Three from review.
`info.video`/`info.audio` are declared — and sent by the mock server and many
panels — as string arrays, but the resolver only read the ffprobe object shape.
Every array response therefore lost the provider's codec, so those source rows
showed no codec fact and the "dub may differ" warning could never fire.
`readStreamInfo` now accepts both, and states nothing when the provider stated
nothing.
The FTS-empty fallback scan matched only the FIRST token, which is fine for a
one-word short title but not for "I Am": every catalog row containing the word
"i" came back for TypeScript to throw away — a full scan of a large catalog on
the single database worker, just to open a detail page. Every token must now
appear.
`writePin` reported success when the canonical write landed but retiring the
old alias failed. Lookups read aliases before the canonical key, so reopening
the movie before enrichment would start the source the user just replaced,
with the icon promising otherwise.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
`setActiveSource` marked the source tried, but discovery calls it the moment
the page opens and a pin or the picker can call it before anything plays. So
opening a movie burned the route copy's turn: if a pinned alternative then
failed, failover skipped a healthy untouched source — and with only one
alternative, reported the options exhausted.
Selection and attempt are now separate. `setActiveSource` selects;
`markPlaying` also spends the turn, and only the three places that really
start playback call it. `runFailover` additionally retires whatever is on
screen before picking, so the failing source is spent however it got there —
relying on the start paths alone would leave one hole per path, and the cost
of missing it is a ping-pong between two sources.
One existing spec asserted the old behaviour (a route copy burned by a switch
it never played); it now plays first, so it still covers what it meant to —
that the tried set survives a rediscovery.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Two follow-ups to the previous round, both in code it introduced.
A width that matched exactly one sub-HD format ignored the height entirely,
so 640x480 came back as 360p — a measurement the numbers contradict. The
height now vetoes, but only in the direction that can be wrong: cropping
removes lines, so a SHORTER frame is a letterboxed master of that format and
the width still names it, while a taller one is a different shape and gets
no tag. That keeps the reason width is preferred in the first place.
And picking a source off the error screen cleared the failure state before
the switch resolved, so an alternative that could not be resolved left the
diagnostic on screen while the caption went back to claiming playback. The
flag now clears only once a switch actually starts something.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Two follow-ups to the previous round, both the same rule again.
The 800-wide band still answered from the width alone, so 800x600 and
800x450 were labelled 480p — published with `api` provenance, so read as a
measurement. Sub-HD formats are now matched against known shapes with the
same 5% tolerance the height path uses, and anything unrecognised carries no
tag at all.
And "Playing from ..." survived a playback failure: the inline host stays
mounted while the diagnostic is on screen, so the page named a source for a
stream it had just reported it could not play. The caption now clears on
failure and returns when the engine produces time again.
Splits the route playback spec along the "what it does" / "what it claims"
seam and lifts the repeated active-source stub into one helper.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Three findings, all of them the same rule: never state as fact something the
app has not established.
The "Playing from …" caption appeared as soon as discovery marked a source
active — before Play was pressed, and again after the player was closed. It
now requires a player that is actually running.
Probe answers were cached by URL alone, but the request now carries the
playlist's headers. Two playlists sharing a stream URL could therefore be
told the other's answer, marking a source dead without ever asking it.
And any width below 900 was labelled 480p, published with `api` provenance:
a 640x360 stream stated 480p as a fact, and a 720x576 PAL source likewise.
Widths below HD only resolve with the height — 720 is NTSC 480p or PAL 576p
— so an unrecognised shape now carries no quality tag at all.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Three findings from the latest review pass.
`normalizeTitleKeys` strips bracketed segments as tag noise, so "Dune (1984)"
normalizes to exactly "dune" — an EXACT match for the 2021 film, ranked above
every fuzzy one, with the year never consulted because that tier skipped the
gate. Auto-failover could switch the user to the other film entirely. The
year is now read out of brackets too, and a stated disagreement rejects the
row on either tier.
Playback positions are keyed by (playlist, stream), so watching through a
pinned alternative stores progress under ITS ids while the page loads the
route copy's row. Starting the pin therefore resumed from a position
belonging to a different copy — usually zero. It now loads its own.
And a pin can point at another copy of the film inside the playlist being
viewed, which discovery excludes wholesale: the pinned row was absent from
the list, so nothing showed as pinned and Play ignored the preference. The
pin is now read before discovery, which keeps that one row.
Moves the pin-shaped decisions into the pin module, where the persistence
helpers already live.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Five findings from the round-5 review.
An external player launched for an alternative carries that playlist's ids,
so the page disowned its own session: the primary button never became Stop,
stopping found nothing to stop, and another click opened a second player.
Multi-source now tells playback which source is actually active, and the
matcher accepts either that or the route's own stream.
Stop also has to beat the pin. The primary action consults the pin first —
that is what makes a pin decide where playback starts — but while a session
is running the same button reads Stop, and consulting the pin there made the
control do the opposite of its label.
A pinned source started from the Resume button resolved at zero, because
nothing reports a live position until the first timeupdate. The controller is
now seeded from the persisted position, one-way: a live value always wins,
since the stored one lags it and applying it would rewind.
A pin whose write failed was still shown as pinned, promising a preference
that reopening the movie would not have.
Portal failures in this path logged raw errors. An Xtream error message
carries the stream URL, and that URL is built out of the username and
password, so they now go through the redacting logger.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Second round of Greptile review findings.
**A pin had no behavioural effect.** Loading a stored pin only decorated the
row: Play still started the route's playlist and failover ranking ignored
`isPinned`, so "make this the main source" survived a restart as an icon and
nothing else. The primary action now starts from the pinned source when one is
set, and the pin outranks everything else in failover ranking.
**Failover stopped at the first unresolvable candidate.** An expired account or
a failing `get_vod_info` on the top-ranked source ended the attempt, and since
production calls `failover()` only once — on the original playback failure — a
healthy lower-ranked source was never reached. It now continues through untried
candidates. `switchTo` reports why it stopped so the loop can tell "could not
resolve, try the next one" from "something newer owns the screen"; without that
distinction a superseded switch would have spun forever, because only the
former marks the candidate tried.
**Identity ignored enrichment.** The key was `playlistId:contentId:title`, so
when `get_vod_info` added a TMDB id and release year to an unchanged title the
host saw no change, never reloaded, and kept yearless discovery and title-only
pin keys — a `tmdb:`-keyed pin could never be found. The key now covers every
field that affects matching.
**A server refusing HEAD read as unavailable.** Some stream hosts answer 405 or
501 to HEAD yet serve the media over GET. The probe now retries once with the
ranged GET the main process already supported, instead of caching a working
source as failed and penalising it during failover.
Greptile also flagged a missing token check after the resolve await in
`switchTo`; that guard landed in 4db3a2fd and sits on the line directly below.
Answered on the thread rather than changed.
The host service passed 400 lines again, so the pin, probe, switch-notice and
current-row concerns moved into focused modules beside it.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
A movie that exists in several imported Xtream playlists now shows a
"Sources N" chip on its detail page and in the player. Switching playlist
mid-film keeps the timecode, a preferred source can be pinned per movie, and
a failed stream offers the alternatives instead of a dead end.
The governing rule is that a guess is never presented as a fact. Every
metadata value carries where it came from — `api` (the provider said so),
`parsed` (inferred from the title) or `probe` (we contacted the stream).
Facts render as plain tags, guesses are prefixed `~` in a warning colour, and
an unknown value renders no tag at all plus a "check" affordance. Ranking and
failover read through `factualOnly()`, so a filename claiming 4K is
structurally unable to outrank a source that was actually reached. A probe
that could not complete reports "unknown", never "unavailable".
Scope is deliberately narrow: Xtream to Xtream, movies only, Electron only.
Stalker never reaches the `content` table and M3U is a JSON blob whose search
forces live content; both are additive later, since the candidate type
already carries all three portal kinds. In the PWA every entry point is gated
off and the chip renders nothing.
Auto-failover is opt-in and off by default. Each source is tried at most once
per session, so it terminates structurally, and the switch is never silent —
the toast names the new playlist, offers an undo, and warns that the dub may
differ only when both sides state an audio track as fact.
Notable details:
- Playlist names are routinely the pasted URL, credentials included. They are
never rendered raw; a short host-only label is derived instead.
- Quality is derived from pixel width, not height: a 2.39:1 1080p master is
1920x800, and bucketing that by height would publish "720p" as a fact.
- Switching is a single `inlinePlayback.set()` so the player and engine
survive and re-seek; the carried position is read before the 15s
persistence throttle so it does not rewind.
- Sources from one playlist collapse into a group, since the same film often
appears there several times under different stream ids.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix(playlists): serialize per-playlist collection writes to prevent lost updates
All per-playlist mutations (portal favorites, recently viewed, playlist
meta/favorites updates) used an uncoordinated read -> patch -> replace-whole-row
pattern, so two overlapping mutations on the same playlist were last-write-wins
and silently dropped each other's changes (flagged by Greptile on PR #1253).
Chain every read-modify-write through a per-playlist promise queue
(Map<playlistId, Promise>) inside defer(), covering both the SQLite upsert and
IndexedDB update paths while keeping the Observable-based public API, laziness,
and emitted values unchanged. A failed mutation does not wedge the queue, and
different playlists are not serialized against each other.
Regression coverage: overlapping favorite+recently-viewed adds, two rapid
favorite adds, IndexedDB-path overlap (all three fail on the old code), plus
queue-continues-after-error and cross-playlist independence guards.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix(playlists): close serialization gaps flagged by review bots
Greptile P1 (writers bypassing the queue): route updateManyPlaylists and
updatePlaylistPositions through the same per-playlist write queue. Auto-refresh
batch writes now re-read the stored row inside the queue and preserve
user-owned fields (favorites, recently viewed, position) instead of writing a
pre-refresh snapshot over them; position updates re-read and patch inside the
queue on both storage paths.
Codex P1 (callers precompute stale snapshots): add an atomic
PlaylistsService.transformPlaylistFavorites(playlistId, transform) that applies
the favorites transform to the freshly-read row inside the queue, and convert
every read-then-set call site to it: UnifiedFavoritesDataService M3U
add/remove/clear/reorder and Stalker reorder/clear, GlobalFavoritesService M3U
removal, DashboardDataService M3U removal. Reorders now keep concurrently
added favorites (appended after the dragged order) instead of dropping them.
New coverage: overlapping favorites transforms, auto-refresh batch write vs
queued favorite add, position update vs queued favorite add, and a public
addFavorite race through UnifiedFavoritesDataService; existing specs updated
to the transform-based contract.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix(playlists): share the canonical refresh merge with the auto-refresh batch
The batch path previously spread the stale refresh snapshot over the freshly
read row and pinned only favorites/recently-viewed/position, so a queued
metadata mutation (hiddenGroupTitles, curated EPG sources) finishing before
the refresh write could be reverted. Extract updatePlaylist's merge into
mergeRefreshedPlaylist() and use it for both the single-playlist update flow
and updateManyPlaylists: refresh-owned data (parsed content, count, EPG
detection) comes from the payload, user-owned state comes from the current
row, and manual/disabled EPG configuration is resolved through
resolvePlaylistEpgSourceState instead of being overwritten.
Regression test: queued hiddenGroupTitles meta update overlapping an
auto-refresh batch write keeps both the metadata change and the refreshed
content, including preserved manualEpgUrls.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix(playlists): let the current row decide autoRefresh during batch refresh
The batch path force-set autoRefresh: true after the merge, so disabling
auto-refresh while a refresh was in flight was reverted by the completing
write. Drop the override — mergeRefreshedPlaylist already prefers the current
row's autoRefresh over the snapshot — and add a count fallback to the snapshot
value for rows without a stored copy.
Regression test: disabling auto-refresh concurrently with the batch write
keeps autoRefresh false while still applying the refreshed content.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
* feat(epg): manual EPG mapping for Stalker portals
Extends the manual EPG-to-channel mapping (PR #1165) to Stalker:
- shared key helper buildStalkerEpgMappingKey — playlist-scoped
stalker:{playlistId}:{channelId} keys, mirroring the Xtream scheme
- ITV sidebar: right-click context menu with "Map EPG channel"; after
the dialog closes with a change, the bulk EPG cache is rebuilt so the
panel and row previews reflect the new mapping immediately
- withStalkerEpg().applyMappedItvEpg(): batch-resolves mappings for
rendered channels (one getEpgMappingsBatch IPC per new id set) and
overlays uploaded-XMLTV programs onto bulkItvEpgByChannel; overrides
survive ensureBulkItvEpg reloads
- stream-resolver: mapping check in loadStalkerEpgItems (detail) and
batched prefetch in loadStalkerEpgBatch (previews);
mappingCandidateKeys/prefetchEpgMappings generalized beyond Xtream
- global favorites: stalker branch for the Map-EPG menu entry (item id
extracted from the stalker::{playlistId}::{id} uid)
- docs: manual-mapping section in docs/architecture/stalker-epg.md and
a CLAUDE.md EPG bullet covering the whole mapping feature
- tests: applyMappedItvEpg suite, stalker preview mapping in the
stream-resolver spec, key-builder specs
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix(epg): harden Stalker mapping edges found in adversarial review
- stream-resolver: stalker items resolve only the playlist-scoped
mapping key — bare tvgId/name candidates (tvgId mirrors the raw
provider id) could only produce false matches against unrelated M3U
mappings, and previews would disagree with the detail path
- applyMappedItvEpg: staleness guard after every await so an in-flight
call cannot write portal A's mapped EPG into portal B's state after
a playlist switch (the store is a root singleton)
- applyMappedItvEpg: ids are marked checked only after a successful
lookup — a transient IPC failure no longer suppresses the mapping
for the rest of the session
- live layout: post-dialog refresh re-applies overrides for the
unfiltered channel list, so an active search cannot drop the playing
channel's mapping
- global favorites: new epgMappingChanged output emitted when the
dialog actually changed a mapping; unified live tab reloads its EPG
previews in response
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
* feat(epg): manual EPG-to-channel mapping with mapping fallback in all EPG paths
* fix: epg mapping in live tv list
* fix(epg): harden manual EPG mapping — upgrade safety, perf, playlist-scoped keys
Follow-up fixes on top of the manual EPG-to-channel mapping feature:
- epg-database: dedupe existing epg_programs rows before creating the
unique (channel_id, start, title) index — a plain CREATE UNIQUE INDEX
crashed the EPG worker on upgrade when historical duplicates exist;
replace INSERT OR REPLACE with ON CONFLICT DO UPDATE so the
epg_programs_fts delete trigger is not bypassed (REPLACE skips delete
triggers unless recursive_triggers is on), with a plain-INSERT
fallback when the index cannot be created
- db: add idx_content_epg_channel — the mapping fallback scanned the
whole content table on every single-channel EPG lookup
- keys: scope Xtream mapping keys per playlist via shared
buildXtreamEpgMappingKey (xtream:{playlistId}:{id}) — bare stream ids
collide across portals; the backend fallback now joins categories to
resolve the playlist id
- pwa: hide "Map EPG channel" entries behind the supportsEpgMapping
capability — the menu item was a dead end in the PWA
- parser: parse the XMLTV offset sign from the string — Math.sign(0)
dropped the minutes of ±00:xx offsets
- cleanup: typed window.electron access instead of ad-hoc casts, drop
unused resolveChannelId and dialog data field, shared
EpgMappingDialogComponent.open() for all seven call sites
- dialog UX: minimum-characters search hint, save/remove snackbars,
current mapping shows the EPG channel display name,
takeUntilDestroyed on the search stream
- i18n: fill the new keys in all 17 locales
- tests: cover mapping CRUD/search escaping, the dedup-index guard,
offset parsing and playlist-scoped keys; update stale stream-resolver
specs for the new 50-item limit and 10s timeout
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix(epg): escape backslashes in EPG channel search LIKE pattern
CodeQL js/incomplete-sanitization: a lone trailing backslash in the
search term paired with the closing wildcard under the ESCAPE clause
and corrupted the pattern.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* perf(epg): batch mapping lookups in the viewport preview queue
Expose the existing getEpgMappingsBatch operation over a new
EPG_MAPPING_GET_BATCH IPC channel and use it in resolveManualMappings —
the per-entry lookup issued one IPC round-trip per visible channel on
every scroll event.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* perf(epg): batch mapping prefetch in the collection preview loader
Resolve all candidate mapping keys for an Xtream preview batch with a
single getEpgMappingsBatch IPC call instead of per-channel lookups.
Also fix a worker early-exit: a channel without tvgId/name returned out
of the shared-iterator loop and silently killed one of the three
concurrent preview workers.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---------
Co-authored-by: 4gray <serega05@gmail.com>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
Enables Xtream catch-up/timeshift from the Favorites and Recent surfaces (per-playlist and global), not just Live TV, and adds start-over replay of the currently-airing programme. Carries tv_archive/tv_archive_duration through the favorites and recently-viewed DB projections and maps them onto UnifiedCollectionItem; tv_archive_duration is interpreted as days, matching live-stream-layout.controlledArchiveDays.
Closes#1138.
Co-authored-by: Claude <noreply@anthropic.com>
* fix(favorites): persist custom drag-and-drop order for Xtream favorites
Prepared-statement writes dispatched via drizzle's `.execute()` on the
better-sqlite3 driver return a promise and defer the write to a microtask.
Inside a synchronous `db.transaction(() => ...)` callback (which cannot
await), the transaction commits before that promise settles, so the write
is a silent no-op — no error, no rows changed.
This bit `reorderGlobalFavorites`: the custom favorites order never
persisted for the per-playlist ("This playlist") Xtream scope, which relies
solely on the `favorites.position` column. The global ("All playlists")
scope masked the bug because it also persists an order to the `appState`
`global-favorites-channel-order-v1` key and re-applies it on read.
`removeRecentItemsBatch` had the same latent bug — batch "clear recent
items" silently did nothing.
Switch both writers to synchronous `.run()`. Add regression coverage that
asserts `.run()` (not `.execute()`) is used and would fail on the old
behavior, and document the gotcha in the DB worker architecture doc.
Verified over CDP against a live Electron instance: reorder writes
positions 0..N, and the order survives navigation and a full reload.
Fixes#1137
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
* fix(favorites): scope reorder position writes by playlist
The global favorites reorder wrote the new position filtering only by
content_id, so two Xtream playlists holding a favorite with the same
content_id would clobber each other's persisted order (greptile P1).
Thread playlist_id through the whole reorder path — the renderer builder
(UnifiedCollectionItem already carries playlistId), the IPC contract
(ElectronBridgeFavoriteReorderUpdate + inline payload types), the worker
op — and scope the prepared UPDATE by (contentId, playlistId), matching
the favorites composite unique index.
Tests: favorites.operations.spec asserts the playlistId placeholder and
per-row playlistId payload; preload contract fixture updated.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
* fix(favorites): include playlist_id in workspace global favorites reorder payload
The workspace global-favorites reorder path still sent updates with only
content_id and position. Since the backend UPDATE is now scoped by
(contentId, playlistId), that payload binds an undefined playlist id and
matches no rows — the DB write silently no-ops (flagged by Greptile P1).
Also scope the prepared-statement example in the sqlite-db-worker gotcha
doc by (contentId, playlistId) so it no longer documents the
cross-playlist rewrite this PR fixes (flagged by Codex P3).
Regression spec asserts the reorder payload carries playlist_id per item
(fails on the old payload shape) and that the appState uid order is
still persisted for non-Xtream items.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
- merge origin/master into PR #964 and keep embedded MPV test on the isolated playback sub-entrypoint
- centralize EPG capability through DataService.supportsEpg and update PWA web-e2e expectations
- split BrowserAccessError copy between Electron and PWA diagnostics
