Files
plezy/lib/services/data_aggregation_service.dart
edde746 5f397a99d9 fix(discover): let a refreshed row override a stale local watch patch
Pausing an episode on one device, finishing it on another and pressing
Refresh left the first device showing the old "minutes left". Restarting the
app showed the right value. Two independent defects produce that, and either
alone reproduces the report.

The first is the watch-state overlay. Every local watch event lands in
WatchStateStore as a patch, and WatchStateSnapshot.apply overwrites
viewOffsetMs unconditionally; isNewerThan only ever orders one patch against
another, never against the server row underneath. Nothing expires a patch and
nothing clears the map except a profile switch, so the Mac's own paused
position kept winning over every subsequent fetch until the process died.

A patch exists to bridge the gap between a local action and the next server
read of that item, so it should stop applying once that read happens. The
store now records the watermark at which a successful authoritative response
returned each key, and suppresses an acknowledged session patch at or below
it. Only a watermark is stored, never the observed state: WatchStateSnapshot
cannot hold a container's leaf counts, and keeping max() per key makes the
order two concurrent responses complete irrelevant. Suppression is a
read-time predicate, so nothing mutates during build.

The barrier covers the parentChain too. patchForItem picks the newest of the
item's own entry and its ancestors', so retiring only the item's entry would
let an older season mark win and render watched/0 -- worse than either the
stale value or the fresh one. An authoritative read of a child already
reflects any container mark that preceded it, so the child's observation
judges its ancestors as well; a newer container action still wins.

Provenance decides what may be suppressed at all. WatchStateEvent now carries
serverAcknowledged, defaulting to false so an unclassified emit site degrades
to today's behaviour rather than silently becoming retireable. An offline
write is owed to the server and a read must never retire it, so it stays
until a WatchPatchPromotionNotifier promotion says the queue replayed it. That
channel is deliberately not a WatchStateEvent: OfflineWatchSyncService reacts
to watched/unwatched by purging queued progress, so replaying one there would
delete a newer rewatch. Promotion matches an exact WatchPatchId -- session
minted for live crossings, derived from the persisted (profile, row, revision)
for queued ones so it still joins after a restart.

Report acceptance is not delivery: PlaybackReportSession resolves true for a
same-state startup heartbeat it drops, so acknowledgement now keys on
onDelivered. A MediaBrowser Started saves play count and last-played date but
not the position, so it cannot acknowledge an offset. No report-derived
watched crossing is acknowledged on any backend -- Jellyfin hard-codes its
threshold and Plex never loads the server pref that would tell it the real
one -- so only an awaited explicit markWatched settles one.

The second defect is that a failed Refresh reported success. Plex _fetchHubs
and the Jellyfin hub legs both degrade a failure to an empty list, and the
library prefetch discarded its failures, so a server whose every hub request
failed was recorded as succeeded; DiscoverProvider then kept the previous rows,
set loaded and surfaced nothing. Worse, the background Continue Watching
refresh wiped the row outright on zero success.

Hub legs now report what they degraded through a HubFetchDiagnostics sink,
which keeps partial rows alongside the failure and leaves every existing
caller untouched. Failures ride through the aggregation results, a leg that
could not run because discovery failed contributes that failure rather than a
successful no-op, and loaded-server ids became succeeded - failed - cancelled
so one bad leg no longer caches a server as covered and blocks its retry. The
toolbar awaits a DiscoverRefreshOutcome and shows the existing unableToLoad
snackbar on failure while the retained rows stay on screen. Rollback after a
mid-pass exception is version-guarded, refilters against the current hidden
libraries and no longer publishes a system shelf the pass never committed.

Observations are staged with the pass and flushed only once the same disposal,
generation and exception checks that authorise committing those rows have
passed, so a discarded or rolled-back response can never suppress a patch.

Also fixes a live data-loss race the promotion work would have built on:
upsertProgressAction stamped a millisecond timestamp and updated the row in
place, so a rewatch queued during an in-flight replay was deleted by id.
Revisions are now strictly monotonic per row, replay deletes and retry updates
compare against them, and the upsert resets the retry fields because a new
revision is a new logical action.

close #1829
2026-08-08 09:09:48 +02:00

830 lines
34 KiB
Dart

import 'dart:async';
import '../media/ids.dart';
import '../media/media_hub.dart';
import '../media/media_item.dart';
import '../media/media_item_merge.dart';
import '../media/media_kind.dart';
import '../media/media_library.dart';
import '../media/media_server_client.dart';
import '../exceptions/media_server_exceptions.dart';
import '../utils/app_logger.dart';
import '../utils/external_ids.dart';
import '../utils/global_key_utils.dart';
import '../utils/search_relevance.dart';
import '../utils/media_server_http_client.dart';
import 'local_playback_history.dart';
import 'multi_server_manager.dart';
/// A row a successful response actually returned, paired with the immutable
/// cache scope of the client that fetched it.
///
/// The scope cannot be recovered later: flattening and dedup lose which
/// client produced a row, and a user-scoped backend's public server id is
/// shared between users — so reconciling by public key could suppress a
/// different user's watch state.
typedef ObservedRow = ({MediaItem item, String? clientScope});
/// [observedItems] are the rows the successful responses actually returned,
/// *before* dedup or limiting. Continue Watching dedup can drop one server's
/// copy in favour of another's, and a dropped copy would otherwise leave that
/// key unreconciled even though its server did return it (#1829).
typedef OnDeckAggregationResult = ({
List<MediaItem> items,
List<ObservedRow> observedItems,
Set<String> succeededServerIds,
Set<String> cancelledServerIds,
Set<String> failedServerIds,
});
typedef HubAggregationResult = ({
List<MediaHub> hubs,
List<ObservedRow> observedItems,
Set<String> succeededServerIds,
Set<String> cancelledServerIds,
Set<String> failedServerIds,
});
typedef LibraryAggregationResult = ({
List<MediaLibrary> libraries,
Set<String> succeededServerIds,
Set<String> cancelledServerIds,
Set<String> failedServerIds,
});
typedef SearchAggregationResult = ({
List<MediaItem> items,
Set<String> succeededServerIds,
Set<String> cancelledServerIds,
Set<String> failedServerIds,
});
typedef _FanOutResult<T> = ({
List<T> items,
Set<String> succeededServerIds,
Set<String> cancelledServerIds,
Set<String> failedServerIds,
});
/// Whether [error] is a client-side abort (client teardown mid-request)
/// rather than a genuine server failure. Aggregation reports these servers
/// in `cancelledServerIds` so callers can tell a *disrupted* pass — whose
/// results say nothing about actual content — from a settled failure.
bool _isCancellation(Object error) => error is MediaServerHttpException && error.isCancellation;
Map<String, int> _searchKindCounts(Iterable<MediaItem> items) {
final counts = <String, int>{};
for (final item in items) {
counts.update(item.kind.name, (count) => count + 1, ifAbsent: () => 1);
}
return counts;
}
/// Drop items belonging to a hidden library.
///
/// Items the backend could not attribute to a library ([MediaItem.libraryGlobalKey]
/// is null) are kept: Plex search and `/library/shared/all` return shared and
/// external rows that have no local section, and those are not something the
/// user hid.
List<MediaItem> _withoutHiddenLibraries(List<MediaItem> items, Set<String>? hiddenLibraryKeys) {
if (hiddenLibraryKeys == null || hiddenLibraryKeys.isEmpty) return items;
return items.where((item) {
final libraryKey = item.libraryGlobalKey;
return libraryKey == null || !hiddenLibraryKeys.contains(libraryKey);
}).toList();
}
/// The server-local library ids [serverId] owns within [hiddenLibraryKeys],
/// which hold cross-server `serverId:libraryId` global keys. Backends that
/// cannot attribute a search hit to a library need these to scope the request.
Set<String> _hiddenLibraryIdsOn(String serverId, Set<String>? hiddenLibraryKeys) {
if (hiddenLibraryKeys == null || hiddenLibraryKeys.isEmpty) return const {};
final ids = <String>{};
for (final key in hiddenLibraryKeys) {
final parsed = parseGlobalKey(key);
if (parsed != null && parsed.serverId == serverId) ids.add(parsed.ratingKey);
}
return ids;
}
/// Cross-server aggregation: fans calls out to every online client and
/// merges the results. Single-server operations now go through the
/// [MediaServerClient] interface directly (resolved via
/// [ProviderExtensions.tryGetMediaClientForServer] etc.), so this service
/// only owns the genuinely multi-server flows: home/discover hubs, on-deck,
/// search, and the global library list.
class DataAggregationService {
final MultiServerManager _serverManager;
DataAggregationService(this._serverManager);
/// Online clients, optionally restricted to [serverIds] — delta refreshes
/// fan out to newly-online servers only.
Map<String, MediaServerClient> _clientsFor(Set<String>? serverIds) {
final clients = _serverManager.onlineClients;
if (serverIds == null) return clients;
return {
for (final entry in clients.entries)
if (serverIds.contains(entry.key)) entry.key: entry.value,
};
}
/// Run [fetch] against every client in [clients] and concatenate the results
/// in client order. A per-server failure is swallowed — logged with
/// [failureMessage] and contributing nothing — so one bad server cannot sink
/// the pass; that server is simply absent from `succeededServerIds` and lands
/// in `failedServerIds`. A client-side abort is *not* a failure: it lands in
/// `cancelledServerIds` and is logged at debug level, so callers can tell a
/// disrupted pass from a settled one and torn-down requests do not spam the
/// error log.
Future<_FanOutResult<T>> _fanOut<T>(
Map<String, MediaServerClient> clients, {
required String Function(String serverId) failureMessage,
required Future<List<T>> Function(String serverId, MediaServerClient client) fetch,
}) async {
final cancelledServerIds = <String>{};
final failedServerIds = <String>{};
final futures = clients.entries.map((entry) async {
try {
return (serverId: entry.key, items: await fetch(entry.key, entry.value));
} catch (e, stackTrace) {
if (_isCancellation(e)) {
cancelledServerIds.add(entry.key);
appLogger.d('Cancelled (client abort): ${failureMessage(entry.key)}');
} else {
failedServerIds.add(entry.key);
appLogger.e(failureMessage(entry.key), error: e, stackTrace: stackTrace);
}
return (serverId: null, items: <T>[]);
}
});
final results = await Future.wait(futures);
return (
items: [for (final result in results) ...result.items],
succeededServerIds: {
for (final result in results)
if (result.serverId != null) result.serverId!,
},
cancelledServerIds: cancelledServerIds,
failedServerIds: failedServerIds,
);
}
/// Fetch libraries from all online clients regardless of backend, returning
/// the merged neutral [MediaLibrary]s alongside the ids of the servers whose
/// fetch actually succeeded. [serverIds] restricts the fan-out to those
/// servers.
///
/// A per-server `fetchLibraries()` failure is swallowed (that server simply
/// contributes no libraries) so one unreachable server doesn't sink the whole
/// list. [succeededServerIds] lets callers tell a *failed* fetch apart from a
/// server that genuinely has no libraries — both contribute nothing, so
/// conflating them would let a transient failure be cached as "loaded" and
/// never retried. Servers whose fetch was aborted client-side land in
/// `cancelledServerIds` — a disrupted pass, unlike a settled failure, must
/// never be committed as authoritative.
Future<LibraryAggregationResult> getMediaLibrariesFromAllServers({Set<String>? serverIds}) async {
final clients = _clientsFor(serverIds);
if (clients.isEmpty) {
appLogger.w('No online servers available for fetching libraries (neutral)');
return (
libraries: const <MediaLibrary>[],
succeededServerIds: const <String>{},
cancelledServerIds: const <String>{},
failedServerIds: const <String>{},
);
}
final fetched = await _fanOut<MediaLibrary>(
clients,
failureMessage: (serverId) => 'Failed neutral library fetch from $serverId',
fetch: (_, client) => client.fetchLibraries(),
);
return (
libraries: fetched.items,
succeededServerIds: fetched.succeededServerIds,
cancelledServerIds: fetched.cancelledServerIds,
failedServerIds: fetched.failedServerIds,
);
}
/// Fetch "On Deck" (Continue Watching) from all servers and merge by recency.
/// Items are tagged with server info by the underlying client. Returns
/// neutral [MediaItem]s plus the ids of servers whose fetch succeeded.
/// [serverIds] restricts the fan-out to those servers.
Future<OnDeckAggregationResult> getOnDeckFromAllServers({
int? limit,
Set<String>? hiddenLibraryKeys,
Set<String>? serverIds,
}) async {
final clients = _clientsFor(serverIds);
if (clients.isEmpty) {
return (
items: const <MediaItem>[],
observedItems: const <ObservedRow>[],
succeededServerIds: const <String>{},
cancelledServerIds: const <String>{},
failedServerIds: const <String>{},
);
}
final observedRows = <ObservedRow>[];
final fetched = await _fanOut<MediaItem>(
clients,
failureMessage: (serverId) => 'Failed on-deck fetch from $serverId',
fetch: (_, client) async {
final rows = await client.fetchContinueWatching(count: limit);
// Capture the scope here: after the fan-out flattens and dedup runs,
// there is no way back to the client that produced a row.
final scope = client.cacheServerId;
observedRows.addAll([for (final row in rows) (item: row, clientScope: scope)]);
return rows;
},
);
// Filter out items from hidden libraries
var filteredOnDeck = _withoutHiddenLibraries(fetched.items, hiddenLibraryKeys);
// Sort by most recently viewed, falling back to addedAt for unwatched items.
// Same key as JellyfinClient's continue-watching merge (MediaItem.recencySortKey)
// so per-server and cross-server ordering can't drift apart.
filteredOnDeck.sort((a, b) => b.recencySortKey.compareTo(a.recencySortKey));
filteredOnDeck = await _deduplicateContinueWatching(filteredOnDeck);
// Apply limit if specified
final items = limit != null && limit < filteredOnDeck.length ? filteredOnDeck.sublist(0, limit) : filteredOnDeck;
appLogger.i('Fetched ${items.length} on deck items from all servers');
return (
items: items,
// Pre-dedup, pre-limit: a copy dropped by dedup or trimmed by the limit
// was still authoritatively returned by its server.
observedItems: observedRows,
succeededServerIds: fetched.succeededServerIds,
cancelledServerIds: fetched.cancelledServerIds,
failedServerIds: fetched.failedServerIds,
);
}
/// Merge an [existing] Continue Watching list with [fresh] rows from
/// newly-online servers: same recency ordering and cross-server identity
/// dedup as [getOnDeckFromAllServers], applied to the union.
Future<List<MediaItem>> mergeContinueWatching(List<MediaItem> existing, List<MediaItem> fresh, {int? limit}) async {
final combined = [...existing, ...fresh]..sort((a, b) => b.recencySortKey.compareTo(a.recencySortKey));
final deduped = await _deduplicateContinueWatching(combined);
return limit != null && limit < deduped.length ? deduped.sublist(0, limit) : deduped;
}
Future<List<MediaItem>> _deduplicateContinueWatching(List<MediaItem> items) async {
if (items.length < 2) return items;
final bucketCounts = <String, int>{};
for (final item in items) {
final bucket = _continueWatchingTitleBucket(item);
if (bucket == null) continue;
bucketCounts[bucket] = (bucketCounts[bucket] ?? 0) + 1;
}
final duplicateBuckets = {
for (final entry in bucketCounts.entries)
if (entry.value > 1) entry.key,
};
if (duplicateBuckets.isEmpty) return items;
final externalIdLoads = <String, Future<ExternalIds>>{};
final identityKeysByIndex = <int, Set<String>>{};
final identityKeyLoads = <Future<void>>[];
for (var i = 0; i < items.length; i++) {
if (!duplicateBuckets.contains(_continueWatchingTitleBucket(items[i]))) continue;
final index = i;
identityKeyLoads.add(
_continueWatchingIdentityKeys(items[index], externalIdLoads).then((keys) => identityKeysByIndex[index] = keys),
);
}
await Future.wait(identityKeyLoads);
// Group duplicates instead of greedily dropping them: the first item to
// claim an identity key anchors the group and holds its shelf slot;
// later items sharing a claimed key join as members without claiming
// their own keys (same transitive semantics as the old drop). Each slot
// then shows the member the user most recently played on this device —
// servers sync watch state across guid-linked siblings, so their
// lastViewedAt ties and can't tell the 4K copy from the 1080p one
// (#1492). Without local history the anchor (recency order) stands.
final keyToGroup = <String, int>{};
final groups = <List<MediaItem>>[];
final groupSlots = <int, int>{};
final result = <MediaItem>[];
for (var i = 0; i < items.length; i++) {
final item = items[i];
if (!duplicateBuckets.contains(_continueWatchingTitleBucket(item))) {
result.add(item);
continue;
}
final identityKeys = identityKeysByIndex[i] ?? const <String>{};
if (identityKeys.isEmpty) {
result.add(item);
continue;
}
var joined = false;
for (final key in identityKeys) {
final groupIndex = keyToGroup[key];
if (groupIndex != null) {
groups[groupIndex].add(item);
joined = true;
break;
}
}
if (joined) continue;
final groupIndex = groups.length;
groups.add([item]);
for (final key in identityKeys) {
keyToGroup[key] = groupIndex;
}
groupSlots[result.length] = groupIndex;
result.add(item);
}
if (groupSlots.isEmpty) return result;
final lastPlayed = await LocalPlaybackHistory.snapshot();
for (final slot in groupSlots.entries) {
final members = groups[slot.value];
if (members.length > 1) {
result[slot.key] = _preferLocallyLastPlayed(members, lastPlayed);
}
}
return result;
}
/// The duplicate-group member most recently played on this device (by item
/// or series key), or the anchor — `members.first`, the group's most recent
/// item by [MediaItem.recencySortKey] — when the local history has nothing
/// newer to say.
MediaItem _preferLocallyLastPlayed(List<MediaItem> members, Map<String, int> lastPlayed) {
var winner = members.first;
var winnerLastPlayedAt = 0;
for (final member in members) {
final itemTs = lastPlayed[member.globalKey] ?? 0;
final seriesKey = member.seriesGlobalKey;
final seriesTs = seriesKey != null ? (lastPlayed[seriesKey] ?? 0) : 0;
final lastPlayedAt = itemTs > seriesTs ? itemTs : seriesTs;
if (lastPlayedAt > winnerLastPlayedAt) {
winner = member;
winnerLastPlayedAt = lastPlayedAt;
}
}
return winner;
}
String? _continueWatchingTitleBucket(MediaItem item) {
final scope = _continueWatchingIdentityScope(item);
if (scope == null) return null;
final title = switch (item.kind) {
MediaKind.episode || MediaKind.season => item.grandparentTitle ?? item.parentTitle ?? item.title,
_ => item.title,
};
final normalized = title?.trim().toLowerCase().replaceAll(RegExp(r'\s+'), ' ');
if (normalized == null || normalized.isEmpty) return null;
return '$scope:$normalized';
}
Future<Set<String>> _continueWatchingIdentityKeys(
MediaItem item,
Map<String, Future<ExternalIds>> externalIdLoads,
) async {
final scope = _continueWatchingIdentityScope(item);
if (scope == null) return const {};
final keys = <String>{};
final serverId = item.serverId;
final targetId = _continueWatchingIdentityTargetId(item);
final client = serverId == null ? null : _serverManager.getClient(ServerId(serverId));
if (client != null && targetId != null && targetId.isNotEmpty) {
try {
final cacheKey = buildGlobalKey(ServerId(serverId!), targetId);
final externalIds = await externalIdLoads.putIfAbsent(cacheKey, () => client.fetchExternalIds(targetId));
_addExternalIdentityKeys(keys, scope, externalIds);
} catch (e, stackTrace) {
appLogger.d(
'Failed to resolve Continue Watching identity for ${item.globalKey}',
error: e,
stackTrace: stackTrace,
);
}
}
final stableGuid = _stableMediaGuid(item.guid);
if (stableGuid != null) {
final guidScope = item.kind == MediaKind.episode ? 'episode' : scope;
keys.add('$guidScope:guid:$stableGuid');
}
return keys;
}
String? _continueWatchingIdentityScope(MediaItem item) {
return switch (item.kind) {
MediaKind.episode || MediaKind.season || MediaKind.show => 'show',
MediaKind.movie => 'movie',
_ => null,
};
}
String? _continueWatchingIdentityTargetId(MediaItem item) {
return switch (item.kind) {
MediaKind.episode => item.grandparentId,
MediaKind.season => item.grandparentId ?? item.parentId,
MediaKind.show || MediaKind.movie => item.id,
_ => null,
};
}
void _addExternalIdentityKeys(Set<String> keys, String scope, ExternalIds externalIds) {
final imdb = externalIds.imdb?.trim().toLowerCase();
if (imdb != null && imdb.isNotEmpty) keys.add('$scope:imdb:$imdb');
final tmdb = externalIds.tmdb;
if (tmdb != null) keys.add('$scope:tmdb:$tmdb');
final tvdb = externalIds.tvdb;
if (tvdb != null) keys.add('$scope:tvdb:$tvdb');
final anidb = externalIds.anidb;
if (anidb != null) keys.add('$scope:anidb:$anidb');
}
String? _stableMediaGuid(String? guid) {
final value = guid?.trim();
if (value == null || value.isEmpty) return null;
if (!value.contains('://')) return null;
if (value.contains('agents.none://')) return null;
return value.toLowerCase();
}
/// Fetch recommendation hubs from all servers as neutral [MediaHub]s.
/// When useGlobalHubs is true (default), rich-hub backends use their true
/// home page hubs (Plex's promoted/global hub endpoint).
/// Backends without rich home hubs fall back to per-library hubs so one
/// capped "Latest" response cannot hide whole library types.
/// [serverIds] restricts the fan-out (including the library prefetch) to
/// those servers. Returns the ids of servers whose hub fetch succeeded so
/// callers do not cache transient per-server failures as loaded.
Future<HubAggregationResult> getHubsFromAllServers({
int? limit,
Set<String>? hiddenLibraryKeys,
bool useGlobalHubs = true,
bool includePlaybackHubs = true,
Set<String>? serverIds,
}) async {
final clients = _clientsFor(serverIds);
if (clients.isEmpty) {
appLogger.w('No online servers available for fetching hubs');
return (
hubs: const <MediaHub>[],
observedItems: const <ObservedRow>[],
succeededServerIds: const <String>{},
cancelledServerIds: const <String>{},
failedServerIds: const <String>{},
);
}
// Home layout needs the library list for every client: fallback backends
// build all their rows from per-library hubs, and rich-hub backends
// (Plex) need it to detect visible music libraries, whose hubs the
// global-hub endpoint excludes. One `fetchLibraries` per server, served
// from the per-backend API cache when warm.
final libraryFetch = useGlobalHubs ? await getMediaLibrariesFromAllServers(serverIds: serverIds) : null;
final libraries = libraryFetch == null ? null : _groupLibrariesByServer(libraryFetch.libraries);
final legSucceededServerIds = <String>{};
final legFailedServerIds = <String>{if (libraryFetch != null) ...libraryFetch.failedServerIds};
final legCancelledServerIds = <String>{if (libraryFetch != null) ...libraryFetch.cancelledServerIds};
final globalDiagnosticsByServer = <String, HubFetchDiagnostics>{};
final observedRows = <ObservedRow>[];
final fetched = await _fanOut<MediaHub>(
clients,
failureMessage: (serverId) => 'Failed to fetch hubs from server $serverId',
fetch: (serverId, client) async {
final serverLibraries = libraries?[serverId];
final shouldUseGlobalHubs = useGlobalHubs && client.capabilities.richHubs;
final prefetchDegraded = legFailedServerIds.contains(serverId) || legCancelledServerIds.contains(serverId);
final hubItemLimit = limit ?? defaultHubPreviewLimit;
List<MediaHub> hubs;
if (shouldUseGlobalHubs) {
// Both legs are independent, so start them before awaiting either.
// Spreading `...await a, ...await b` into one list literal evaluates
// them in order, which serialised the music rows behind the global
// hub round trip.
final globalDiagnostics = HubFetchDiagnostics();
globalDiagnosticsByServer[serverId] = globalDiagnostics;
final globalFuture = client.fetchGlobalHubs(
limit: hubItemLimit,
includePlaybackHubs: includePlaybackHubs,
diagnostics: globalDiagnostics,
);
// Plex's promoted/global hub endpoint never includes music
// libraries — append their per-library hubs so music rows
// reach home. A failed prefetch cannot establish that there
// are no visible music libraries, so it is not a successful no-op.
final musicFuture = prefetchDegraded
? null
: _fetchLibraryHubsForClient(
client,
limit: hubItemLimit,
hiddenLibraryKeys: hiddenLibraryKeys,
includePlaybackHubs: includePlaybackHubs,
libraries: serverLibraries ?? const [],
kinds: const {MediaKind.artist},
);
Object? globalError;
StackTrace? globalStackTrace;
List<MediaHub> globalHubs = const [];
try {
globalHubs = await globalFuture;
} catch (error, stackTrace) {
globalDiagnostics.recordFailure(error);
globalError = error;
globalStackTrace = stackTrace;
}
final music = musicFuture == null ? null : await musicFuture;
if (globalHubs.isNotEmpty || (!globalDiagnostics.failed && !globalDiagnostics.cancelled)) {
legSucceededServerIds.add(serverId);
}
if (music != null) {
if (music.succeeded) legSucceededServerIds.add(serverId);
if (music.failed) legFailedServerIds.add(serverId);
if (music.cancelled) legCancelledServerIds.add(serverId);
}
if (globalError != null) Error.throwWithStackTrace(globalError, globalStackTrace!);
hubs = [...globalHubs, if (music != null) ...music.hubs];
} else {
// A fallback backend cannot run a hub leg without the libraries its
// prefetch failed to discover.
if (useGlobalHubs && prefetchDegraded) return const <MediaHub>[];
final libraryHubs = await _fetchLibraryHubsForClient(
client,
limit: hubItemLimit,
hiddenLibraryKeys: hiddenLibraryKeys,
includePlaybackHubs: includePlaybackHubs,
libraries: useGlobalHubs ? serverLibraries : null,
);
if (libraryHubs.succeeded || (!libraryHubs.failed && !libraryHubs.cancelled)) {
legSucceededServerIds.add(serverId);
}
if (libraryHubs.failed) legFailedServerIds.add(serverId);
if (libraryHubs.cancelled) legCancelledServerIds.add(serverId);
hubs = libraryHubs.hubs;
}
final processed = _postProcessHubs(hubs, serverId: ServerId(serverId), hiddenLibraryKeys: hiddenLibraryKeys);
final scope = client.cacheServerId;
observedRows.addAll([
for (final hub in processed)
for (final item in hub.items) (item: item, clientScope: scope),
]);
return processed;
},
);
for (final entry in globalDiagnosticsByServer.entries) {
if (entry.value.failed) legFailedServerIds.add(entry.key);
if (entry.value.cancelled) legCancelledServerIds.add(entry.key);
}
final all = fetched.items;
final hubs = limit != null && limit < all.length ? all.sublist(0, limit) : all;
return (
hubs: hubs,
// Pre-limit, so a row trimmed off the tail still counts as observed.
observedItems: observedRows,
succeededServerIds: legSucceededServerIds,
cancelledServerIds: {...fetched.cancelledServerIds, ...legCancelledServerIds},
failedServerIds: {...fetched.failedServerIds, ...legFailedServerIds},
);
}
/// Per-library hub fetch for a single client. Filters to visible libraries
/// of [kinds] (movie/show/clip/artist by default — clip covers Jellyfin
/// musicvideos/homevideos, #1476; artist brings music rows to home) and
/// concatenates the results. The rich-hub music append passes
/// `{MediaKind.artist}` to fetch only what the global endpoint misses.
///
/// `succeeded` stays false when no library leg was attempted. This prevents
/// the optional rich-hub music append from masking a failed global leg; the
/// fallback caller separately recognizes a clean zero-library result as an
/// authoritative no-op.
Future<({List<MediaHub> hubs, bool succeeded, bool failed, bool cancelled})> _fetchLibraryHubsForClient(
MediaServerClient client, {
required int limit,
Set<String>? hiddenLibraryKeys,
required bool includePlaybackHubs,
List<MediaLibrary>? libraries,
Set<MediaKind> kinds = const {MediaKind.movie, MediaKind.show, MediaKind.clip, MediaKind.artist},
}) async {
final libs = libraries ?? await client.fetchLibraries();
final visible = libs.where((l) {
if (!kinds.contains(l.kind)) return false;
if (l.hidden) return false;
if (hiddenLibraryKeys != null && hiddenLibraryKeys.contains(l.globalKey)) return false;
return true;
}).toList();
// Sliding window rather than batches of three separated by a barrier: a
// batch waits for its slowest member before the next one starts, so wall
// time was the sum of per-batch maxima and one slow library stalled every
// library queued behind it (#1784). Starting the next request the moment
// any slot frees keeps the same peak concurrency with no head-of-line
// blocking. Results are written back by index so hub order stays the
// library order regardless of completion order.
const concurrency = 3;
final results = List<List<MediaHub>>.filled(visible.length, const []);
var next = 0;
var succeeded = false;
var failed = false;
var cancelled = false;
Future<void> worker() async {
while (true) {
final index = next++;
if (index >= visible.length) return;
final library = visible[index];
final diagnostics = HubFetchDiagnostics();
try {
final hubs = await client.fetchLibraryHubs(
library.id,
libraryName: library.title,
limit: limit,
includePlaybackHubs: includePlaybackHubs,
libraryKind: library.kind,
diagnostics: diagnostics,
);
results[index] = hubs;
if (hubs.isNotEmpty || (!diagnostics.failed && !diagnostics.cancelled)) succeeded = true;
} catch (e, st) {
if (_isCancellation(e)) {
cancelled = true;
appLogger.d('Cancelled library hub fetch for ${library.globalKey}');
} else {
failed = true;
appLogger.e('Failed to fetch library hubs for ${library.globalKey}', error: e, stackTrace: st);
}
}
if (diagnostics.failed) failed = true;
if (diagnostics.cancelled) cancelled = true;
}
}
await Future.wait([for (var i = 0; i < concurrency && i < visible.length; i++) worker()]);
return (hubs: [for (final list in results) ...list], succeeded: succeeded, failed: failed, cancelled: cancelled);
}
/// Filter hidden-library items and drop empty hubs.
List<MediaHub> _postProcessHubs(List<MediaHub> hubs, {required ServerId serverId, Set<String>? hiddenLibraryKeys}) {
var filtered = hubs;
if (hiddenLibraryKeys != null && hiddenLibraryKeys.isNotEmpty) {
filtered = filtered
.map((hub) {
final filteredItems = hub.items.where((item) {
final libraryId = item.libraryId;
if (libraryId == null) return true;
final globalKey = buildGlobalKey(ServerId(serverId), libraryId);
return !hiddenLibraryKeys.contains(globalKey);
}).toList();
if (filteredItems.isEmpty) return null;
return hub.copyWith(items: filteredItems, size: filteredItems.length);
})
.whereType<MediaHub>()
.toList();
}
return filtered;
}
/// Search across all online servers (Plex + Jellyfin). Per-server outcomes
/// distinguish authoritative empty results from failed or cancelled legs.
///
/// [hiddenLibraryKeys] excludes results the user has hidden, matching every
/// other aggregated surface. Backends whose search rows carry no library id
/// cannot be filtered here; they must scope the search server-side instead.
Future<SearchAggregationResult> searchAcrossServers(
String query, {
int? limit,
Set<String>? hiddenLibraryKeys,
AbortController? abort,
}) async {
if (query.trim().isEmpty) {
return (
items: const <MediaItem>[],
succeededServerIds: const <String>{},
cancelledServerIds: const <String>{},
failedServerIds: const <String>{},
);
}
abort?.throwIfAborted();
final clients = _serverManager.onlineClients;
if (clients.isEmpty) {
return (
items: const <MediaItem>[],
succeededServerIds: const <String>{},
cancelledServerIds: const <String>{},
failedServerIds: const <String>{},
);
}
final resultLimit = limit ?? defaultMediaSearchLimit;
final fetchLimit = resultLimit < defaultMediaSearchLimit ? defaultMediaSearchLimit : resultLimit;
final fetched = await _fanOut<MediaItem>(
clients,
failureMessage: (serverId) => 'Search failed on $serverId',
fetch: (serverId, client) async {
final stopwatch = Stopwatch()..start();
final items = await client.searchItems(
query,
limit: fetchLimit,
abort: abort,
excludedLibraryIds: _hiddenLibraryIdsOn(serverId, hiddenLibraryKeys),
);
appLogger.i(
'Search completed on $serverId in ${stopwatch.elapsedMilliseconds}ms: '
'${items.length} results ${_searchKindCounts(items)}',
);
return items;
},
);
abort?.throwIfAborted();
// Before ranking, so hidden results cannot spend the `resultLimit` budget
// and silently shrink what the user sees.
final visible = _withoutHiddenLibraries(fetched.items, hiddenLibraryKeys);
final items = rankMediaSearchResults(visible, query, limit: resultLimit);
appLogger.i(
'Search aggregation completed: ${items.length} results '
'(${fetched.succeededServerIds.length} succeeded, ${fetched.cancelledServerIds.length} cancelled, '
'${fetched.failedServerIds.length} failed) ${_searchKindCounts(items)}',
);
return (
items: items,
succeededServerIds: fetched.succeededServerIds,
cancelledServerIds: fetched.cancelledServerIds,
failedServerIds: fetched.failedServerIds,
);
}
/// Reverse external-id lookup fanned out to every online server (see
/// [MediaServerClient.findByExternalIds]). One request wave per tap on an
/// Explore catalog item; per-server failures are logged and skipped.
///
/// Every server contributes every copy it holds, not one apiece: the same
/// movie routinely sits in a 4K library and an HD library on one server
/// (#1754). Results are deduped by global key and ordered best-first with
/// [compareLibraryCopies] so the chooser is stable across repeated passes.
///
/// Because per-server failures are dropped here, a caller holding earlier
/// results must merge rather than replace (see [mergeLibraryCopies]) — a
/// degraded wave is not evidence that a copy went away.
Future<List<MediaItem>> findByExternalIdsAcrossServers(
ExternalIds ids, {
required MediaKind kind,
List<String> titles = const [],
int? year,
String? plexGuid,
ExternalSeasonRef? season,
}) async {
if (!ids.hasAny && plexGuid == null) return [];
final clients = _serverManager.onlineClients;
if (clients.isEmpty) return [];
final futures = clients.entries.map((entry) async {
try {
return await entry.value.findByExternalIds(
ids,
kind: kind,
titles: titles,
year: year,
plexGuid: plexGuid,
season: season,
);
} catch (e, st) {
appLogger.w('External-id lookup failed on ${entry.key}', error: e, stackTrace: st);
return const <MediaItem>[];
}
});
return mergeLibraryCopies(const [], (await Future.wait(futures)).expand((items) => items));
}
/// Group libraries by server (internal aggregation helper).
Map<String, List<MediaLibrary>> _groupLibrariesByServer(List<MediaLibrary> libraries) {
final grouped = <String, List<MediaLibrary>>{};
for (final library in libraries) {
final serverId = library.serverId;
if (serverId != null) {
grouped.putIfAbsent(serverId, () => []).add(library);
}
}
return grouped;
}
}