building this source reports the same string and the
// channel is the only thing that differs between them.
//
// versionCode is minutes since 2020-01-01 at BUILD time. It is the
// value the platform decides installs by, so it must be monotonic by
// construction.
//
// This comment used to say versionCode was a commit count and that it
// was "monotonic forever". It was neither — a commit count runs ahead
// on `dev`, so a dev build outranked the `main` release meant to
// replace it and Android refused the install as a downgrade. Worth
// knowing the claim was here, stated as a reassurance, while the bug
// it denied was live.
val versionNameOverride =
(project.findProperty("MINSTREL_VERSION_NAME") as String?)?.takeIf { it.isNotBlank() }
val versionCodeOverride =
@@ -61,9 +72,13 @@ android {
getDefaultProguardFile("proguard-android-optimize.txt"),
"proguard-rules.pro",
)
// Signed with the release key or not at all. Falling back to the
// debug key made a missing secret into a published APK that no
// install could ever update (family idea #5103, practice 2). An
// unsigned build installs nowhere, so the gap shows at once.
signingConfig =
if (System.getenv("ANDROID_KEYSTORE_PATH").isNullOrEmpty()) {
signingConfigs.getByName("debug")
null
} else {
signingConfigs.getByName("release")
}
+27
View File
@@ -8,6 +8,10 @@
<uses-permission android:name="android.permission.FOREGROUND_SERVICE" />
<uses-permission android:name="android.permission.FOREGROUND_SERVICE_MEDIA_PLAYBACK" />
<uses-permission android:name="android.permission.POST_NOTIFICATIONS" />
<!-- Notifications when the app is closed (M489 #5347): the delivery
service, and starting it again after a reboot or an update. -->
<uses-permission android:name="android.permission.FOREGROUND_SERVICE_SPECIAL_USE" />
<uses-permission android:name="android.permission.RECEIVE_BOOT_COMPLETED" />
<!-- In-app self-update. REQUEST_INSTALL_PACKAGES lets us hand an APK to the
platform installer at all; UPDATE_PACKAGES_WITHOUT_USER_ACTION (API 31+)
is what lets that install happen with NO confirm dialog. The platform
@@ -57,6 +61,29 @@
</intent-filter>
</service>
<!-- Keeps the process alive so notifications arrive with the app
closed. specialUse: dataSync is stopped after six hours on
Android 15, and shortService after three minutes. -->
<service
android:name=".notifications.delivery.DeliveryService"
android:exported="false"
android:foregroundServiceType="specialUse">
<property
android:name="android.app.PROPERTY_SPECIAL_USE_FGS_SUBTYPE"
android:value="Maintains the connection to the user's own Minstrel server that delivers their notifications, in place of a third-party push service." />
</service>
<!-- Starts delivery after a reboot or an update; both broadcasts may
start a foreground service from the background. -->
<receiver
android:name=".notifications.delivery.BootReceiver"
android:exported="true">
<intent-filter>
<action android:name="android.intent.action.BOOT_COMPLETED" />
<action android:name="android.intent.action.MY_PACKAGE_REPLACED" />
</intent-filter>
</receiver>
<!-- The FileProvider that used to live here existed solely to expose the
downloaded update APK as a content:// URI for the old ACTION_VIEW
install intent. A PackageInstaller session takes a stream instead,
@@ -1,10 +1,14 @@
package com.fabledsword.minstrel
import android.Manifest
import android.content.Intent
import android.content.pm.PackageManager
import android.os.Build
import android.os.Bundle
import androidx.activity.ComponentActivity
import androidx.activity.compose.setContent
import androidx.activity.enableEdgeToEdge
import androidx.activity.result.contract.ActivityResultContracts
import androidx.compose.foundation.layout.Box
import androidx.compose.foundation.layout.fillMaxSize
import androidx.compose.material3.CircularProgressIndicator
@@ -16,9 +20,12 @@ import androidx.compose.runtime.LaunchedEffect
import androidx.compose.runtime.getValue
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.core.content.ContextCompat
import androidx.hilt.navigation.compose.hiltViewModel
import androidx.lifecycle.compose.collectAsStateWithLifecycle
import androidx.lifecycle.lifecycleScope
import androidx.navigation.compose.rememberNavController
import com.fabledsword.minstrel.auth.AuthStore
import com.fabledsword.minstrel.auth.ui.AuthGateViewModel
import com.fabledsword.minstrel.cache.CachedTrackIds
import com.fabledsword.minstrel.connectivity.LocalServerHealth
@@ -27,7 +34,10 @@ import com.fabledsword.minstrel.connectivity.NetworkStatusController
import com.fabledsword.minstrel.nav.DetailSeedCache
import com.fabledsword.minstrel.nav.LocalDetailSeedCache
import com.fabledsword.minstrel.nav.MinstrelNavGraph
import com.fabledsword.minstrel.nav.Notifications
import com.fabledsword.minstrel.nav.NowPlaying
import com.fabledsword.minstrel.notifications.delivery.deliveryWanted
import com.fabledsword.minstrel.notifications.ui.routeForLink
import com.fabledsword.minstrel.shared.widgets.LocalCachedTrackIds
import com.fabledsword.minstrel.theme.MinstrelTheme
import com.fabledsword.minstrel.theme.ThemePreferenceViewModel
@@ -35,6 +45,10 @@ import dagger.hilt.android.AndroidEntryPoint
import kotlinx.coroutines.flow.MutableStateFlow
import kotlinx.coroutines.flow.StateFlow
import kotlinx.coroutines.flow.asStateFlow
import kotlinx.coroutines.flow.combine
import kotlinx.coroutines.flow.first
import kotlinx.coroutines.flow.map
import kotlinx.coroutines.launch
import javax.inject.Inject
@AndroidEntryPoint
@@ -42,41 +56,72 @@ class MainActivity : ComponentActivity() {
@Inject lateinit var seedCache: DetailSeedCache
@Inject lateinit var cachedTrackIds: CachedTrackIds
@Inject lateinit var serverHealth: NetworkStatusController
@Inject lateinit var authStore: AuthStore
// Flipped to true when the user taps the media notification (or
// any other entry point that asks for the full player). The App
// composable observes this, navigates to NowPlaying once the
// NavHost is ready, then calls back to reset the flag so the
// navigation doesn't re-fire on the next recomposition.
private val pendingOpenNowPlaying = MutableStateFlow(false)
// Set when the user taps a notification: the media one asks for the full
// player, a Minstrel notice for what it is about. The App composable
// navigates there once the NavHost is ready, then calls back to clear it
// so the navigation doesn't re-fire on the next recomposition.
private val pendingRoute = MutableStateFlow<Any?>(null)
// The answer needs no handling: the system remembers it, and the
// notification settings screen reads it on every resume.
private val askToNotify = registerForActivityResult(ActivityResultContracts.RequestPermission()) { }
override fun onCreate(savedInstanceState: Bundle?) {
super.onCreate(savedInstanceState)
enableEdgeToEdge()
consumeOpenNowPlayingIntent(intent)
consumeRouteIntent(intent)
askToNotifyOnceWanted()
setContent {
App(
seedCache = seedCache,
cachedTrackIds = cachedTrackIds,
serverHealth = serverHealth,
pendingOpenNowPlaying = pendingOpenNowPlaying.asStateFlow(),
onOpenedNowPlaying = { pendingOpenNowPlaying.value = false },
pendingRoute = pendingRoute.asStateFlow(),
onOpenedRoute = { pendingRoute.value = null },
)
}
}
override fun onNewIntent(intent: Intent) {
super.onNewIntent(intent)
consumeOpenNowPlayingIntent(intent)
consumeRouteIntent(intent)
}
private fun consumeOpenNowPlayingIntent(intent: Intent?) {
if (intent?.getBooleanExtra(EXTRA_OPEN_NOW_PLAYING, false) == true) {
pendingOpenNowPlaying.value = true
private fun consumeRouteIntent(intent: Intent?) {
if (intent == null) return
if (intent.getBooleanExtra(EXTRA_OPEN_NOW_PLAYING, false)) {
pendingRoute.value = NowPlaying
// Strip the extra so a subsequent config-change recreation
// doesn't re-trigger the navigation.
intent.removeExtra(EXTRA_OPEN_NOW_PLAYING)
}
intent.getStringExtra(EXTRA_NOTIFICATION_LINK)?.let { link ->
// A notice the app has no screen for, or a pile of them, opens
// the inbox.
pendingRoute.value = routeForLink(link) ?: Notifications
intent.removeExtra(EXTRA_NOTIFICATION_LINK)
}
}
/**
* Android 13+ asks before an app may post notifications (M489 #5347).
* Asked once delivery is wanted (signed in, background delivery on),
* each launch until answered: after a second "no" the system stops
* showing the prompt by itself, and Settings → Notifications links to
* the system page.
*/
private fun askToNotifyOnceWanted() {
if (Build.VERSION.SDK_INT < Build.VERSION_CODES.TIRAMISU) return
lifecycleScope.launch {
val signedIn = authStore.sessionCookie.map { !it.isNullOrEmpty() }
combine(signedIn, authStore.backgroundDelivery, ::deliveryWanted).first { it }
val permission = Manifest.permission.POST_NOTIFICATIONS
val granted = ContextCompat.checkSelfPermission(this@MainActivity, permission) ==
PackageManager.PERMISSION_GRANTED
if (!granted) askToNotify.launch(permission)
}
}
companion object {
@@ -84,6 +129,10 @@ class MainActivity : ComponentActivity() {
* so a media-notification tap lands on the full NowPlaying screen
* instead of whatever shell route MainActivity last rendered. */
const val EXTRA_OPEN_NOW_PLAYING = "com.fabledsword.minstrel.action.OPEN_NOW_PLAYING"
/** PendingIntent extra on a Minstrel notice: the web path it links to,
* or empty for a pile, which opens the inbox. */
const val EXTRA_NOTIFICATION_LINK = "com.fabledsword.minstrel.action.NOTIFICATION_LINK"
}
}
@@ -92,15 +141,15 @@ private fun App(
seedCache: DetailSeedCache,
cachedTrackIds: CachedTrackIds,
serverHealth: NetworkStatusController,
pendingOpenNowPlaying: StateFlow<Boolean>,
onOpenedNowPlaying: () -> Unit,
pendingRoute: StateFlow<Any?>,
onOpenedRoute: () -> Unit,
themeVm: ThemePreferenceViewModel = hiltViewModel(),
gate: AuthGateViewModel = hiltViewModel(),
) {
val theme by themeVm.themeMode.collectAsStateWithLifecycle()
val cached by cachedTrackIds.ids.collectAsStateWithLifecycle()
val health: ServerHealth by serverHealth.state.collectAsStateWithLifecycle()
val pending by pendingOpenNowPlaying.collectAsStateWithLifecycle()
val pending by pendingRoute.collectAsStateWithLifecycle()
MinstrelTheme(darkOverride = theme.toDarkOverride()) {
CompositionLocalProvider(
LocalDetailSeedCache provides seedCache,
@@ -119,16 +168,14 @@ private fun App(
// Queue / unauthenticated) bypass the shell entirely.
val navController = rememberNavController()
// Honour a pending notification-tap once the NavHost is
// mounted. launchSingleTop avoids stacking copies of
// NowPlaying if the user taps the notification while
// already on it; the callback clears the flag so a later
// recomposition (config change, theme switch) doesn't
// re-navigate.
// mounted. launchSingleTop avoids stacking copies of a
// screen if the user taps the notification while already
// on it; the callback clears it so a later recomposition
// (config change, theme switch) doesn't re-navigate.
LaunchedEffect(pending, navController) {
if (pending) {
navController.navigate(NowPlaying) { launchSingleTop = true }
onOpenedNowPlaying()
}
val route = pending ?: return@LaunchedEffect
navController.navigate(route) { launchSingleTop = true }
onOpenedRoute()
}
MinstrelNavGraph(
navController = navController,
@@ -16,6 +16,7 @@ import com.fabledsword.minstrel.diagnostics.DiagnosticsUploader
import com.fabledsword.minstrel.events.EventsStream
import com.fabledsword.minstrel.events.LiveEventsDispatcher
import com.fabledsword.minstrel.metadata.FreshnessSweeper
import com.fabledsword.minstrel.notifications.delivery.DeliveryLauncher
import com.fabledsword.minstrel.player.AudioPrefetcher
import com.fabledsword.minstrel.player.CoverPrefetcher
import com.fabledsword.minstrel.player.PlayEventsReporter
@@ -75,6 +76,14 @@ class MinstrelApplication :
*/
@Suppress("unused") @Inject lateinit var mutationReplayer: MutationReplayer
/**
* Same construct-the-singleton trick — DeliveryLauncher starts and stops
* the background-delivery service and runs the notification catch-up on
* every nudge and reconnect (M489 #5347). Without this @Inject no phone
* notification would ever be posted.
*/
@Suppress("unused") @Inject lateinit var deliveryLauncher: DeliveryLauncher
/**
* Same construct-the-singleton trick — PlayEventsReporter's init
* block subscribes to PlayerController.uiState and reports the
@@ -15,10 +15,14 @@ import androidx.compose.material3.HorizontalDivider
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.OutlinedButton
import androidx.compose.material3.Scaffold
import androidx.compose.material3.SnackbarHost
import androidx.compose.material3.SnackbarHostState
import androidx.compose.material3.Text
import androidx.compose.material3.TextButton
import androidx.compose.runtime.Composable
import androidx.compose.runtime.LaunchedEffect
import androidx.compose.runtime.getValue
import androidx.compose.runtime.remember
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.text.style.TextOverflow
@@ -42,6 +46,12 @@ fun AdminQuarantineScreen(
viewModel: AdminQuarantineViewModel = hiltViewModel(),
) {
val state by viewModel.uiState.collectAsStateWithLifecycle()
val snackbarHostState = remember { SnackbarHostState() }
LaunchedEffect(Unit) {
viewModel.transientMessages.collect { msg ->
snackbarHostState.showSnackbar(msg)
}
}
Scaffold(
contentWindowInsets = ShellContentWindowInsets,
modifier = Modifier.fillMaxSize(),
@@ -53,6 +63,7 @@ fun AdminQuarantineScreen(
onBack = { navController.popBackStack() },
)
},
snackbarHost = { SnackbarHost(snackbarHostState) },
) { inner ->
PullToRefreshScaffold(
onRefresh = { viewModel.refresh().join() },
@@ -10,10 +10,13 @@ import com.fabledsword.minstrel.events.EventsStream
import com.fabledsword.minstrel.models.AdminQuarantineItemRef
import dagger.hilt.android.lifecycle.HiltViewModel
import kotlinx.coroutines.Job
import kotlinx.coroutines.channels.Channel
import kotlinx.coroutines.flow.Flow
import kotlinx.coroutines.flow.MutableStateFlow
import kotlinx.coroutines.flow.StateFlow
import kotlinx.coroutines.flow.asStateFlow
import kotlinx.coroutines.flow.filter
import kotlinx.coroutines.flow.receiveAsFlow
import kotlinx.coroutines.launch
import javax.inject.Inject
@@ -34,6 +37,15 @@ class AdminQuarantineViewModel @Inject constructor(
private val internal = MutableStateFlow<AdminQuarantineUiState>(AdminQuarantineUiState.Loading)
val uiState: StateFlow<AdminQuarantineUiState> = internal.asStateFlow()
/**
* One-shot messages for the screen's snackbar. A failed action has to say
* why: the row quietly reappearing reads as a glitch, and for a Delete
* file refused by a read-only library it hides the one thing the
* operator can fix (#3918).
*/
private val transientMessagesChannel = Channel<String>(Channel.BUFFERED)
val transientMessages: Flow<String> = transientMessagesChannel.receiveAsFlow()
init {
refresh()
viewModelScope.launch {
@@ -86,8 +98,9 @@ class AdminQuarantineViewModel @Inject constructor(
try {
action(trackId)
} catch (
@Suppress("TooGenericExceptionCaught", "SwallowedException") e: Throwable,
@Suppress("TooGenericExceptionCaught") e: Throwable,
) {
transientMessagesChannel.trySend(ErrorCopy.fromThrowable(e))
refresh()
}
}
@@ -50,7 +50,14 @@ class BaseUrlInterceptor @Inject constructor(
.port(baseUrl.port)
.build()
} ?: original.url
return chain.proceed(original.newBuilder().url(rewritten).build())
return chain.proceed(
original.newBuilder()
.url(rewritten)
// Lets CleartextGuardInterceptor tell server requests from
// external fetches once the placeholder host is gone.
.tag(MinstrelServerRequest::class.java, MinstrelServerRequest)
.build(),
)
}
companion object {
@@ -0,0 +1,84 @@
package com.fabledsword.minstrel.api
import okhttp3.Interceptor
import okhttp3.Response
import java.io.IOException
import java.net.Inet4Address
import java.net.Inet6Address
import java.net.InetAddress
/**
* Marks a request as bound for the Minstrel server, set by
* [BaseUrlInterceptor] when it retargets the placeholder host. Those are the
* requests that carry the session cookie and the password.
*/
object MinstrelServerRequest
/**
* Plain `http://` to the Minstrel server is allowed only when the connection
* actually lands on a private address (family security baseline #5105,
* practice 13).
*
* Cleartext stays permitted app-wide for LAN servers and UPnP
* (network_security_config.xml, #2439), but a password or session cookie sent
* over plain HTTP to a public address can be read by anyone on the path.
*
* Checked per connection, on the address the socket really reached, not on
* the URL when it was typed: a name that resolved to the home network when it
* was entered resolves to a public address once the phone leaves home, and
* that is exactly when the password would go out in the clear. A network
* interceptor runs after the connection is made and before any request byte
* is written, so nothing is sent.
*/
class CleartextGuardInterceptor(
// The policy is a parameter so a test can refuse loopback, the only
// address a test server can listen on.
private val allows: (InetAddress) -> Boolean = CleartextPolicy::allows,
) : Interceptor {
override fun intercept(chain: Interceptor.Chain): Response {
val request = chain.request()
if (request.isHttps || request.tag(MinstrelServerRequest::class.java) == null) {
return chain.proceed(request)
}
val address = chain.connection()?.route()?.socketAddress?.address
if (address != null && !allows(address)) {
throw CleartextToPublicHostException(request.url.host)
}
return chain.proceed(request)
}
}
/** The server was reached over plain HTTP at a public address, and refused. */
class CleartextToPublicHostException(host: String) :
IOException("refusing plain http:// to $host: it is a public address")
/** Which addresses plain HTTP may reach: the home network, never the internet. */
object CleartextPolicy {
private const val CGNAT_FIRST_OCTET = 100
private const val CGNAT_SECOND_MASK = 0xC0
private const val CGNAT_SECOND_PREFIX = 64
private const val ULA_MASK = 0xFE
private const val ULA_PREFIX = 0xFC
private const val BYTE = 0xFF
fun allows(address: InetAddress): Boolean =
address.isLoopbackAddress ||
address.isSiteLocalAddress || // 10/8, 172.16/12, 192.168/16
address.isLinkLocalAddress || // 169.254/16, fe80::/10
address.isAnyLocalAddress ||
isCarrierGradeNat(address) ||
isUniqueLocal(address)
// 100.64/10. Tailscale and other overlay VPNs hand these out; the overlay
// encrypts the traffic itself.
private fun isCarrierGradeNat(address: InetAddress): Boolean {
if (address !is Inet4Address) return false
val b = address.address
return (b[0].toInt() and BYTE) == CGNAT_FIRST_OCTET &&
(b[1].toInt() and CGNAT_SECOND_MASK) == CGNAT_SECOND_PREFIX
}
// fc00::/7, IPv6's private range.
private fun isUniqueLocal(address: InetAddress): Boolean =
address is Inet6Address && (address.address[0].toInt() and ULA_MASK) == ULA_PREFIX
}
@@ -37,18 +37,36 @@ object ErrorCopy {
* as connection failures.
*/
fun fromThrowable(t: Throwable): String = when (t) {
is HttpException -> messageFor(codeFromHttp(t))
is HttpException -> fromHttp(t)
is CleartextToPublicHostException -> messageFor("cleartext_public")
is IOException -> messageFor("connection_refused")
else -> TABLE.getValue("unknown")
}
private fun codeFromHttp(e: HttpException): String {
/**
* Codes whose server message carries specifics the operator needs in
* order to act — which directory, which uid — that fixed copy cannot say.
* For these the message follows the copy (#3918). Kept to a named set on
* purpose: most server messages are internal detail. Mirrors web's
* errors.ts.
*/
private val DETAIL_CODES = setOf("library_not_writable", "file_delete_failed")
private fun fromHttp(e: HttpException): String {
val body = bodyFromHttp(e)
val copy = messageFor(body.code.ifEmpty { "unknown" })
return if (body.code in DETAIL_CODES && body.message.isNotBlank()) {
"$copy ${body.message}"
} else {
copy
}
}
private fun bodyFromHttp(e: HttpException): Body {
val raw = runCatching { e.response()?.errorBody()?.string() }.getOrNull()
?: return "unknown"
val code = runCatching { json.decodeFromString<Envelope>(raw).error?.code }
.getOrNull()
.orEmpty()
return code.ifEmpty { "unknown" }
?: return Body()
return runCatching { json.decodeFromString<Envelope>(raw).error }
.getOrNull() ?: Body()
}
private val TABLE: Map<String, String> = mapOf(
@@ -58,6 +76,7 @@ object ErrorCopy {
"forbidden" to "You don't have permission to do that.",
"not_authorized" to "You don't have permission to do that.",
"invalid_credentials" to "Wrong username or password.",
"rate_limited" to "Too many attempts. Wait a few minutes and try again.",
"wrong_password" to "Current password is incorrect.",
"password_too_short" to "Password must be at least 8 characters.",
"username_invalid" to "That username isn't valid.",
@@ -83,6 +102,9 @@ object ErrorCopy {
"mbid_required" to "An MBID is required for this lookup.",
"system_playlist_readonly" to "System playlists can't be edited directly.",
"connection_refused" to "Couldn't reach the server. Check the URL and try again.",
"cleartext_public" to
"This server is on the internet, so its URL must start with https://. " +
"Plain http:// only works on your home network.",
"lidarr_unreachable" to
"Lidarr is unreachable right now. Try again, or check Admin → Integrations.",
"lidarr_disabled" to "Lidarr integration is not enabled.",
@@ -99,6 +121,8 @@ object ErrorCopy {
"request_not_pending" to "This request is no longer pending.",
"request_not_found" to "That request no longer exists.",
"track_not_found" to "That track no longer exists.",
"library_not_writable" to "The music library isn't writable by the server.",
"file_delete_failed" to "The file couldn't be deleted.",
"album_not_found" to "That album no longer exists.",
"artist_not_found" to "That artist no longer exists.",
"playlist_not_found" to "That playlist no longer exists.",
@@ -70,6 +70,9 @@ object NetworkModule {
.addInterceptor(auth)
.addInterceptor(baseUrl)
.addInterceptor(logging)
// A network interceptor, so it sees the address the connection
// really reached and runs before any request byte is written.
.addNetworkInterceptor(CleartextGuardInterceptor())
.connectTimeout(CONNECT_TIMEOUT_SECONDS, TimeUnit.SECONDS)
.readTimeout(READ_TIMEOUT_SECONDS, TimeUnit.SECONDS)
.build()
@@ -29,11 +29,20 @@ interface CastApi {
* Request body. [expSeconds] is clamped server-side to [60, 86400];
* the 21_600 default (6h) is long enough to play through any typical
* track without re-minting mid-playback.
*
* [level] asks for the leveled stream (M464 #5001): the track rendered at
* the user's loudness gain, which the server works out from their setting.
* [asAlbum] says the track plays among its album in order, which picks
* album gain in auto mode. [prerender] says the speaker will fetch it
* soon, so the server renders it ahead.
*/
@Serializable
data class StreamTokenRequest(
val trackId: String,
val expSeconds: Int = 21_600,
val level: Boolean = false,
val asAlbum: Boolean = false,
val prerender: Boolean = false,
)
/**
@@ -53,4 +62,6 @@ data class StreamTokenResponse(
val url: String,
val mime: String = "audio/mpeg",
val title: String = "",
/** [url] is the leveled stream; false when leveling is off or changes nothing. */
val leveled: Boolean = false,
)
@@ -3,6 +3,7 @@ package com.fabledsword.minstrel.api.endpoints
import com.fabledsword.minstrel.models.wire.ListenBrainzStatusWire
import com.fabledsword.minstrel.models.wire.MyProfileWire
import com.fabledsword.minstrel.models.wire.SystemPlaylistsStatusWire
import com.fabledsword.minstrel.settings.data.NormalizationPrefs
import kotlinx.serialization.SerialName
import kotlinx.serialization.Serializable
import retrofit2.http.Body
@@ -59,6 +60,14 @@ interface MeApi {
*/
@PUT("api/me/listenbrainz")
suspend fun setListenBrainz(@Body body: ListenBrainzPutBody): ListenBrainzStatusWire
/** The caller's loudness-normalization preference, or the defaults if never set. */
@GET("api/me/normalization")
suspend fun getNormalization(): NormalizationPrefs
/** Replaces the whole preference; returns what the server stored. */
@PUT("api/me/normalization")
suspend fun putNormalization(@Body body: NormalizationPrefs): NormalizationPrefs
}
/**
@@ -0,0 +1,90 @@
package com.fabledsword.minstrel.api.endpoints
import kotlinx.serialization.SerialName
import kotlinx.serialization.Serializable
import retrofit2.http.Body
import retrofit2.http.GET
import retrofit2.http.POST
import retrofit2.http.PUT
import retrofit2.http.Path
import retrofit2.http.Query
/**
* The notifications inbox and its per-user settings (M489). The server
* renders each notice's title, body and link, so the app shows them as given.
*/
interface NotificationsApi {
@GET("api/me/notifications")
suspend fun list(@Query("limit") limit: Int): NotificationsPageWire
/** 204; 404 when the notice is gone, which a replay treats as done. */
@POST("api/me/notifications/{id}/read")
suspend fun markRead(@Path("id") id: String)
@POST("api/me/notifications/read-all")
suspend fun readAll(@Body body: ReadAllBody)
@GET("api/me/notification-settings")
suspend fun getSettings(): NotificationSettingsWire
@PUT("api/me/notification-settings")
suspend fun putSettings(@Body body: PutNotificationSettingsBody): NotificationSettingsWire
}
@Serializable
data class NotificationWire(
val id: String,
val kind: String,
val title: String,
val body: String,
val link: String,
@SerialName("created_at") val createdAt: String,
@SerialName("read_at") val readAt: String? = null,
)
@Serializable
data class NotificationsPageWire(
val items: List<NotificationWire>,
@SerialName("unread_count") val unreadCount: Long,
@SerialName("next_before") val nextBefore: String? = null,
)
/**
* `upTo` limits "mark all read" to what existed when the user asked, so a
* replay landing later leaves newer notices unread. No default on purpose:
* the app's Json drops default-valued fields.
*/
@Serializable
data class ReadAllBody(@SerialName("up_to") val upTo: String?)
@Serializable
data class NotificationKindSettingWire(
val kind: String,
@SerialName("admin_only") val adminOnly: Boolean,
val inbox: Boolean,
val phone: Boolean,
val email: Boolean,
)
@Serializable
data class NotificationSettingsWire(
val kinds: List<NotificationKindSettingWire>,
@SerialName("email_available") val emailAvailable: Boolean,
@SerialName("email_unavailable_reason") val emailUnavailableReason: String? = null,
)
/**
* One kind's change. Untouched channels stay null and, being equal to their
* default, are left out of the JSON, so the server changes only the channel
* the user touched.
*/
@Serializable
data class NotificationSettingChangeWire(
val kind: String,
val inbox: Boolean? = null,
val phone: Boolean? = null,
val email: Boolean? = null,
)
@Serializable
data class PutNotificationSettingsBody(val kinds: List<NotificationSettingChangeWire>)
@@ -0,0 +1,15 @@
package com.fabledsword.minstrel.api.endpoints
import com.fabledsword.minstrel.models.wire.ReplayGainResponseWire
import retrofit2.http.GET
import retrofit2.http.Query
/** The player's loudness lookup (#4997), kept apart from the browse surface in [LibraryApi]. */
interface ReplayGainApi {
/**
* ReplayGain values for up to 200 comma-separated track ids. An id
* missing from `items` has not been measured yet.
*/
@GET("api/tracks/replay-gain")
suspend fun getReplayGain(@Query("ids") ids: String): ReplayGainResponseWire
}
@@ -4,12 +4,19 @@ import com.fabledsword.minstrel.cache.audiocache.CacheSettings
import com.fabledsword.minstrel.cache.db.dao.AuthSessionDao
import com.fabledsword.minstrel.cache.db.entities.AuthSessionEntity
import com.fabledsword.minstrel.di.ApplicationScope
import com.fabledsword.minstrel.settings.data.NormalizationPrefs
import kotlinx.coroutines.CoroutineScope
import kotlinx.coroutines.Deferred
import kotlinx.coroutines.async
import kotlinx.coroutines.flow.MutableStateFlow
import kotlinx.coroutines.flow.StateFlow
import kotlinx.coroutines.flow.asStateFlow
import kotlinx.coroutines.launch
import kotlinx.coroutines.sync.Mutex
import kotlinx.coroutines.sync.withLock
import kotlinx.coroutines.withTimeoutOrNull
import kotlinx.serialization.json.Json
import timber.log.Timber
import javax.inject.Inject
import javax.inject.Singleton
@@ -27,6 +34,14 @@ import javax.inject.Singleton
* in-memory state changes synchronously so the next interceptor read
* sees the new value immediately; the DAO write coroutine catches up
* shortly after.
*
* **The session cookie is the exception** (M462 #4985): it is persisted
* through [SessionVault] (Keystore-encrypted), not the Room row. On the
* first launch after the upgrade, a cookie still in the row is moved into
* the vault and the column cleared, so nobody is signed out by the change.
* If the Keystore cannot be used on a device, the cookie stays in the row
* as before rather than being lost. [awaitSessionHydrated] lets a caller
* that needs a definitive answer (the auth gate) wait for this.
*/
// AuthStore is the single-row facade over auth_session (de-facto
// app_preferences — see entity comment). It legitimately owns one
@@ -39,6 +54,7 @@ import javax.inject.Singleton
@Singleton
class AuthStore @Inject constructor(
private val dao: AuthSessionDao,
private val vault: SessionVault,
@ApplicationScope private val scope: CoroutineScope,
) {
private val sessionCookieState = MutableStateFlow<String?>(null)
@@ -62,18 +78,39 @@ class AuthStore @Inject constructor(
private val diagnosticsOptOutState = MutableStateFlow(false)
val diagnosticsOptOut: StateFlow<Boolean> = diagnosticsOptOutState.asStateFlow()
private val normalizationState = MutableStateFlow(NormalizationPrefs.DEFAULT)
val normalization: StateFlow<NormalizationPrefs> = normalizationState.asStateFlow()
// Background delivery (M489 #5347): the device's choice, on by default.
// The shade's high-water mark lives in the same row but is read and
// written through the DAO by NotificationSync, awaited, never cached here.
private val backgroundDeliveryState = MutableStateFlow(true)
val backgroundDelivery: StateFlow<Boolean> = backgroundDeliveryState.asStateFlow()
private val json = Json { ignoreUnknownKeys = true }
// Serialises every cookie persist with the one-time hydration, so a
// sign-in or a 401 that lands while hydration runs is never overwritten
// by the stale value hydration read.
private val cookieLock = Mutex()
// Set by setSessionCookie. Once something has written the cookie this
// process, that value wins over whatever hydration finds on disk.
@Volatile private var cookieTouched = false
private val cookieHydration: Deferred<Unit> = scope.async { hydrateSessionCookie() }
init {
scope.launch {
dao.observe().collect { row ->
sessionCookieState.value = row?.sessionCookie
baseUrlState.value = row?.baseUrl ?: DEFAULT_BASE_URL
userJsonState.value = row?.userJson
themeModeState.value = row?.themeMode
clientIdState.value = row?.clientId
cacheSettingsState.value = decodeCacheSettings(row?.cacheSettingsJson)
diagnosticsOptOutState.value = row?.diagnosticsOptOut ?: false
normalizationState.value = decodeNormalization(row?.normalizationJson)
backgroundDeliveryState.value = row?.backgroundDelivery ?: true
}
}
}
@@ -85,9 +122,57 @@ class AuthStore @Inject constructor(
}.getOrDefault(CacheSettings.DEFAULT)
}
private fun decodeNormalization(raw: String?): NormalizationPrefs {
if (raw.isNullOrEmpty()) return NormalizationPrefs.DEFAULT
return runCatching {
json.decodeFromString(NormalizationPrefs.serializer(), raw)
}.getOrDefault(NormalizationPrefs.DEFAULT)
}
/**
* Suspends until the stored session cookie has been loaded into
* [sessionCookie], or [HYDRATION_DEADLINE_MS] passes (rule 156: a wedged
* Keystore must not leave the start screen spinning). Returns false on
* the deadline; the caller then decides from whatever has loaded, and a
* late hydration still lands in [sessionCookie].
*/
suspend fun awaitSessionHydrated(): Boolean {
val done = withTimeoutOrNull(HYDRATION_DEADLINE_MS) { cookieHydration.await() } != null
if (!done) Timber.w("auth store: session hydration passed its deadline; deciding without it")
return done
}
fun setSessionCookie(value: String?) {
cookieTouched = true
sessionCookieState.value = value
scope.launch { persistCookie(value) }
scope.launch { cookieLock.withLock { storeCookie(value) } }
}
private suspend fun hydrateSessionCookie() = cookieLock.withLock {
val legacy = runCatching { dao.get()?.sessionCookie }.getOrNull()
if (cookieTouched) return@withLock
// A cookie in the row is the newer one when both exist: the row is
// only written when the vault failed, and an install upgrading from
// before the vault has nothing in the vault yet.
val cookie = legacy ?: vault.read()
// Best-effort: if moving it fails, the session still loads this time
// and the move is retried on the next launch. Hydration must never
// throw, or awaitSessionHydrated would leave the auth gate stuck.
if (legacy != null) {
runCatching { storeCookie(legacy) }
.onFailure { Timber.w(it, "auth store: could not move the session cookie into the vault") }
}
sessionCookieState.value = cookie
}
// Vault first; the Room row only when the Keystore is unusable, so a
// broken Keystore degrades to the old storage rather than a sign-out.
private suspend fun storeCookie(value: String?) {
if (vault.write(value)) {
if (dao.get()?.sessionCookie != null) dao.setSessionCookie(null)
} else {
persistLegacyCookie(value)
}
}
fun setBaseUrl(value: String) {
@@ -121,7 +206,24 @@ class AuthStore @Inject constructor(
scope.launch { persistDiagnosticsOptOut(value) }
}
private suspend fun persistCookie(value: String?) {
fun setNormalization(value: NormalizationPrefs) {
normalizationState.value = value
val encoded = json.encodeToString(NormalizationPrefs.serializer(), value)
scope.launch { persistNormalization(encoded) }
}
fun setBackgroundDelivery(value: Boolean) {
backgroundDeliveryState.value = value
scope.launch {
if (dao.get() == null) {
dao.upsert(currentEntity().copy(backgroundDelivery = value))
} else {
dao.setBackgroundDelivery(value)
}
}
}
private suspend fun persistLegacyCookie(value: String?) {
if (dao.get() == null) {
dao.upsert(currentEntity().copy(sessionCookie = value))
} else {
@@ -177,9 +279,19 @@ class AuthStore @Inject constructor(
}
}
private suspend fun persistNormalization(json: String) {
if (dao.get() == null) {
dao.upsert(currentEntity().copy(normalizationJson = json))
} else {
dao.setNormalizationJson(json)
}
}
private fun currentEntity(): AuthSessionEntity = AuthSessionEntity(
id = ROW_ID,
sessionCookie = sessionCookieState.value,
// Never copied into the row: the cookie lives in the vault, and
// persistLegacyCookie sets it explicitly on the fallback path.
sessionCookie = null,
baseUrl = baseUrlState.value,
userJson = userJsonState.value,
themeMode = themeModeState.value,
@@ -189,10 +301,19 @@ class AuthStore @Inject constructor(
cacheSettingsState.value,
),
diagnosticsOptOut = diagnosticsOptOutState.value,
normalizationJson = json.encodeToString(
NormalizationPrefs.serializer(),
normalizationState.value,
),
backgroundDelivery = backgroundDeliveryState.value,
)
companion object {
const val DEFAULT_BASE_URL: String = "http://localhost:8080"
// Generous on purpose: hydration is one local row read and one
// Keystore decrypt, normally milliseconds. This only bounds "never".
const val HYDRATION_DEADLINE_MS: Long = 10_000
private const val ROW_ID = 0
}
}
@@ -0,0 +1,147 @@
package com.fabledsword.minstrel.auth
import android.content.Context
import android.security.keystore.KeyGenParameterSpec
import android.security.keystore.KeyProperties
import dagger.Binds
import dagger.Module
import dagger.hilt.InstallIn
import dagger.hilt.android.qualifiers.ApplicationContext
import dagger.hilt.components.SingletonComponent
import timber.log.Timber
import java.security.KeyStore
import java.util.Base64
import javax.crypto.Cipher
import javax.crypto.KeyGenerator
import javax.crypto.SecretKey
import javax.crypto.spec.GCMParameterSpec
import javax.inject.Inject
import javax.inject.Singleton
/**
* Where the session cookie lives at rest (M462 #4985).
*
* The cookie is a bearer credential: anyone holding it is signed in as the
* user until the server expires or revokes it. It used to sit in plain text
* in the Room `auth_session` row, readable from any copy of the app's data
* directory (a rooted device, an adb backup of a debuggable build, a
* forensic image). Now only ciphertext is stored, under an AES key that
* lives in the Android Keystore and never leaves it, so a copy of the
* files alone yields nothing usable.
*/
interface SessionVault {
/** The stored cookie, or null when none is stored or it can't be decrypted. */
fun read(): String?
/**
* Stores [value], or clears the stored cookie when null. Returns false
* when the Keystore could not be used, so the caller can fall back
* rather than lose the session.
*/
fun write(value: String?): Boolean
}
/**
* AES-GCM sealing of a short string, framed as base64(iv || ciphertext+tag).
* Kept apart from the Keystore so the framing can be unit-tested on the JVM
* with an ordinary key; the Android Keystore has no JVM implementation.
*/
internal object SealedBox {
private const val TRANSFORMATION = "AES/GCM/NoPadding"
private const val TAG_BITS = 128
private const val IV_BYTES = 12
// Binds a sealed value to its purpose: a blob sealed for something else
// under the same key will not open as a session cookie.
private val AAD = "minstrel-session-cookie-v1".toByteArray(Charsets.UTF_8)
fun seal(key: SecretKey, plaintext: String): String {
val cipher = Cipher.getInstance(TRANSFORMATION)
// No IV passed: the provider generates a fresh random one. Keystore
// keys refuse a caller-chosen IV for encryption by default.
cipher.init(Cipher.ENCRYPT_MODE, key)
cipher.updateAAD(AAD)
val sealed = cipher.iv + cipher.doFinal(plaintext.toByteArray(Charsets.UTF_8))
return Base64.getEncoder().encodeToString(sealed)
}
fun open(key: SecretKey, sealed: String): String {
val bytes = Base64.getDecoder().decode(sealed)
require(bytes.size > IV_BYTES) { "sealed value too short" }
val cipher = Cipher.getInstance(TRANSFORMATION)
cipher.init(Cipher.DECRYPT_MODE, key, GCMParameterSpec(TAG_BITS, bytes, 0, IV_BYTES))
cipher.updateAAD(AAD)
return String(cipher.doFinal(bytes, IV_BYTES, bytes.size - IV_BYTES), Charsets.UTF_8)
}
}
/**
* [SessionVault] backed by a Keystore AES key and a private prefs file that
* holds only the sealed value.
*
* A value that will not open (the key was wiped by a factory-reset of the
* Keystore, or the file was restored onto another device; app backup is off,
* but a vendor transfer tool may still copy files) is discarded and reported
* as absent. The user signs in again, which is the right outcome for a
* credential that no longer verifies.
*/
@Singleton
class KeystoreSessionVault @Inject constructor(
@ApplicationContext context: Context,
) : SessionVault {
private val prefs = context.getSharedPreferences(PREFS_NAME, Context.MODE_PRIVATE)
override fun read(): String? {
val sealed = prefs.getString(KEY_COOKIE, null) ?: return null
return runCatching { SealedBox.open(key(), sealed) }
.onFailure {
Timber.w(it, "session vault: stored cookie would not decrypt; discarding it")
prefs.edit().remove(KEY_COOKIE).commit()
}
.getOrNull()
}
override fun write(value: String?): Boolean = runCatching {
val editor = prefs.edit()
if (value == null) {
editor.remove(KEY_COOKIE)
} else {
editor.putString(KEY_COOKIE, SealedBox.seal(key(), value))
}
editor.commit()
}.onFailure {
Timber.w(it, "session vault: Keystore unavailable; cookie not stored in the vault")
}.getOrDefault(false)
private fun key(): SecretKey {
val keyStore = KeyStore.getInstance(ANDROID_KEYSTORE).apply { load(null) }
(keyStore.getKey(KEY_ALIAS, null) as? SecretKey)?.let { return it }
val generator = KeyGenerator.getInstance(KeyProperties.KEY_ALGORITHM_AES, ANDROID_KEYSTORE)
generator.init(
KeyGenParameterSpec.Builder(
KEY_ALIAS,
KeyProperties.PURPOSE_ENCRYPT or KeyProperties.PURPOSE_DECRYPT,
)
.setBlockModes(KeyProperties.BLOCK_MODE_GCM)
.setEncryptionPaddings(KeyProperties.ENCRYPTION_PADDING_NONE)
.setKeySize(KEY_BITS)
.build(),
)
return generator.generateKey()
}
private companion object {
const val ANDROID_KEYSTORE = "AndroidKeyStore"
const val KEY_ALIAS = "minstrel_session_cookie"
const val KEY_BITS = 256
const val PREFS_NAME = "session_vault"
const val KEY_COOKIE = "sealed_cookie"
}
}
@Module
@InstallIn(SingletonComponent::class)
abstract class SessionVaultModule {
@Binds
abstract fun bindSessionVault(impl: KeystoreSessionVault): SessionVault
}
@@ -16,10 +16,11 @@ import javax.inject.Inject
/**
* Computes the initial startDestination for the root NavHost based on
* persisted auth state. Sits on top of AuthSessionDao directly rather
* than AuthStore's StateFlow because the StateFlow defaults to null
* until Room's first async emission — we need a definitive answer
* before drawing any nav graph.
* persisted auth state. Reads the row from AuthSessionDao directly, and
* the session cookie only after [AuthStore.awaitSessionHydrated]: both
* StateFlows default to null until their async load lands, and we need a
* definitive answer before drawing any nav graph. The cookie is no longer
* in the row (it lives in the Keystore-backed SessionVault, #4985).
*
* - no row at all → ServerUrl (first launch)
* - row with baseUrl, no cookie → Login (URL configured, not yet signed in)
@@ -31,6 +32,7 @@ import javax.inject.Inject
@HiltViewModel
class AuthGateViewModel @Inject constructor(
private val dao: AuthSessionDao,
private val authStore: AuthStore,
) : ViewModel() {
private val internal = MutableStateFlow<Any?>(null)
@@ -38,12 +40,13 @@ class AuthGateViewModel @Inject constructor(
init {
viewModelScope.launch {
authStore.awaitSessionHydrated()
val signedIn = !authStore.sessionCookie.value.isNullOrEmpty()
val row = dao.get()
internal.value = when {
row == null -> ServerUrl
row.baseUrl == AuthStore.DEFAULT_BASE_URL && row.sessionCookie.isNullOrEmpty() ->
ServerUrl
row.sessionCookie.isNullOrEmpty() -> Login
row.baseUrl == AuthStore.DEFAULT_BASE_URL && !signedIn -> ServerUrl
!signedIn -> Login
else -> Home
}
}
@@ -3,6 +3,8 @@ package com.fabledsword.minstrel.cache.db
import androidx.room.Database
import androidx.room.RoomDatabase
import androidx.room.TypeConverters
import androidx.room.migration.Migration
import androidx.sqlite.db.SupportSQLiteDatabase
import com.fabledsword.minstrel.cache.db.dao.AudioCacheIndexDao
import com.fabledsword.minstrel.cache.db.dao.AuthSessionDao
import com.fabledsword.minstrel.cache.db.dao.CachedAlbumDao
@@ -11,6 +13,7 @@ import com.fabledsword.minstrel.cache.db.dao.CachedHistorySnapshotDao
import com.fabledsword.minstrel.cache.db.dao.CachedHomeIndexDao
import com.fabledsword.minstrel.cache.db.dao.CachedLikeDao
import com.fabledsword.minstrel.cache.db.dao.CachedMutationDao
import com.fabledsword.minstrel.cache.db.dao.CachedNotificationDao
import com.fabledsword.minstrel.cache.db.dao.CachedPlaylistDao
import com.fabledsword.minstrel.cache.db.dao.CachedResumeStateDao
import com.fabledsword.minstrel.cache.db.dao.CachedPlaylistTrackDao
@@ -26,6 +29,8 @@ import com.fabledsword.minstrel.cache.db.entities.CachedHistorySnapshotEntity
import com.fabledsword.minstrel.cache.db.entities.CachedHomeIndexEntity
import com.fabledsword.minstrel.cache.db.entities.CachedLikeEntity
import com.fabledsword.minstrel.cache.db.entities.CachedMutationEntity
import com.fabledsword.minstrel.cache.db.entities.CachedNotificationEntity
import com.fabledsword.minstrel.cache.db.entities.CachedNotificationSettingsEntity
import com.fabledsword.minstrel.cache.db.entities.CachedPlaylistEntity
import com.fabledsword.minstrel.cache.db.entities.CachedResumeStateEntity
import com.fabledsword.minstrel.cache.db.entities.CachedPlaylistTrackEntity
@@ -64,14 +69,30 @@ import com.fabledsword.minstrel.cache.db.entities.SyncMetadataEntity
CachedHistorySnapshotEntity::class,
AuthSessionEntity::class,
DiagnosticEventEntity::class,
CachedNotificationEntity::class,
CachedNotificationSettingsEntity::class,
],
// v12: + auth_session.backgroundDelivery and notifiedUpTo, the device's
// background-delivery choice and its shade high-water mark (M489 #5347).
// v11: + cached_notifications and cached_notification_settings, the
// notifications inbox and its settings (M489). MIGRATION_10_11 creates
// both; nothing to backfill, the first refresh fills them.
// v10: + cached_tracks.trackGain/trackPeak and cached_albums.albumGain/
// albumPeak, the ReplayGain values the player levels by (M464 #5000).
// MIGRATION_9_10 also rewinds the sync cursor, so the next sync re-sends
// every row and an existing cache gains its values.
// v9: + auth_session.normalizationJson, the loudness-normalization
// preference (M464 #4998). The first schema step with an explicit
// Migration (MIGRATION_8_9): a destructive rebuild would also wipe this
// row — the server address and theme — and the queued offline writes,
// which is too much to lose for one added column.
// v8: + cached_tracks.missing, the server's missing-file mark (#2704),
// so cache-first surfaces stop offering files that cannot stream.
// v7: + diagnostic_events table (M9) and the diagnosticsOptOut column
// on auth_session. Pre-v1 destructive fallback rebuilds on mismatch —
// which is exactly right here: the next sync refills every row with the
// new column populated, so there is nothing to migrate by hand.
version = 8,
version = 12,
exportSchema = true,
)
@TypeConverters(MinstrelTypeConverters::class)
@@ -91,4 +112,58 @@ abstract class AppDatabase : RoomDatabase() {
abstract fun cachedHistorySnapshotDao(): CachedHistorySnapshotDao
abstract fun authSessionDao(): AuthSessionDao
abstract fun diagnosticEventDao(): DiagnosticEventDao
abstract fun cachedNotificationDao(): CachedNotificationDao
}
/** v8 → v9: add the nullable normalization preference column (#4998). */
val MIGRATION_8_9: Migration = object : Migration(8, 9) {
override fun migrate(db: SupportSQLiteDatabase) {
db.execSQL("ALTER TABLE auth_session ADD COLUMN normalizationJson TEXT")
}
}
/**
* v9 → v10: the gain columns (#5000). Rows synced before this carry no gains,
* and the sync is incremental, so it would never re-send them: the cursor goes
* back to 0 and the next sync is a full one, upserting every row in place.
*/
val MIGRATION_9_10: Migration = object : Migration(9, 10) {
override fun migrate(db: SupportSQLiteDatabase) {
db.execSQL("ALTER TABLE cached_tracks ADD COLUMN trackGain REAL")
db.execSQL("ALTER TABLE cached_tracks ADD COLUMN trackPeak REAL")
db.execSQL("ALTER TABLE cached_albums ADD COLUMN albumGain REAL")
db.execSQL("ALTER TABLE cached_albums ADD COLUMN albumPeak REAL")
db.execSQL("UPDATE sync_metadata SET cursor = 0")
}
}
/**
* v10 → v11: the notifications inbox cache and its settings (M489). The SQL
* matches what Room generates for the two entities; Room checks it on open.
*/
val MIGRATION_10_11: Migration = object : Migration(10, 11) {
override fun migrate(db: SupportSQLiteDatabase) {
db.execSQL(
"CREATE TABLE IF NOT EXISTS `cached_notifications` (" +
"`id` TEXT NOT NULL, `kind` TEXT NOT NULL, `title` TEXT NOT NULL, " +
"`body` TEXT NOT NULL, `link` TEXT NOT NULL, `createdAt` INTEGER NOT NULL, " +
"`readAt` INTEGER, PRIMARY KEY(`id`))",
)
db.execSQL(
"CREATE TABLE IF NOT EXISTS `cached_notification_settings` (" +
"`id` INTEGER NOT NULL, `json` TEXT NOT NULL, PRIMARY KEY(`id`))",
)
}
}
/**
* v11 → v12: background delivery (M489 #5347). The device choice defaults on,
* as the entity's column default says; the high-water mark starts empty, so
* the first catch-up sets it without announcing anything.
*/
val MIGRATION_11_12: Migration = object : Migration(11, 12) {
override fun migrate(db: SupportSQLiteDatabase) {
db.execSQL("ALTER TABLE auth_session ADD COLUMN backgroundDelivery INTEGER NOT NULL DEFAULT 1")
db.execSQL("ALTER TABLE auth_session ADD COLUMN notifiedUpTo INTEGER")
}
}
@@ -10,6 +10,7 @@ import com.fabledsword.minstrel.cache.db.dao.CachedHistorySnapshotDao
import com.fabledsword.minstrel.cache.db.dao.CachedHomeIndexDao
import com.fabledsword.minstrel.cache.db.dao.CachedLikeDao
import com.fabledsword.minstrel.cache.db.dao.CachedMutationDao
import com.fabledsword.minstrel.cache.db.dao.CachedNotificationDao
import com.fabledsword.minstrel.cache.db.dao.DiagnosticEventDao
import com.fabledsword.minstrel.cache.db.dao.CachedPlaylistDao
import com.fabledsword.minstrel.cache.db.dao.CachedPlaylistTrackDao
@@ -37,6 +38,7 @@ object DatabaseModule {
// launch, so users lose only the unsynced mutation queue
// (acceptable while we're iterating). Replace with explicit
// Migration entries before the first tagged release.
.addMigrations(MIGRATION_8_9, MIGRATION_9_10, MIGRATION_10_11, MIGRATION_11_12)
.fallbackToDestructiveMigration(dropAllTables = true)
.build()
@@ -111,5 +113,10 @@ object DatabaseModule {
fun provideDiagnosticEventDao(db: AppDatabase): DiagnosticEventDao =
db.diagnosticEventDao()
@Provides
@Singleton
fun provideCachedNotificationDao(db: AppDatabase): CachedNotificationDao =
db.cachedNotificationDao()
private const val DATABASE_NAME = "minstrel.db"
}
@@ -6,6 +6,7 @@ import androidx.room.OnConflictStrategy
import androidx.room.Query
import com.fabledsword.minstrel.cache.db.entities.AuthSessionEntity
import kotlinx.coroutines.flow.Flow
import kotlinx.datetime.Instant
@Dao
interface AuthSessionDao {
@@ -46,4 +47,16 @@ interface AuthSessionDao {
/** Partial update: change only the per-device diagnostics opt-out. */
@Query("UPDATE auth_session SET diagnosticsOptOut = :optOut WHERE id = 0")
suspend fun setDiagnosticsOptOut(optOut: Boolean)
/** Partial update: change only the serialized normalization preference. */
@Query("UPDATE auth_session SET normalizationJson = :json WHERE id = 0")
suspend fun setNormalizationJson(json: String?)
/** Partial update: change only the background-delivery choice. */
@Query("UPDATE auth_session SET backgroundDelivery = :enabled WHERE id = 0")
suspend fun setBackgroundDelivery(enabled: Boolean)
/** Partial update: change only the shade's high-water mark. */
@Query("UPDATE auth_session SET notifiedUpTo = :upTo WHERE id = 0")
suspend fun setNotifiedUpTo(upTo: Instant?)
}
@@ -31,4 +31,8 @@ interface CachedMutationDao {
@Query("DELETE FROM cached_mutations")
suspend fun clear()
/** Whether a write of [kind] is still waiting to be replayed. */
@Query("SELECT EXISTS(SELECT 1 FROM cached_mutations WHERE kind = :kind)")
suspend fun hasPending(kind: String): Boolean
}
@@ -0,0 +1,51 @@
package com.fabledsword.minstrel.cache.db.dao
import androidx.room.Dao
import androidx.room.Insert
import androidx.room.OnConflictStrategy
import androidx.room.Query
import androidx.room.Transaction
import com.fabledsword.minstrel.cache.db.entities.CachedNotificationEntity
import com.fabledsword.minstrel.cache.db.entities.CachedNotificationSettingsEntity
import kotlinx.coroutines.flow.Flow
import kotlinx.datetime.Instant
@Dao
interface CachedNotificationDao {
@Query("SELECT * FROM cached_notifications ORDER BY createdAt DESC, id DESC")
fun observeAll(): Flow<List<CachedNotificationEntity>>
@Query("SELECT COUNT(*) FROM cached_notifications WHERE readAt IS NULL")
fun observeUnreadCount(): Flow<Int>
@Query("SELECT * FROM cached_notifications")
suspend fun getAll(): List<CachedNotificationEntity>
@Query("DELETE FROM cached_notifications")
suspend fun clear()
@Insert(onConflict = OnConflictStrategy.REPLACE)
suspend fun insertAll(rows: List<CachedNotificationEntity>)
/** The newest page replaces the cache whole: a notice gone server-side goes here too. */
@Transaction
suspend fun replaceAll(rows: List<CachedNotificationEntity>) {
clear()
insertAll(rows)
}
@Query("UPDATE cached_notifications SET readAt = :at WHERE id = :id AND readAt IS NULL")
suspend fun markRead(id: String, at: Instant)
@Query("UPDATE cached_notifications SET readAt = :at WHERE readAt IS NULL")
suspend fun markAllRead(at: Instant)
@Query("SELECT * FROM cached_notification_settings WHERE id = 1")
fun observeSettings(): Flow<CachedNotificationSettingsEntity?>
@Query("SELECT * FROM cached_notification_settings WHERE id = 1")
suspend fun getSettings(): CachedNotificationSettingsEntity?
@Insert(onConflict = OnConflictStrategy.REPLACE)
suspend fun upsertSettings(row: CachedNotificationSettingsEntity)
}
@@ -38,6 +38,27 @@ interface CachedTrackDao {
@Insert(onConflict = OnConflictStrategy.REPLACE)
suspend fun upsertAll(rows: List<CachedTrackEntity>)
/**
* ReplayGain values for [ids] (M464 #5000): the track's own from its row,
* the album's from its album row. A track not in the cache has no row.
*/
@Query(
"SELECT t.id AS id, t.trackGain AS trackGain, t.trackPeak AS trackPeak, " +
"a.albumGain AS albumGain, a.albumPeak AS albumPeak " +
"FROM cached_tracks t LEFT JOIN cached_albums a ON a.id = t.albumId " +
"WHERE t.id IN (:ids)",
)
suspend fun replayGains(ids: List<String>): List<CachedReplayGain>
@Query("DELETE FROM cached_tracks WHERE id IN (:ids)")
suspend fun deleteByIds(ids: List<String>)
}
/** One row of [CachedTrackDao.replayGains]. */
data class CachedReplayGain(
val id: String,
val trackGain: Float?,
val trackPeak: Float?,
val albumGain: Float?,
val albumPeak: Float?,
)
@@ -1,7 +1,9 @@
package com.fabledsword.minstrel.cache.db.entities
import androidx.room.ColumnInfo
import androidx.room.Entity
import androidx.room.PrimaryKey
import kotlinx.datetime.Instant
/**
* Single-row table holding the user's session cookie, configured
@@ -43,4 +45,23 @@ data class AuthSessionEntity(
* choice lives here. Default false = honor the account flag.
*/
val diagnosticsOptOut: Boolean = false,
/**
* JSON-encoded NormalizationPrefs (settings/data), the last value seen
* from the server or set here (M464 #4998). Null = never fetched; the
* defaults apply. Kept so offline playback still levels.
*/
val normalizationJson: String? = null,
/**
* "Notifications when the app is closed" (M489 #5347): keep the
* delivery foreground service running. A device choice, on by default.
* The column default matches MIGRATION_11_12's, which Room checks.
*/
@ColumnInfo(defaultValue = "1")
val backgroundDelivery: Boolean = true,
/**
* The newest notice this device has announced in the shade. A catch-up
* announces only what is newer, so a reboot or a reconnect never
* re-announces a backlog. Cleared on sign-out.
*/
val notifiedUpTo: Instant? = null,
)
@@ -18,5 +18,8 @@ data class CachedAlbumEntity(
val releaseDate: String? = null,
val coverPath: String? = null,
val mbid: String? = null,
// ReplayGain 2.0 album values (M464); null until every track is measured.
val albumGain: Float? = null,
val albumPeak: Float? = null,
val fetchedAt: Instant = Clock.System.now(),
)
@@ -0,0 +1,21 @@
package com.fabledsword.minstrel.cache.db.entities
import androidx.room.Entity
import androidx.room.PrimaryKey
import kotlinx.datetime.Instant
/**
* One notice from the user's inbox (M489), kept so the Notifications screen
* and the bell's badge work offline, and so a read made offline shows at once.
* The newest page is cached; older notices are the server's to keep.
*/
@Entity(tableName = "cached_notifications")
data class CachedNotificationEntity(
@PrimaryKey val id: String,
val kind: String,
val title: String,
val body: String,
val link: String,
val createdAt: Instant,
val readAt: Instant?,
)
@@ -0,0 +1,18 @@
package com.fabledsword.minstrel.cache.db.entities
import androidx.room.Entity
import androidx.room.PrimaryKey
/**
* Single-row copy of the user's notification settings (M489), stored as the
* wire JSON, so the settings screen opens offline and a toggle shows at once.
*/
@Entity(tableName = "cached_notification_settings")
data class CachedNotificationSettingsEntity(
@PrimaryKey val id: Int = SINGLETON_ID,
val json: String,
) {
companion object {
const val SINGLETON_ID = 1
}
}
@@ -26,5 +26,9 @@ data class CachedTrackEntity(
val fileFormat: String? = null,
val genre: String? = null,
val missing: Boolean = false,
// ReplayGain 2.0 track values (M464), kept so cached audio levels
// offline. Null until the server has measured the track.
val trackGain: Float? = null,
val trackPeak: Float? = null,
val fetchedAt: Instant = Clock.System.now(),
)
@@ -1,7 +1,9 @@
package com.fabledsword.minstrel.cache.mutations
import com.fabledsword.minstrel.api.endpoints.NotificationSettingChangeWire
import com.fabledsword.minstrel.cache.db.dao.CachedMutationDao
import com.fabledsword.minstrel.cache.db.entities.CachedMutationEntity
import com.fabledsword.minstrel.settings.data.NormalizationPrefs
import kotlinx.coroutines.channels.BufferOverflow
import kotlinx.coroutines.flow.MutableSharedFlow
import kotlinx.coroutines.flow.SharedFlow
@@ -41,6 +43,22 @@ object MutationKind {
// an undo collapses to the latest intent instead of replaying as two
// opposed calls whose order decides the outcome.
const val SUGGESTION_SNOOZE_TOGGLE: String = "suggestion_snooze_toggle"
// M464 #4998 loudness-normalization preference. The payload is the whole
// preference, a target state like the toggles above, so queued changes
// collapse to the last one and an older one can never be replayed last.
const val NORMALIZATION_SET: String = "normalization_set"
// M489 notifications inbox. A read is a one-way action (read never goes
// back to unread), so neither read kind needs collapsing. Read-all
// carries the moment the user asked, so a late replay leaves newer
// notices unread.
const val NOTIFICATION_READ: String = "notification_read"
const val NOTIFICATIONS_READ_ALL: String = "notifications_read_all"
// M489 per-kind channel setting. One row per (kind, channel) target
// state, collapsed on that pair, so the newest choice is the one sent.
const val NOTIFICATION_SETTING_SET: String = "notification_setting_set"
}
/**
@@ -85,6 +103,7 @@ data class RequestCreatePayload(
* This matches `feedback_offline_first_for_server_writes` — writes
* never go fire-and-forget.
*/
@Suppress("TooManyFunctions") // one enqueue per mutation kind, like the replayer's dispatchers
@Singleton
class MutationQueue @Inject constructor(
private val dao: CachedMutationDao,
@@ -177,6 +196,31 @@ class MutationQueue @Inject constructor(
),
)
/** Queues the user's whole normalization preference for replay. */
suspend fun enqueueNormalizationSet(prefs: NormalizationPrefs): Long = insertUserDriven(
MutationKind.NORMALIZATION_SET,
json.encodeToString(NormalizationPrefs.serializer(), prefs),
)
suspend fun enqueueNotificationRead(id: String): Long = insertUserDriven(
MutationKind.NOTIFICATION_READ,
json.encodeToString(NotificationReadPayload.serializer(), NotificationReadPayload(id)),
)
suspend fun enqueueNotificationsReadAll(upToIso: String?): Long = insertUserDriven(
MutationKind.NOTIFICATIONS_READ_ALL,
json.encodeToString(
NotificationsReadAllPayload.serializer(),
NotificationsReadAllPayload(upToIso),
),
)
suspend fun enqueueNotificationSettingSet(payload: NotificationSettingPayload): Long =
insertUserDriven(
MutationKind.NOTIFICATION_SETTING_SET,
json.encodeToString(NotificationSettingPayload.serializer(), payload),
)
suspend fun enqueueRequestCancel(requestId: String): Long = insertUserDriven(
MutationKind.REQUEST_CANCEL,
json.encodeToString(
@@ -322,3 +366,36 @@ data class PlaybackErrorReportPayload(
val detail: String? = null,
val clientId: String,
)
/** Persisted payload for `MutationKind.NOTIFICATION_READ` (M489). */
@Serializable
data class NotificationReadPayload(val id: String)
/**
* Persisted payload for `MutationKind.NOTIFICATIONS_READ_ALL` (M489).
* `upToIso` is the newest notice the user could see when they asked; null
* when the inbox was empty on the device, which marks everything.
*/
@Serializable
data class NotificationsReadAllPayload(val upToIso: String?)
/**
* Persisted payload for `MutationKind.NOTIFICATION_SETTING_SET` (M489): one
* kind's one channel, as a target state. `channel` is "inbox" | "phone" |
* "email". No defaults, so every field is always written.
*/
@Serializable
data class NotificationSettingPayload(
val kind: String,
val channel: String,
val value: Boolean,
)
/** The wire change for one queued channel setting, or null for an unknown channel. */
internal fun notificationSettingChange(p: NotificationSettingPayload): NotificationSettingChangeWire? =
when (p.channel) {
"inbox" -> NotificationSettingChangeWire(kind = p.kind, inbox = p.value)
"phone" -> NotificationSettingChangeWire(kind = p.kind, phone = p.value)
"email" -> NotificationSettingChangeWire(kind = p.kind, email = p.value)
else -> null
}
@@ -7,6 +7,10 @@ import com.fabledsword.minstrel.api.endpoints.DiscoverApi
import com.fabledsword.minstrel.api.endpoints.EventsApi
import com.fabledsword.minstrel.api.endpoints.FlagRequest
import com.fabledsword.minstrel.api.endpoints.LikesApi
import com.fabledsword.minstrel.api.endpoints.MeApi
import com.fabledsword.minstrel.api.endpoints.NotificationsApi
import com.fabledsword.minstrel.api.endpoints.PutNotificationSettingsBody
import com.fabledsword.minstrel.api.endpoints.ReadAllBody
import com.fabledsword.minstrel.api.endpoints.PlaybackErrorReportRequest
import com.fabledsword.minstrel.api.endpoints.PlaybackErrorsApi
import com.fabledsword.minstrel.api.endpoints.PlaylistsApi
@@ -22,6 +26,7 @@ import com.fabledsword.minstrel.cache.db.dao.CachedMutationDao
import com.fabledsword.minstrel.cache.db.entities.CachedMutationEntity
import com.fabledsword.minstrel.di.ApplicationScope
import com.fabledsword.minstrel.likes.data.LikesRepository
import com.fabledsword.minstrel.settings.data.NormalizationPrefs
import com.fabledsword.minstrel.models.wire.CreateRequestBody
import kotlinx.coroutines.CoroutineScope
import kotlinx.coroutines.flow.distinctUntilChanged
@@ -79,6 +84,8 @@ class MutationReplayer @Inject constructor(
private val eventsApi: EventsApi = retrofit.create()
private val requestsApi: RequestsApi = retrofit.create()
private val playbackErrorsApi: PlaybackErrorsApi = retrofit.create()
private val meApi: MeApi = retrofit.create()
private val notificationsApi: NotificationsApi = retrofit.create()
private val mutex = Mutex()
@@ -166,10 +173,23 @@ class MutationReplayer @Inject constructor(
MutationKind.REQUEST_CANCEL -> dispatchRequestCancel(row.payload)
MutationKind.PLAYBACK_ERROR_REPORT -> dispatchPlaybackErrorReport(row.payload)
MutationKind.SUGGESTION_SNOOZE_TOGGLE -> dispatchSuggestionSnoozeToggle(row.payload)
MutationKind.NORMALIZATION_SET -> dispatchNormalizationSet(row.payload)
MutationKind.NOTIFICATION_READ,
MutationKind.NOTIFICATIONS_READ_ALL,
MutationKind.NOTIFICATION_SETTING_SET,
-> dispatchNotification(row)
// Unknown kind — drop so a stale schema entry can't wedge the queue.
else -> Outcome.DROP
}
/** The notifications inbox's kinds (M489), split out to keep [dispatch] simple. */
private suspend fun dispatchNotification(row: CachedMutationEntity): Outcome = when (row.kind) {
MutationKind.NOTIFICATION_READ -> dispatchNotificationRead(row.payload)
MutationKind.NOTIFICATIONS_READ_ALL -> dispatchNotificationsReadAll(row.payload)
MutationKind.NOTIFICATION_SETTING_SET -> dispatchNotificationSettingSet(row.payload)
else -> Outcome.DROP
}
private suspend fun dispatchLikeToggle(payload: String): Outcome {
val decoded = json.decodeFromString(LikeTogglePayload.serializer(), payload)
val kindPath = when (decoded.entityType) {
@@ -279,6 +299,37 @@ class MutationReplayer @Inject constructor(
return Outcome.SENT
}
/**
* Sends the queued normalization preference. The device already shows
* it, so the server's echo is not written back: a change made since the
* row was queued would be a newer row, and the collapse keeps only that.
*/
private suspend fun dispatchNormalizationSet(payload: String): Outcome {
meApi.putNormalization(json.decodeFromString(NormalizationPrefs.serializer(), payload))
return Outcome.SENT
}
/** A 404 (the notice was trimmed or already gone) is a 4xx, so DROP: nothing left to do. */
private suspend fun dispatchNotificationRead(payload: String): Outcome {
val decoded = json.decodeFromString(NotificationReadPayload.serializer(), payload)
notificationsApi.markRead(decoded.id)
return Outcome.SENT
}
private suspend fun dispatchNotificationsReadAll(payload: String): Outcome {
val decoded = json.decodeFromString(NotificationsReadAllPayload.serializer(), payload)
notificationsApi.readAll(ReadAllBody(upTo = decoded.upToIso))
return Outcome.SENT
}
/** An unknown channel can only come from a corrupt row: DROP it. */
private suspend fun dispatchNotificationSettingSet(payload: String): Outcome {
val decoded = json.decodeFromString(NotificationSettingPayload.serializer(), payload)
val change = notificationSettingChange(decoded) ?: return Outcome.DROP
notificationsApi.putSettings(PutNotificationSettingsBody(listOf(change)))
return Outcome.SENT
}
private suspend fun dispatchPlaybackErrorReport(payload: String): Outcome {
val decoded = json.decodeFromString(PlaybackErrorReportPayload.serializer(), payload)
playbackErrorsApi.report(
@@ -303,7 +354,8 @@ class MutationReplayer @Inject constructor(
/**
* Row ids of desired-state toggles superseded by a later toggle for the same
* entity. Applies to every kind whose payload encodes a TARGET state rather
* than an action — like-toggles and suggestion snoozes (#2374) — because
* than an action — like-toggles, suggestion snoozes (#2374) and the
* normalization preference (#4998) — because
* replaying a stale one last would invert the final state.
*
* Top-level and pure so it can be unit-tested without standing up a Retrofit
@@ -340,5 +392,14 @@ private fun toggleKeyOf(row: CachedMutationEntity, json: Json): String? = when (
json.decodeFromString(SuggestionSnoozeTogglePayload.serializer(), row.payload)
}.getOrNull()?.let { "${row.kind}:${it.mbid}" }
MutationKind.NOTIFICATION_SETTING_SET -> runCatching {
json.decodeFromString(NotificationSettingPayload.serializer(), row.payload)
}.getOrNull()?.let { "${row.kind}:${it.kind}:${it.channel}" }
// One preference per user, so every normalization row shares one key.
MutationKind.NORMALIZATION_SET -> runCatching {
json.decodeFromString(NormalizationPrefs.serializer(), row.payload)
}.getOrNull()?.let { row.kind }
else -> null
}
@@ -206,6 +206,8 @@ private fun SyncAlbumWire.toEntity(): CachedAlbumEntity = CachedAlbumEntity(
releaseDate = releaseDate,
coverPath = coverArtPath,
mbid = mbid,
albumGain = albumGain,
albumPeak = albumPeak,
)
private fun SyncTrackWire.toEntity(): CachedTrackEntity = CachedTrackEntity(
@@ -220,4 +222,6 @@ private fun SyncTrackWire.toEntity(): CachedTrackEntity = CachedTrackEntity(
fileFormat = fileFormat,
genre = genre,
missing = missing,
trackGain = trackGain,
trackPeak = trackPeak,
)
@@ -1,14 +1,18 @@
package com.fabledsword.minstrel.events
import com.fabledsword.minstrel.auth.AuthStore
import com.fabledsword.minstrel.connectivity.ConnectivityObserver
import com.fabledsword.minstrel.di.ApplicationScope
import kotlinx.coroutines.CoroutineScope
import kotlinx.coroutines.Job
import kotlinx.coroutines.channels.BufferOverflow
import kotlinx.coroutines.delay
import kotlinx.coroutines.flow.MutableSharedFlow
import kotlinx.coroutines.flow.MutableStateFlow
import kotlinx.coroutines.flow.SharedFlow
import kotlinx.coroutines.flow.StateFlow
import kotlinx.coroutines.flow.asSharedFlow
import kotlinx.coroutines.flow.asStateFlow
import kotlinx.coroutines.flow.distinctUntilChanged
import kotlinx.coroutines.flow.map
import kotlinx.coroutines.launch
@@ -28,9 +32,6 @@ import javax.inject.Singleton
private const val SSE_PATH = "/api/events/stream"
private const val EVENTS_BUFFER_CAPACITY = 64
private const val BASE_BACKOFF_MS = 1_000L
private const val MAX_BACKOFF_MS = 30_000L
private const val BACKOFF_FACTOR = 2
/**
* Long-lived SSE subscription to `GET /api/events/stream`. Exposes
@@ -45,11 +46,16 @@ private const val BACKOFF_FACTOR = 2
* - No client-side timeout — the server emits 15s heartbeats which
* okhttp-sse handles transparently.
* - Reconnect-with-backoff: if the stream drops mid-session (server
* restart, network blip) it reconnects with exponential backoff
* (1s → 2s → … → 30s cap), reset to 1s on a successful open. Only
* restart, network blip) it reconnects after [ReconnectBackoff]'s
* jittered wait (2s doubling to 5 min), reset on a successful open.
* A network coming up reconnects at once: the callback is a hint, and
* the only test of whether the server is reachable is trying it. Only
* reconnects while still signed in; a sign-out cancels the pending
* retry. Without this a single blip silently kills cross-device
* reactivity until the next app launch.
* - [connected] says whether a stream is open. Background delivery
* (M489 #5347) catches up on every rising edge: nothing replays a
* frame sent while the stream was down.
*
* The URL passes through the placeholder host that
* `BaseUrlInterceptor` rewrites — same mechanism the rest of the
@@ -62,6 +68,7 @@ class EventsStream @Inject constructor(
@ApplicationScope private val scope: CoroutineScope,
private val okHttpClient: OkHttpClient,
private val json: Json,
private val connectivity: ConnectivityObserver,
) {
private val factory = EventSources.createFactory(okHttpClient)
@@ -72,10 +79,13 @@ class EventsStream @Inject constructor(
)
val events: SharedFlow<LiveEvent> = emitter.asSharedFlow()
private val connectedState = MutableStateFlow(false)
val connected: StateFlow<Boolean> = connectedState.asStateFlow()
private var currentSource: EventSource? = null
@Volatile private var signedIn = false
private var reconnectJob: Job? = null
private var backoffMs = BASE_BACKOFF_MS
private var backoffMs = ReconnectBackoff.BASE_MS
init {
scope.launch {
@@ -85,13 +95,30 @@ class EventsStream @Inject constructor(
.collect { isSignedIn ->
signedIn = isSignedIn
if (isSignedIn) {
backoffMs = BASE_BACKOFF_MS
backoffMs = ReconnectBackoff.BASE_MS
connect()
} else {
disconnect()
}
}
}
scope.launch {
connectivity.online.collect { up -> if (up) reconnectNow() }
}
}
/**
* Cuts a pending backoff short: reconnects at once and starts the ladder
* over. For a network that has just come up, or the app coming to the
* foreground, where waiting out a five-minute backoff would leave the
* server unheard from for nothing. Does nothing while a stream is open or
* opening.
*/
@Synchronized
fun reconnectNow() {
if (!signedIn || reconnectJob?.isActive != true) return
backoffMs = ReconnectBackoff.BASE_MS
connect()
}
@Synchronized
@@ -106,20 +133,22 @@ class EventsStream @Inject constructor(
reconnectJob?.cancel()
currentSource?.cancel()
currentSource = null
connectedState.value = false
}
/**
* Schedule a reconnect after the current backoff, then double it
* (capped). No-op when signed out — sign-out's [disconnect]
* Schedule a reconnect after the current backoff, jittered, then
* double it (capped). No-op when signed out — sign-out's [disconnect]
* cancels the pending job. A successful [Listener.onOpen] resets
* the backoff to the floor.
*/
@Synchronized
private fun scheduleReconnect() {
connectedState.value = false
if (!signedIn) return
reconnectJob?.cancel()
val waitMs = backoffMs
backoffMs = (backoffMs * BACKOFF_FACTOR).coerceAtMost(MAX_BACKOFF_MS)
val waitMs = ReconnectBackoff.jittered(backoffMs)
backoffMs = ReconnectBackoff.next(backoffMs)
reconnectJob = scope.launch {
delay(waitMs)
if (signedIn) connect()
@@ -136,7 +165,8 @@ class EventsStream @Inject constructor(
private inner class Listener : EventSourceListener() {
override fun onOpen(eventSource: EventSource, response: Response) {
backoffMs = BASE_BACKOFF_MS
backoffMs = ReconnectBackoff.BASE_MS
connectedState.value = true
}
override fun onEvent(
@@ -5,6 +5,7 @@ import androidx.lifecycle.LifecycleOwner
import androidx.lifecycle.ProcessLifecycleOwner
import com.fabledsword.minstrel.di.ApplicationScope
import com.fabledsword.minstrel.likes.data.LikesRepository
import com.fabledsword.minstrel.notifications.data.NotificationsRepository
import kotlinx.coroutines.CoroutineScope
import kotlinx.coroutines.launch
import javax.inject.Inject
@@ -30,6 +31,7 @@ import javax.inject.Singleton
class LiveEventsDispatcher @Inject constructor(
private val eventsStream: EventsStream,
private val likes: LikesRepository,
private val notifications: NotificationsRepository,
@ApplicationScope private val scope: CoroutineScope,
) : DefaultLifecycleObserver {
@@ -49,6 +51,8 @@ class LiveEventsDispatcher @Inject constructor(
"artist.liked",
"artist.unliked",
-> refreshLikes()
// M489: a contentless nudge; the inbox refetches its newest page.
"notification.created" -> refreshNotifications()
}
// Other kinds (playlist.*, quarantine.*, request.status_changed,
// scan.*) reach screen-scoped subscribers via EventsStream
@@ -58,9 +62,18 @@ class LiveEventsDispatcher @Inject constructor(
override fun onStart(owner: LifecycleOwner) {
// App returned to the foreground. SSE will catch up but might
// not have reconnected yet; flush the cross-screen refreshes
// not have reconnected yet (it may be deep in its backoff, so it
// is told to try now); flush the cross-screen refreshes
// defensively.
eventsStream.reconnectNow()
refreshLikes()
refreshNotifications()
}
private fun refreshNotifications() {
scope.launch {
runCatching { notifications.refresh() }
}
}
private fun refreshLikes() {
@@ -0,0 +1,31 @@
package com.fabledsword.minstrel.events
import kotlin.random.Random
/**
* How long [EventsStream] waits before reconnecting (M489 #5347).
*
* The stream is wanted around the clock once notifications arrive with the app
* closed, so a server that is down overnight must not cost a radio wakeup
* every few seconds: Roundtable's flat 2s came to about 43,000 of them. The
* wait doubles from [BASE_MS] to [MAX_MS]. The jitter spreads the moment every
* phone on the server reconnects, which is when the server has just come back
* and can least take a spike.
*
* A network coming up (a hint, never a gate) or the app coming to the
* foreground reconnects at once and starts the ladder over.
*/
internal object ReconnectBackoff {
const val BASE_MS = 2_000L
const val MAX_MS = 300_000L
private const val JITTER = 0.25
/** The wait after [currentMs], before jitter. */
fun next(currentMs: Long): Long = (currentMs * 2).coerceAtMost(MAX_MS)
/** [ms] moved by up to a quarter either way. */
fun jittered(ms: Long, random: Random = Random.Default): Long {
val spread = ms * JITTER
return (ms + random.nextDouble(-spread, spread)).toLong()
}
}
@@ -1,15 +1,26 @@
package com.fabledsword.minstrel.models
/**
* Wire shape returned by `GET /api/client/version`. Mirrors
* the Flutter client's `UpdateInfo`.
* The server-bundled APK, as reported by `GET /api/client/version`.
*
* `version` is the server-bundled APK version (may have a leading
* "v" from the git tag); `apkUrl` is server-relative (e.g.
* `/api/client/apk`); `sizeBytes` is the download size.
* Three values that are deliberately kept apart:
*
* - [version] is a LABEL for people — "YYYY.MM.DD.HHMM", derived from the
* build's commit, so two channels carrying the same code read the same.
* Display this; never decide on it when [code] is present.
* - [code] is the ORDERING KEY, and is the same value Android itself
* installs by. It answers "may this be installed over that?", which the
* name cannot. Null when the server predates the field.
* - [channel] is a SIBLING FIELD, never a suffix inside the name. Reported
* verbatim rather than validated, so an unexpected value is shown rather
* than dropped.
*
* [apkUrl] is server-relative (e.g. `/api/client/apk`).
*/
data class UpdateInfo(
val version: String,
val code: Long?,
val channel: String?,
val apkUrl: String,
val sizeBytes: Long,
)
@@ -0,0 +1,18 @@
package com.fabledsword.minstrel.models.wire
import kotlinx.serialization.SerialName
import kotlinx.serialization.Serializable
/** One track's entry in `GET /api/tracks/replay-gain` (#4997). Decode-only. */
@Serializable
data class ReplayGainWire(
@SerialName("track_gain") val trackGain: Float? = null,
@SerialName("track_peak") val trackPeak: Float? = null,
@SerialName("album_gain") val albumGain: Float? = null,
@SerialName("album_peak") val albumPeak: Float? = null,
)
@Serializable
data class ReplayGainResponseWire(
val items: Map<String, ReplayGainWire> = emptyMap(),
)
@@ -28,6 +28,10 @@ data class SyncAlbumWire(
@SerialName("release_date") val releaseDate: String? = null,
@SerialName("cover_art_path") val coverArtPath: String? = null,
val mbid: String? = null,
// The album's ReplayGain 2.0 values (#4997): dB to -18 LUFS and a linear
// peak. Null until every track on the album is measured.
@SerialName("album_gain") val albumGain: Float? = null,
@SerialName("album_peak") val albumPeak: Float? = null,
)
@Serializable
@@ -51,6 +55,9 @@ data class SyncTrackWire(
// its tracks stay playable, which is the correct reading of "this server
// has nothing to say about missing files".
val missing: Boolean = false,
// The track's ReplayGain 2.0 values (#4997); null until it is measured.
@SerialName("track_gain") val trackGain: Float? = null,
@SerialName("track_peak") val trackPeak: Float? = null,
)
/**
@@ -4,12 +4,26 @@ import kotlinx.serialization.SerialName
import kotlinx.serialization.Serializable
/**
* Wire shape for `GET /api/client/version`. Defaults match Flutter:
* apk_url falls back to `/api/client/apk` if the server omits it.
* Wire shape for `GET /api/client/version`.
*
* `apkUrl` falls back to `/api/client/apk` if the server omits it.
*
* [code] MUST stay nullable, and this is not a style preference. The app's
* Json is configured with `coerceInputValues = true`, which replaces a JSON
* null with the declared default for a NON-nullable property — so writing
* `val code: Long = 0` would turn "this server reports no ordering key" into
* "this build's ordering key is 0", silently, with no error anywhere. A
* nullable type is what keeps absent distinguishable from zero, and the
* distinction is the whole reason the field exists.
*
* A server predating the ordering key sends neither [code] nor [channel];
* both arrive null and the caller falls back to comparing names.
*/
@Serializable
data class UpdateInfoWire(
val version: String = "",
val code: Long? = null,
val channel: String? = null,
@SerialName("apk_url") val apkUrl: String = "/api/client/apk",
@SerialName("size_bytes") val sizeBytes: Long = 0,
)
@@ -25,6 +25,8 @@ import com.fabledsword.minstrel.home.ui.HomeScreen
import com.fabledsword.minstrel.library.ui.AlbumDetailScreen
import com.fabledsword.minstrel.library.ui.ArtistDetailScreen
import com.fabledsword.minstrel.library.ui.LibraryScreen
import com.fabledsword.minstrel.notifications.ui.NotificationSettingsScreen
import com.fabledsword.minstrel.notifications.ui.NotificationsScreen
import com.fabledsword.minstrel.player.ui.NowPlayingScreen
import com.fabledsword.minstrel.player.ui.QueueScreen
import com.fabledsword.minstrel.playlists.ui.PlaylistDetailScreen
@@ -59,6 +61,7 @@ fun MinstrelNavGraph(
) {
inShellTopLevel(navController, expandPlayer)
inShellDetail(navController, expandPlayer)
inShellNotifications(navController, expandPlayer)
outsideShell(navController)
}
}
@@ -200,6 +203,27 @@ private fun NavGraphBuilder.inShellDetail(
}
}
/** The notifications inbox and its settings (M489). */
private fun NavGraphBuilder.inShellNotifications(
navController: NavHostController,
expandPlayer: () -> Unit,
) {
composable<Notifications> {
WithAnimatedScope {
ShellScaffold(onExpandPlayer = expandPlayer) {
NotificationsScreen(navController = navController)
}
}
}
composable<NotificationSettings> {
WithAnimatedScope {
ShellScaffold(onExpandPlayer = expandPlayer) {
NotificationSettingsScreen(navController = navController)
}
}
}
}
private fun NavGraphBuilder.outsideShell(navController: NavHostController) {
composable<NowPlaying>(
// Slide up from the bottom on enter; back down on dismiss.
@@ -13,6 +13,8 @@ import kotlinx.serialization.Serializable
@Serializable data object Settings
@Serializable data object Admin
@Serializable data object Requests
@Serializable data object Notifications
@Serializable data object NotificationSettings
// ── In-shell detail / push-on-top destinations ────────────────────────
@@ -0,0 +1,104 @@
package com.fabledsword.minstrel.notifications.data
import com.fabledsword.minstrel.api.endpoints.NotificationSettingsWire
import com.fabledsword.minstrel.api.endpoints.NotificationsApi
import com.fabledsword.minstrel.api.endpoints.PutNotificationSettingsBody
import com.fabledsword.minstrel.cache.db.dao.CachedMutationDao
import com.fabledsword.minstrel.cache.db.dao.CachedNotificationDao
import com.fabledsword.minstrel.cache.db.entities.CachedNotificationSettingsEntity
import com.fabledsword.minstrel.cache.mutations.MutationKind
import com.fabledsword.minstrel.cache.mutations.MutationQueue
import com.fabledsword.minstrel.cache.mutations.NotificationSettingPayload
import com.fabledsword.minstrel.cache.mutations.notificationSettingChange
import kotlinx.coroutines.CancellationException
import kotlinx.coroutines.flow.Flow
import kotlinx.coroutines.flow.map
import kotlinx.serialization.json.Json
import retrofit2.Retrofit
import retrofit2.create
import timber.log.Timber
import javax.inject.Inject
import javax.inject.Singleton
/** A notification channel, as the server names it. */
enum class NotificationChannel(val wire: String) { INBOX("inbox"), PHONE("phone"), EMAIL("email") }
/**
* Per-user notification settings (M489), following snippet #5107: the device
* copy (a cached JSON row) shows a change at once, the PUT is best effort,
* and a failed or out-of-order one is queued for the MutationReplayer.
*/
@Singleton
class NotificationSettingsRepository @Inject constructor(
retrofit: Retrofit,
private val dao: CachedNotificationDao,
private val mutationDao: CachedMutationDao,
private val mutationQueue: MutationQueue,
private val json: Json,
) {
private val api: NotificationsApi = retrofit.create()
val settings: Flow<NotificationSettingsWire?> = dao.observeSettings().map { it?.let(::decode) }
/**
* Takes the server's settings unless a change made here is still queued:
* the server has not seen it, and its older value would undo the change
* on screen until the replay lands. Throws on a network failure.
*/
suspend fun refresh() {
val server = api.getSettings()
if (mutationDao.hasPending(MutationKind.NOTIFICATION_SETTING_SET)) return
save(server)
}
suspend fun set(kind: String, channel: NotificationChannel, value: Boolean) {
current()?.let { save(it.withChannel(kind, channel, value)) }
val payload = NotificationSettingPayload(kind = kind, channel = channel.wire, value = value)
// An earlier change still queued would replay after this PUT; queue
// behind it instead, and the replayer sends the newest per channel.
if (mutationDao.hasPending(MutationKind.NOTIFICATION_SETTING_SET)) {
mutationQueue.enqueueNotificationSettingSet(payload)
return
}
val change = notificationSettingChange(payload) ?: return
try {
save(api.putSettings(PutNotificationSettingsBody(listOf(change))))
} catch (e: CancellationException) {
throw e
} catch (
@Suppress("TooGenericExceptionCaught") e: Throwable,
) {
Timber.i(e, "notification settings: PUT failed; queued")
mutationQueue.enqueueNotificationSettingSet(payload)
}
}
private suspend fun current(): NotificationSettingsWire? = dao.getSettings()?.let(::decode)
/** Null for an unreadable row: the next refresh writes a good one. */
private fun decode(row: CachedNotificationSettingsEntity): NotificationSettingsWire? =
runCatching { json.decodeFromString(NotificationSettingsWire.serializer(), row.json) }.getOrNull()
private suspend fun save(s: NotificationSettingsWire) {
val encoded = json.encodeToString(NotificationSettingsWire.serializer(), s)
dao.upsertSettings(CachedNotificationSettingsEntity(json = encoded))
}
}
internal fun NotificationSettingsWire.withChannel(
kind: String,
channel: NotificationChannel,
value: Boolean,
): NotificationSettingsWire = copy(
kinds = kinds.map { k ->
if (k.kind != kind) {
k
} else {
when (channel) {
NotificationChannel.INBOX -> k.copy(inbox = value)
NotificationChannel.PHONE -> k.copy(phone = value)
NotificationChannel.EMAIL -> k.copy(email = value)
}
}
},
)
@@ -0,0 +1,132 @@
package com.fabledsword.minstrel.notifications.data
import com.fabledsword.minstrel.api.endpoints.NotificationWire
import com.fabledsword.minstrel.api.endpoints.NotificationsApi
import com.fabledsword.minstrel.api.endpoints.ReadAllBody
import com.fabledsword.minstrel.cache.db.dao.CachedNotificationDao
import com.fabledsword.minstrel.cache.db.entities.CachedNotificationEntity
import com.fabledsword.minstrel.cache.mutations.MutationQueue
import kotlinx.coroutines.CancellationException
import kotlinx.coroutines.flow.Flow
import kotlinx.coroutines.flow.distinctUntilChanged
import kotlinx.datetime.Clock
import kotlinx.datetime.Instant
import retrofit2.HttpException
import retrofit2.Retrofit
import retrofit2.create
import timber.log.Timber
import javax.inject.Inject
import javax.inject.Singleton
import kotlin.time.Duration.Companion.milliseconds
/**
* The notifications inbox (M489). The newest page lives in Room so the bell's
* badge and the Notifications screen work offline. Reads follow rule 100: the
* device marks the row at once, the call is best effort, and a failed call is
* queued for the MutationReplayer.
*/
@Singleton
class NotificationsRepository @Inject constructor(
retrofit: Retrofit,
private val dao: CachedNotificationDao,
private val mutationQueue: MutationQueue,
) {
private val api: NotificationsApi = retrofit.create()
val notifications: Flow<List<CachedNotificationEntity>> = dao.observeAll().distinctUntilChanged()
/** Unread notices in the cached page, which is what the badge shows. */
val unreadCount: Flow<Int> = dao.observeUnreadCount().distinctUntilChanged()
/**
* Replaces the cache with the server's newest page. A read made on this
* device and not yet replayed stays read: a read never goes back to
* unread, so the device's mark wins over the server's older view.
* Throws on a network failure; the cache then stands.
*/
suspend fun refresh() {
val page = api.list(limit = PAGE_SIZE)
val localReads = dao.getAll().associate { it.id to it.readAt }
dao.replaceAll(page.items.mapNotNull { it.toEntity(localReads[it.id]) })
}
suspend fun markRead(id: String) {
dao.markRead(id, Clock.System.now())
try {
api.markRead(id)
} catch (e: CancellationException) {
throw e
} catch (e: HttpException) {
// A 4xx (404: trimmed or gone server-side) leaves nothing to mark;
// a 5xx is the server's trouble, so the read waits in the queue.
if (e.code() >= HTTP_SERVER_ERROR) {
mutationQueue.enqueueNotificationRead(id)
} else {
Timber.i(e, "notifications: mark read refused (%d)", e.code())
}
} catch (
@Suppress("TooGenericExceptionCaught") e: Throwable,
) {
Timber.i(e, "notifications: mark read failed; queued")
mutationQueue.enqueueNotificationRead(id)
}
}
/**
* Marks everything the device holds read. The server is asked to mark
* only what existed up to the newest notice shown here, so a replay that
* lands later leaves anything newer unread.
*/
suspend fun markAllRead() {
val upTo = readAllCutoff(dao.getAll().map { it.createdAt })?.toString()
dao.markAllRead(Clock.System.now())
try {
api.readAll(ReadAllBody(upTo = upTo))
} catch (e: CancellationException) {
throw e
} catch (
@Suppress("TooGenericExceptionCaught") e: Throwable,
) {
Timber.i(e, "notifications: mark all read failed; queued")
mutationQueue.enqueueNotificationsReadAll(upTo)
}
}
/** The page the device holds, newest first. */
suspend fun cachedPage(): List<CachedNotificationEntity> = dao.getAll()
/** On sign-out: the next account on this device must not see these. */
suspend fun clearLocal() {
dao.clear()
}
companion object {
/** The newest page the device keeps; older notices stay on the server. */
const val PAGE_SIZE = 50
private const val HTTP_SERVER_ERROR = 500
}
}
/**
* The newest notice shown, rounded up a millisecond. Room keeps milliseconds
* and the server microseconds, so the stored time can sit just before the
* server's; without rounding up, the newest notice would stay unread.
*/
internal fun readAllCutoff(createdAts: List<Instant>): Instant? =
createdAts.maxOrNull()?.plus(1.milliseconds)
/** Null for a row whose timestamp will not parse, rather than failing the page. */
internal fun NotificationWire.toEntity(localReadAt: Instant?): CachedNotificationEntity? {
val created = runCatching { Instant.parse(createdAt) }.getOrNull() ?: return null
val serverRead = readAt?.let { runCatching { Instant.parse(it) }.getOrNull() }
return CachedNotificationEntity(
id = id,
kind = kind,
title = title,
body = body,
link = link,
createdAt = created,
readAt = serverRead ?: localReadAt,
)
}
@@ -0,0 +1,46 @@
package com.fabledsword.minstrel.notifications.delivery
import android.content.BroadcastReceiver
import android.content.Context
import android.content.Intent
import com.fabledsword.minstrel.auth.AuthStore
import com.fabledsword.minstrel.di.ApplicationScope
import dagger.hilt.android.AndroidEntryPoint
import kotlinx.coroutines.CoroutineScope
import kotlinx.coroutines.launch
import javax.inject.Inject
/**
* Starts background delivery after a reboot, and after the app updates itself
* (M489 #5347). Without it, notifications stop until the user next opens the
* app, which is exactly when they least need telling.
*
* Both broadcasts are among the few exemptions allowed to start a foreground
* service from the background, so the start happens here, inside the
* receiver's window, rather than whenever [DeliveryLauncher] next observes it.
*/
@AndroidEntryPoint
class BootReceiver : BroadcastReceiver() {
@Inject lateinit var authStore: AuthStore
@Inject @ApplicationScope lateinit var scope: CoroutineScope
override fun onReceive(context: Context, intent: Intent) {
if (intent.action != Intent.ACTION_BOOT_COMPLETED && intent.action != Intent.ACTION_MY_PACKAGE_REPLACED) return
val pending = goAsync()
scope.launch {
try {
// Bounded (rule 156): the session load gives up after its
// own deadline and answers from what it has.
authStore.awaitSessionHydrated()
val signedIn = !authStore.sessionCookie.value.isNullOrEmpty()
if (deliveryWanted(signedIn, authStore.backgroundDelivery.value)) {
DeliveryService.start(context)
}
} finally {
pending.finish()
}
}
}
}
@@ -0,0 +1,69 @@
package com.fabledsword.minstrel.notifications.delivery
import android.content.Context
import com.fabledsword.minstrel.auth.AuthStore
import com.fabledsword.minstrel.di.ApplicationScope
import com.fabledsword.minstrel.events.EventsStream
import dagger.hilt.android.qualifiers.ApplicationContext
import kotlinx.coroutines.CoroutineScope
import kotlinx.coroutines.flow.combine
import kotlinx.coroutines.flow.distinctUntilChanged
import kotlinx.coroutines.flow.filter
import kotlinx.coroutines.flow.map
import kotlinx.coroutines.launch
import javax.inject.Inject
import javax.inject.Singleton
/**
* Phone notifications (M489 #5347). Activated by force-@Inject in
* MinstrelApplication, and the single owner of [DeliveryService]'s lifetime:
* it runs exactly while signed in with "Notifications when the app is closed"
* on. Signing out stops it, so a logged-out device holds no connection.
*
* The catch-up runs here, in the process, whether or not the service does:
* with background delivery off, notifications still reach the shade while the
* process is alive (the app open, or music playing).
*/
@Singleton
class DeliveryLauncher @Inject constructor(
@ApplicationContext private val context: Context,
authStore: AuthStore,
eventsStream: EventsStream,
sync: NotificationSync,
@ApplicationScope scope: CoroutineScope,
) {
init {
val signedIn = authStore.sessionCookie.map { !it.isNullOrEmpty() }.distinctUntilChanged()
scope.launch {
combine(signedIn, authStore.backgroundDelivery, ::deliveryWanted)
.distinctUntilChanged()
.collect { wanted -> if (wanted) DeliveryService.start(context) else DeliveryService.stop(context) }
}
// A nudge and a reconnect mean the same thing: go and look. The
// stream only runs while signed in, so neither fires signed out.
scope.launch {
eventsStream.events
.filter { it.kind == NOTIFICATION_CREATED }
.collect { sync.catchUp() }
}
scope.launch {
eventsStream.connected.filter { it }.collect { sync.catchUp() }
}
// A sign-out, not the signed-out start a cold launch begins with
// before the session loads: that would wipe the mark on every start.
scope.launch {
var wasSignedIn = false
signedIn.collect { now ->
if (wasSignedIn && !now) sync.forget()
wasSignedIn = now
}
}
}
private companion object {
const val NOTIFICATION_CREATED = "notification.created"
}
}
@@ -0,0 +1,94 @@
package com.fabledsword.minstrel.notifications.delivery
import com.fabledsword.minstrel.api.endpoints.NotificationSettingsWire
import com.fabledsword.minstrel.cache.db.entities.CachedNotificationEntity
import kotlinx.datetime.Instant
/**
* The decisions behind phone notifications (M489 #5347), kept free of Android
* so they can be tested as plain functions.
*/
/** A handful each get a line; more than this become one line with a count. */
internal const val INDIVIDUAL_LIMIT = 3
/** What a catch-up announces, and the high-water mark to keep afterwards. */
internal data class CatchUp(
val announce: List<CachedNotificationEntity>,
val upTo: Instant?,
)
/**
* Picks what to announce from the newest page.
*
* - With no mark yet (a fresh sign-in on this device), the mark is set to the
* newest notice and nothing is announced: the user is looking at the app,
* and a backlog in the shade is noise. An empty inbox sets the mark to the
* start of time, so the very first notice is announced.
* - Otherwise: unread notices newer than the mark, whose kind has the phone
* on, oldest first so the newest lands on top of the shade.
* - The mark moves to the newest notice in the page, announced or not, so a
* kind with the phone off is never announced later.
*
* A coalesced notice (tracks missing, now 5) moves its time forward when its
* count grows, so it is announced again with the new count.
*/
internal fun planCatchUp(
page: List<CachedNotificationEntity>,
mark: Instant?,
phoneOn: (String) -> Boolean,
): CatchUp {
val newest = page.maxOfOrNull { it.createdAt }
if (mark == null) return CatchUp(emptyList(), newest ?: Instant.DISTANT_PAST)
val fresh = page
.filter { it.readAt == null && it.createdAt > mark && phoneOn(it.kind) }
.sortedBy { it.createdAt }
val upTo = if (newest != null && newest > mark) newest else mark
return CatchUp(fresh, upTo)
}
/** How a catch-up reaches the shade. */
internal sealed interface Announcement {
data class Each(val items: List<CachedNotificationEntity>) : Announcement
/** Coming back from a day offline is one line, not forty buzzes. */
data class Pile(val count: Int, val channel: ShadeChannel) : Announcement
}
internal fun announcementFor(fresh: List<CachedNotificationEntity>): Announcement? = when {
fresh.isEmpty() -> null
fresh.size <= INDIVIDUAL_LIMIT -> Announcement.Each(fresh)
// A pile goes where its notices would: library health only when every
// one of them is, so a listener's own news is never filed under it.
fresh.all { shadeChannelFor(it.kind) == ShadeChannel.LIBRARY_HEALTH } ->
Announcement.Pile(fresh.size, ShadeChannel.LIBRARY_HEALTH)
else -> Announcement.Pile(fresh.size, ShadeChannel.YOUR_REQUESTS)
}
/** The phone channel per kind, as the user set it. No settings cached: on, the default for every kind. */
internal fun phoneOnFor(settings: NotificationSettingsWire?): (String) -> Boolean {
val byKind = settings?.kinds?.associateBy { it.kind }.orEmpty()
return { kind -> byKind[kind]?.let { it.inbox && it.phone } ?: true }
}
/**
* Android channels, so either audience can be silenced in system settings as
* well as in Minstrel's. The ids are permanent: renaming one orphans the
* user's system-level choice for it.
*/
internal enum class ShadeChannel(val id: String, val title: String, val description: String) {
YOUR_REQUESTS("your_requests", "Your requests", "Approved, declined, and new music arriving"),
LIBRARY_HEALTH("library_health", "Library health", "For admins: requests to review and library problems"),
}
private val requesterKinds = setOf("request_approved", "request_rejected", "request_completed")
/** Everything a listener can receive is theirs; the rest is admin work. */
internal fun shadeChannelFor(kind: String): ShadeChannel =
if (kind in requesterKinds) ShadeChannel.YOUR_REQUESTS else ShadeChannel.LIBRARY_HEALTH
/** The delivery service runs exactly while signed in with background delivery on. */
internal fun deliveryWanted(signedIn: Boolean, enabled: Boolean): Boolean = signedIn && enabled
/** One line for a pile. */
internal fun pileText(count: Int): String = "$count new notifications"
@@ -0,0 +1,124 @@
package com.fabledsword.minstrel.notifications.delivery
import android.app.Notification
import android.app.NotificationChannel
import android.app.NotificationManager
import android.app.PendingIntent
import android.app.Service
import android.content.Context
import android.content.Intent
import android.content.pm.ServiceInfo
import android.os.Build
import android.os.IBinder
import androidx.core.app.NotificationCompat
import androidx.core.app.ServiceCompat
import com.fabledsword.minstrel.MainActivity
import com.fabledsword.minstrel.R
import timber.log.Timber
/**
* Keeps Minstrel's process alive so notifications arrive with the app closed
* (M489 #5347, Roundtable's DeliveryService).
*
* **Why a held connection instead of push.** The APK installs from the
* Minstrel server itself, not the Play Store, so nothing pushes the app onto
* Firebase; the alternatives cost more (UnifiedPush's only embedded
* distributor is Firebase, and ntfy is a second app to install and set up).
*
* **This service holds nothing itself.** [com.fabledsword.minstrel.events.EventsStream]
* stays connected for as long as the process lives and someone is signed in,
* and [DeliveryLauncher] runs the catch-up on every nudge and reconnect. All
* this does is keep the process from being reclaimed. While music plays the
* player's own foreground service does that too; this one still runs, so
* delivery does not depend on playback.
*
* **Typed `specialUse`, and that is not arbitrary.** Android 15 stops a
* `dataSync` service after six hours, and `shortService` is capped at three
* minutes. `specialUse` is the only type that may run as long as the user
* wants it to.
*
* The quiet notice in the shade is the honest price of not using Google's
* push, and the settings screen says so.
*/
class DeliveryService : Service() {
override fun onBind(intent: Intent?): IBinder? = null
override fun onStartCommand(intent: Intent?, flags: Int, startId: Int): Int {
if (intent?.action == ACTION_STOP) {
ServiceCompat.stopForeground(this, ServiceCompat.STOP_FOREGROUND_REMOVE)
stopSelf()
return START_NOT_STICKY
}
ensureChannel()
val type = if (Build.VERSION.SDK_INT >= Build.VERSION_CODES.UPSIDE_DOWN_CAKE) {
ServiceInfo.FOREGROUND_SERVICE_TYPE_SPECIAL_USE
} else {
0
}
ServiceCompat.startForeground(this, NOTIFICATION_ID, ongoing(), type)
// If the system reclaims the process, bring it back: the point is to
// be running when nothing else is.
return START_STICKY
}
private fun ongoing(): Notification {
val open = PendingIntent.getActivity(
this,
0,
Intent(this, MainActivity::class.java),
PendingIntent.FLAG_IMMUTABLE,
)
return NotificationCompat.Builder(this, CHANNEL_ID)
.setSmallIcon(R.drawable.ic_notification)
.setContentTitle("Minstrel")
.setContentText("Listening for notifications")
.setOngoing(true)
.setContentIntent(open)
// MIN: at the bottom of the shade, no sound, no heads-up. It has
// to exist; it does not have to be loud.
.setPriority(NotificationCompat.PRIORITY_MIN)
.build()
}
private fun ensureChannel() {
val manager = getSystemService(NotificationManager::class.java) ?: return
manager.createNotificationChannel(
NotificationChannel(CHANNEL_ID, "Background connection", NotificationManager.IMPORTANCE_MIN)
.apply { description = "The quiet notice shown while Minstrel listens for notifications" },
)
}
companion object {
const val ACTION_STOP = "com.fabledsword.minstrel.DELIVERY_STOP"
private const val CHANNEL_ID = "background_connection"
private const val NOTIFICATION_ID = 4801
/**
* Starting a foreground service from the background is refused on
* Android 12+ outside the exemptions (boot, an update, the app on
* screen). A refusal is logged, not thrown: the app's next start, or
* the next boot, tries again.
*/
fun start(context: Context) {
try {
context.startForegroundService(Intent(context, DeliveryService::class.java))
} catch (
@Suppress("TooGenericExceptionCaught") e: RuntimeException,
) {
Timber.w(e, "delivery service: start refused")
}
}
fun stop(context: Context) {
try {
context.startService(Intent(context, DeliveryService::class.java).setAction(ACTION_STOP))
} catch (
@Suppress("TooGenericExceptionCaught") e: RuntimeException,
) {
// Not running and not startable from here: nothing to stop.
Timber.i(e, "delivery service: stop not delivered")
}
}
}
}
@@ -0,0 +1,86 @@
package com.fabledsword.minstrel.notifications.delivery
import android.annotation.SuppressLint
import android.app.NotificationChannel
import android.app.NotificationManager
import android.app.PendingIntent
import android.content.Context
import android.content.Intent
import androidx.core.app.NotificationCompat
import androidx.core.app.NotificationManagerCompat
import com.fabledsword.minstrel.MainActivity
import com.fabledsword.minstrel.R
import com.fabledsword.minstrel.cache.db.entities.CachedNotificationEntity
import dagger.hilt.android.qualifiers.ApplicationContext
import javax.inject.Inject
import javax.inject.Singleton
/**
* Puts an [Announcement] in the shade (M489 #5347). The words are the
* server's: the same title and body the inbox shows. Tapping one opens what it
* is about; tapping a pile opens the inbox.
*/
@Singleton
class NotificationPoster @Inject constructor(
@ApplicationContext private val context: Context,
) {
// Checked just below: posting is skipped when the user has said no.
@SuppressLint("MissingPermission")
internal fun post(announcement: Announcement) {
val manager = NotificationManagerCompat.from(context)
// Denied POST_NOTIFICATIONS, or every channel silenced: say nothing.
// The inbox still has it all.
if (!manager.areNotificationsEnabled()) return
ensureChannels()
when (announcement) {
is Announcement.Each -> announcement.items.forEach { manager.notify(it.id.hashCode(), single(it)) }
is Announcement.Pile -> manager.notify(PILE_ID, pile(announcement))
}
}
private fun single(item: CachedNotificationEntity) =
NotificationCompat.Builder(context, shadeChannelFor(item.kind).id)
.setSmallIcon(R.drawable.ic_notification)
.setContentTitle(item.title)
.setContentText(item.body)
.setStyle(NotificationCompat.BigTextStyle().bigText(item.body))
.setWhen(item.createdAt.toEpochMilliseconds())
.setShowWhen(true)
.setAutoCancel(true)
.setContentIntent(openIntent(item.link, item.id.hashCode()))
.build()
private fun pile(p: Announcement.Pile) =
NotificationCompat.Builder(context, p.channel.id)
.setSmallIcon(R.drawable.ic_notification)
.setContentTitle("Minstrel")
.setContentText(pileText(p.count))
.setAutoCancel(true)
.setContentIntent(openIntent(link = null, requestCode = PILE_ID))
.build()
/** Each notice gets its own request code, so their links stay apart. */
private fun openIntent(link: String?, requestCode: Int): PendingIntent =
PendingIntent.getActivity(
context,
requestCode,
Intent(context, MainActivity::class.java)
.addFlags(Intent.FLAG_ACTIVITY_SINGLE_TOP or Intent.FLAG_ACTIVITY_CLEAR_TOP)
.putExtra(MainActivity.EXTRA_NOTIFICATION_LINK, link.orEmpty()),
PendingIntent.FLAG_IMMUTABLE or PendingIntent.FLAG_UPDATE_CURRENT,
)
private fun ensureChannels() {
val manager = context.getSystemService(NotificationManager::class.java) ?: return
ShadeChannel.entries.forEach { c ->
manager.createNotificationChannel(
NotificationChannel(c.id, c.title, NotificationManager.IMPORTANCE_DEFAULT)
.apply { description = c.description },
)
}
}
private companion object {
const val PILE_ID = 4802
}
}
@@ -0,0 +1,79 @@
package com.fabledsword.minstrel.notifications.delivery
import androidx.lifecycle.Lifecycle
import androidx.lifecycle.ProcessLifecycleOwner
import com.fabledsword.minstrel.cache.db.dao.AuthSessionDao
import com.fabledsword.minstrel.notifications.data.NotificationSettingsRepository
import com.fabledsword.minstrel.notifications.data.NotificationsRepository
import kotlinx.coroutines.CancellationException
import kotlinx.coroutines.Dispatchers
import kotlinx.coroutines.flow.first
import kotlinx.coroutines.sync.Mutex
import kotlinx.coroutines.sync.withLock
import kotlinx.coroutines.withContext
import timber.log.Timber
import javax.inject.Inject
import javax.inject.Singleton
/**
* Decides what the shade says by asking the server, never by trusting a frame
* (M489 #5347, Roundtable's NotificationSync).
*
* A `notification.created` frame and a reconnect both mean "go and look", so
* both run [catchUp]. The frame carries nothing, so nothing can be announced
* twice: the answer is the server's unread state and this device's
* high-water mark, which survive a reboot. A notice raised while the phone was
* unreachable is not lost either: the event bus replays nothing, but the row
* is still there, unread.
*/
@Singleton
class NotificationSync @Inject constructor(
private val notifications: NotificationsRepository,
private val settings: NotificationSettingsRepository,
private val prefs: AuthSessionDao,
private val poster: NotificationPoster,
) {
// One catch-up at a time: a nudge arriving with a reconnect must not
// read the mark before the other has moved it.
private val mutex = Mutex()
/**
* Announces what is new since the mark, then moves it. Failure is quiet on
* purpose: this runs on a reconnect, exactly when a request is most likely
* to lose a race with a network still settling, and the next trigger tries
* again. While the app is on screen nothing is posted (the bell shows it),
* but the mark still moves, so it is not announced later either.
*/
suspend fun catchUp() {
mutex.withLock { catchUpLocked() }
}
private suspend fun catchUpLocked() {
try {
notifications.refresh()
} catch (e: CancellationException) {
throw e
} catch (
@Suppress("TooGenericExceptionCaught") e: Throwable,
) {
Timber.i(e, "notification catch-up: refresh failed; next trigger retries")
return
}
val plan = planCatchUp(
page = notifications.cachedPage(),
mark = prefs.get()?.notifiedUpTo,
phoneOn = phoneOnFor(settings.settings.first()),
)
if (!appOnScreen()) announcementFor(plan.announce)?.let(poster::post)
plan.upTo?.let { prefs.setNotifiedUpTo(it) }
}
/** On sign-out: the next account starts with no mark, so it announces no backlog. */
suspend fun forget() {
mutex.withLock { prefs.setNotifiedUpTo(null) }
}
private suspend fun appOnScreen(): Boolean = withContext(Dispatchers.Main) {
ProcessLifecycleOwner.get().lifecycle.currentState.isAtLeast(Lifecycle.State.STARTED)
}
}
@@ -0,0 +1,27 @@
package com.fabledsword.minstrel.notifications.ui
import com.fabledsword.minstrel.nav.Admin
import com.fabledsword.minstrel.nav.AdminQuarantine
import com.fabledsword.minstrel.nav.AdminRequests
import com.fabledsword.minstrel.nav.AlbumDetail
import com.fabledsword.minstrel.nav.ArtistDetail
import com.fabledsword.minstrel.nav.Requests
/**
* The app screen for a notice's link. The server writes web paths, the same
* for every client; this maps them onto the app's routes. Admin pages the app
* has no screen for (missing files, duplicates, playback errors) open the
* Admin landing; anything unknown opens nothing.
*/
internal fun routeForLink(link: String): Any? {
val parts = link.trim('/').split('/').filter { it.isNotEmpty() }
return when {
parts.size == 2 && parts[0] == "albums" -> AlbumDetail(parts[1])
parts.size == 2 && parts[0] == "artists" -> ArtistDetail(parts[1])
parts == listOf("requests") -> Requests
parts == listOf("admin", "requests") -> AdminRequests
parts == listOf("admin", "quarantine") -> AdminQuarantine
parts.firstOrNull() == "admin" -> Admin
else -> null
}
}
@@ -0,0 +1,312 @@
@file:Suppress("TooManyFunctions") // Compose screen + private row composables
package com.fabledsword.minstrel.notifications.ui
import android.Manifest
import android.content.Context
import android.content.Intent
import android.os.Build
import android.provider.Settings
import androidx.activity.compose.rememberLauncherForActivityResult
import androidx.activity.result.contract.ActivityResultContracts
import androidx.compose.foundation.clickable
import androidx.compose.foundation.layout.Arrangement
import androidx.compose.foundation.layout.Box
import androidx.compose.foundation.layout.Column
import androidx.compose.foundation.layout.Row
import androidx.compose.foundation.layout.fillMaxSize
import androidx.compose.foundation.layout.fillMaxWidth
import androidx.compose.foundation.layout.padding
import androidx.compose.foundation.layout.width
import androidx.compose.foundation.rememberScrollState
import androidx.compose.foundation.selection.toggleable
import androidx.compose.foundation.verticalScroll
import androidx.compose.material3.Checkbox
import androidx.compose.material3.Icon
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.Scaffold
import androidx.compose.material3.Switch
import androidx.compose.material3.Text
import androidx.compose.runtime.Composable
import androidx.compose.runtime.getValue
import androidx.compose.runtime.mutableStateOf
import androidx.compose.runtime.remember
import androidx.compose.runtime.setValue
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.platform.LocalContext
import androidx.compose.ui.semantics.Role
import androidx.compose.ui.semantics.contentDescription
import androidx.compose.ui.semantics.semantics
import androidx.compose.ui.text.style.TextAlign
import androidx.compose.ui.unit.dp
import androidx.core.app.NotificationManagerCompat
import androidx.hilt.navigation.compose.hiltViewModel
import androidx.lifecycle.compose.LifecycleResumeEffect
import androidx.lifecycle.compose.collectAsStateWithLifecycle
import androidx.navigation.NavHostController
import com.composables.icons.lucide.ChevronRight
import com.composables.icons.lucide.Lucide
import com.fabledsword.minstrel.api.endpoints.NotificationKindSettingWire
import com.fabledsword.minstrel.api.endpoints.NotificationSettingsWire
import com.fabledsword.minstrel.nav.NotificationSettings
import com.fabledsword.minstrel.notifications.data.NotificationChannel
import com.fabledsword.minstrel.shared.widgets.LoadingCentered
import com.fabledsword.minstrel.shared.widgets.MinstrelTopAppBar
import com.fabledsword.minstrel.shared.widgets.ShellContentWindowInsets
private val CELL_WIDTH = 56.dp
/**
* Per-user notification settings (M489): a row per kind with a box for each
* channel. Rows read as a menu (rule 188); anything that needs explaining
* is one short line.
*/
@Composable
fun NotificationSettingsScreen(
navController: NavHostController,
viewModel: NotificationSettingsViewModel = hiltViewModel(),
) {
val settings by viewModel.settings.collectAsStateWithLifecycle()
val isAdmin by viewModel.isAdmin.collectAsStateWithLifecycle()
val backgroundDelivery by viewModel.backgroundDelivery.collectAsStateWithLifecycle()
Scaffold(
contentWindowInsets = ShellContentWindowInsets,
modifier = Modifier.fillMaxSize(),
topBar = {
MinstrelTopAppBar(
title = "Notifications",
navController = navController,
currentRouteName = NotificationSettings::class.qualifiedName,
onBack = { navController.popBackStack() },
)
},
) { inner ->
val s = settings
if (s == null) {
LoadingCentered(modifier = Modifier.padding(inner))
} else {
SettingsBody(
settings = s,
isAdmin = isAdmin,
deviceRows = { DeviceRows(backgroundDelivery, viewModel::setBackgroundDelivery) },
onSet = viewModel::set,
modifier = Modifier.padding(inner),
)
}
}
}
@Composable
private fun SettingsBody(
settings: NotificationSettingsWire,
isAdmin: Boolean,
deviceRows: @Composable () -> Unit,
onSet: (String, NotificationChannel, Boolean) -> Unit,
modifier: Modifier = Modifier,
) {
Column(
modifier = modifier
.fillMaxSize()
.verticalScroll(rememberScrollState())
.padding(vertical = 8.dp),
) {
deviceRows()
HeaderRow()
settings.kinds.filter { !it.adminOnly }.forEach { KindRow(it, settings.emailAvailable, onSet) }
val adminKinds = settings.kinds.filter { it.adminOnly }
if (adminKinds.isNotEmpty()) {
SectionLabel("Library health")
adminKinds.forEach { KindRow(it, settings.emailAvailable, onSet) }
}
if (!settings.emailAvailable) {
Muted(emailUnavailableLine(settings.emailUnavailableReason, isAdmin))
}
}
}
@Composable
private fun HeaderRow() {
Row(
modifier = Modifier.fillMaxWidth().padding(horizontal = 16.dp, vertical = 4.dp),
verticalAlignment = Alignment.CenterVertically,
) {
Box(Modifier.weight(1f))
NotificationChannel.entries.forEach { c ->
Text(
text = channelLabel(c),
modifier = Modifier.width(CELL_WIDTH),
style = MaterialTheme.typography.labelMedium,
color = MaterialTheme.colorScheme.onSurfaceVariant,
textAlign = TextAlign.Center,
)
}
}
}
@Composable
private fun KindRow(
row: NotificationKindSettingWire,
emailAvailable: Boolean,
onSet: (String, NotificationChannel, Boolean) -> Unit,
) {
val label = kindLabel(row.kind)
Row(
modifier = Modifier.fillMaxWidth().padding(horizontal = 16.dp),
verticalAlignment = Alignment.CenterVertically,
) {
Text(label, modifier = Modifier.weight(1f), style = MaterialTheme.typography.bodyLarge)
NotificationChannel.entries.forEach { c ->
val enabled = channelEnabled(row, c, emailAvailable)
Box(Modifier.width(CELL_WIDTH), contentAlignment = Alignment.Center) {
Checkbox(
checked = channelValue(row, c) && enabled,
onCheckedChange = { onSet(row.kind, c, it) },
enabled = enabled,
modifier = Modifier.semantics { contentDescription = "$label: ${channelLabel(c)}" },
)
}
}
}
}
/**
* What this device does, above the per-kind grid that follows the account:
* whether notifications arrive with the app closed (M489 #5347), and whether
* the system lets Minstrel post them at all.
*/
@Composable
private fun DeviceRows(backgroundDelivery: Boolean, onBackgroundDelivery: (Boolean) -> Unit) {
val context = LocalContext.current
// Re-read on every resume: the user may come back from system settings.
var phoneAllowed by remember { mutableStateOf(true) }
LifecycleResumeEffect(Unit) {
phoneAllowed = NotificationManagerCompat.from(context).areNotificationsEnabled()
onPauseOrDispose { }
}
val askToNotify = rememberLauncherForActivityResult(ActivityResultContracts.RequestPermission()) { granted ->
phoneAllowed = granted
}
BackgroundDeliveryRow(
checked = backgroundDelivery,
onChange = { on ->
onBackgroundDelivery(on)
// Android 13+: turning it on is the moment to ask, if not yet allowed.
if (on && !phoneAllowed && Build.VERSION.SDK_INT >= Build.VERSION_CODES.TIRAMISU) {
askToNotify.launch(Manifest.permission.POST_NOTIFICATIONS)
}
},
)
if (!phoneAllowed) {
PhoneBlockedRow(onClick = { openAppNotificationSettings(context) })
}
}
@Composable
private fun BackgroundDeliveryRow(checked: Boolean, onChange: (Boolean) -> Unit) {
Row(
modifier = Modifier
.fillMaxWidth()
.toggleable(value = checked, role = Role.Switch, onValueChange = onChange)
.padding(horizontal = 16.dp, vertical = 12.dp),
verticalAlignment = Alignment.CenterVertically,
horizontalArrangement = Arrangement.spacedBy(8.dp),
) {
Column(Modifier.weight(1f)) {
Text("Notifications when the app is closed", style = MaterialTheme.typography.bodyLarge)
// The ongoing notice is the price of not using Google's push;
// said plainly rather than left to be discovered.
Muted("Keeps a quiet notice in the shade, in place of Google's push", padded = false)
}
Switch(checked = checked, onCheckedChange = null)
}
}
@Composable
private fun PhoneBlockedRow(onClick: () -> Unit) {
Row(
modifier = Modifier
.fillMaxWidth()
.clickable(onClick = onClick)
.padding(horizontal = 16.dp, vertical = 12.dp),
verticalAlignment = Alignment.CenterVertically,
horizontalArrangement = Arrangement.spacedBy(8.dp),
) {
Column(Modifier.weight(1f)) {
Text("Phone alerts are off", style = MaterialTheme.typography.bodyLarge)
Muted("Allow notifications for Minstrel in system settings", padded = false)
}
Icon(Lucide.ChevronRight, contentDescription = null, tint = MaterialTheme.colorScheme.onSurfaceVariant)
}
}
@Composable
private fun SectionLabel(text: String) {
Text(
text = text,
modifier = Modifier.padding(start = 16.dp, end = 16.dp, top = 16.dp, bottom = 4.dp),
style = MaterialTheme.typography.titleSmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
@Composable
private fun Muted(text: String, padded: Boolean = true) {
Text(
text = text,
modifier = if (padded) Modifier.padding(horizontal = 16.dp, vertical = 12.dp) else Modifier,
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
private fun openAppNotificationSettings(context: Context) {
val intent = Intent(Settings.ACTION_APP_NOTIFICATION_SETTINGS)
.putExtra(Settings.EXTRA_APP_PACKAGE, context.packageName)
.addFlags(Intent.FLAG_ACTIVITY_NEW_TASK)
runCatching { context.startActivity(intent) }
}
/** Phone and email ride on the inbox; email also needs an address and SMTP. */
internal fun channelEnabled(
row: NotificationKindSettingWire,
channel: NotificationChannel,
emailAvailable: Boolean,
): Boolean = when (channel) {
NotificationChannel.INBOX -> true
NotificationChannel.PHONE -> row.inbox
NotificationChannel.EMAIL -> row.inbox && emailAvailable
}
internal fun channelValue(row: NotificationKindSettingWire, channel: NotificationChannel): Boolean =
when (channel) {
NotificationChannel.INBOX -> row.inbox
NotificationChannel.PHONE -> row.phone
NotificationChannel.EMAIL -> row.email
}
internal fun channelLabel(c: NotificationChannel): String = when (c) {
NotificationChannel.INBOX -> "Inbox"
NotificationChannel.PHONE -> "Phone"
NotificationChannel.EMAIL -> "Email"
}
/** Short row labels, the same as the web's. Unknown kinds show their key. */
internal fun kindLabel(kind: String): String = when (kind) {
"request_approved" -> "Request approved"
"request_rejected" -> "Request declined"
"request_completed" -> "New music arrived"
"request_pending" -> "Requests to review"
"quarantine_flagged" -> "Tracks flagged"
"scan_failed" -> "Library scan failed"
"tracks_missing" -> "Tracks gone missing"
"duplicates_found" -> "Duplicates to review"
"playback_errors" -> "Playback errors"
else -> kind
}
internal fun emailUnavailableLine(reason: String?, isAdmin: Boolean): String = when {
reason == "no_address" -> "Email is off: add an email address in your profile"
isAdmin -> "Email is off: SMTP isn't set up on the server"
else -> "Email is off: this server doesn't send email"
}
@@ -0,0 +1,48 @@
package com.fabledsword.minstrel.notifications.ui
import androidx.lifecycle.ViewModel
import androidx.lifecycle.viewModelScope
import com.fabledsword.minstrel.api.endpoints.NotificationSettingsWire
import com.fabledsword.minstrel.auth.AuthController
import com.fabledsword.minstrel.auth.AuthStore
import com.fabledsword.minstrel.notifications.data.NotificationChannel
import com.fabledsword.minstrel.notifications.data.NotificationSettingsRepository
import dagger.hilt.android.lifecycle.HiltViewModel
import kotlinx.coroutines.flow.SharingStarted
import kotlinx.coroutines.flow.StateFlow
import kotlinx.coroutines.flow.map
import kotlinx.coroutines.flow.stateIn
import kotlinx.coroutines.launch
import javax.inject.Inject
private const val SHARE_STOP_TIMEOUT_MS = 5_000L
@HiltViewModel
class NotificationSettingsViewModel @Inject constructor(
private val repository: NotificationSettingsRepository,
authController: AuthController,
private val authStore: AuthStore,
) : ViewModel() {
/** "Notifications when the app is closed": this device's choice (M489 #5347). */
val backgroundDelivery: StateFlow<Boolean> = authStore.backgroundDelivery
val settings: StateFlow<NotificationSettingsWire?> = repository.settings
.stateIn(viewModelScope, SharingStarted.WhileSubscribed(SHARE_STOP_TIMEOUT_MS), null)
val isAdmin: StateFlow<Boolean> = authController.currentUser
.map { it?.isAdmin == true }
.stateIn(viewModelScope, SharingStarted.WhileSubscribed(SHARE_STOP_TIMEOUT_MS), false)
init {
// Offline, the device's copy stands.
viewModelScope.launch { runCatching { repository.refresh() } }
}
fun set(kind: String, channel: NotificationChannel, value: Boolean) {
viewModelScope.launch { repository.set(kind, channel, value) }
}
fun setBackgroundDelivery(on: Boolean) {
authStore.setBackgroundDelivery(on)
}
}
@@ -0,0 +1,155 @@
package com.fabledsword.minstrel.notifications.ui
import androidx.compose.foundation.background
import androidx.compose.foundation.clickable
import androidx.compose.foundation.layout.Arrangement
import androidx.compose.foundation.layout.Box
import androidx.compose.foundation.layout.Row
import androidx.compose.foundation.layout.fillMaxSize
import androidx.compose.foundation.layout.fillMaxWidth
import androidx.compose.foundation.layout.padding
import androidx.compose.foundation.layout.size
import androidx.compose.foundation.lazy.LazyColumn
import androidx.compose.foundation.lazy.items
import androidx.compose.foundation.shape.CircleShape
import androidx.compose.material3.HorizontalDivider
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.Scaffold
import androidx.compose.material3.Text
import androidx.compose.material3.TextButton
import androidx.compose.runtime.Composable
import androidx.compose.runtime.getValue
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.semantics.contentDescription
import androidx.compose.ui.semantics.semantics
import androidx.compose.ui.text.font.FontWeight
import androidx.compose.ui.text.style.TextOverflow
import androidx.compose.ui.unit.dp
import androidx.hilt.navigation.compose.hiltViewModel
import androidx.lifecycle.compose.collectAsStateWithLifecycle
import androidx.navigation.NavHostController
import com.composables.icons.lucide.Bell
import com.composables.icons.lucide.Lucide
import com.fabledsword.minstrel.cache.db.entities.CachedNotificationEntity
import com.fabledsword.minstrel.nav.Notifications
import com.fabledsword.minstrel.shared.widgets.EmptyState
import com.fabledsword.minstrel.shared.widgets.LoadingCentered
import com.fabledsword.minstrel.shared.widgets.MinstrelTopAppBar
import com.fabledsword.minstrel.shared.widgets.PullToRefreshScaffold
import com.fabledsword.minstrel.shared.widgets.ShellContentWindowInsets
import kotlinx.datetime.Clock
import kotlinx.datetime.Instant
import kotlin.time.Duration.Companion.days
import kotlin.time.Duration.Companion.hours
import kotlin.time.Duration.Companion.minutes
/**
* The notifications inbox (M489): one row per notice, its title on one line
* and how long ago. Tapping a row marks it read and opens what it is about.
*/
@Composable
fun NotificationsScreen(
navController: NavHostController,
viewModel: NotificationsViewModel = hiltViewModel(),
) {
val items by viewModel.items.collectAsStateWithLifecycle()
val anyUnread = items?.any { it.readAt == null } == true
Scaffold(
contentWindowInsets = ShellContentWindowInsets,
modifier = Modifier.fillMaxSize(),
topBar = {
MinstrelTopAppBar(
title = "Notifications",
navController = navController,
currentRouteName = Notifications::class.qualifiedName,
onBack = { navController.popBackStack() },
actions = {
if (anyUnread) {
TextButton(onClick = viewModel::markAllRead) { Text("Mark all read") }
}
},
)
},
) { inner ->
PullToRefreshScaffold(
onRefresh = { viewModel.refresh().join() },
modifier = Modifier.fillMaxSize().padding(inner),
) {
val list = items
when {
list == null -> LoadingCentered()
list.isEmpty() -> EmptyState(
title = "Nothing waiting for you",
body = "Requests, new music and anything needing a look will gather here.",
icon = Lucide.Bell,
)
else -> LazyColumn(modifier = Modifier.fillMaxSize()) {
items(list, key = { it.id }) { item ->
NotificationRow(
item = item,
onClick = {
viewModel.open(item)
routeForLink(item.link)?.let { navController.navigate(it) }
},
)
HorizontalDivider()
}
}
}
}
}
}
@Composable
private fun NotificationRow(item: CachedNotificationEntity, onClick: () -> Unit) {
val unread = item.readAt == null
Row(
modifier = Modifier
.fillMaxWidth()
.clickable(onClick = onClick)
.padding(horizontal = 16.dp, vertical = 14.dp)
.semantics(mergeDescendants = true) {
if (unread) contentDescription = "Unread: ${item.title}"
},
verticalAlignment = Alignment.CenterVertically,
horizontalArrangement = Arrangement.spacedBy(12.dp),
) {
Box(
modifier = Modifier
.size(8.dp)
.then(
if (unread) {
Modifier.background(MaterialTheme.colorScheme.onSurface, CircleShape)
} else {
Modifier
},
),
)
Text(
text = item.title,
modifier = Modifier.weight(1f),
style = MaterialTheme.typography.bodyLarge,
fontWeight = if (unread) FontWeight.Medium else FontWeight.Normal,
color = if (unread) MaterialTheme.colorScheme.onSurface else MaterialTheme.colorScheme.onSurfaceVariant,
maxLines = 1,
overflow = TextOverflow.Ellipsis,
)
Text(
text = ago(item.createdAt),
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
}
/** "just now", "12m", "5h", "3d": the same coarse steps as the web inbox. */
internal fun ago(at: Instant, now: Instant = Clock.System.now()): String {
val d = now - at
return when {
d >= 1.days -> "${d.inWholeDays}d"
d >= 1.hours -> "${d.inWholeHours}h"
d >= 1.minutes -> "${d.inWholeMinutes}m"
else -> "just now"
}
}
@@ -0,0 +1,49 @@
package com.fabledsword.minstrel.notifications.ui
import androidx.lifecycle.ViewModel
import androidx.lifecycle.viewModelScope
import com.fabledsword.minstrel.cache.db.entities.CachedNotificationEntity
import com.fabledsword.minstrel.connectivity.NetworkStatusController
import com.fabledsword.minstrel.connectivity.recoveries
import com.fabledsword.minstrel.notifications.data.NotificationsRepository
import dagger.hilt.android.lifecycle.HiltViewModel
import kotlinx.coroutines.Job
import kotlinx.coroutines.flow.SharingStarted
import kotlinx.coroutines.flow.StateFlow
import kotlinx.coroutines.flow.map
import kotlinx.coroutines.flow.stateIn
import kotlinx.coroutines.launch
import javax.inject.Inject
private const val SHARE_STOP_TIMEOUT_MS = 5_000L
/** `items` is null until the cache has been read once, so the screen can tell "loading" from "empty". */
@HiltViewModel
class NotificationsViewModel @Inject constructor(
private val repository: NotificationsRepository,
networkStatus: NetworkStatusController,
) : ViewModel() {
val items: StateFlow<List<CachedNotificationEntity>?> = repository.notifications
.map<List<CachedNotificationEntity>, List<CachedNotificationEntity>?> { it }
.stateIn(viewModelScope, SharingStarted.WhileSubscribed(SHARE_STOP_TIMEOUT_MS), null)
init {
refresh()
// Back online: the cached page may be stale (#1245's recovery idiom).
viewModelScope.launch { networkStatus.recoveries().collect { refresh() } }
}
/** Offline, the cached page stands. */
fun refresh(): Job = viewModelScope.launch {
runCatching { repository.refresh() }
}
fun open(item: CachedNotificationEntity) {
if (item.readAt != null) return
viewModelScope.launch { repository.markRead(item.id) }
}
fun markAllRead() {
viewModelScope.launch { repository.markAllRead() }
}
}
@@ -93,6 +93,8 @@ class MinstrelForwardingPlayer(
* one. Diagnostics-only; see [TransportObservation].
*/
val onTransport: (TransportObservation) -> Unit = {},
/** The renderer's 1-based queue position, every poll. */
val onRendererTrack: (trackNumber: Int) -> Unit = {},
)
private val scope = CoroutineScope(SupervisorJob() + Dispatchers.IO)
@@ -622,6 +624,7 @@ class MinstrelForwardingPlayer(
trackNumber = info.track,
)
syncLocalCursorToRemote(sonosTrack = info.track, trackUri = info.trackUri)
events.onRendererTrack(info.track)
val transport = active.avTransport.getTransportInfo()
when (transport.state) {
TransportState.PLAYING -> {
@@ -12,8 +12,10 @@ import androidx.media3.session.SessionCommand
import com.fabledsword.minstrel.MainActivity
import com.fabledsword.minstrel.likes.data.LikesRepository
import com.fabledsword.minstrel.likes.data.LikesRepository.Companion.ENTITY_TRACK
import com.fabledsword.minstrel.settings.data.NormalizationRepository
import com.google.common.collect.ImmutableList
import dagger.hilt.android.AndroidEntryPoint
import kotlinx.coroutines.CancellationException
import kotlinx.coroutines.CoroutineScope
import kotlinx.coroutines.Dispatchers
import kotlinx.coroutines.ExperimentalCoroutinesApi
@@ -25,6 +27,7 @@ import kotlinx.coroutines.flow.flatMapLatest
import kotlinx.coroutines.flow.flowOf
import kotlinx.coroutines.flow.onStart
import kotlinx.coroutines.launch
import timber.log.Timber
import javax.inject.Inject
/**
@@ -66,6 +69,8 @@ class MinstrelPlayerService : MediaSessionService() {
@Inject lateinit var likesRepository: LikesRepository
@Inject lateinit var normalizationRepository: NormalizationRepository
private val serviceScope = CoroutineScope(SupervisorJob() + Dispatchers.Main.immediate)
private var mediaSession: MediaSession? = null
@@ -82,6 +87,21 @@ class MinstrelPlayerService : MediaSessionService() {
.build()
mediaSession = session
serviceScope.launch { observeLikeState(session, player) }
serviceScope.launch { refreshNormalization() }
}
// Takes up a leveling preference changed on another device (M464 #5000).
// Offline, the device's copy stands, which is the one playback reads.
private suspend fun refreshNormalization() {
try {
normalizationRepository.refresh()
} catch (e: CancellationException) {
throw e
} catch (
@Suppress("TooGenericExceptionCaught") e: Throwable,
) {
Timber.d(e, "normalization refresh skipped")
}
}
/**
@@ -14,6 +14,8 @@ import androidx.media3.session.MediaController
import androidx.media3.session.SessionToken
import com.fabledsword.minstrel.di.ApplicationScope
import com.fabledsword.minstrel.models.TrackRef
import com.fabledsword.minstrel.player.gain.GainAudioProcessor
import com.fabledsword.minstrel.player.gain.ReplayGainStore
import com.fabledsword.minstrel.playlists.data.PlaylistsRepository
import com.fabledsword.minstrel.playlists.data.toPlayableTrackRefs
import com.fabledsword.minstrel.shared.resolveServerUrl
@@ -69,6 +71,7 @@ class PlayerController @Inject constructor(
private val playerFactory: PlayerFactory,
private val activeUpnpHolder: com.fabledsword.minstrel.player.output.ActiveUpnpHolder,
private val remoteState: RemotePlayerState,
private val replayGains: ReplayGainStore,
) {
/**
@@ -277,6 +280,11 @@ class PlayerController @Inject constructor(
queueRefs = playable.tracks
val items = playable.tracks.map { it.toMediaItem(source) }
val startIndex = playable.initialIndex
// Gains for the first tracks, so the first one levels from its first
// sample rather than ramping in once the lookup lands.
replayGains.request(
playable.tracks.drop(startIndex).take(GAIN_PREFETCH).map { it.id },
)
// Drift #562 cold-boot resume calls this from a non-Main suspend
// context after awaitReady() unblocks (ResumeController launches
// on Dispatchers.Default by the time it reaches us). MediaController
@@ -792,7 +800,13 @@ class PlayerController @Inject constructor(
// scrubber a real total even when the wrapped ExoPlayer is
// paused under UPnP (it never probes a duration in that state).
if (durationSec > 0) setDurationMs(durationSec.toLong() * MS_PER_SECOND)
if (source != null) setExtras(sourceExtras(source))
// Album and track position let the gain processor tell an
// album played in order from a mix (M464 #5000).
trackNumber?.let { setTrackNumber(it) }
discNumber?.let { setDiscNumber(it) }
val extras = GainAudioProcessor.albumExtras(albumId)
if (source != null) extras.putAll(sourceExtras(source))
setExtras(extras)
// Point the notification / lock-screen art at the SAME album
// cover the in-app surfaces use (TrackRef.coverUrl ->
// /api/albums/{id}/cover). Without this, Media3 falls back to
@@ -860,6 +874,7 @@ class PlayerController @Inject constructor(
const val MINSTREL_SOURCE_KEY: String = "minstrel_source"
private const val MS_PER_SECOND = 1_000L
private const val MAX_INTERPOLATION_DRIFT_MS = 5_000L
private const val GAIN_PREFETCH = 20
}
}
@@ -4,6 +4,7 @@ import android.content.Context
import androidx.media3.common.AudioAttributes
import androidx.media3.common.C
import androidx.media3.common.Player
import androidx.media3.common.audio.AudioProcessor
import androidx.media3.common.util.BitmapLoader
import androidx.media3.database.StandaloneDatabaseProvider
import androidx.media3.datasource.DataSourceBitmapLoader
@@ -12,11 +13,18 @@ import androidx.media3.datasource.cache.CacheDataSource
import androidx.media3.datasource.cache.LeastRecentlyUsedCacheEvictor
import androidx.media3.datasource.cache.SimpleCache
import androidx.media3.datasource.okhttp.OkHttpDataSource
import androidx.media3.exoplayer.DefaultRenderersFactory
import androidx.media3.exoplayer.ExoPlayer
import androidx.media3.exoplayer.audio.AudioSink
import androidx.media3.exoplayer.audio.DefaultAudioSink
import androidx.media3.exoplayer.source.DefaultMediaSourceFactory
import androidx.media3.session.CacheBitmapLoader
import com.fabledsword.minstrel.auth.AuthStore
import com.fabledsword.minstrel.cache.audiocache.CacheConfig
import com.fabledsword.minstrel.player.gain.GainAudioProcessor
import com.fabledsword.minstrel.player.gain.ReplayGainStore
import com.fabledsword.minstrel.player.output.ActiveUpnpHolder
import com.fabledsword.minstrel.player.output.SonosQueueLoader
import dagger.hilt.android.qualifiers.ApplicationContext
import kotlinx.coroutines.channels.BufferOverflow
import kotlinx.coroutines.flow.MutableSharedFlow
@@ -56,6 +64,9 @@ class PlayerFactory @Inject constructor(
private val activeUpnpHolder: ActiveUpnpHolder,
private val remoteState: RemotePlayerState,
private val serverHealth: com.fabledsword.minstrel.connectivity.NetworkStatusController,
private val authStore: AuthStore,
private val replayGains: ReplayGainStore,
private val sonosQueue: SonosQueueLoader,
) {
private val cacheDir: File = File(context.cacheDir, "audio_cache").apply { mkdirs() }
@@ -120,6 +131,7 @@ class PlayerFactory @Inject constructor(
onStalled = { trackId -> stallEventsInternal.tryEmit(trackId) },
onQueueTruncated = { queueRepairInternal.tryEmit(Unit) },
onTransport = { transportInternal.tryEmit(it) },
onRendererTrack = { sonosQueue.onRendererTrack(it) },
),
)
}
@@ -142,7 +154,28 @@ class PlayerFactory @Inject constructor(
val mediaSourceFactory = DefaultMediaSourceFactory(context)
.setDataSourceFactory(cacheDataSource)
return ExoPlayer.Builder(context)
// Loudness normalization (M464 #5000) runs inside the audio sink so
// each track's gain starts on its first sample. Audio offload would
// bypass the sink's processors; ExoPlayer leaves it off unless asked
// (TrackSelectionParameters.audioOffloadPreferences), and nothing here
// asks, so leveling always applies.
val gainProcessor = GainAudioProcessor(
prefs = { authStore.normalization.value },
store = replayGains,
)
val renderersFactory = object : DefaultRenderersFactory(context) {
override fun buildAudioSink(
context: Context,
enableFloatOutput: Boolean,
enableAudioOutputPlaybackParams: Boolean,
): AudioSink = DefaultAudioSink.Builder(context)
.setEnableFloatOutput(enableFloatOutput)
.setEnableAudioOutputPlaybackParameters(enableAudioOutputPlaybackParams)
.setAudioProcessors(arrayOf<AudioProcessor>(gainProcessor))
.build()
}
val exo = ExoPlayer.Builder(context, renderersFactory)
.setMediaSourceFactory(mediaSourceFactory)
.setAudioAttributes(
AudioAttributes.Builder()
@@ -153,6 +186,17 @@ class PlayerFactory @Inject constructor(
)
.setHandleAudioBecomingNoisy(true)
.build()
// The processor finds a track's neighbours in play order, which
// depends on shuffle mode.
gainProcessor.shuffleEnabled = exo.shuffleModeEnabled
exo.addListener(
object : Player.Listener {
override fun onShuffleModeEnabledChanged(shuffleModeEnabled: Boolean) {
gainProcessor.shuffleEnabled = shuffleModeEnabled
}
},
)
return exo
}
/**
@@ -19,6 +19,13 @@ import javax.inject.Singleton
class StreamTokenProvider @Inject constructor(retrofit: Retrofit) {
private val api: CastApi = retrofit.create()
suspend fun mint(trackId: String): StreamTokenResponse =
api.streamToken(StreamTokenRequest(trackId = trackId))
/**
* Mints a URL for a speaker, leveled when the user's setting calls for
* it (M464 #5002). The server answers with the plain stream when it does
* not, so every speaker URL asks.
*/
suspend fun mint(trackId: String, asAlbum: Boolean = false, prerender: Boolean = false): StreamTokenResponse =
api.streamToken(
StreamTokenRequest(trackId = trackId, level = true, asAlbum = asAlbum, prerender = prerender),
)
}
@@ -0,0 +1,139 @@
package com.fabledsword.minstrel.player.gain
import android.os.Bundle
import androidx.media3.common.C
import androidx.media3.common.MediaItem
import androidx.media3.common.Player
import androidx.media3.common.Timeline
import androidx.media3.common.audio.AudioProcessor
import androidx.media3.common.audio.BaseAudioProcessor
import com.fabledsword.minstrel.settings.data.NormalizationBoost
import com.fabledsword.minstrel.settings.data.NormalizationMode
import com.fabledsword.minstrel.settings.data.NormalizationPrefs
import java.nio.ByteBuffer
/**
* Applies each track's loudness gain inside ExoPlayer's audio sink (M464
* #5000), so the level changes on the exact sample a track starts, gapless
* transitions included. `Player.setVolume` could do neither: it stops at 1,
* and set from a transition callback it lands about two seconds late
* because the next track is already buffered (androidx/media#418).
*
* Media3 1.11 flushes the sink's processors at every item boundary with the
* playlist [Timeline] and the new item's period, which is how this knows
* which track it is processing and who its neighbours are.
*
* Boosts above unity are held under -1 dBFS by a peak limiter when the user
* chose one. In headroom mode the gain already stops short of the track's
* true peak, and the final clamp only guards against a bad measurement.
*
* Runs on the playback thread. [prefs] and [store] are read per buffer, so a
* changed preference or a gain that arrives mid-track applies at once,
* through a short ramp rather than a step.
*/
class GainAudioProcessor(
private val prefs: () -> NormalizationPrefs,
private val store: ReplayGainStore,
) : BaseAudioProcessor() {
/** Mirrors the player's shuffle mode, so neighbours are found in play order. */
@Volatile var shuffleEnabled: Boolean = false
private var currentId: String? = null
private var asAlbum = false
private val stage = GainStage()
override fun onConfigure(inputAudioFormat: AudioProcessor.AudioFormat): AudioProcessor.AudioFormat =
when (inputAudioFormat.encoding) {
C.ENCODING_PCM_16BIT, C.ENCODING_PCM_FLOAT -> inputAudioFormat
else -> AudioProcessor.AudioFormat.NOT_SET
}
override fun onFlush(streamMetadata: AudioProcessor.StreamMetadata) {
identify(streamMetadata)
// A new stream starts at its own level: ramping in from the previous
// track's gain would swell or dip its first moments.
stage.reset(inputAudioFormat.sampleRate, inputAudioFormat.channelCount, targetGain())
}
override fun onReset() {
currentId = null
asAlbum = false
}
override fun queueInput(inputBuffer: ByteBuffer) {
val size = inputBuffer.remaining()
if (size == 0) return
val out = replaceOutputBuffer(size)
val p = prefs()
val target = targetGain(p)
val limiting = p.mode != NormalizationMode.OFF && p.boost == NormalizationBoost.LIMITER
if (stage.isUnity(target, limiting)) {
out.put(inputBuffer)
} else {
stage.process(
input = inputBuffer,
out = out,
channels = inputAudioFormat.channelCount,
isFloat = inputAudioFormat.encoding == C.ENCODING_PCM_FLOAT,
target = target,
limiting = limiting,
)
}
out.flip()
}
private fun targetGain(p: NormalizationPrefs = prefs()): Float {
val id = currentId ?: return 1f
return GainMath.dbToLinear(GainMath.gainDb(p, store.get(id), asAlbum))
}
// Which track this stream is, and whether it is playing as part of its
// album. Also asks the store for the gains of the tracks coming up, so
// each is known before it starts.
private fun identify(meta: AudioProcessor.StreamMetadata) {
currentId = null
asAlbum = false
val uid = meta.periodUid
if (uid == null || meta.timeline.isEmpty) return
val timeline = meta.timeline
val index = timeline.getPeriodByUid(uid, Timeline.Period()).windowIndex
val window = Timeline.Window()
fun itemAt(i: Int): MediaItem? =
if (i == C.INDEX_UNSET) null else timeline.getWindow(i, window).mediaItem
fun nextOf(i: Int) = timeline.getNextWindowIndex(i, Player.REPEAT_MODE_OFF, shuffleEnabled)
val cur = itemAt(index)
val prev = itemAt(timeline.getPreviousWindowIndex(index, Player.REPEAT_MODE_OFF, shuffleEnabled))
val next = itemAt(nextOf(index))
if (cur != null) {
currentId = cur.mediaId
asAlbum = GainMath.playingAsAlbum(prev?.albumPosition(), cur.albumPosition(), next?.albumPosition())
}
val upcoming = mutableListOf<String>()
var i = index
while (i != C.INDEX_UNSET && upcoming.size < LOOKAHEAD) {
itemAt(i)?.let { upcoming += it.mediaId }
i = nextOf(i)
}
store.request(upcoming)
}
companion object {
/** MediaMetadata extras key holding the track's album id. */
const val EXTRA_ALBUM_ID = "minstrel.album_id"
private const val LOOKAHEAD = 20
/** Puts what [albumPosition] reads into a MediaItem's metadata extras. */
fun albumExtras(albumId: String, into: Bundle = Bundle()): Bundle =
into.apply { putString(EXTRA_ALBUM_ID, albumId) }
}
}
private fun MediaItem.albumPosition(): AlbumPosition = AlbumPosition(
albumId = mediaMetadata.extras?.getString(GainAudioProcessor.EXTRA_ALBUM_ID),
discNumber = mediaMetadata.discNumber,
trackNumber = mediaMetadata.trackNumber,
)
@@ -0,0 +1,96 @@
package com.fabledsword.minstrel.player.gain
import com.fabledsword.minstrel.settings.data.NormalizationBoost
import com.fabledsword.minstrel.settings.data.NormalizationMode
import com.fabledsword.minstrel.settings.data.NormalizationPrefs
import kotlin.math.log10
import kotlin.math.min
import kotlin.math.pow
/**
* One track's ReplayGain 2.0 values from the server (#4997): dB to the
* -18 LUFS reference, and linear true peaks. A null field has not been
* measured yet.
*/
data class ReplayGain(
val trackGain: Float?,
val trackPeak: Float?,
val albumGain: Float?,
val albumPeak: Float?,
) {
companion object {
val NONE = ReplayGain(null, null, null, null)
}
}
/** Where a queue item sits in its album, for the auto-mode album rule. */
data class AlbumPosition(val albumId: String?, val discNumber: Int?, val trackNumber: Int?)
/**
* Loudness-normalization math (M464 #5000). The same rules as the web
* player's `web/src/lib/player/gain.ts`, so a track levels the same on
* every device.
*/
object GainMath {
private const val REFERENCE_LUFS = -18
/** Headroom mode raises a quiet track only until its true peak reaches this. */
const val PEAK_CEILING_DBTP = -1f
/**
* No track is raised more than this, whatever its measurement says: a
* near-silent track would otherwise come out as amplified noise.
*/
const val MAX_BOOST_DB = 12f
private const val DISC_STRIDE = 1000
// Amplitude decibels: 20 dB per factor of ten.
private const val DB_PER_DECADE = 20f
/**
* The gain to apply, in dB. 0 when leveling is off or the track has not
* been measured: an unmeasured track plays as mastered.
*/
fun gainDb(prefs: NormalizationPrefs, g: ReplayGain?, asAlbum: Boolean): Float {
if (prefs.mode == NormalizationMode.OFF || g == null) return 0f
val wantAlbum = prefs.mode == NormalizationMode.ALBUM ||
(prefs.mode == NormalizationMode.AUTO && asAlbum)
// Album gain falls back to track gain while the album is still being
// measured; track gain never falls back to album gain.
val useAlbum = wantAlbum && g.albumGain != null
val gain = if (useAlbum) g.albumGain else g.trackGain
val peak = if (useAlbum) g.albumPeak else g.trackPeak
return gain?.let { leveled(prefs, it, peak) } ?: 0f
}
private fun leveled(prefs: NormalizationPrefs, gain: Float, peak: Float?): Float {
var db = gain + (prefs.targetLufs - REFERENCE_LUFS)
if (prefs.boost == NormalizationBoost.HEADROOM && peak != null && peak > 0f) {
db = min(db, PEAK_CEILING_DBTP - DB_PER_DECADE * log10(peak))
}
return min(db, MAX_BOOST_DB)
}
/**
* Whether the current item is being played as part of its album, in
* order: a neighbour in play order is from the same album and sits on
* the right side of it. That is when album gain keeps the album's own
* dynamics (a quiet intro stays quiet); anywhere else track gain levels
* the mix.
*/
fun playingAsAlbum(prev: AlbumPosition?, cur: AlbumPosition, next: AlbumPosition?): Boolean {
val curOrder = order(cur)
if (cur.albumId == null || curOrder == null) return false
val prevOrder = prev?.takeIf { it.albumId == cur.albumId }?.let { order(it) }
val nextOrder = next?.takeIf { it.albumId == cur.albumId }?.let { order(it) }
return (prevOrder != null && prevOrder < curOrder) || (nextOrder != null && nextOrder > curOrder)
}
// Disc-major track order. A track with no number has no place in the
// order and is never evidence of album play.
private fun order(p: AlbumPosition): Int? =
p.trackNumber?.let { (p.discNumber ?: 1) * DISC_STRIDE + it }
fun dbToLinear(db: Float): Float = 10f.pow(db / DB_PER_DECADE)
}
@@ -0,0 +1,82 @@
package com.fabledsword.minstrel.player.gain
import java.nio.ByteBuffer
import kotlin.math.abs
import kotlin.math.exp
import kotlin.math.max
import kotlin.math.roundToInt
/**
* The sample arithmetic of [GainAudioProcessor], apart from Media3 so it can
* be tested on the JVM: a gain that ramps toward its target, then an
* optional peak limiter, then a clamp to full scale. Interleaved 16-bit or
* float PCM in, the same format out.
*/
internal class GainStage {
private var gain = 1f
private var envelope = 0f
private var rampCoeff = 1f
private var releaseCoeff = 0f
private var frame = FloatArray(2)
/** Sets the sample rate and channel count, and starts the gain at [startGain]. */
fun reset(sampleRate: Int, channels: Int, startGain: Float) {
val rate = sampleRate.coerceAtLeast(1).toFloat()
rampCoeff = 1f - exp(-1f / (RAMP_SECONDS * rate))
releaseCoeff = exp(-1f / (RELEASE_SECONDS * rate))
if (frame.size < channels) frame = FloatArray(channels)
gain = startGain
envelope = 0f
}
/** True when processing would copy the input unchanged. */
fun isUnity(target: Float, limiting: Boolean): Boolean = gain == 1f && target == 1f && !limiting
/**
* Processes every whole frame of [input] into [out]. A trailing partial
* frame is dropped, as Media3's own processors do.
*/
@Suppress("LongParameterList") // the PCM layout is four facts; a holder type would only rename them
fun process(
input: ByteBuffer,
out: ByteBuffer,
channels: Int,
isFloat: Boolean,
target: Float,
limiting: Boolean,
) {
val bytesPerFrame = channels * if (isFloat) FLOAT_BYTES else PCM16_BYTES
val frames = input.remaining() / bytesPerFrame
repeat(frames) {
gain += (target - gain) * rampCoeff
var peak = 0f
for (c in 0 until channels) {
val s = (if (isFloat) input.float else input.short / PCM16_SCALE) * gain
frame[c] = s
peak = max(peak, abs(s))
}
var reduce = 1f
if (limiting) {
envelope = max(peak, envelope * releaseCoeff)
if (envelope > LIMIT_CEILING) reduce = LIMIT_CEILING / envelope
}
for (c in 0 until channels) {
val v = (frame[c] * reduce).coerceIn(-1f, 1f)
if (isFloat) out.putFloat(v) else out.putShort((v * PCM16_MAX).roundToInt().toShort())
}
}
input.position(input.limit())
}
companion object {
/** -1 dBFS, the ceiling the web player's limiter holds too. */
const val LIMIT_CEILING = 0.8913f
private const val RAMP_SECONDS = 0.05f
private const val RELEASE_SECONDS = 0.25f
private const val PCM16_SCALE = 32768f
private const val PCM16_MAX = 32767f
private const val PCM16_BYTES = 2
private const val FLOAT_BYTES = 4
}
}
@@ -0,0 +1,140 @@
package com.fabledsword.minstrel.player.gain
import com.fabledsword.minstrel.api.endpoints.ReplayGainApi
import com.fabledsword.minstrel.cache.db.dao.CachedTrackDao
import com.fabledsword.minstrel.connectivity.NetworkStatusController
import com.fabledsword.minstrel.connectivity.ServerHealth
import com.fabledsword.minstrel.di.ApplicationScope
import kotlinx.coroutines.CancellationException
import kotlinx.coroutines.CoroutineScope
import kotlinx.coroutines.launch
import kotlinx.coroutines.withTimeoutOrNull
import retrofit2.Retrofit
import retrofit2.create
import timber.log.Timber
import java.util.concurrent.ConcurrentHashMap
import javax.inject.Inject
import javax.inject.Singleton
/**
* The gains the player levels by (M464 #5000), looked up per track and
* held for the life of the process. The audio thread reads them with [get];
* [request] fills them in the background.
*
* Sources, most to least preferred:
* 1. the library cache, synced with the gains, so cached audio levels
* offline;
* 2. the server's replay-gain lookup, for a track not cached yet or
* measured since the last sync;
* 3. none: the track plays as mastered, and the player picks the gain up
* the moment it lands.
*/
@Singleton
class ReplayGainStore internal constructor(
private val scope: CoroutineScope,
private val trackDao: CachedTrackDao,
private val api: ReplayGainApi,
private val serverHealth: () -> ServerHealth,
private val clock: () -> Long,
) {
@Inject constructor(
@ApplicationScope scope: CoroutineScope,
trackDao: CachedTrackDao,
retrofit: Retrofit,
network: NetworkStatusController,
) : this(
scope = scope,
trackDao = trackDao,
api = retrofit.create(),
serverHealth = { network.state.value },
clock = System::currentTimeMillis,
)
private val gains = ConcurrentHashMap<String, ReplayGain>()
private val inflight: MutableSet<String> = ConcurrentHashMap.newKeySet()
// When the server last said it had nothing for a track. Not cached for
// good: the backfill may measure it minutes later.
private val missedAt = ConcurrentHashMap<String, Long>()
/** The track's gains, or null while they are unknown. Safe from any thread. */
fun get(trackId: String): ReplayGain? = gains[trackId]
/** Starts loading gains for any of [trackIds] not already known or loading. */
fun request(trackIds: Collection<String>) {
val now = clock()
val wanted = trackIds.filter { id ->
!gains.containsKey(id) &&
(missedAt[id]?.let { now - it > MISS_RETRY_MS } ?: true) &&
inflight.add(id)
}
if (wanted.isEmpty()) return
scope.launch {
try {
load(wanted)
} finally {
inflight.removeAll(wanted.toSet())
}
}
}
internal suspend fun load(ids: List<String>) {
fromCache(ids)
val rest = ids.filter { !gains.containsKey(it) }
val health = serverHealth()
if (rest.isEmpty() || health == ServerHealth.Offline || health == ServerHealth.ServerDown) return
for (batch in rest.chunked(MAX_IDS_PER_REQUEST)) {
// A failed lookup leaves the batch unknown; the next request asks again.
if (!fromServer(batch)) return
}
}
private suspend fun fromCache(ids: List<String>) {
val rows = try {
trackDao.replayGains(ids)
} catch (e: CancellationException) {
throw e
} catch (
@Suppress("TooGenericExceptionCaught") e: Throwable,
) {
Timber.w(e, "replay gain: cache read failed")
emptyList()
}
for (row in rows) {
// A cached row without a track gain is not an answer: the server
// may have measured the track since this device last synced.
if (row.trackGain == null) continue
gains[row.id] = ReplayGain(row.trackGain, row.trackPeak, row.albumGain, row.albumPeak)
}
}
/** Asks the server about [batch]; false when it could not be asked. */
private suspend fun fromServer(batch: List<String>): Boolean {
val res = try {
withTimeoutOrNull(REQUEST_TIMEOUT_MS) { api.getReplayGain(batch.joinToString(",")) }
} catch (e: CancellationException) {
throw e
} catch (
@Suppress("TooGenericExceptionCaught") e: Throwable,
) {
Timber.w(e, "replay gain: lookup failed")
null
} ?: return false
val now = clock()
for (id in batch) {
val w = res.items[id]
if (w == null) {
missedAt[id] = now
} else {
gains[id] = ReplayGain(w.trackGain, w.trackPeak, w.albumGain, w.albumPeak)
}
}
return true
}
private companion object {
const val MAX_IDS_PER_REQUEST = 200 // the endpoint's limit
const val REQUEST_TIMEOUT_MS = 10_000L
const val MISS_RETRY_MS = 5 * 60_000L
}
}
@@ -4,6 +4,8 @@ import com.fabledsword.minstrel.di.ApplicationScope
import com.fabledsword.minstrel.models.TrackRef
import com.fabledsword.minstrel.player.RemotePlayerState
import com.fabledsword.minstrel.player.StreamTokenProvider
import com.fabledsword.minstrel.player.gain.AlbumPosition
import com.fabledsword.minstrel.player.gain.GainMath
import com.fabledsword.minstrel.player.output.upnp.AVTransportClient
import com.fabledsword.minstrel.player.output.upnp.SoapFaultException
import com.fabledsword.minstrel.player.output.upnp.bareUdn
@@ -24,6 +26,11 @@ import javax.inject.Singleton
* with its own failure modes, and it had grown large enough to hide one:
* every write here is a SOAP call that can fail individually, and until
* [verifyQueueLength] nothing ever read the result back.
*
* Every URL sent is a leveled one (M464 #5002): the server renders the track
* at the user's loudness gain, or hands back the plain stream when leveling
* is off. Renders are slow enough to matter, so the track playing and the
* one after it are rendered ahead; the rest render when the speaker asks.
*/
@Singleton
class SonosQueueLoader @Inject constructor(
@@ -32,6 +39,13 @@ class SonosQueueLoader @Inject constructor(
private val activeUpnpHolder: ActiveUpnpHolder,
private val remoteState: RemotePlayerState,
) {
// The queue as last sent to the renderer, for looking up the track after
// the one it is playing.
@Volatile private var sent: List<TrackRef> = emptyList()
// The renderer track the next one was last prerendered for.
@Volatile private var prerenderedAfter = 0
suspend fun load(
transport: AVTransportClient,
route: OutputRoute,
@@ -46,8 +60,10 @@ class SonosQueueLoader @Inject constructor(
"UPnP select: add %d initial tracks (currentIndex=%d, totalQueue=%d)",
initialBatch.size, currentIndex, queue.size,
)
initialBatch.forEachIndexed { idx, ref ->
val token = streamTokens.mint(ref.id)
sent = queue
prerenderedAfter = currentIndex + 1
initialBatch.indices.forEach { idx ->
val token = mint(queue, idx, prerender = idx == currentIndex)
transport.addURIToQueue(
uri = token.url,
mime = token.mime,
@@ -64,12 +80,12 @@ class SonosQueueLoader @Inject constructor(
Timber.w("UPnP select: Play")
transport.play()
Timber.w("UPnP select: initial done; backgrounding remainder")
val remaining = queue.drop(initialEnd)
// Verify even when there is no tail to append: the initial batch is
// sent the same way and can be dropped the same way.
scope.launch {
if (remaining.isNotEmpty()) {
extendQueueOnSonos(transport, route, remaining, initialEnd)
prerender(queue, currentIndex + 1)
if (initialEnd < queue.size) {
extendQueueOnSonos(transport, route, queue, initialEnd)
}
verifyQueueLength(transport, route, queue)
}
@@ -84,15 +100,16 @@ class SonosQueueLoader @Inject constructor(
private suspend fun extendQueueOnSonos(
transport: AVTransportClient,
route: OutputRoute,
tracks: List<TrackRef>,
queue: List<TrackRef>,
startPosition: Int,
) {
val count = queue.size - startPosition
Timber.w(
"UPnP extend: appending %d tracks starting at position %d",
tracks.size, startPosition + 1,
count, startPosition + 1,
)
val succeeded = appendTracksToQueue(transport, route, tracks, startPosition)
Timber.w("UPnP extend: done (%d / %d appended)", succeeded, tracks.size)
val succeeded = appendTracksToQueue(transport, route, queue, startPosition)
Timber.w("UPnP extend: done (%d / %d appended)", succeeded, count)
}
/**
@@ -136,18 +153,17 @@ class SonosQueueLoader @Inject constructor(
Timber.w("UPnP verify: renderer holds %d tracks, queue intact", nrTracks)
return
}
val missing = fullQueue.drop(nrTracks)
Timber.w(
"UPnP verify: %s holds %d of %d tracks; appending %d missing (round %d)",
route.name, nrTracks, fullQueue.size, missing.size, round + 1,
route.name, nrTracks, fullQueue.size, fullQueue.size - nrTracks, round + 1,
)
appendTracksToQueue(transport, route, missing, nrTracks)
appendTracksToQueue(transport, route, fullQueue, nrTracks)
}
Timber.w("UPnP verify: gave up repairing queue length on %s", route.name)
}
/**
* Append [tracks] at [startPosition] (0-based), returning how many landed.
* Append [queue] from [startPosition] (0-based) on, returning how many landed.
* Tolerates individual AddURIToQueue failures — log and continue so some
* tracks loaded is better than zero tracks loaded — and stops early after
* [EXTEND_ABORT_AFTER_FAILURES] consecutive ones.
@@ -155,20 +171,20 @@ class SonosQueueLoader @Inject constructor(
private suspend fun appendTracksToQueue(
transport: AVTransportClient,
route: OutputRoute,
tracks: List<TrackRef>,
queue: List<TrackRef>,
startPosition: Int,
): Int {
var consecutiveFailures = 0
var succeeded = 0
var aborted = false
for ((i, ref) in tracks.withIndex()) {
for (i in 0 until queue.size - startPosition) {
if (aborted) break
if (activeUpnpHolder.active.value?.routeId != route.id) {
Timber.w("UPnP extend: cancelled at offset %d (route changed)", i)
aborted = true
} else {
val outcome = runCatching {
val token = streamTokens.mint(ref.id)
val token = mint(queue, startPosition + i)
transport.addURIToQueue(
uri = token.url,
mime = token.mime,
@@ -220,6 +236,7 @@ class SonosQueueLoader @Inject constructor(
newQueue: List<TrackRef>,
): Boolean {
val newIds = newQueue.map { it.id }
sent = newQueue
if (oldIds == newIds) return true
val prefixLen = commonPrefixLength(oldIds, newIds)
val suffixLen = commonSuffixLength(
@@ -271,8 +288,7 @@ class SonosQueueLoader @Inject constructor(
prefixLen + 1,
)
for (i in 0 until addedCount) {
val ref = newQueue[prefixLen + i]
val token = streamTokens.mint(ref.id)
val token = mint(newQueue, prefixLen + i)
transport.addURIToQueue(
uri = token.url,
mime = token.mime,
@@ -283,6 +299,27 @@ class SonosQueueLoader @Inject constructor(
}
}
/**
* Called on every poll with the renderer's 1-based track number. When it
* moves, the track after it is rendered ahead, so the speaker's fetch of
* it finds the render ready.
*/
fun onRendererTrack(trackNumber: Int) {
if (trackNumber <= 0 || trackNumber == prerenderedAfter) return
prerenderedAfter = trackNumber
val queue = sent
scope.launch { prerender(queue, trackNumber) }
}
private suspend fun prerender(queue: List<TrackRef>, index: Int) {
if (index !in queue.indices) return
runCatching { mint(queue, index, prerender = true) }
.onFailure { Timber.w(it, "UPnP prerender failed for %s", queue[index].id) }
}
private suspend fun mint(queue: List<TrackRef>, index: Int, prerender: Boolean = false) =
streamTokens.mint(queue[index].id, asAlbum = playingAsAlbum(queue, index), prerender = prerender)
private fun commonPrefixLength(a: List<String>, b: List<String>): Int {
val limit = minOf(a.size, b.size)
for (i in 0 until limit) {
@@ -299,18 +336,33 @@ class SonosQueueLoader @Inject constructor(
return limit
}
private companion object {
companion object {
/**
* Whether [queue]'s track at [index] plays among its album in order,
* judged by its neighbours in the renderer's queue, which plays
* straight through.
*/
internal fun playingAsAlbum(queue: List<TrackRef>, index: Int): Boolean =
GainMath.playingAsAlbum(
prev = queue.getOrNull(index - 1)?.let(::position),
cur = position(queue[index]),
next = queue.getOrNull(index + 1)?.let(::position),
)
private fun position(t: TrackRef) =
AlbumPosition(t.albumId.ifEmpty { null }, t.discNumber, t.trackNumber)
// Abort the append loop after this many consecutive AddURIToQueue
// failures; Sonos rate-limits burst adds and a wall of failures means
// it has stopped accepting, not that the next one might land.
const val EXTEND_ABORT_AFTER_FAILURES = 3
const val EXTEND_THROTTLE_MS = 50L
private const val EXTEND_ABORT_AFTER_FAILURES = 3
private const val EXTEND_THROTTLE_MS = 50L
// Verify/repair passes after a queue load. Two: one to catch the
// common case (a rate-limit burst dropped a chunk), one to catch a
// repair that itself got rate-limited. Beyond that the renderer is
// refusing for a reason retrying won't fix, and the stall watchdog
// becomes the backstop.
const val VERIFY_ROUNDS = 2
private const val VERIFY_ROUNDS = 2
}
}
@@ -109,8 +109,11 @@ private fun MiniCover(coverUrl: String, contentDescription: String) {
* NowPlayingScreen via [onExpandClick].
*
* Layout (Column):
* - Slim seek slider at the top (4dp track)
* - Row: cover | title/artist column | like | prev | play/pause | next
* - Slim seek slider pinned at the top (4dp track)
* - Row: cover | title/artist column | like | prev | play/pause | next.
* Weighted so it fills the rest of the fixed-height bar and centres its
* own content; otherwise the row keeps its intrinsic 48dp and the
* leftover height collects at the bottom as dead surface.
*
* No kebab on the mini bar (operator 2026-06-01): the full kebab
* surface lives on NowPlayingScreen, and dropping it from the mini
@@ -164,6 +167,12 @@ fun MiniPlayer(
durationMs = state.durationMs,
)
MiniRow(
// Take whatever the progress fill leaves. Without this the
// Column stacks 4dp + the row's intrinsic 48dp from the top
// and the remaining 28dp of an 80dp bar sits empty
// underneath — the content looked top-aligned rather than
// centred, with a dead strip above the gesture bar.
modifier = Modifier.weight(1f),
track = track,
isPlaying = state.isPlaying,
isUpnpLoading = state.isUpnpLoading,
@@ -205,6 +214,7 @@ private fun MiniProgressFill(positionMs: Long, durationMs: Long) {
@Composable
@Suppress("LongParameterList")
private fun MiniRow(
modifier: Modifier,
track: TrackRef,
isPlaying: Boolean,
isUpnpLoading: Boolean,
@@ -216,7 +226,7 @@ private fun MiniRow(
onToggleLike: () -> Unit,
) {
Row(
modifier = Modifier
modifier = modifier
.fillMaxWidth()
.padding(horizontal = 12.dp),
verticalAlignment = Alignment.CenterVertically,
@@ -0,0 +1,32 @@
package com.fabledsword.minstrel.player.ui
import kotlin.math.roundToInt
/**
* Drag offset to row delta, rounded to the nearest row boundary. The Android
* half of web's `offsetToDelta` (web/src/lib/components/queue-row-math.ts):
* the two clients are meant to agree, and QueueDragMathTest mirrors web's
* cases so both are held to it (#2436).
*
* Ties round toward positive infinity, as JS `Math.round` does: half a row
* down moves one row, half a row up stays put. A row not yet measured
* (height 0) moves nothing.
*/
internal fun queueDragDelta(offsetPx: Float, rowHeightPx: Int): Int =
if (rowHeightPx > 0) (offsetPx / rowHeightPx).roundToInt() else 0
/**
* The queue index a row dragged from [index] lands on, clamped to the queue.
* A drag past either end stops at the first or last row.
*/
internal fun queueDragTarget(
index: Int,
offsetPx: Float,
rowHeightPx: Int,
queueSize: Int,
): Int {
// coerceIn throws on an empty range; a row can't be dragged in an empty
// queue, but the composable's keys can briefly outlive the list.
if (queueSize <= 0) return index
return (index + queueDragDelta(offsetPx, rowHeightPx)).coerceIn(0, queueSize - 1)
}
@@ -48,7 +48,6 @@ import com.fabledsword.minstrel.shared.formatDuration
import com.fabledsword.minstrel.shared.widgets.LikeButton
import com.fabledsword.minstrel.shared.widgets.ServerImage
import com.fabledsword.minstrel.theme.LocalActionColors
import kotlin.math.roundToInt
/*
* A single queue row, split out of QueueScreen.kt when swipe-to-remove (#2435)
@@ -278,8 +277,7 @@ private fun Modifier.queueReorderDrag(
onOffsetChange(offset)
},
onDragEnd = {
val delta = if (rowHeightPx > 0) (offset / rowHeightPx).roundToInt() else 0
val target = (index + delta).coerceIn(0, queueSize - 1)
val target = queueDragTarget(index, offset, rowHeightPx, queueSize)
if (target != index) onMove(index, target)
offset = 0f
onOffsetChange(0f)
@@ -0,0 +1,55 @@
package com.fabledsword.minstrel.settings.data
import kotlinx.serialization.SerialName
import kotlinx.serialization.Serializable
/**
* The user's loudness-normalization preference (M464 #4998). The server
* holds it, so this device, the web player and casts all level the same
* way; a copy is kept on the device so offline playback still applies it.
* Field names and values match `GET/PUT /api/me/normalization`.
*
* No constructor defaults, on purpose: the app's Json leaves out any field
* equal to its default (encodeDefaults is off), and the server refuses a PUT
* missing a field. Without defaults every field is always sent.
*/
@Serializable
data class NormalizationPrefs(
val mode: NormalizationMode,
@SerialName("target_lufs") val targetLufs: Int,
val boost: NormalizationBoost,
) {
companion object {
/** The targets the server accepts, quietest first. */
val TARGETS: List<Int> = listOf(-18, -16, -14)
/** Must match library.DefaultNormalizationPrefs on the server. */
val DEFAULT: NormalizationPrefs = NormalizationPrefs(
mode = NormalizationMode.AUTO,
targetLufs = -18,
boost = NormalizationBoost.HEADROOM,
)
}
}
@Serializable
enum class NormalizationMode {
/** Tracks play at their mastered volume. */
@SerialName("off") OFF,
/** Album gain while an album plays in order, track gain otherwise. */
@SerialName("auto") AUTO,
@SerialName("track") TRACK,
@SerialName("album") ALBUM,
}
@Serializable
enum class NormalizationBoost {
/** Quiet tracks are raised only as far as their true peak allows. */
@SerialName("headroom") HEADROOM,
/** Quiet tracks are raised all the way and a limiter holds the peaks. */
@SerialName("limiter") LIMITER,
}
@@ -0,0 +1,67 @@
package com.fabledsword.minstrel.settings.data
import com.fabledsword.minstrel.api.endpoints.MeApi
import com.fabledsword.minstrel.auth.AuthStore
import com.fabledsword.minstrel.cache.db.dao.CachedMutationDao
import com.fabledsword.minstrel.cache.mutations.MutationKind
import com.fabledsword.minstrel.cache.mutations.MutationQueue
import kotlinx.coroutines.CancellationException
import kotlinx.coroutines.flow.StateFlow
import retrofit2.Retrofit
import retrofit2.create
import timber.log.Timber
import javax.inject.Inject
import javax.inject.Singleton
/**
* The user's loudness-normalization preference (M464 #4998). The device's
* copy lives in [AuthStore] so playback reads it offline; the server's copy
* is what other devices and casts use.
*
* Writes follow rule 100: the device shows the change at once, the PUT is
* best effort, and a failed PUT is queued for the MutationReplayer.
*/
@Singleton
class NormalizationRepository @Inject constructor(
retrofit: Retrofit,
private val authStore: AuthStore,
private val mutationQueue: MutationQueue,
private val mutationDao: CachedMutationDao,
) {
private val api: MeApi = retrofit.create()
val prefs: StateFlow<NormalizationPrefs> = authStore.normalization
/**
* Takes the server's value, unless a change made here is still queued:
* the server has not seen it yet, and taking its older value would undo
* the change on screen until the replay landed. Throws on a network
* failure; the device's copy then stands.
*/
suspend fun refresh() {
val server = api.getNormalization()
if (mutationDao.hasPending(MutationKind.NORMALIZATION_SET)) return
authStore.setNormalization(server)
}
suspend fun set(next: NormalizationPrefs) {
authStore.setNormalization(next)
// An earlier change still queued would be replayed after this PUT
// and undo it. Queue behind it instead: the replayer sends only the
// newest queued preference.
if (mutationDao.hasPending(MutationKind.NORMALIZATION_SET)) {
mutationQueue.enqueueNormalizationSet(next)
return
}
try {
api.putNormalization(next)
} catch (e: CancellationException) {
throw e
} catch (
@Suppress("TooGenericExceptionCaught") e: Throwable,
) {
Timber.i(e, "normalization: PUT failed; queued for replay")
mutationQueue.enqueueNormalizationSet(next)
}
}
}
@@ -9,7 +9,7 @@ import com.fabledsword.minstrel.update.data.ApkInstaller
import com.fabledsword.minstrel.update.data.InstallStage
import com.fabledsword.minstrel.update.data.UpdateRepository
import com.fabledsword.minstrel.update.data.isBusy
import com.fabledsword.minstrel.update.data.isVersionNewer
import com.fabledsword.minstrel.update.data.isUpdateAvailable
import com.fabledsword.minstrel.update.data.message
import com.fabledsword.minstrel.update.data.stage
import dagger.hilt.android.lifecycle.HiltViewModel
@@ -37,6 +37,14 @@ sealed interface UpdateCheckResult {
data class AboutUiState(
val installedVersion: String = BuildConfig.VERSION_NAME,
// The value the platform installs by, and therefore the one the update
// check must decide on. Held in state rather than read inline so a test
// can drive the comparison without a BuildConfig.
val installedCode: Long = BuildConfig.VERSION_CODE.toLong(),
// A debug build is signed with this machine's debug key, so the server's
// release-signed APK can never install over it (family idea #5103,
// practice 9). It updates from Android Studio instead.
val selfUpdates: Boolean = !BuildConfig.DEBUG,
val isChecking: Boolean = false,
val installStage: InstallStage = InstallStage.IDLE,
val installMessage: String? = null,
@@ -45,8 +53,9 @@ data class AboutUiState(
/**
* Backs the About card's update controls. "Check for updates" calls
* [UpdateRepository.getLatest], compares versus the build's
* VERSION_NAME via [isVersionNewer], and reports the terminal state.
* [UpdateRepository.getLatest], compares versus this build via
* [isUpdateAvailable] — on the ordering key where the server reports one,
* on the name otherwise — and reports the terminal state.
* When an update is available, [install] downloads the APK via
* [ApkInstaller] and installs it — routing the user to the "install
* unknown apps" settings page first when that permission hasn't been
@@ -62,13 +71,21 @@ class AboutCardViewModel @Inject constructor(
val state: StateFlow<AboutUiState> = internal.asStateFlow()
fun checkForUpdates() {
if (internal.value.isChecking) return
if (internal.value.isChecking || !internal.value.selfUpdates) return
viewModelScope.launch {
internal.update { it.copy(isChecking = true, installMessage = null) }
val installed = internal.value.installedVersion
val installedCode = internal.value.installedCode
val result = runCatching { repository.getLatest() }
.map { latest ->
if (isVersionNewer(latest.version, installed)) {
if (
isUpdateAvailable(
serverCode = latest.code,
serverName = latest.version,
installedCode = installedCode,
installedName = installed,
)
) {
UpdateCheckResult.UpdateAvailable(latest)
} else {
UpdateCheckResult.Latest
@@ -0,0 +1,143 @@
package com.fabledsword.minstrel.settings.ui
import androidx.compose.foundation.layout.Arrangement
import androidx.compose.foundation.layout.Column
import androidx.compose.foundation.layout.Row
import androidx.compose.foundation.layout.Spacer
import androidx.compose.foundation.layout.fillMaxWidth
import androidx.compose.foundation.layout.padding
import androidx.compose.foundation.layout.size
import androidx.compose.material3.ElevatedCard
import androidx.compose.material3.ExperimentalMaterial3Api
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.SegmentedButton
import androidx.compose.material3.SegmentedButtonDefaults
import androidx.compose.material3.SingleChoiceSegmentedButtonRow
import androidx.compose.material3.Switch
import androidx.compose.material3.Text
import androidx.compose.runtime.Composable
import androidx.compose.runtime.getValue
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.unit.dp
import androidx.hilt.navigation.compose.hiltViewModel
import androidx.lifecycle.compose.collectAsStateWithLifecycle
import com.fabledsword.minstrel.settings.data.NormalizationBoost
import com.fabledsword.minstrel.settings.data.NormalizationMode
import com.fabledsword.minstrel.settings.data.NormalizationPrefs
/**
* Volume leveling (M464 #4998). The choice is stored on the server, so the
* web player and casts follow it too.
*/
@Composable
fun NormalizationCard(viewModel: NormalizationViewModel = hiltViewModel()) {
val prefs by viewModel.prefs.collectAsStateWithLifecycle()
NormalizationCardContent(prefs = prefs, onChange = viewModel::update)
}
@OptIn(ExperimentalMaterial3Api::class)
@Composable
internal fun NormalizationCardContent(
prefs: NormalizationPrefs,
onChange: (NormalizationPrefs) -> Unit,
) {
ElevatedCard(modifier = Modifier.fillMaxWidth()) {
Column(
modifier = Modifier.padding(16.dp),
verticalArrangement = Arrangement.spacedBy(8.dp),
) {
Text(
text = "Volume leveling",
style = MaterialTheme.typography.titleMedium,
color = MaterialTheme.colorScheme.onSurface,
)
Hint(modeHint(prefs.mode))
val modes = NormalizationMode.entries
SingleChoiceSegmentedButtonRow(modifier = Modifier.fillMaxWidth()) {
modes.forEachIndexed { index, mode ->
SegmentedButton(
selected = mode == prefs.mode,
onClick = { onChange(prefs.copy(mode = mode)) },
shape = SegmentedButtonDefaults.itemShape(
index = index,
count = modes.size,
),
) { Text(modeLabel(mode)) }
}
}
if (prefs.mode != NormalizationMode.OFF) {
TargetRow(prefs = prefs, onChange = onChange)
BoostRow(prefs = prefs, onChange = onChange)
}
}
}
}
@OptIn(ExperimentalMaterial3Api::class)
@Composable
private fun TargetRow(prefs: NormalizationPrefs, onChange: (NormalizationPrefs) -> Unit) {
Text(
text = "Target loudness",
style = MaterialTheme.typography.bodyLarge,
color = MaterialTheme.colorScheme.onSurface,
)
val targets = NormalizationPrefs.TARGETS
SingleChoiceSegmentedButtonRow(modifier = Modifier.fillMaxWidth()) {
targets.forEachIndexed { index, lufs ->
SegmentedButton(
selected = lufs == prefs.targetLufs,
onClick = { onChange(prefs.copy(targetLufs = lufs)) },
shape = SegmentedButtonDefaults.itemShape(
index = index,
count = targets.size,
),
) { Text("$lufs LUFS") }
}
}
}
@Composable
private fun BoostRow(prefs: NormalizationPrefs, onChange: (NormalizationPrefs) -> Unit) {
Row(verticalAlignment = Alignment.CenterVertically) {
Column(modifier = Modifier.weight(1f)) {
Text(
text = "Boost quiet tracks fully",
style = MaterialTheme.typography.bodyLarge,
color = MaterialTheme.colorScheme.onSurface,
)
Hint("A limiter catches the peaks.")
}
Spacer(Modifier.size(12.dp))
Switch(
checked = prefs.boost == NormalizationBoost.LIMITER,
onCheckedChange = { on ->
val boost = if (on) NormalizationBoost.LIMITER else NormalizationBoost.HEADROOM
onChange(prefs.copy(boost = boost))
},
)
}
}
@Composable
private fun Hint(text: String) {
Text(
text = text,
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
private fun modeLabel(mode: NormalizationMode): String = when (mode) {
NormalizationMode.OFF -> "Off"
NormalizationMode.AUTO -> "Auto"
NormalizationMode.TRACK -> "Track"
NormalizationMode.ALBUM -> "Album"
}
private fun modeHint(mode: NormalizationMode): String = when (mode) {
NormalizationMode.OFF -> "Tracks play at their mastered volume."
NormalizationMode.AUTO -> "Album gain for whole albums, track gain otherwise."
NormalizationMode.TRACK -> "Every track at the same loudness."
NormalizationMode.ALBUM -> "Albums keep their own quiet and loud tracks."
}
@@ -0,0 +1,44 @@
package com.fabledsword.minstrel.settings.ui
import androidx.lifecycle.ViewModel
import androidx.lifecycle.viewModelScope
import com.fabledsword.minstrel.settings.data.NormalizationPrefs
import com.fabledsword.minstrel.settings.data.NormalizationRepository
import dagger.hilt.android.lifecycle.HiltViewModel
import kotlinx.coroutines.CancellationException
import kotlinx.coroutines.flow.StateFlow
import kotlinx.coroutines.launch
import timber.log.Timber
import javax.inject.Inject
/**
* Backs the Volume leveling card. Shows the device's copy straight away and
* refreshes it from the server on open, so a change made on the web shows
* here too.
*/
@HiltViewModel
class NormalizationViewModel @Inject constructor(
private val repository: NormalizationRepository,
) : ViewModel() {
val prefs: StateFlow<NormalizationPrefs> = repository.prefs
init {
viewModelScope.launch {
try {
repository.refresh()
} catch (e: CancellationException) {
throw e
} catch (
@Suppress("TooGenericExceptionCaught") e: Throwable,
) {
// Offline: the device's copy stands.
Timber.i(e, "normalization: refresh failed")
}
}
}
fun update(next: NormalizationPrefs) {
viewModelScope.launch { repository.set(next) }
}
}
@@ -57,7 +57,7 @@ class PasswordViewModel @Inject constructor(
try {
repository.changePassword(current = s.current, next = s.next)
internal.update {
PasswordUiState(message = "Password changed.")
PasswordUiState(message = "Password changed. Your other devices have been signed out.")
}
} catch (
@Suppress("TooGenericExceptionCaught") e: Throwable,
@@ -45,6 +45,7 @@ import androidx.compose.ui.unit.dp
import androidx.hilt.navigation.compose.hiltViewModel
import androidx.lifecycle.compose.collectAsStateWithLifecycle
import androidx.navigation.NavHostController
import com.composables.icons.lucide.Bell
import com.composables.icons.lucide.ChevronRight
import com.composables.icons.lucide.ListMusic
import com.composables.icons.lucide.LogOut
@@ -52,6 +53,7 @@ import com.composables.icons.lucide.Lucide
import com.composables.icons.lucide.Shield
import com.fabledsword.minstrel.BuildConfig
import com.fabledsword.minstrel.nav.Admin
import com.fabledsword.minstrel.nav.NotificationSettings
import com.fabledsword.minstrel.nav.Requests
import com.fabledsword.minstrel.nav.Settings as SettingsRoute
import com.fabledsword.minstrel.nav.ServerUrl
@@ -98,6 +100,7 @@ fun SettingsScreen(
themeMode = themeMode,
onPickTheme = themeVm::setThemeMode,
onNavToRequests = { navController.navigate(Requests) },
onNavToNotifications = { navController.navigate(NotificationSettings) },
onNavToAdmin = { navController.navigate(Admin) },
onToggleDiagnostics = viewModel::setDiagnosticsOptOut,
onSignOutClick = { showSignOutConfirm = true },
@@ -121,6 +124,7 @@ private fun SettingsList(
themeMode: ThemeMode,
onPickTheme: (ThemeMode) -> Unit,
onNavToRequests: () -> Unit,
onNavToNotifications: () -> Unit,
onNavToAdmin: () -> Unit,
onToggleDiagnostics: (Boolean) -> Unit,
onSignOutClick: () -> Unit,
@@ -144,6 +148,12 @@ private fun SettingsList(
subtitle = "Track what you've asked Minstrel to add",
onClick = onNavToRequests,
)
NavTile(
icon = Lucide.Bell,
title = "Notifications",
subtitle = "What reaches you, and where",
onClick = onNavToNotifications,
)
if (state.isAdmin) {
NavTile(
icon = Lucide.Shield,
@@ -161,6 +171,7 @@ private fun SettingsList(
onToggle = onToggleDiagnostics,
)
}
NormalizationCard()
AppearanceCard(themeMode = themeMode, onPick = onPickTheme)
StorageCard()
AboutCard()
@@ -382,6 +393,14 @@ private fun AboutCard(viewModel: AboutCardViewModel = hiltViewModel()) {
@Composable
private fun UpdateControls(state: AboutUiState, viewModel: AboutCardViewModel) {
if (!state.selfUpdates) {
Text(
text = "Debug build: updates install from Android Studio.",
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
return
}
UpdateCheckLine(result = state.result)
Button(
onClick = viewModel::checkForUpdates,
@@ -4,6 +4,7 @@ import androidx.lifecycle.ViewModel
import androidx.lifecycle.viewModelScope
import com.fabledsword.minstrel.auth.AuthController
import com.fabledsword.minstrel.auth.AuthStore
import com.fabledsword.minstrel.notifications.data.NotificationsRepository
import dagger.hilt.android.lifecycle.HiltViewModel
import kotlinx.coroutines.flow.MutableStateFlow
import kotlinx.coroutines.flow.StateFlow
@@ -28,6 +29,7 @@ data class SettingsState(
class SettingsViewModel @Inject constructor(
private val authController: AuthController,
private val authStore: AuthStore,
private val notifications: NotificationsRepository,
) : ViewModel() {
private val transient = MutableStateFlow(TransientState())
@@ -70,6 +72,8 @@ class SettingsViewModel @Inject constructor(
viewModelScope.launch {
transient.update { it.copy(isSigningOut = true) }
authController.signOut()
// The inbox is this account's; the next one on the device must not see it.
runCatching { notifications.clearLocal() }
transient.update { it.copy(isSigningOut = false, signedOut = true) }
}
}
@@ -3,6 +3,7 @@ package com.fabledsword.minstrel.shared.widgets
import androidx.lifecycle.ViewModel
import androidx.lifecycle.viewModelScope
import com.fabledsword.minstrel.auth.AuthController
import com.fabledsword.minstrel.notifications.data.NotificationsRepository
import dagger.hilt.android.lifecycle.HiltViewModel
import kotlinx.coroutines.flow.SharingStarted
import kotlinx.coroutines.flow.StateFlow
@@ -22,6 +23,7 @@ private const val SHARE_STOP_TIMEOUT_MS = 5_000L
@HiltViewModel
class AppBarActionsViewModel @Inject constructor(
authController: AuthController,
notifications: NotificationsRepository,
) : ViewModel() {
val isAdmin: StateFlow<Boolean> = authController.currentUser
.map { it?.isAdmin == true }
@@ -30,4 +32,12 @@ class AppBarActionsViewModel @Inject constructor(
started = SharingStarted.WhileSubscribed(SHARE_STOP_TIMEOUT_MS),
initialValue = false,
)
/** Unread notices for the bell's badge (M489). */
val unreadCount: StateFlow<Int> = notifications.unreadCount
.stateIn(
scope = viewModelScope,
started = SharingStarted.WhileSubscribed(SHARE_STOP_TIMEOUT_MS),
initialValue = 0,
)
}
@@ -1,10 +1,13 @@
package com.fabledsword.minstrel.shared.widgets
import androidx.compose.foundation.layout.Row
import androidx.compose.material3.Badge
import androidx.compose.material3.BadgedBox
import androidx.compose.material3.DropdownMenu
import androidx.compose.material3.DropdownMenuItem
import androidx.compose.material3.Icon
import androidx.compose.material3.IconButton
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.Text
import androidx.compose.runtime.Composable
import androidx.compose.runtime.getValue
@@ -14,6 +17,7 @@ import androidx.compose.runtime.setValue
import androidx.hilt.navigation.compose.hiltViewModel
import androidx.lifecycle.compose.collectAsStateWithLifecycle
import androidx.navigation.NavHostController
import com.composables.icons.lucide.Bell
import com.composables.icons.lucide.House
import com.composables.icons.lucide.LibraryBig
import com.composables.icons.lucide.Lucide
@@ -23,6 +27,7 @@ import com.fabledsword.minstrel.nav.Admin
import com.fabledsword.minstrel.nav.Discover
import com.fabledsword.minstrel.nav.Home
import com.fabledsword.minstrel.nav.Library
import com.fabledsword.minstrel.nav.Notifications
import com.fabledsword.minstrel.nav.Playlists
import com.fabledsword.minstrel.nav.Search as SearchRoute
import com.fabledsword.minstrel.nav.Settings as SettingsRoute
@@ -50,6 +55,7 @@ fun MainAppBarActions(
viewModel: AppBarActionsViewModel = hiltViewModel(),
) {
val isAdmin by viewModel.isAdmin.collectAsStateWithLifecycle()
val unread by viewModel.unreadCount.collectAsStateWithLifecycle()
Row {
if (currentRouteName != Home::class.qualifiedName) {
IconButton(onClick = { navController.navigate(Home) }) {
@@ -66,10 +72,47 @@ fun MainAppBarActions(
Icon(Lucide.SearchIcon, contentDescription = "Search")
}
}
if (currentRouteName != Notifications::class.qualifiedName) {
NotificationsBell(unread = unread, onClick = { navController.navigate(Notifications) })
}
OverflowMenu(navController = navController, isAdmin = isAdmin)
}
}
/** The inbox bell with its unread badge (M489): the count to 9, then "9+". */
@Composable
private fun NotificationsBell(unread: Int, onClick: () -> Unit) {
val label = badgeLabel(unread)
IconButton(onClick = onClick) {
// Inverse of the bar, not error red: an unread count is news, not a
// fault, and the house style keeps the accent off general chrome.
BadgedBox(
badge = {
if (label.isNotEmpty()) {
Badge(
containerColor = MaterialTheme.colorScheme.onSurface,
contentColor = MaterialTheme.colorScheme.surface,
) { Text(label) }
}
},
) {
Icon(
Lucide.Bell,
contentDescription = if (unread > 0) "Notifications, $unread unread" else "Notifications",
)
}
}
}
/** Badge text: nothing at zero, the count up to nine, then "9+". */
internal fun badgeLabel(count: Int): String = when {
count <= 0 -> ""
count > MAX_BADGE -> "$MAX_BADGE+"
else -> count.toString()
}
private const val MAX_BADGE = 9
@Composable
private fun OverflowMenu(navController: NavHostController, isAdmin: Boolean) {
var expanded by remember { mutableStateOf(false) }
@@ -19,7 +19,8 @@ private const val POLL_INTERVAL_MS = 24 * 60 * 60 * 1000L
/**
* Drives the shell's soft "update available" banner. Polls
* `/api/client/version` at launch + every 24h and, when the bundled
* APK is strictly newer than this build, exposes its [UpdateInfo] so
* APK outranks this build — by ordering key where the server reports one,
* by name otherwise — exposes its [UpdateInfo] so
* [com.fabledsword.minstrel.update.ui.UpdateBanner] can nudge an
* install. Mirrors Flutter's `ClientUpdateController`.
*
@@ -28,6 +29,9 @@ private const val POLL_INTERVAL_MS = 24 * 60 * 60 * 1000L
* restart re-shows it, which is acceptable nudging for v1 (matches
* Flutter). Server 404 / network errors stay silent. Constructed at
* launch via the construct-the-singleton trick in `MinstrelApplication`.
*
* A debug build never polls: it is signed with a local debug key, so the
* server's release-signed APK could never install over it (#5103).
*/
@Singleton
class UpdateBannerController @Inject constructor(
@@ -44,6 +48,10 @@ class UpdateBannerController @Inject constructor(
}.stateIn(scope, SharingStarted.Eagerly, null)
init {
if (!BuildConfig.DEBUG) startPolling()
}
private fun startPolling() {
scope.launch {
while (true) {
runOnce()
@@ -58,6 +66,13 @@ class UpdateBannerController @Inject constructor(
private suspend fun runOnce() {
val info = runCatching { repository.getLatest() }.getOrNull() ?: return
latest.value = info.takeIf { isVersionNewer(it.version, BuildConfig.VERSION_NAME) }
latest.value = info.takeIf {
isUpdateAvailable(
serverCode = it.code,
serverName = it.version,
installedCode = BuildConfig.VERSION_CODE.toLong(),
installedName = BuildConfig.VERSION_NAME,
)
}
}
}
@@ -21,10 +21,39 @@ class UpdateRepository @Inject constructor(retrofit: Retrofit) {
private fun UpdateInfoWire.toDomain(): UpdateInfo = UpdateInfo(
version = version,
code = code,
channel = channel,
apkUrl = apkUrl,
sizeBytes = sizeBytes,
)
/**
* True when [server] should be offered over the installed build.
*
* **Decide on the ordering key whenever the server sends one.** That is the
* same value Android's package installer compares, so an offer made this way
* implies an install the platform will actually accept. The app used to
* compare NAMES while the platform installed by `versionCode`, with nothing
* keeping the two orderings consistent — so it could offer a build Android
* then refused as a downgrade, or stay quiet about one it would have taken.
*
* Name comparison survives only as the fallback for a server that predates
* the field. A null code means "this server cannot tell me" — never "zero" —
* because treating absent as zero would rank every such server as infinitely
* old and offer its build to everyone, forever.
*/
fun isUpdateAvailable(
serverCode: Long?,
serverName: String,
installedCode: Long,
installedName: String,
): Boolean =
if (serverCode != null) {
serverCode > installedCode
} else {
isVersionNewer(serverName, installedName)
}
/**
* True when [server] is strictly newer than [installed]. Mirrors
* Flutter's `isVersionNewer` — splits both strings on `.`, parses
@@ -0,0 +1,21 @@
<?xml version="1.0" encoding="utf-8"?>
<!-- Status-bar icon for Minstrel's own notifications (M489 #5347): Lucide's
bell, stroked in white. The system tints it; only the alpha is used. -->
<vector xmlns:android="http://schemas.android.com/apk/res/android"
android:width="24dp"
android:height="24dp"
android:viewportWidth="24"
android:viewportHeight="24">
<path
android:pathData="M6,8a6,6 0,0 1,12 0c0,7 3,9 3,9H3s3,-2 3,-9"
android:strokeColor="#FFFFFFFF"
android:strokeWidth="2"
android:strokeLineCap="round"
android:strokeLineJoin="round" />
<path
android:pathData="M10.3,21a1.94,1.94 0,0 0,3.4 0"
android:strokeColor="#FFFFFFFF"
android:strokeWidth="2"
android:strokeLineCap="round"
android:strokeLineJoin="round" />
</vector>
@@ -17,8 +17,12 @@
hostnames rather than CIDR ranges, and both sets of hosts above are unknowable
until runtime. So a permissive base-config is an honest description of our
situation — the gain over the manifest attribute is that the reasoning now
lives somewhere, and there is one place to tighten if a future settings screen
can distinguish a LAN server from a WAN one.
lives somewhere.
The LAN/WAN line this file cannot draw is drawn in code instead:
api/CleartextGuard.kt refuses plain http:// to the Minstrel server whenever
the connection lands on a public address (family baseline #5105, practice
13). LAN servers and UPnP speakers are unaffected.
Worth stating because it looks worse than it is: this is NOT a tamper risk for
the in-app updater. An APK altered in transit and re-signed is rejected by the
@@ -1,6 +1,7 @@
package com.fabledsword.minstrel.api
import com.fabledsword.minstrel.auth.AuthStore
import com.fabledsword.minstrel.auth.FakeSessionVault
import com.fabledsword.minstrel.cache.db.dao.AuthSessionDao
import io.mockk.coEvery
import io.mockk.every
@@ -43,7 +44,7 @@ class AuthCookieInterceptorTest {
coEvery { setSessionCookie(any()) } returns Unit
coEvery { setBaseUrl(any()) } returns Unit
}
authStore = AuthStore(dao, TestScope(UnconfinedTestDispatcher()))
authStore = AuthStore(dao, FakeSessionVault(), TestScope(UnconfinedTestDispatcher()))
// BaseUrlInterceptor rewrites placeholder.invalid → mock server.
// AuthCookieInterceptor scopes its attach + clear behavior to
@@ -1,6 +1,7 @@
package com.fabledsword.minstrel.api
import com.fabledsword.minstrel.auth.AuthStore
import com.fabledsword.minstrel.auth.FakeSessionVault
import com.fabledsword.minstrel.cache.db.dao.AuthSessionDao
import io.mockk.coEvery
import io.mockk.every
@@ -38,7 +39,7 @@ class BaseUrlInterceptorTest {
coEvery { setSessionCookie(any()) } returns Unit
coEvery { setBaseUrl(any()) } returns Unit
}
authStore = AuthStore(dao, TestScope(UnconfinedTestDispatcher()))
authStore = AuthStore(dao, FakeSessionVault(), TestScope(UnconfinedTestDispatcher()))
}
@AfterEach
@@ -0,0 +1,89 @@
package com.fabledsword.minstrel.api
import okhttp3.OkHttpClient
import okhttp3.Request
import okhttp3.mockwebserver.MockResponse
import okhttp3.mockwebserver.MockWebServer
import org.junit.jupiter.api.AfterEach
import org.junit.jupiter.api.BeforeEach
import org.junit.jupiter.api.Test
import org.junit.jupiter.api.assertThrows
import java.net.InetAddress
import kotlin.test.assertEquals
import kotlin.test.assertFalse
import kotlin.test.assertTrue
/** Plain http:// to the Minstrel server only on a private address (#5105, practice 13). */
class CleartextGuardTest {
private lateinit var server: MockWebServer
@BeforeEach
fun setup() {
server = MockWebServer().apply { start() }
}
@AfterEach
fun teardown() {
server.shutdown()
}
private fun ip(s: String) = InetAddress.getByName(s)
@Test
fun `home network and overlay addresses may use plain http`() {
for (a in listOf(
"127.0.0.1", "10.1.2.3", "172.16.0.9", "172.31.255.1", "192.168.1.20",
"169.254.3.4", "100.64.0.1", "100.127.255.254", "::1", "fd12:3456::1", "fe80::1",
)) {
assertTrue(CleartextPolicy.allows(ip(a)), "$a should be allowed")
}
}
@Test
fun `public addresses may not`() {
for (a in listOf(
"8.8.8.8", "172.32.0.1", "100.128.0.1", "100.63.255.255", "203.0.113.5",
"2001:db8::1", "2606:4700::1111",
)) {
assertFalse(CleartextPolicy.allows(ip(a)), "$a should be refused")
}
}
private fun client(allows: Boolean) = OkHttpClient.Builder()
.addNetworkInterceptor(CleartextGuardInterceptor { allows })
.build()
private fun serverRequest() = Request.Builder()
.url(server.url("/api/auth/login"))
.tag(MinstrelServerRequest::class.java, MinstrelServerRequest)
.build()
@Test
fun `a refused server request sends nothing`() {
server.enqueue(MockResponse().setResponseCode(200))
assertThrows<CleartextToPublicHostException> {
client(allows = false).newCall(serverRequest()).execute()
}
assertEquals(0, server.requestCount, "the request reached the server")
}
@Test
fun `an allowed server request goes through`() {
server.enqueue(MockResponse().setResponseCode(200))
client(allows = true).newCall(serverRequest()).execute().use { assertEquals(200, it.code) }
assertEquals(1, server.requestCount)
}
@Test
fun `requests not bound for the Minstrel server are left alone`() {
server.enqueue(MockResponse().setResponseCode(200))
client(allows = false).newCall(Request.Builder().url(server.url("/art.jpg")).build())
.execute().use { assertEquals(200, it.code) }
}
@Test
fun `the refusal has its own message`() {
val msg = ErrorCopy.fromThrowable(CleartextToPublicHostException("music.example.com"))
assertTrue(msg.contains("https://"), msg)
}
}
@@ -0,0 +1,60 @@
package com.fabledsword.minstrel.api
import okhttp3.MediaType.Companion.toMediaType
import okhttp3.ResponseBody.Companion.toResponseBody
import org.junit.jupiter.api.Assertions.assertEquals
import org.junit.jupiter.api.Test
import retrofit2.HttpException
import retrofit2.Response
import java.io.IOException
class ErrorCopyTest {
private fun httpError(status: Int, body: String): HttpException =
HttpException(
Response.error<Unit>(status, body.toResponseBody("application/json".toMediaType())),
)
@Test
fun libraryNotWritableAppendsTheServerDetail() {
val detail = "Minstrel runs as uid 1000, gid 1000 and cannot delete from /music/A " +
"(read-only file system). The library mount must be writable by that user. " +
"Nothing was deleted."
val e = httpError(409, """{"error":{"code":"library_not_writable","message":"$detail"}}""")
assertEquals(
"${ErrorCopy.messageFor("library_not_writable")} $detail",
ErrorCopy.fromThrowable(e),
)
}
@Test
fun detailCodeWithoutAMessageShowsTheCopyAlone() {
val e = httpError(409, """{"error":{"code":"library_not_writable","message":""}}""")
assertEquals(ErrorCopy.messageFor("library_not_writable"), ErrorCopy.fromThrowable(e))
}
// Server messages are usually internal detail; appending them for every
// code would leak driver errors into snackbars. This pins the scope.
@Test
fun otherCodesNeverCarryTheServerMessage() {
val e = httpError(404, """{"error":{"code":"track_not_found","message":"pgx: no rows"}}""")
assertEquals(ErrorCopy.messageFor("track_not_found"), ErrorCopy.fromThrowable(e))
}
@Test
fun anUnparseableBodyFallsBackToUnknown() {
val e = httpError(500, "not json")
assertEquals(ErrorCopy.messageFor("unknown"), ErrorCopy.fromThrowable(e))
}
@Test
fun transportFailureMapsToConnectionRefused() {
assertEquals(
ErrorCopy.messageFor("connection_refused"),
ErrorCopy.fromThrowable(IOException("refused")),
)
}
}
@@ -0,0 +1,111 @@
package com.fabledsword.minstrel.auth
import com.fabledsword.minstrel.cache.db.dao.AuthSessionDao
import com.fabledsword.minstrel.cache.db.entities.AuthSessionEntity
import io.mockk.coEvery
import io.mockk.coVerify
import io.mockk.every
import io.mockk.mockk
import kotlinx.coroutines.ExperimentalCoroutinesApi
import kotlinx.coroutines.flow.flowOf
import kotlinx.coroutines.test.StandardTestDispatcher
import kotlinx.coroutines.test.TestScope
import kotlinx.coroutines.test.UnconfinedTestDispatcher
import kotlinx.coroutines.test.advanceUntilIdle
import kotlinx.coroutines.test.runTest
import org.junit.jupiter.api.Test
import kotlin.test.assertEquals
import kotlin.test.assertNull
/**
* The session cookie moved out of the Room row into a Keystore-backed vault
* (M462 #4985). What matters is that nobody is signed out by it: an install
* upgrading with a cookie in the row keeps it, a device whose Keystore will
* not work keeps the old storage, and a sign-in or 401 that lands while the
* one-time move runs is never overwritten by the value it read.
*/
@OptIn(ExperimentalCoroutinesApi::class)
class AuthStoreSessionVaultTest {
/** A DAO whose single row holds [legacyCookie] in its sessionCookie column. */
private class RowDao(var legacyCookie: String?) {
val dao: AuthSessionDao = mockk {
every { observe() } returns flowOf(null)
coEvery { get() } answers {
AuthSessionEntity(baseUrl = "http://music.local", sessionCookie = legacyCookie)
}
coEvery { upsert(any()) } answers { legacyCookie = firstArg<AuthSessionEntity>().sessionCookie }
coEvery { setSessionCookie(any()) } answers { legacyCookie = firstArg() }
}
}
@Test
fun `an upgrading install keeps its session, moved out of the row into the vault`() = runTest {
val row = RowDao(legacyCookie = "session=abc")
val vault = FakeSessionVault()
val store = AuthStore(row.dao, vault, TestScope(UnconfinedTestDispatcher(testScheduler)))
store.awaitSessionHydrated()
assertEquals("session=abc", store.sessionCookie.value)
assertEquals("session=abc", vault.stored)
assertNull(row.legacyCookie, "the plain-text copy must be cleared from the row")
}
@Test
fun `a cookie already in the vault is loaded without touching the row`() = runTest {
val row = RowDao(legacyCookie = null)
val vault = FakeSessionVault(stored = "session=xyz")
val store = AuthStore(row.dao, vault, TestScope(UnconfinedTestDispatcher(testScheduler)))
store.awaitSessionHydrated()
assertEquals("session=xyz", store.sessionCookie.value)
assertEquals(0, vault.writes)
coVerify(exactly = 0) { row.dao.setSessionCookie(any()) }
}
@Test
fun `when the Keystore will not work the cookie stays in the row instead of being lost`() = runTest {
val row = RowDao(legacyCookie = "session=abc")
val vault = FakeSessionVault(available = false)
val store = AuthStore(row.dao, vault, TestScope(UnconfinedTestDispatcher(testScheduler)))
store.awaitSessionHydrated()
assertEquals("session=abc", store.sessionCookie.value)
assertEquals("session=abc", row.legacyCookie)
store.setSessionCookie("session=new")
assertEquals("session=new", row.legacyCookie, "fallback writes go to the row")
}
@Test
fun `sign-in and sign-out write the vault, and never the row`() = runTest {
val row = RowDao(legacyCookie = null)
val vault = FakeSessionVault()
val store = AuthStore(row.dao, vault, TestScope(UnconfinedTestDispatcher(testScheduler)))
store.awaitSessionHydrated()
store.setSessionCookie("session=new")
assertEquals("session=new", vault.stored)
assertNull(row.legacyCookie)
store.setSessionCookie(null)
assertNull(vault.stored)
assertNull(store.sessionCookie.value)
}
@Test
fun `a sign-out that lands before hydration finishes is not undone by it`() = runTest {
val row = RowDao(legacyCookie = "session=stale")
val vault = FakeSessionVault()
val scope = TestScope(StandardTestDispatcher(testScheduler))
// Hydration is queued but has not run; a 401 clears the session first.
val store = AuthStore(row.dao, vault, scope)
store.setSessionCookie(null)
scope.advanceUntilIdle()
assertNull(store.sessionCookie.value, "hydration must not resurrect the stale cookie")
assertNull(vault.stored)
}
}
@@ -0,0 +1,23 @@
package com.fabledsword.minstrel.auth
/**
* In-memory [SessionVault] for JVM tests: the Android Keystore has no JVM
* implementation. [available] = false stands in for a device whose Keystore
* refuses to work, so callers' fallback paths can be exercised.
*/
class FakeSessionVault(
var stored: String? = null,
var available: Boolean = true,
) : SessionVault {
var writes = 0
private set
override fun read(): String? = if (available) stored else null
override fun write(value: String?): Boolean {
if (!available) return false
writes++
stored = value
return true
}
}
@@ -0,0 +1,48 @@
package com.fabledsword.minstrel.auth
import org.junit.jupiter.api.Test
import org.junit.jupiter.api.assertThrows
import java.util.Base64
import javax.crypto.KeyGenerator
import javax.crypto.SecretKey
import kotlin.test.assertEquals
import kotlin.test.assertFalse
import kotlin.test.assertNotEquals
/**
* The sealing half of the session vault, with an ordinary JVM AES key in
* place of the Keystore one (the Keystore has no JVM implementation).
*/
class SealedBoxTest {
private fun newKey(): SecretKey = KeyGenerator.getInstance("AES").apply { init(256) }.generateKey()
@Test
fun `round-trips a cookie`() {
val key = newKey()
assertEquals("session=abc", SealedBox.open(key, SealedBox.seal(key, "session=abc")))
}
@Test
fun `the stored form does not contain the cookie, and differs every time`() {
val key = newKey()
val first = SealedBox.seal(key, "session=abc")
val second = SealedBox.seal(key, "session=abc")
assertNotEquals(first, second, "a fresh IV per seal")
val decoded = String(Base64.getDecoder().decode(first), Charsets.ISO_8859_1)
assertFalse(decoded.contains("session=abc"), "plaintext visible in the sealed value")
}
@Test
fun `a tampered value does not open`() {
val key = newKey()
val bytes = Base64.getDecoder().decode(SealedBox.seal(key, "session=abc"))
bytes[bytes.size - 1] = (bytes[bytes.size - 1].toInt() xor 1).toByte()
assertThrows<Exception> { SealedBox.open(key, Base64.getEncoder().encodeToString(bytes)) }
}
@Test
fun `a value sealed under another key does not open`() {
val sealed = SealedBox.seal(newKey(), "session=abc")
assertThrows<Exception> { SealedBox.open(newKey(), sealed) }
}
}
@@ -1,6 +1,9 @@
package com.fabledsword.minstrel.cache.mutations
import com.fabledsword.minstrel.api.endpoints.NotificationSettingChangeWire
import com.fabledsword.minstrel.cache.db.entities.CachedMutationEntity
import com.fabledsword.minstrel.settings.data.NormalizationMode
import com.fabledsword.minstrel.settings.data.NormalizationPrefs
import kotlinx.serialization.json.Json
import org.junit.jupiter.api.Test
import kotlin.test.assertEquals
@@ -129,4 +132,56 @@ class SupersededToggleIdsTest {
fun `an empty queue collapses nothing`() {
assertTrue(supersededToggleIds(emptyList(), json).isEmpty())
}
private fun normalizationRow(id: Long, mode: NormalizationMode) = CachedMutationEntity(
id = id,
kind = MutationKind.NORMALIZATION_SET,
payload = json.encodeToString(
NormalizationPrefs.serializer(),
NormalizationPrefs.DEFAULT.copy(mode = mode),
),
)
// There is one preference per user, so any two queued changes to it
// collapse, and only the newest is sent (#4998).
@Test
fun `only the newest queued normalization change survives`() {
val rows = listOf(
normalizationRow(1, NormalizationMode.TRACK),
snoozeRow(2, "mb-a", desiredSnoozed = true),
normalizationRow(3, NormalizationMode.OFF),
normalizationRow(4, NormalizationMode.ALBUM),
)
assertEquals(setOf(1L, 3L), supersededToggleIds(rows, json))
}
private fun settingRow(id: Long, kind: String, channel: String, value: Boolean) = CachedMutationEntity(
id = id,
kind = MutationKind.NOTIFICATION_SETTING_SET,
payload = json.encodeToString(
NotificationSettingPayload.serializer(),
NotificationSettingPayload(kind, channel, value),
),
)
@Test
fun `notification settings collapse per kind and channel, never across them`() {
val rows = listOf(
settingRow(1, "request_completed", "email", false),
settingRow(2, "request_completed", "phone", false),
settingRow(3, "request_completed", "email", true),
settingRow(4, "request_approved", "email", false),
)
// Only the older email toggle for request_completed is superseded.
assertEquals(setOf(1L), supersededToggleIds(rows, json))
}
@Test
fun `a queued setting becomes a one-channel change, and an unknown channel none`() {
assertEquals(
NotificationSettingChangeWire(kind = "k", phone = true),
notificationSettingChange(NotificationSettingPayload("k", "phone", true)),
)
assertEquals(null, notificationSettingChange(NotificationSettingPayload("k", "pager", true)))
}
}
@@ -0,0 +1,24 @@
package com.fabledsword.minstrel.events
import org.junit.jupiter.api.Test
import kotlin.random.Random
import kotlin.test.assertEquals
import kotlin.test.assertTrue
class ReconnectBackoffTest {
@Test
fun `doubles from two seconds to a five minute cap`() {
val ladder = generateSequence(ReconnectBackoff.BASE_MS) { ReconnectBackoff.next(it) }.take(10).toList()
assertEquals(listOf(2_000L, 4_000L, 8_000L, 16_000L, 32_000L, 64_000L, 128_000L, 256_000L), ladder.take(8))
assertEquals(300_000L, ladder[8])
assertEquals(300_000L, ladder[9])
}
@Test
fun `jitter stays within a quarter either way and does spread`() {
val random = Random(42)
val waits = List(1_000) { ReconnectBackoff.jittered(100_000L, random) }
assertTrue(waits.all { it in 75_000L..125_000L }, "out of bounds: ${waits.minOrNull()}..${waits.maxOrNull()}")
assertTrue(waits.toSet().size > 100, "jitter should spread reconnects")
}
}
@@ -0,0 +1,107 @@
package com.fabledsword.minstrel.notifications.data
import com.fabledsword.minstrel.api.endpoints.NotificationKindSettingWire
import com.fabledsword.minstrel.api.endpoints.NotificationSettingChangeWire
import com.fabledsword.minstrel.api.endpoints.NotificationSettingsWire
import com.fabledsword.minstrel.api.endpoints.NotificationsApi
import com.fabledsword.minstrel.api.endpoints.PutNotificationSettingsBody
import com.fabledsword.minstrel.cache.db.dao.CachedMutationDao
import com.fabledsword.minstrel.cache.db.dao.CachedNotificationDao
import com.fabledsword.minstrel.cache.db.entities.CachedNotificationSettingsEntity
import com.fabledsword.minstrel.cache.mutations.MutationKind
import com.fabledsword.minstrel.cache.mutations.MutationQueue
import com.fabledsword.minstrel.cache.mutations.NotificationSettingPayload
import io.mockk.coEvery
import io.mockk.coVerify
import io.mockk.every
import io.mockk.mockk
import io.mockk.slot
import kotlinx.coroutines.test.runTest
import kotlinx.serialization.json.Json
import org.junit.jupiter.api.Test
import retrofit2.Retrofit
import java.io.IOException
import kotlin.test.assertEquals
import kotlin.test.assertFalse
/** Settings follow snippet #5107: shown at once, never overwritten or reordered by an older change. */
class NotificationSettingsRepositoryTest {
private val json = Json { ignoreUnknownKeys = true }
private val api: NotificationsApi = mockk(relaxed = true)
private val dao: CachedNotificationDao = mockk(relaxed = true)
private val mutationDao: CachedMutationDao = mockk()
private val queue: MutationQueue = mockk(relaxed = true)
private val retrofit: Retrofit = mockk {
every { create(NotificationsApi::class.java) } returns api
}
private val repo = NotificationSettingsRepository(retrofit, dao, mutationDao, queue, json)
private val settings = NotificationSettingsWire(
kinds = listOf(
NotificationKindSettingWire("request_completed", false, inbox = true, phone = true, email = true),
),
emailAvailable = true,
)
private fun cached() {
coEvery { dao.getSettings() } returns CachedNotificationSettingsEntity(
json = json.encodeToString(NotificationSettingsWire.serializer(), settings),
)
}
@Test
fun `a toggle is shown at once and sends only that channel`() = runTest {
cached()
coEvery { mutationDao.hasPending(MutationKind.NOTIFICATION_SETTING_SET) } returns false
val saved = mutableListOf<CachedNotificationSettingsEntity>()
coEvery { dao.upsertSettings(capture(saved)) } returns Unit
val body = slot<PutNotificationSettingsBody>()
coEvery { api.putSettings(capture(body)) } returns settings
repo.set("request_completed", NotificationChannel.EMAIL, false)
val optimistic = json.decodeFromString(NotificationSettingsWire.serializer(), saved.first().json)
assertFalse(optimistic.kinds.single().email)
assertEquals(
listOf(NotificationSettingChangeWire(kind = "request_completed", email = false)),
body.captured.kinds,
)
coVerify(exactly = 0) { queue.enqueueNotificationSettingSet(any()) }
}
@Test
fun `a toggle that cannot reach the server is queued`() = runTest {
cached()
coEvery { mutationDao.hasPending(MutationKind.NOTIFICATION_SETTING_SET) } returns false
coEvery { api.putSettings(any()) } throws IOException("offline")
repo.set("request_completed", NotificationChannel.PHONE, false)
coVerify {
queue.enqueueNotificationSettingSet(NotificationSettingPayload("request_completed", "phone", false))
}
}
@Test
fun `a toggle behind a queued one queues too`() = runTest {
cached()
coEvery { mutationDao.hasPending(MutationKind.NOTIFICATION_SETTING_SET) } returns true
repo.set("request_completed", NotificationChannel.INBOX, false)
coVerify(exactly = 0) { api.putSettings(any()) }
coVerify {
queue.enqueueNotificationSettingSet(NotificationSettingPayload("request_completed", "inbox", false))
}
}
@Test
fun `refresh leaves a queued change on screen`() = runTest {
coEvery { api.getSettings() } returns settings
coEvery { mutationDao.hasPending(MutationKind.NOTIFICATION_SETTING_SET) } returns true
repo.refresh()
coVerify(exactly = 0) { dao.upsertSettings(any()) }
}
}
@@ -0,0 +1,136 @@
package com.fabledsword.minstrel.notifications.data
import com.fabledsword.minstrel.api.endpoints.NotificationWire
import com.fabledsword.minstrel.api.endpoints.NotificationsApi
import com.fabledsword.minstrel.api.endpoints.NotificationsPageWire
import com.fabledsword.minstrel.api.endpoints.ReadAllBody
import com.fabledsword.minstrel.cache.db.dao.CachedNotificationDao
import com.fabledsword.minstrel.cache.db.entities.CachedNotificationEntity
import com.fabledsword.minstrel.cache.mutations.MutationQueue
import io.mockk.coEvery
import io.mockk.coVerify
import io.mockk.every
import io.mockk.mockk
import io.mockk.slot
import kotlinx.coroutines.test.runTest
import kotlinx.datetime.Instant
import okhttp3.ResponseBody.Companion.toResponseBody
import org.junit.jupiter.api.Test
import retrofit2.HttpException
import retrofit2.Response
import retrofit2.Retrofit
import java.io.IOException
import kotlin.test.assertEquals
import kotlin.test.assertNull
/**
* Reads are offline-first (rule 100): the device marks the row at once, and a
* read the server has not taken is queued, except when the server says the
* notice is gone.
*/
class NotificationsRepositoryTest {
private val api: NotificationsApi = mockk(relaxed = true)
private val dao: CachedNotificationDao = mockk(relaxed = true)
private val queue: MutationQueue = mockk(relaxed = true)
private val retrofit: Retrofit = mockk {
every { create(NotificationsApi::class.java) } returns api
}
private val repo = NotificationsRepository(retrofit, dao, queue)
private fun httpError(status: Int) = HttpException(Response.error<Unit>(status, "".toResponseBody()))
private fun row(id: String, created: String, readAt: Instant? = null) = CachedNotificationEntity(
id = id,
kind = "request_completed",
title = "t",
body = "b",
link = "/requests",
createdAt = Instant.parse(created),
readAt = readAt,
)
@Test
fun `a read is shown at once and sent`() = runTest {
repo.markRead("n1")
coVerify { dao.markRead("n1", any()) }
coVerify { api.markRead("n1") }
coVerify(exactly = 0) { queue.enqueueNotificationRead(any()) }
}
@Test
fun `a read that cannot reach the server is queued`() = runTest {
coEvery { api.markRead("n1") } throws IOException("offline")
repo.markRead("n1")
coVerify { dao.markRead("n1", any()) }
coVerify { queue.enqueueNotificationRead("n1") }
}
@Test
fun `a notice the server no longer has is not queued`() = runTest {
coEvery { api.markRead("n1") } throws httpError(404)
repo.markRead("n1")
coVerify(exactly = 0) { queue.enqueueNotificationRead(any()) }
}
@Test
fun `a server error queues the read for later`() = runTest {
coEvery { api.markRead("n1") } throws httpError(503)
repo.markRead("n1")
coVerify { queue.enqueueNotificationRead("n1") }
}
@Test
fun `mark all read sends the newest notice shown, rounded up, and queues it offline`() = runTest {
coEvery { dao.getAll() } returns listOf(
row("a", "2026-10-08T10:00:00.123Z"),
row("b", "2026-10-08T11:00:00.456Z"),
)
val body = slot<ReadAllBody>()
coEvery { api.readAll(capture(body)) } throws IOException("offline")
repo.markAllRead()
coVerify { dao.markAllRead(any()) }
assertEquals("2026-10-08T11:00:00.457Z", body.captured.upTo)
coVerify { queue.enqueueNotificationsReadAll("2026-10-08T11:00:00.457Z") }
}
@Test
fun `refresh keeps a read made here that the server has not seen yet`() = runTest {
val readHere = Instant.parse("2026-10-08T12:00:00Z")
coEvery { dao.getAll() } returns listOf(row("a", "2026-10-08T10:00:00Z", readAt = readHere))
coEvery { api.list(any()) } returns NotificationsPageWire(
items = listOf(
NotificationWire("a", "request_completed", "t", "b", "/requests", "2026-10-08T10:00:00Z", null),
NotificationWire("b", "request_completed", "t", "b", "/requests", "2026-10-08T11:00:00Z", null),
),
unreadCount = 2,
)
val saved = slot<List<CachedNotificationEntity>>()
coEvery { dao.replaceAll(capture(saved)) } returns Unit
repo.refresh()
val byId = saved.captured.associateBy { it.id }
assertEquals(readHere, byId.getValue("a").readAt)
assertNull(byId.getValue("b").readAt)
}
@Test
fun `a notice with an unreadable timestamp is skipped, not the whole page`() {
val bad = NotificationWire("x", "k", "t", "b", "/", "yesterday", null)
assertNull(bad.toEntity(null))
}
@Test
fun `no cutoff when nothing is shown`() {
assertNull(readAllCutoff(emptyList()))
}
}
@@ -0,0 +1,128 @@
package com.fabledsword.minstrel.notifications.delivery
import com.fabledsword.minstrel.api.endpoints.NotificationKindSettingWire
import com.fabledsword.minstrel.api.endpoints.NotificationSettingsWire
import com.fabledsword.minstrel.cache.db.entities.CachedNotificationEntity
import kotlinx.datetime.Instant
import org.junit.jupiter.api.Test
import kotlin.test.assertEquals
import kotlin.test.assertFalse
import kotlin.test.assertIs
import kotlin.test.assertNull
import kotlin.test.assertTrue
class DeliveryPlanTest {
private val t0 = Instant.parse("2026-10-08T12:00:00Z")
private fun notice(
id: String,
minutes: Long,
kind: String = "request_approved",
read: Boolean = false,
) = CachedNotificationEntity(
id = id,
kind = kind,
title = "t-$id",
body = "b-$id",
link = "/requests",
createdAt = Instant.fromEpochMilliseconds(t0.toEpochMilliseconds() + minutes * 60_000),
readAt = if (read) t0 else null,
)
private val allOn: (String) -> Boolean = { true }
@Test
fun `a first look sets the mark to the newest and announces nothing`() {
val page = listOf(notice("b", 5), notice("a", 1))
val plan = planCatchUp(page, mark = null, phoneOn = allOn)
assertTrue(plan.announce.isEmpty())
assertEquals(page[0].createdAt, plan.upTo)
}
@Test
fun `a first look at an empty inbox still lets the first notice through`() {
val first = planCatchUp(emptyList(), mark = null, phoneOn = allOn)
assertEquals(Instant.DISTANT_PAST, first.upTo)
val next = planCatchUp(listOf(notice("a", 1)), mark = first.upTo, phoneOn = allOn)
assertEquals(listOf("a"), next.announce.map { it.id })
}
@Test
fun `announces unread notices newer than the mark, oldest first, and moves the mark`() {
val page = listOf(notice("c", 9), notice("b", 6), notice("old", 2), notice("read", 7, read = true))
val plan = planCatchUp(page, mark = notice("m", 3).createdAt, phoneOn = allOn)
assertEquals(listOf("b", "c"), plan.announce.map { it.id })
assertEquals(page[0].createdAt, plan.upTo)
}
@Test
fun `nothing new leaves the mark where it was`() {
val mark = notice("m", 10).createdAt
val plan = planCatchUp(listOf(notice("a", 1)), mark = mark, phoneOn = allOn)
assertTrue(plan.announce.isEmpty())
assertEquals(mark, plan.upTo)
}
@Test
fun `a kind with the phone off is passed over, and not announced later either`() {
val page = listOf(notice("h", 5, kind = "tracks_missing"), notice("r", 4))
val plan = planCatchUp(page, mark = t0, phoneOn = { it != "tracks_missing" })
assertEquals(listOf("r"), plan.announce.map { it.id })
assertEquals(page[0].createdAt, plan.upTo, "the mark passes it")
}
@Test
fun `phone prefs come from the cached settings, with every kind on when none are cached`() {
val settings = NotificationSettingsWire(
kinds = listOf(
NotificationKindSettingWire("request_approved", false, inbox = true, phone = false, email = true),
NotificationKindSettingWire("request_completed", false, inbox = false, phone = true, email = true),
NotificationKindSettingWire("request_rejected", false, inbox = true, phone = true, email = true),
),
emailAvailable = true,
)
val phoneOn = phoneOnFor(settings)
assertFalse(phoneOn("request_approved"))
assertFalse(phoneOn("request_completed"), "phone rides on the inbox")
assertTrue(phoneOn("request_rejected"))
assertTrue(phoneOnFor(null)("tracks_missing"))
}
@Test
fun `three get a line each, more become one line with the count`() {
assertNull(announcementFor(emptyList()))
val three = (1..3).map { notice("n$it", it.toLong()) }
assertEquals(Announcement.Each(three), announcementFor(three))
val five = (1..5).map { notice("n$it", it.toLong()) }
val pile = assertIs<Announcement.Pile>(announcementFor(five))
assertEquals(5, pile.count)
assertEquals(ShadeChannel.YOUR_REQUESTS, pile.channel)
assertEquals("5 new notifications", pileText(pile.count))
}
@Test
fun `a pile of admin notices goes to library health, a mixed one does not`() {
val admin = (1..4).map { notice("a$it", it.toLong(), kind = "tracks_missing") }
assertEquals(ShadeChannel.LIBRARY_HEALTH, assertIs<Announcement.Pile>(announcementFor(admin)).channel)
val mixed = admin + notice("r", 9)
assertEquals(ShadeChannel.YOUR_REQUESTS, assertIs<Announcement.Pile>(announcementFor(mixed)).channel)
}
@Test
fun `listener kinds are their requests, the rest is library health`() {
listOf("request_approved", "request_rejected", "request_completed").forEach {
assertEquals(ShadeChannel.YOUR_REQUESTS, shadeChannelFor(it))
}
listOf("request_pending", "quarantine_flagged", "scan_failed", "tracks_missing").forEach {
assertEquals(ShadeChannel.LIBRARY_HEALTH, shadeChannelFor(it))
}
}
@Test
fun `the service runs only signed in with background delivery on`() {
assertTrue(deliveryWanted(signedIn = true, enabled = true))
assertFalse(deliveryWanted(signedIn = true, enabled = false))
assertFalse(deliveryWanted(signedIn = false, enabled = true))
}
}

Some files were not shown because too many files have changed in this diff Show More