Files
silo-server/internal/sections/fetcher.go
T
430224a1b9 perf: cut home-screen, Continue Watching, and Latest latency; cache shared home rails (#292)
* perf(jellycompat,sections): bound resume scan, batch leaf detail progress, widen section concurrency

Three low-risk fixes from the section-fetch performance investigation
(docs/superpowers/plans/2026-07-03-section-fetch-performance.md):

- jellycompat: bound loadProgressPage at resumeScanMaxRows=300 so a single
  request never pages through more than that many in-progress rows. The cap is
  unconditional: it also covers the sparse-visible-set case (a heavy watcher
  whose recent rows are mostly dismissed/superseded, or a Series/Season-only
  request that matches no leaf in-progress row), where the page never fills and
  the loop would otherwise scan the entire history — previously an O(history)
  scan reaching tens of seconds. In the common case the loop exits far earlier,
  so the cap only bounds the pathological worst case; 300 leaves ample headroom
  to fill a ~20-item Continue Watching page. Beyond the cap the reported total
  is a clamped lower bound. Covered by TestLoadProgressPage_BoundsScanForSparseVisibleSet.
- jellycompat: batch the leaf-item (movie/episode) progress lookup in
  GetItemDetailsByIDs via ListProgressWithCompletedHistory instead of a
  per-item GetProgressWithCompletedHistory (~100 sequential queries for a
  50-item detail page). Series keep the per-item episode-rollup path (they own
  no progress row). Output is unchanged; a batch-lookup failure is now logged
  rather than silently dropping played state for the whole page.
- sections: raise fetchAllMaxConcurrency 4 -> 6 to cut FetchAll wave count for
  large home layouts, staying within the default 20-conn pool.

Part of the home/continue-watching latency work.

AI-use disclosure: implemented with AI (Claude) assistance.

* perf(jellycompat): keep Latest browse on the cross-library fast path under isPlayed

/Items/Latest with isPlayed=false is the highest-frequency compat browse
(~10.8k calls/day). The played overlay can't be pushed into SQL, so browse
over-fetches and filters locally. The cross-library recently_added fast path
(BrowseRecentlyAddedAcrossLibraries: one ~1ms index walk per library) was
gated on Offset==0, so a heavy watcher who had already seen the newest items
needed a 2nd chunk and fell through to BrowsePage — a whole-catalog
MIN(first_seen_at) + GROUP BY HashAggregate over ~147k movies measured at
~755ms per call (0.8-1.6s observed end-to-end).

Fetch the entire over-fetch budget (maxScannedRows) in a single merged
fast-path walk instead of paging into BrowsePage, so the loop fills from one
call. The clamp caveat (MaxLimit=1000 leaves a fall-through only for
requestedLimit>200, off the Latest hot path) is documented inline.

Part of the home/browse latency work.

AI-use disclosure: implemented with AI (Claude) assistance.

* fix(jellycompat): scope resume scan cap to resume path and bound the fast-path loop

Addresses PR #292 review feedback:

- Codex (P2): the resumeScanMaxRows cap was applied unconditionally in the
  general loop, which also paginates the completed (watched-items) list. Gate it
  on resumeFiltered so the completed path keeps exact TotalRecordCount and deep
  StartIndex pagination. Covered by TestLoadProgressPage_CompletedScanNotCapped.
- CodeRabbit (Critical): the earlier raw-offset fast-path loop — the default
  Continue Watching shape and the sections-fallback route — had the same
  unbounded-scan bug and was not covered by the cap (the existing test forces
  EnableTotalRecordCount=true, routing around it). Bound it with the same
  resumeScanMaxRows guard. Covered by
  TestLoadProgressPage_BoundsFastPathScanForSparseVisibleSet.
- CodeRabbit (Minor): tag the doc's fenced example blocks as text to satisfy
  markdownlint MD040.

AI-use disclosure: implemented with AI (Claude) assistance.

* perf(sections): cache shared user-agnostic home rails per access scope

Home-screen rails that are identical for everyone who can see the same
libraries (recently added, recently released, genre, trending on server,
most watched, new to library, critically acclaimed, award winners, format
showcase, seasonal, mood, trending discover, admin-curated lists, and
library collections) were rebuilt from Postgres once per request, per user.
Only the overlay on top of each row (watched flags, play position, presigned
poster URLs) is actually per-user.

Insert a process-global resolved-list cache at the FetchOne choke point in
internal/sections. Each cacheable row is built once per access scope, held
with a 15m TTL, and refreshed in the background 3m before expiry;
singleflight collapses cold-miss stampedes into a single build. The per-user
overlay still runs fresh in buildSectionsResponse, so no profile state is
ever shared. Random and per-user rows (continue watching, next up,
recommendations, hidden gems, forgotten favorites, activity feed, user
collections) bypass the cache.

The access-scope key captures every access boundary the fetch path enforces
-- section identity (type + id + config hash) + item limit + accessible and
disabled libraries + max content rating + excluded media types + name prefix
+ allowed-content-id allowlist -- and nothing per-user, so entries are safely
shared. Empty membership is never cached (avoids freezing a transiently empty
rail); background refreshes are bounded by a timeout.

Scale (analytical, derived from the cache behavior -- not a measured
latency): for the user-agnostic rows, Postgres section-query volume collapses
from O(rows x concurrent requests) to O(rows x distinct access scopes) per
15m refresh window, because most users share a handful of access scopes.
Illustrative -- 40 cacheable rows on a home screen, 1000 concurrent users
falling into ~5 distinct access scopes:

  - before: ~40 x 1000 = ~40,000 section queries per wave of home loads
  - after:  ~40 x 5    = ~200 builds per 15m window (plus one background
            refresh per row per scope), i.e. a warm home load runs zero
            section queries for these rows.

That is a ~99% reduction in shared section-query load at that concurrency;
the win grows with concurrency and shrinks as access-scope diversity rises.

Design/plan doc added under docs/superpowers/plans/.

* perf(jellycompat): serve per-library Latest via the cached recently-added section

A jellyfin-compat per-library /Items/Latest rail is the same user-agnostic
list as the native "recently added" library rail -- both order by
mil.first_seen_at DESC. It was rebuilt on every request through
directContentService.BrowseItems, missing the resolved-list cache entirely.

Route per-library Latest for movies and series libraries through the native
section fetch instead, so it reuses the shared cache. HandleLatest resolves
the library's type once, and for a movies/series library builds a synthetic
SectionRecentlyAdded with the same type + config + limit + access scope the
native rail uses and calls FetchOne; the per-user overlay (favorites,
progress, episode targets, presign) is extracted into buildLatestItemDTOs and
shared by both the native and BrowseItems paths, so no overlay logic is
duplicated. Cached *models.MediaItem values are read-only -- LocalizeItemModels
deep-copies before any presign mutation.

To let the two surfaces share one entry, resolvedListCacheKey no longer
includes the arbitrary section ID: every cacheable section type derives its
membership from type + config + limit + scope, never from its own ID (audited
all 14 cacheable types plus the library-collection path; the sole s.ID read
lives in the non-cacheable user-collection branch). A native recently-added
rail and the compat Latest for the same library + scope now collapse to ONE
cache entry, built once and reused. Access-scope isolation is unchanged --
the removed ID never carried access information, and every access boundary
(libraries, rating cap, excluded types, content allow-list, name prefix)
still keys the entry.

Guardrails: the native path is restricted to movies and series libraries;
every other library type (ebook, music, manga, mixed) is ignored and keeps
its exact BrowseItems behavior -- important because an unfiltered
recently-added fetch would otherwise surface non-video items to Jellyfin
clients that only expect video. Deeper pages, played-filter and
backdrop-required requests, a client asking for a type other than the
library's own, and any FetchOne error also fall back to BrowseItems.

Chosen over an alternative that gave the synthetic section a deterministic ID
(which kept two separate cache entries): both returned identical data with
similar complexity, so the shared-entry design won.

* fix(sections,jellycompat): post-review fixes for the shared-list cache and Latest path

Consolidates fixes from the branch's adversarial review and PR #292 review
comments into one commit:

- Latest fast path: fall back to BrowseItems when a request carries a genre,
  name-prefix, or person filter (the synthetic recently-added section cannot
  express these, so serving it unfiltered would return a wrong, broader set).
  Eligibility is decided by latestFastPathEligible and covered by a test.
- Clamp the /Items/Latest page size to compatBrowseMaxLimit before building the
  section, matching the BrowseItems fallback, so a large client Limit can't drive
  an oversized recently-added fetch or explode the shared cache key with unbounded
  ItemLimit values.
- Evict expired entries from the process-global resolvedListCache: resolvedListSet
  sweeps expired keys at most once per minute, bounding the map to scopes seen
  within one TTL window. Covered by TestResolvedListCacheEvictsExpiredEntries.
- Log a short digest of the cache key (resolvedListLogKey) instead of the raw key
  in the background-refresh panic/error paths, since the key embeds
  user-controlled access-scope fields such as NamePrefix.

Skipped review comments (verified already fixed or stale against current code):
the resume fast-path scan bound and watched-items cap (04d2e795) and the docs
fence-language tags (already addressed).

Build, vet, and go test -race pass for internal/sections and internal/jellycompat.

* perf(plugins): cache plugin installations in-memory, invalidated on lifecycle change

## Problem
Every poster/image on a warm home rail re-read plugin_installations from
Postgres to answer "is this plugin enabled?" and to acquire the plugin client
(Source A: metadata chain buildProviders enabled-check; Source B: ensureClient
-> loadInstallation). Plugin-resolved image URLs are never URL-cached, so the
plugin source and the DB read behind it fired again on every identical warm
request; 100% of images in the target library are plugin-backed.

## Solution
- Guarded in-memory installation cache (map[int]*Installation + RWMutex) in
  plugins.Service. loadInstallation reads through it; the requireEnabled gate
  stays after the cache read so ErrInstallationDisabled semantics are unchanged.
  invalidateInstallationCache clears it and is self-registered as a lifecycle
  hook, so Service.OnLifecycleChange wipes it on install/enable/disable/update/
  uninstall.
- A generation counter closes an invalidate-vs-repopulate race: captured before
  installations.GetByID and re-checked under the write lock, so a row fetched
  before a lifecycle mutation is never written into a freshly cleared cache
  (would otherwise resurrect a just-disabled plugin).
- Route the metadata chain enabled-check through the same cache via a structural
  InstallationEnabledChecker interface (nil-safe: falls back to the pool query
  when no checker is injected), wired in cmd/silo/main.go.

## Post-review fix (auto-update reliability blocker)
AutoUpdateService mutated installations (new InstallPath, old dir deleted) on
the default auto update policy without firing OnLifecycleChange, leaving the
cache stale and breaking plugins with "stored plugin manifest mismatch" until
restart. It now takes an onChange callback wired to Service.OnLifecycleChange
and fires it once per Check run that mutated a row.

## Verification
go build/vet, go test ./internal/plugins/... ./internal/metadata/... (-race).
Tests: cache hit/invalidation, racing-invalidation guard, IsInstallationEnabled,
auto-update fires onChange.

## AI-use disclosure
Implemented with AI assistance (Claude).

* perf(jellycompat): batch per-item presign, and enrich series on the cached Latest path

## Problem
List rails presigned each item's poster/backdrop/logo/still image individually
(~160 singular resolver calls for a 40-item page where 4 batched calls suffice),
and ItemsHandler carried a near-verbatim duplicate of the batch presigner.

## Solution (batching)
Promote the batch presigner to a shared package-level presignCompatListItems
(presign_list.go) with a generic collectImagePaths[T]; convert the per-item
loops (cached home/Latest rail, favorites, batch loaders, userdata favorites) to
one batched PresignImageURLsWithExpiry per image type per page; batch the
season/episode collections; delete the three duplicate presign helpers. URL
output is unchanged (verified byte-for-byte).

## Post-review fix (series Latest data-parity regression)
The native cached Latest fast path built items via compatListItemsFromModels +
buildLatestItemDTOs and never ran the series watch-state rollup, so a series
library's Latest lost Played / UnplayedItemCount and page 1 disagreed with the
BrowseItems fallback. enrichSeriesUserData is promoted to the ContentService
interface and called on the native path (reused, not duplicated).

## Verification
go build/vet, go test ./internal/jellycompat/... ./internal/catalog/...
Tests: bounded presign invocation counts + per-item URL mapping; series rollup
populated on the native Latest path.

## AI-use disclosure
Implemented with AI assistance (Claude).

* perf(sections): gate personalized rails out of the shared cache; widen refresh lead

## Problem
1. The shared home-rail cache whitelisted custom_filter/genre sections by TYPE
   alone, but those route through fetchFiltered -> ParseQueryDefinition and can
   carry personalized (per-profile) rules/sorts (watched, favorited,
   in_watchlist, in_progress, last_watched; sorts progress/date_viewed/plays).
   Their membership is per-profile yet the cache key excludes userID/profileID,
   so a personalized rail built for one profile was served to others in the same
   access scope for up to 15m -- a cross-profile watchlist/watch-state leak.
2. The background-refresh lead was tuned so steady traffic is served a warm
   entry from a longer soft window.

## Solution
- Add QueryDefinition.IsPersonalized() (reusing the existing
  QueryFieldRequiresProfile/QuerySortRequiresProfile helpers).
  isCacheableSectionType now parses the section QueryDefinition and refuses to
  cache custom_filter/genre when personalized; non-personalized definitions stay
  cacheable. Seasonal/mood/trending build their definitions server-side and stay
  unconditionally cacheable.
- resolvedListRefreshLead 3m -> 10m (soft threshold builtAt+5min instead of
  builtAt+12min).

## Verification
go build/vet, go test ./internal/sections/... ./internal/catalog/... (-race).
Test: personalized custom_filter/genre not cacheable; non-personalized are.

## AI-use disclosure
Implemented with AI assistance (Claude).

* fix(sections,metadata): post-review fixes for shared cache and plugin chain staleness

Addresses three review findings on PR #292:

- sections: canonicalize section config JSON before hashing so configs
  differing only in whitespace/field order share a cache entry (native +
  jellycompat rail sharing). Added TestHashSectionConfigCanonicalizes.
- metadata: invalidate the resolved-chain cache on plugin lifecycle
  changes; the installation-enabled check already reads the invalidated
  plugin cache, but resolveChainCached could serve a stale provider chain
  for up to chainCacheTTL after a provider's availability changed.
- jellycompat: move ctx to the first parameter of presignCompatListItems
  for consistency with the other presign helpers.

Skipped the episode-image presign batching nitpick: the resolver already
dedupes+singleflights, so it is a Minor perf-only item not worth the
two-pass refactor risk in this pass.

* fix(sections,jellycompat): harden shared rail cache and Latest fast path per review

Addresses the eight findings from the deep review of this PR:

- Detach the blocking cold-miss rebuild from the singleflight leader's
  request context (context.WithoutCancel + the shared 30s build timeout)
  so one client disconnect no longer fails every collapsed waiter and
  leaves the entry uncached.
- Stop client-controlled values minting unbounded cache entries: the
  compat Latest fast path now always fetches a fixed 100-row budget and
  slices to the requested limit (one entry per scope+library instead of
  one per Limit value), and an unrecognized MaxOfficialRating string
  disqualifies the fast path instead of entering the global cache key.
- Add release_date to the sections item projection/scan so movies served
  via the Latest fast path keep PremiereDate (Jellyfin default-set field)
  in parity with the BrowseItems fallback.
- Fall back to per-item progress lookups when the batched leaf progress
  query fails, restoring one-item-at-a-time degradation instead of
  blanking played state for the whole page.
- Derive cache eligibility from a single source of truth: fetchSection
  and isCacheableSectionType now share the userAgnosticSectionFetcher
  table, whose no-userID/profileID signature makes a fetcher drop out of
  the cacheable set at compile time if it ever gains per-profile inputs.
- Decide Latest fast-path eligibility off the actual browse params the
  fallback would receive, so any filter later added to buildBrowseParams
  automatically disqualifies the cached path; share one
  compatDefaultBrowseLimit constant between both paths.
- Extract AccessFilter.WriteAccessScopeCacheKey as the shared, security-
  critical serializer for all access-scoped caches (resolved-list,
  editorial candidates, audiobook groups); the editorial key now captures
  ExcludedMediaTypes, which its loaders already applied in SQL.
- Strip leaked agent-transcript markup from the section-fetch plan doc.

go build ./..., go vet, gofmt clean; go test -race on
internal/sections, internal/catalog, internal/jellycompat passes
(TestBeginWebOperation* failures are the known pre-existing flakes).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Quick <31828688+Quick104@users.noreply.github.com>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-05 02:01:30 -04:00

3356 lines
108 KiB
Go

package sections
import (
"context"
"encoding/json"
"fmt"
"log/slog"
"maps"
"sort"
"strconv"
"strings"
"sync"
"time"
"github.com/jackc/pgx/v5"
"github.com/jackc/pgx/v5/pgxpool"
"golang.org/x/sync/singleflight"
"github.com/Silo-Server/silo-server/internal/catalog"
"github.com/Silo-Server/silo-server/internal/models"
"github.com/Silo-Server/silo-server/internal/overlays"
"github.com/Silo-Server/silo-server/internal/recommendations"
"github.com/Silo-Server/silo-server/internal/sections/recipes"
"github.com/Silo-Server/silo-server/internal/userstore"
)
// SectionWithItems is a resolved section populated with query results.
type SectionWithItems struct {
ResolvedSection
Items []*models.MediaItem `json:"items"`
TotalCount int `json:"total_count"`
ItemMeta map[string]SectionItemMeta
}
// SectionItemMeta carries optional per-item metadata for richer section UIs.
type SectionItemMeta struct {
SeriesID *string
SeriesTitle string
SeasonNumber *int
EpisodeNumber *int
Badges []string
PositionSeconds *float64
DurationSeconds *float64
ProgressUpdatedAt *string
ItemSource string // "in_progress" or "next_up"
SortTimestamp time.Time // when the preceding episode was completed (for ordering)
}
const recentSeasonPremiereBadgeWindowDays = 14
const editorialCandidateCacheTTL = 24 * time.Hour
// fetchAllMaxConcurrency bounds how many sections FetchAll resolves in parallel.
// Each concurrent section holds a pgxpool connection, so this is also per-request
// pool pressure: with the default pool of 20 conns (database.max_connections),
// N concurrent home requests can hold up to N*fetchAllMaxConcurrency connections.
// Raised from 4 to 6 to cut the wave count for large home layouts (e.g. 32
// sections: 8 waves → 6) while keeping headroom for a few concurrent home loads
// before the pool saturates. Do not raise further without widening the pool or
// load-testing pool contention.
const fetchAllMaxConcurrency = 6
const slowSectionFetchThreshold = 500 * time.Millisecond
const slowAggregateFetchThreshold = time.Second
type recommendationReader interface {
GetForYouMain(ctx context.Context, userID int, profileID string, limit int, filter catalog.AccessFilter) (*recommendations.ForYouRow, error)
GetBecauseYouWatched(ctx context.Context, userID int, profileID, sourceItemID string, limit int, filter catalog.AccessFilter) ([]recommendations.ScoredItem, error)
GetSimilarUsersLiked(ctx context.Context, userID int, profileID string, limit int, filter catalog.AccessFilter) ([]recommendations.ScoredItem, error)
GetTasteMatchRow(ctx context.Context, userID int, profileID, genre string, limit int, filter catalog.AccessFilter) (*recommendations.ForYouRow, error)
}
// trendingSnapshotGetter is the read side of the trending snapshot table.
// Satisfied by *TrendingSnapshotRepository.
type trendingSnapshotGetter interface {
Get(ctx context.Context, source, window string) (TrendingSnapshot, bool, error)
}
// Fetcher runs section queries against the database.
type Fetcher struct {
pool *pgxpool.Pool
progressFilter *catalog.ContinueWatchingProgressFilter
StoreProvider userstore.UserStoreProvider
CollectionRepo *catalog.LibraryCollectionRepository
RecommendationRepo *recommendations.Repo // retained for non-reader call sites
RecommendationReader recommendationReader
NextUpRepo *catalog.NextUpRepository
AudiobookNextRepo *catalog.AudiobookNextRepository
// TrendingSnapshots reads the persisted external-trending snapshots that
// back the trending_discover section. Nil renders that section empty.
// Snapshots are produced out-of-band by TrendingRefresher, so the read path
// never calls the upstream provider.
TrendingSnapshots trendingSnapshotGetter
candidateCacheMu sync.Mutex
candidateCache *editorialCandidateCache
candidateGroup singleflight.Group
// Clock returns the current time. Defaults to recipes.RealClock{}.
// Tests inject recipes.FixedClock for deterministic seasonal/editorial behavior.
Clock recipes.Clock
}
// NewFetcher creates a new section Fetcher.
func NewFetcher(pool *pgxpool.Pool) *Fetcher {
return &Fetcher{
pool: pool,
progressFilter: catalog.NewContinueWatchingProgressFilter(pool),
Clock: recipes.RealClock{},
}
}
type editorialCandidateLoader func(context.Context, string, *int, []int, catalog.AccessFilter) ([]string, error)
type editorialCandidateCache struct {
mu sync.RWMutex
entries map[string]editorialCandidateCacheEntry
}
type editorialCandidateCacheEntry struct {
candidates []string
expiresAt time.Time
}
func newEditorialCandidateCache() *editorialCandidateCache {
return &editorialCandidateCache{entries: make(map[string]editorialCandidateCacheEntry)}
}
func (c *editorialCandidateCache) get(key string, now time.Time) ([]string, bool) {
if c == nil {
return nil, false
}
c.mu.RLock()
entry, ok := c.entries[key]
c.mu.RUnlock()
if !ok || !now.Before(entry.expiresAt) {
return nil, false
}
return append([]string(nil), entry.candidates...), true
}
func (c *editorialCandidateCache) set(key string, candidates []string, expiresAt time.Time) {
if c == nil {
return
}
c.mu.Lock()
c.entries[key] = editorialCandidateCacheEntry{
candidates: append([]string(nil), candidates...),
expiresAt: expiresAt,
}
c.mu.Unlock()
}
func (f *Fetcher) ensureEditorialCandidateCache() *editorialCandidateCache {
f.candidateCacheMu.Lock()
defer f.candidateCacheMu.Unlock()
if f.candidateCache == nil {
f.candidateCache = newEditorialCandidateCache()
}
return f.candidateCache
}
func (f *Fetcher) now() time.Time {
if f != nil && f.Clock != nil {
return f.Clock.Now()
}
return time.Now()
}
func (f *Fetcher) cachedEditorialCandidates(ctx context.Context, subjectType string, libraryID *int, libraryIDs []int, filter catalog.AccessFilter, ttl time.Duration, loader editorialCandidateLoader) ([]string, error) {
if ttl <= 0 {
return loader(ctx, subjectType, libraryID, libraryIDs, filter)
}
cache := f.ensureEditorialCandidateCache()
key := editorialCandidateCacheKey(subjectType, libraryID, libraryIDs, filter)
now := f.now()
if candidates, ok := cache.get(key, now); ok {
return candidates, nil
}
value, err, _ := f.candidateGroup.Do(key, func() (any, error) {
now := f.now()
if candidates, ok := cache.get(key, now); ok {
return candidates, nil
}
candidates, err := loader(ctx, subjectType, libraryID, libraryIDs, filter)
if err != nil {
return nil, err
}
cache.set(key, candidates, now.Add(ttl))
return append([]string(nil), candidates...), nil
})
if err != nil {
return nil, err
}
candidates, _ := value.([]string)
return append([]string(nil), candidates...), nil
}
func editorialCandidateCacheKey(subjectType string, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) string {
var b strings.Builder
b.WriteString("subject=")
b.WriteString(strings.ToLower(strings.TrimSpace(subjectType)))
b.WriteString("|library=")
if libraryID == nil {
b.WriteString("all")
} else {
b.WriteString(strconv.Itoa(*libraryID))
}
b.WriteString("|libraries=")
writeOptionalSortedInts(&b, libraryIDs)
// Serialize the full access scope through the shared catalog helper. The
// candidate loaders push rating AND excluded media types (via
// ApplySectionAccessFilter) into their SQL, so the key must capture every
// boundary; the shared helper guarantees it stays in sync with the other
// access-scoped caches when AccessFilter grows.
filter.WriteAccessScopeCacheKey(&b)
return b.String()
}
func writeOptionalSortedInts(b *strings.Builder, values []int) {
if values == nil {
b.WriteString("<nil>")
return
}
if len(values) == 0 {
b.WriteString("<empty>")
return
}
writeSortedInts(b, values)
}
func writeSortedInts(b *strings.Builder, values []int) {
if len(values) == 0 {
return
}
sorted := append([]int(nil), values...)
sort.Ints(sorted)
for i, value := range sorted {
if i > 0 {
b.WriteByte(',')
}
b.WriteString(strconv.Itoa(value))
}
}
// FetchOne runs one section query and returns it with items.
func (f *Fetcher) FetchOne(ctx context.Context, resolved ResolvedSection, libraryID *int, libraryIDs []int, userID int, profileID string, filter catalog.AccessFilter) (SectionWithItems, error) {
start := time.Now()
var result SectionWithItems
var err error
defer func() {
f.logSlowSectionFetch(resolved, libraryID, libraryIDs, result, time.Since(start), err)
}()
if resolved.SectionType == SectionContinueWatching {
result, err = f.fetchContinueWatchingSection(ctx, resolved, libraryID, libraryIDs, userID, profileID, filter)
return result, err
}
if resolved.SectionType == SectionNextUp {
result, err = f.fetchNextUpSection(ctx, resolved, libraryID, libraryIDs, userID, profileID, filter)
return result, err
}
if resolved.SectionType == SectionNextInSeries {
result, err = f.fetchNextInSeriesSection(ctx, resolved, libraryID, libraryIDs, userID, profileID, filter)
return result, err
}
if resolved.SectionType == SectionEditorialSpotlight {
var items []*models.MediaItem
var total int
var title string
items, total, title, err = f.fetchEditorialSpotlightWithTitle(ctx, resolved, libraryID, libraryIDs, filter)
if err != nil {
return SectionWithItems{}, err
}
if title != "" {
resolved.Title = title
}
result = SectionWithItems{
ResolvedSection: resolved,
Items: items,
TotalCount: total,
}
return result, nil
}
var items []*models.MediaItem
var total int
if f.isCacheableSectionType(resolved) {
// User-agnostic rows share one process-global entry per access scope.
// getOrRefresh returns a defensive slice copy, so reordering or
// truncating result.Items below (or in a caller) cannot corrupt the
// cached entry; the pointed-to *MediaItem must not be mutated in place
// (see cloneMediaItems). The per-user overlay still runs fresh in
// buildSectionsResponse.
key := resolvedListCacheKey(resolved, libraryID, libraryIDs, filter)
items, total, err = getOrRefresh(ctx, key, f.now(), func(loadCtx context.Context) ([]*models.MediaItem, int, error) {
return f.fetchSection(loadCtx, resolved, libraryID, libraryIDs, userID, profileID, filter)
})
} else {
items, total, err = f.fetchSection(ctx, resolved, libraryID, libraryIDs, userID, profileID, filter)
}
if err != nil {
return SectionWithItems{}, err
}
// Apply seasonal title override when the active theme has a custom name.
// Done here (rather than inside fetchSeasonalThemed) so the section's
// stored Title is preserved as the fallback used by callers that bypass
// SectionWithItems construction.
if resolved.SectionType == SectionSeasonalThemed && len(items) > 0 {
if title := f.seasonalTitleOverride(resolved); title != "" {
resolved.Title = title
}
}
result = SectionWithItems{
ResolvedSection: resolved,
Items: items,
TotalCount: total,
}
return result, nil
}
func (f *Fetcher) logSlowSectionFetch(resolved ResolvedSection, libraryID *int, libraryIDs []int, result SectionWithItems, duration time.Duration, err error) {
if duration < slowSectionFetchThreshold {
return
}
attrs := []any{
"section_id", resolved.ID,
"type", resolved.SectionType,
"title", resolved.Title,
"item_count", len(result.Items),
"total_count", result.TotalCount,
"duration_ms", duration.Milliseconds(),
}
if libraryID != nil {
attrs = append(attrs, "library_id", *libraryID)
}
if libraryIDs != nil {
attrs = append(attrs, "library_ids", append([]int(nil), libraryIDs...))
}
if err != nil {
attrs = append(attrs, "error", err)
}
slog.Warn("slow section fetch", attrs...)
}
// seasonalTitleOverride returns the per-theme display name override for a
// seasonal section, or "" when no override applies.
func (f *Fetcher) seasonalTitleOverride(resolved ResolvedSection) string {
if len(resolved.Config) == 0 {
return ""
}
var p recipes.SeasonalThemedParams
if err := json.Unmarshal(resolved.Config, &p); err != nil {
return ""
}
if len(p.ThemeTitles) == 0 {
return ""
}
now := time.Now()
if f.Clock != nil {
now = f.Clock.Now()
}
return recipes.SeasonalTitleOverride(p, now)
}
func (f *Fetcher) fetchContinueWatchingSection(ctx context.Context, resolved ResolvedSection, libraryID *int, libraryIDs []int, userID int, profileID string, filter catalog.AccessFilter) (SectionWithItems, error) {
if f.StoreProvider == nil || userID <= 0 || profileID == "" {
return SectionWithItems{
ResolvedSection: resolved,
Items: []*models.MediaItem{},
TotalCount: 0,
ItemMeta: map[string]SectionItemMeta{},
}, nil
}
store, err := f.StoreProvider.ForUser(ctx, userID)
if err != nil {
return SectionWithItems{}, fmt.Errorf("getting user store: %w", err)
}
limit := resolved.ItemLimit
if limit <= 0 {
limit = 20
}
continueType := ContinueTypeFromConfig(resolved.Config)
cfgFilters := ParseConfigFilters(resolved.Config)
configLibraryIDs := cfgFilters.LibraryIDs()
effectiveLibID := libraryID
effectiveLibraryIDs := libraryIDs
if effectiveLibID == nil && len(configLibraryIDs) > 0 {
effectiveLibraryIDs = configLibraryIDs
}
dismissals := f.listContinueWatchingDismissals(ctx, store, profileID)
orderedItems := make([]*models.MediaItem, 0, limit)
itemMeta := make(map[string]SectionItemMeta)
// Ebook resume points live in ebook_reader_progress rather than the
// watch-progress store, so reading sections pull from that table and skip
// the next-up handling below.
if continueType == ContinueTypeReading {
orderedItems, err = f.collectContinueProgressItems(ctx, store, profileID, dismissals, continueType, effectiveLibID, effectiveLibraryIDs, filter, limit, orderedItems, itemMeta,
func(pageLimit, offset int) ([]userstore.WatchProgress, error) {
entries, err := f.listEbookContinueWatchingProgress(ctx, userID, profileID, pageLimit, offset)
if err != nil {
return nil, fmt.Errorf("listing ebook reader progress: %w", err)
}
return entries, nil
})
if err != nil {
return SectionWithItems{}, err
}
// Manga chapters are ebook items linked to a series; collapse multiple
// in-progress chapters of the same manga to a single card (keeping the
// most recently read), mirroring the episode→series collapse. Resolve
// the linkage into itemMeta so the shared collapse can group by series.
if len(orderedItems) > 1 {
f.applyMangaChapterSeriesMeta(ctx, orderedItems, itemMeta)
orderedItems = collapseContinueWatchingSeriesCandidates(orderedItems, itemMeta)
}
return SectionWithItems{
ResolvedSection: resolved,
Items: orderedItems,
TotalCount: len(orderedItems),
ItemMeta: itemMeta,
}, nil
}
orderedItems, err = f.collectContinueProgressItems(ctx, store, profileID, dismissals, continueType, effectiveLibID, effectiveLibraryIDs, filter, limit, orderedItems, itemMeta,
func(pageLimit, offset int) ([]userstore.WatchProgress, error) {
entries, err := store.ListProgress(ctx, profileID, "in_progress", pageLimit, offset)
if err != nil {
return nil, fmt.Errorf("listing progress: %w", err)
}
return entries, nil
})
if err != nil {
return SectionWithItems{}, err
}
nextUpMode := ""
if ContinueTypeAllowsNextUp(continueType) && !resolved.SuppressNextUp {
nextUpMode, _ = store.GetSetting(ctx, "next_up_mode")
if nextUpMode == "" {
nextUpMode = "combined"
}
}
if ContinueTypeAllowsNextUp(continueType) && nextUpMode == "combined" {
nextUpItems, nextUpMeta, nextUpErr := f.FetchNextUpItems(ctx, userID, profileID, effectiveLibID, effectiveLibraryIDs, filter, limit)
if nextUpErr != nil {
slog.Error("fetching next-up items", "error", nextUpErr)
} else {
orderedItems = append(orderedItems, nextUpItems...)
maps.Copy(itemMeta, nextUpMeta)
}
}
if nextUpMode == "combined" && len(orderedItems) > 1 {
orderedItems = collapseContinueWatchingSeriesCandidates(orderedItems, itemMeta)
}
if nextUpMode == "combined" && len(orderedItems) > 1 {
sort.SliceStable(orderedItems, func(i, j int) bool {
left := itemMeta[orderedItems[i].ContentID].SortTimestamp
right := itemMeta[orderedItems[j].ContentID].SortTimestamp
return left.After(right)
})
}
if orderedItems == nil {
orderedItems = []*models.MediaItem{}
}
return SectionWithItems{
ResolvedSection: resolved,
Items: orderedItems,
TotalCount: len(orderedItems),
ItemMeta: itemMeta,
}, nil
}
// collectContinueProgressItems pages through in-progress entries from
// listPage, drops dismissed entries, resolves the remainder to media items,
// and appends them to orderedItems until limit is reached, the source is
// exhausted, or continueProgressMaxScanned entries have been scanned. Paging
// past the requested limit matters because dismissal filtering and
// library/access scoping happen after the source query: trimming at the
// source LIMIT would under-fill the section whenever dismissals exist.
func (f *Fetcher) collectContinueProgressItems(
ctx context.Context,
store userstore.UserStore,
profileID string,
dismissals catalog.HomeDismissalIndex,
continueType ContinueType,
libraryID *int,
libraryIDs []int,
filter catalog.AccessFilter,
limit int,
orderedItems []*models.MediaItem,
itemMeta map[string]SectionItemMeta,
listPage func(pageLimit, offset int) ([]userstore.WatchProgress, error),
) ([]*models.MediaItem, error) {
// LIMIT/OFFSET pages over a live, updated_at-ordered source: a progress
// write between page fetches shifts rows across page boundaries, so the
// same content_id can be returned on two consecutive pages. Track what has
// already been appended so a card never appears twice in one section.
seen := make(map[string]struct{}, limit)
for _, item := range orderedItems {
seen[item.ContentID] = struct{}{}
}
for offset := 0; len(orderedItems) < limit && offset < continueProgressMaxScanned; offset += continueProgressPageSize {
pageLimit := continueProgressPageSize
if remaining := continueProgressMaxScanned - offset; remaining < pageLimit {
pageLimit = remaining
}
progressEntries, err := listPage(pageLimit, offset)
if err != nil {
return nil, err
}
if len(progressEntries) == 0 {
break
}
rawProgressCount := len(progressEntries)
progressEntries = dismissals.FilterProgress(progressEntries)
pageItems, pageMeta, err := f.fetchContinueProgressItems(ctx, store, profileID, progressEntries, continueType, libraryID, libraryIDs, filter)
if err != nil {
return nil, err
}
orderedItems = appendUniqueContinueItems(orderedItems, pageItems, seen, limit)
maps.Copy(itemMeta, pageMeta)
if rawProgressCount < pageLimit {
break
}
}
return orderedItems, nil
}
// appendUniqueContinueItems appends pageItems to orderedItems up to limit,
// skipping content IDs already recorded in seen and recording every appended
// ID. Pages come from a live LIMIT/OFFSET scan, so without this guard an item
// shifted across a page boundary by a concurrent progress write would surface
// twice in the same section.
func appendUniqueContinueItems(orderedItems, pageItems []*models.MediaItem, seen map[string]struct{}, limit int) []*models.MediaItem {
for _, item := range pageItems {
if len(orderedItems) >= limit {
break
}
if _, dup := seen[item.ContentID]; dup {
continue
}
seen[item.ContentID] = struct{}{}
orderedItems = append(orderedItems, item)
}
return orderedItems
}
// ebookContinueWatchingProgressQuery lists in-progress ebook resume points for
// Continue Reading. Like the video path (pgstore ListProgress "in_progress"),
// rows the user hid via user_history_hidden_items are excluded while reading
// activity after hidden_before counts again.
var ebookContinueWatchingProgressQuery = fmt.Sprintf(`
SELECT content_id, progress::double precision, updated_at
FROM ebook_reader_progress
WHERE user_id = $1
AND profile_id = $2
AND progress > 0
AND progress < %s
AND NOT EXISTS (
SELECT 1
FROM user_history_hidden_items hhi
WHERE hhi.user_id = ebook_reader_progress.user_id
AND hhi.profile_id = ebook_reader_progress.profile_id
AND hhi.media_item_id = ebook_reader_progress.content_id
AND ebook_reader_progress.updated_at <= hhi.hidden_before
)
ORDER BY updated_at DESC, content_id ASC
LIMIT $3 OFFSET $4
`, catalog.EbookFinishedProgressThresholdSQL)
func (f *Fetcher) listEbookContinueWatchingProgress(ctx context.Context, userID int, profileID string, limit, offset int) ([]userstore.WatchProgress, error) {
if f.pool == nil || userID <= 0 || profileID == "" || limit <= 0 {
return nil, nil
}
rows, err := f.pool.Query(ctx, ebookContinueWatchingProgressQuery, userID, profileID, limit, offset)
if err != nil {
return nil, err
}
defer rows.Close()
entries := make([]userstore.WatchProgress, 0)
for rows.Next() {
var contentID string
var progress float64
var updatedAt time.Time
if err := rows.Scan(&contentID, &progress, &updatedAt); err != nil {
return nil, err
}
entries = append(entries, userstore.WatchProgress{
MediaItemID: contentID,
PositionSeconds: progress,
DurationSeconds: 1,
UpdatedAt: updatedAt.UTC().Format(time.RFC3339),
})
}
if err := rows.Err(); err != nil {
return nil, err
}
return entries, nil
}
const (
continueProgressPageSize = 100
continueProgressMaxScanned = 1000
)
func (f *Fetcher) fetchContinueProgressItems(ctx context.Context, store userstore.UserStore, profileID string, entries []userstore.WatchProgress, continueType ContinueType, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]*models.MediaItem, map[string]SectionItemMeta, error) {
if len(entries) == 0 {
return nil, map[string]SectionItemMeta{}, nil
}
contentIDs := make([]string, 0, len(entries))
for _, entry := range entries {
contentIDs = append(contentIDs, entry.MediaItemID)
}
mediaItems, err := f.fetchItemsByContentIDs(ctx, contentIDs, libraryID, libraryIDs, filter)
if err != nil {
return nil, nil, err
}
episodeItems, episodeMeta, err := f.fetchEpisodeTargetsByContentIDs(ctx, contentIDs, libraryID, libraryIDs, filter)
if err != nil {
return nil, nil, err
}
itemByID := make(map[string]*models.MediaItem, len(mediaItems)+len(episodeItems))
for _, item := range mediaItems {
itemByID[item.ContentID] = item
}
for _, item := range episodeItems {
itemByID[item.ContentID] = item
}
matchingEntries := make([]userstore.WatchProgress, 0, len(entries))
hasEpisodeEntries := false
for _, entry := range entries {
item, ok := itemByID[entry.MediaItemID]
if !ok || !ContinueTypeMatchesItem(continueType, item.Type) {
continue
}
if item.Type == "episode" {
hasEpisodeEntries = true
}
matchingEntries = append(matchingEntries, entry)
}
if ContinueTypeAllowsNextUp(continueType) && hasEpisodeEntries {
supersededEpisodeProgress, err := f.progressFilter.SupersededEpisodeProgressIDs(ctx, store, profileID, matchingEntries)
if err != nil {
return nil, nil, err
}
matchingEntries = catalog.FilterSupersededProgress(matchingEntries, supersededEpisodeProgress)
}
itemMeta := make(map[string]SectionItemMeta, len(matchingEntries))
orderedItems := make([]*models.MediaItem, 0, len(matchingEntries))
for _, entry := range matchingEntries {
item, ok := itemByID[entry.MediaItemID]
if !ok {
continue
}
meta := episodeMeta[entry.MediaItemID]
position := entry.PositionSeconds
duration := entry.DurationSeconds
meta.PositionSeconds = &position
meta.DurationSeconds = &duration
progressUpdatedAt := entry.UpdatedAt
meta.ProgressUpdatedAt = &progressUpdatedAt
meta.ItemSource = "in_progress"
if updatedAt, parseErr := time.Parse(time.RFC3339, entry.UpdatedAt); parseErr == nil {
meta.SortTimestamp = updatedAt
}
itemMeta[entry.MediaItemID] = meta
orderedItems = append(orderedItems, item)
}
return orderedItems, itemMeta, nil
}
func (f *Fetcher) fetchNextUpSection(ctx context.Context, resolved ResolvedSection, libraryID *int, libraryIDs []int, userID int, profileID string, filter catalog.AccessFilter) (SectionWithItems, error) {
emptyResult := SectionWithItems{
ResolvedSection: resolved,
Items: []*models.MediaItem{},
TotalCount: 0,
ItemMeta: map[string]SectionItemMeta{},
}
if f.StoreProvider == nil || userID <= 0 || profileID == "" {
return emptyResult, nil
}
store, err := f.StoreProvider.ForUser(ctx, userID)
if err != nil {
return SectionWithItems{}, fmt.Errorf("getting user store: %w", err)
}
// Only resolve if user preference is "separate"
nextUpMode, _ := store.GetSetting(ctx, "next_up_mode")
if nextUpMode != "separate" {
return emptyResult, nil
}
limit := resolved.ItemLimit
if limit <= 0 {
limit = 20
}
items, meta, err := f.FetchNextUpItems(ctx, userID, profileID, libraryID, libraryIDs, filter, limit)
if err != nil {
return SectionWithItems{}, err
}
if items == nil {
items = []*models.MediaItem{}
}
if meta == nil {
meta = map[string]SectionItemMeta{}
}
return SectionWithItems{
ResolvedSection: resolved,
Items: items,
TotalCount: len(items),
ItemMeta: meta,
}, nil
}
// FetchNextUpItems returns the next unwatched episode per series for the given user profile.
// It delegates to NextUpRepository for the optimized LATERAL JOIN query, then resolves
// the full MediaItem objects via fetchEpisodeTargetsByContentIDs.
func (f *Fetcher) FetchNextUpItems(ctx context.Context, userID int, profileID string, libraryID *int, libraryIDs []int, filter catalog.AccessFilter, limit int) ([]*models.MediaItem, map[string]SectionItemMeta, error) {
if f.NextUpRepo == nil {
return nil, nil, nil
}
results, err := f.NextUpRepo.ListNextUp(ctx, catalog.NextUpQuery{
UserID: userID,
ProfileID: profileID,
LibraryID: libraryID,
LibraryIDs: libraryIDs,
Limit: limit,
})
if err != nil {
return nil, nil, err
}
if len(results) == 0 {
return nil, nil, nil
}
results = f.filterNextUpDismissals(ctx, userID, profileID, results)
if len(results) == 0 {
return nil, nil, nil
}
contentIDs := make([]string, len(results))
resultByID := make(map[string]catalog.NextUpResult, len(results))
for i, res := range results {
contentIDs[i] = res.ContentID
resultByID[res.ContentID] = res
}
episodeItems, episodeMeta, err := f.fetchEpisodeTargetsByContentIDs(ctx, contentIDs, libraryID, libraryIDs, filter)
if err != nil {
return nil, nil, fmt.Errorf("fetching episode targets for next-up: %w", err)
}
itemByID := make(map[string]*models.MediaItem, len(episodeItems))
for _, item := range episodeItems {
itemByID[item.ContentID] = item
}
meta := make(map[string]SectionItemMeta, len(results))
orderedItems := make([]*models.MediaItem, 0, len(results))
for _, res := range results {
m, ok := episodeMeta[res.ContentID]
if !ok {
continue
}
m.ItemSource = "next_up"
m.SortTimestamp = res.CompletedAt
meta[res.ContentID] = m
item, ok := itemByID[res.ContentID]
if !ok {
continue
}
orderedItems = append(orderedItems, item)
}
return orderedItems, meta, nil
}
// fetchNextInSeriesSection surfaces the next unstarted audiobook per series
// the profile has finished a book of. Candidates come from AudiobookNextRepo
// with library scoping pushed into the SQL — otherwise finished series whose
// next book lives in another library would consume the candidate limit and
// starve a library-scoped section. Content-rating access filtering happens
// while resolving candidate IDs to items, so the candidate query still
// over-fetches to compensate for that post-resolution trim.
func (f *Fetcher) fetchNextInSeriesSection(ctx context.Context, resolved ResolvedSection, libraryID *int, libraryIDs []int, userID int, profileID string, filter catalog.AccessFilter) (SectionWithItems, error) {
emptyResult := SectionWithItems{
ResolvedSection: resolved,
Items: []*models.MediaItem{},
TotalCount: 0,
ItemMeta: map[string]SectionItemMeta{},
}
if f.AudiobookNextRepo == nil || userID <= 0 || profileID == "" {
return emptyResult, nil
}
limit := resolved.ItemLimit
if limit <= 0 {
limit = 20
}
candidateLimit := limit * 3
if candidateLimit > 100 {
candidateLimit = 100
}
results, err := f.AudiobookNextRepo.ListNextInSeries(ctx, catalog.NextInSeriesQuery{
UserID: userID,
ProfileID: profileID,
Limit: candidateLimit,
LibraryID: libraryID,
LibraryIDs: effectiveFetchLibraryIDs(libraryIDs, filter),
DisabledLibraryIDs: filter.DisabledLibraryIDs,
})
if err != nil {
return SectionWithItems{}, err
}
if len(results) == 0 {
return emptyResult, nil
}
contentIDs := make([]string, len(results))
for i, res := range results {
contentIDs[i] = res.ContentID
}
items, err := f.fetchItemsByContentIDs(ctx, contentIDs, libraryID, libraryIDs, filter)
if err != nil {
return SectionWithItems{}, fmt.Errorf("fetching next-in-series items: %w", err)
}
itemByID := make(map[string]*models.MediaItem, len(items))
for _, item := range items {
itemByID[item.ContentID] = item
}
meta := make(map[string]SectionItemMeta, len(results))
ordered := make([]*models.MediaItem, 0, limit)
for _, res := range results {
if len(ordered) >= limit {
break
}
item, ok := itemByID[res.ContentID]
if !ok {
continue
}
m := SectionItemMeta{
SeriesTitle: res.SeriesName,
ItemSource: "next_in_series",
SortTimestamp: res.LastFinishedAt,
}
if res.SeriesIndex != nil {
if n := int(*res.SeriesIndex); float64(n) == *res.SeriesIndex {
m.Badges = append(m.Badges, fmt.Sprintf("Book %d", n))
}
}
meta[res.ContentID] = m
ordered = append(ordered, item)
}
return SectionWithItems{
ResolvedSection: resolved,
Items: ordered,
TotalCount: len(ordered),
ItemMeta: meta,
}, nil
}
func collapseContinueWatchingSeriesCandidates(items []*models.MediaItem, meta map[string]SectionItemMeta) []*models.MediaItem {
selectedBySeries := make(map[string]int)
result := make([]*models.MediaItem, 0, len(items))
for _, item := range items {
if item == nil {
continue
}
itemMeta := meta[item.ContentID]
if itemMeta.SeriesID == nil || *itemMeta.SeriesID == "" {
result = append(result, item)
continue
}
seriesID := *itemMeta.SeriesID
selectedIndex, ok := selectedBySeries[seriesID]
if !ok {
selectedBySeries[seriesID] = len(result)
result = append(result, item)
continue
}
current := result[selectedIndex]
if continueWatchingCandidatePreferred(itemMeta, meta[current.ContentID]) {
result[selectedIndex] = item
}
}
return result
}
func continueWatchingCandidatePreferred(candidate, current SectionItemMeta) bool {
candidateInProgress := candidate.ItemSource == "in_progress"
currentInProgress := current.ItemSource == "in_progress"
if candidateInProgress != currentInProgress {
return candidateInProgress
}
if candidateInProgress {
if !candidate.SortTimestamp.Equal(current.SortTimestamp) {
return candidate.SortTimestamp.After(current.SortTimestamp)
}
return episodeOrdinal(candidate) > episodeOrdinal(current)
}
return candidate.SortTimestamp.After(current.SortTimestamp)
}
func episodeOrdinal(meta SectionItemMeta) int {
season := 0
if meta.SeasonNumber != nil {
season = *meta.SeasonNumber
}
episode := 0
if meta.EpisodeNumber != nil {
episode = *meta.EpisodeNumber
}
return season*100000 + episode
}
func (f *Fetcher) listContinueWatchingDismissals(ctx context.Context, store userstore.UserStore, profileID string) catalog.HomeDismissalIndex {
dismissals, err := store.ListHomeDismissals(ctx, profileID, userstore.HomeSurfaceContinueWatching)
if err != nil {
slog.Error("listing continue watching dismissals", "profile_id", profileID, "error", err)
return catalog.HomeDismissalIndex{}
}
return catalog.NewHomeDismissalIndex(dismissals)
}
func (f *Fetcher) filterNextUpDismissals(ctx context.Context, userID int, profileID string, results []catalog.NextUpResult) []catalog.NextUpResult {
if len(results) == 0 || f.StoreProvider == nil || userID <= 0 || profileID == "" {
return results
}
store, err := f.StoreProvider.ForUser(ctx, userID)
if err != nil {
slog.Error("getting user store for next-up dismissals", "profile_id", profileID, "error", err)
return results
}
dismissals, err := store.ListHomeDismissals(ctx, profileID, userstore.HomeSurfaceNextUp)
if err != nil {
slog.Error("listing next-up dismissals", "profile_id", profileID, "error", err)
return results
}
if len(dismissals) == 0 {
return results
}
dismissedIDs := make(map[string]struct{}, len(dismissals))
for _, dismissal := range dismissals {
dismissedIDs[dismissal.MediaItemID] = struct{}{}
}
filtered := make([]catalog.NextUpResult, 0, len(results))
for _, result := range results {
if _, dismissed := dismissedIDs[result.ContentID]; dismissed {
continue
}
filtered = append(filtered, result)
}
return filtered
}
// FetchAll runs all section queries in parallel and returns sections with items.
func (f *Fetcher) FetchAll(ctx context.Context, resolved []ResolvedSection, libraryID *int, libraryIDs []int, userID int, profileID string, filter catalog.AccessFilter) []SectionWithItems {
start := time.Now()
results := fetchAllWithRunner(ctx, resolved, fetchAllMaxConcurrency, func(ctx context.Context, sec ResolvedSection) (SectionWithItems, error) {
return f.FetchOne(ctx, sec, libraryID, libraryIDs, userID, profileID, filter)
})
duration := time.Since(start)
if duration >= slowAggregateFetchThreshold {
attrs := []any{
"section_count", len(resolved),
"duration_ms", duration.Milliseconds(),
}
if libraryID != nil {
attrs = append(attrs, "library_id", *libraryID)
}
if libraryIDs != nil {
attrs = append(attrs, "library_ids", append([]int(nil), libraryIDs...))
}
slog.Warn("slow aggregate section fetch", attrs...)
}
return results
}
type sectionFetchRunner func(context.Context, ResolvedSection) (SectionWithItems, error)
func fetchAllWithRunner(ctx context.Context, resolved []ResolvedSection, maxConcurrency int, runner sectionFetchRunner) []SectionWithItems {
if maxConcurrency <= 0 {
maxConcurrency = 1
}
results := make([]SectionWithItems, len(resolved))
sem := make(chan struct{}, maxConcurrency)
var wg sync.WaitGroup
for i, sec := range resolved {
i, sec := i, sec
wg.Add(1)
go func() {
defer wg.Done()
sem <- struct{}{}
defer func() { <-sem }()
result, err := runner(ctx, sec)
if err != nil {
slog.Error("fetching section items", "section_id", sec.ID, "type", sec.SectionType, "error", err)
result = SectionWithItems{
ResolvedSection: sec,
Items: []*models.MediaItem{},
}
}
results[i] = result
}()
}
wg.Wait()
return results
}
// FetchItemsByContentIDs resolves content IDs to full MediaItem objects,
// applying the provided access filter across libraries.
func (f *Fetcher) FetchItemsByContentIDs(ctx context.Context, contentIDs []string, filter catalog.AccessFilter) ([]*models.MediaItem, error) {
return f.fetchItemsByContentIDs(ctx, contentIDs, nil, nil, filter)
}
// FetchEpisodesByContentIDs resolves episode content IDs to MediaItem objects
// with their parent series metadata. Returns the items and per-episode metadata
// (series title, season/episode numbers). Non-episode content IDs are silently ignored.
func (f *Fetcher) FetchEpisodesByContentIDs(ctx context.Context, contentIDs []string, filter catalog.AccessFilter) ([]*models.MediaItem, map[string]SectionItemMeta, error) {
return f.fetchEpisodeTargetsByContentIDs(ctx, contentIDs, nil, nil, filter)
}
// ListOverlaySummaries batches file lookups for section cards and derives the
// compact overlay summary per content ID.
func (f *Fetcher) ListOverlaySummaries(ctx context.Context, contentIDs []string, filter catalog.AccessFilter) (map[string]*models.OverlaySummary, error) {
summaries := make(map[string]*models.OverlaySummary, len(contentIDs))
if len(contentIDs) == 0 {
return summaries, nil
}
rows, err := f.pool.Query(ctx, `
SELECT content_id, episode_id, file_path, resolution, codec_audio, audio_tracks, hdr, video_tracks,
codec_video, audio_channels, container, subtitle_tracks, external_subtitles, edition_key
FROM media_files
WHERE (content_id = ANY($1) OR episode_id = ANY($1)) AND missing_since IS NULL
ORDER BY content_id ASC, episode_id ASC, id ASC
`, contentIDs)
if err != nil {
return nil, fmt.Errorf("querying overlay summaries: %w", err)
}
defer rows.Close()
requested := make(map[string]struct{}, len(contentIDs))
for _, contentID := range contentIDs {
requested[contentID] = struct{}{}
}
grouped := make(map[string][]*models.MediaFile, len(contentIDs))
for rows.Next() {
var contentID string
var episodeID *string
var filePath string
var resolution *string
var codecAudio *string
var audioTracksJSON []byte
var hdr bool
var videoTracksJSON []byte
var codecVideo *string
var audioChannels *int
var container *string
var subtitleTracksJSON []byte
var externalSubtitlesJSON []byte
var editionKey *string
if err := rows.Scan(
&contentID, &episodeID, &filePath, &resolution, &codecAudio, &audioTracksJSON, &hdr, &videoTracksJSON,
&codecVideo, &audioChannels, &container, &subtitleTracksJSON, &externalSubtitlesJSON, &editionKey,
); err != nil {
return nil, fmt.Errorf("scanning overlay summary row: %w", err)
}
file := &models.MediaFile{
ContentID: contentID,
FilePath: filePath,
HDR: hdr,
}
if episodeID != nil {
file.EpisodeID = *episodeID
}
if resolution != nil {
file.Resolution = *resolution
}
if codecAudio != nil {
file.CodecAudio = *codecAudio
}
if codecVideo != nil {
file.CodecVideo = *codecVideo
}
if audioChannels != nil {
file.AudioChannels = *audioChannels
}
if container != nil {
file.Container = *container
}
if editionKey != nil {
file.EditionKey = *editionKey
}
if len(audioTracksJSON) > 0 {
if err := json.Unmarshal(audioTracksJSON, &file.AudioTracks); err != nil {
return nil, fmt.Errorf("unmarshaling overlay audio tracks: %w", err)
}
}
if len(videoTracksJSON) > 0 {
if err := json.Unmarshal(videoTracksJSON, &file.VideoTracks); err != nil {
return nil, fmt.Errorf("unmarshaling overlay video tracks: %w", err)
}
}
if len(subtitleTracksJSON) > 0 {
if err := json.Unmarshal(subtitleTracksJSON, &file.SubtitleTracks); err != nil {
return nil, fmt.Errorf("unmarshaling overlay subtitle tracks: %w", err)
}
}
if len(externalSubtitlesJSON) > 0 {
if err := json.Unmarshal(externalSubtitlesJSON, &file.ExternalSubtitles); err != nil {
return nil, fmt.Errorf("unmarshaling overlay external subtitles: %w", err)
}
}
groupKey := contentID
if episodeID != nil {
if _, ok := requested[*episodeID]; ok {
groupKey = *episodeID
}
}
grouped[groupKey] = append(grouped[groupKey], file)
}
if err := rows.Err(); err != nil {
return nil, fmt.Errorf("iterating overlay summary rows: %w", err)
}
for contentID, files := range grouped {
files = catalog.FilterMediaFilesByAccess(files, filter)
if summary := overlays.BuildSummary(files); summary != nil {
summaries[contentID] = summary
}
}
return summaries, nil
}
// userAgnosticSectionFetch is the signature shared by every fetch helper whose
// membership is identical for all profiles with the same access scope: no
// userID, no profileID. The signature is the enforcement mechanism for the
// shared resolved-list cache — a helper cannot start reading per-profile state
// without changing its signature, dropping out of userAgnosticSectionFetcher,
// and forcing an explicit new cacheability decision at compile time.
type userAgnosticSectionFetch func(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]*models.MediaItem, int, error)
// userAgnosticSectionFetcher returns the fetch function for section types whose
// shared item list may be cached across profiles, or nil for every other type.
// It is the single source of truth for that decision: fetchSection dispatches
// through it AND isCacheableSectionType derives cache eligibility from it, so
// the two can never disagree.
//
// Deliberately absent despite having the user-agnostic signature:
// - SectionRandom — user-agnostic but must reshuffle per request;
// - SectionEditorialSpotlight — has its own candidate cache + rotation and a
// dedicated FetchOne branch;
// - SectionGenre / SectionCustomFilter — fetchFiltered shares the signature,
// but a user-supplied QueryDefinition can carry personalized rules, so
// cacheability needs the dynamic IsPersonalized check in
// isCacheableSectionType (dispatch stays in fetchSection's switch);
// - SectionCollection — user vs library collections are decided by config,
// handled dynamically in isCacheableSectionType.
func (f *Fetcher) userAgnosticSectionFetcher(t SectionType) userAgnosticSectionFetch {
switch t {
case SectionRecentlyAdded:
return f.fetchRecentlyAdded
case SectionRecentlyReleased:
return f.fetchRecentlyReleased
case SectionCriticallyAcclaimed:
return f.fetchCriticallyAcclaimed
case SectionAwardWinners:
// fetchAwardWinners derives everything from the section config; adapt
// it to the shared signature.
return func(ctx context.Context, s ResolvedSection, _ *int, _ []int, _ catalog.AccessFilter) ([]*models.MediaItem, int, error) {
return f.fetchAwardWinners(ctx, s)
}
case SectionFormatShowcase:
return f.fetchFormatShowcase
case SectionSeasonalThemed:
return f.fetchSeasonalThemed
case SectionMoodCollection:
return f.fetchMoodCollection
case SectionTrendingOnServer:
return f.fetchTrending
case SectionNewToLibrary:
return f.fetchNewToLibrary
case SectionMostWatched:
return f.fetchMostWatched
case SectionTrendingDiscover:
return f.fetchTrendingDiscover
case SectionAdminCuratedList:
return f.fetchAdminCuratedList
default:
return nil
}
}
func (f *Fetcher) fetchSection(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, userID int, profileID string, filter catalog.AccessFilter) ([]*models.MediaItem, int, error) {
if fetch := f.userAgnosticSectionFetcher(s.SectionType); fetch != nil {
return fetch(ctx, s, libraryID, libraryIDs, filter)
}
switch s.SectionType {
case SectionGenre, SectionCustomFilter:
return f.fetchFiltered(ctx, s, libraryID, libraryIDs, filter)
case SectionRandom:
return f.fetchRandom(ctx, s, libraryID, libraryIDs, filter)
case SectionCollection:
return f.fetchCollection(ctx, s, libraryID, libraryIDs, userID, profileID, filter)
case SectionRecommendedForYou, SectionBecauseYouWatched, SectionSimilarUsersLiked, SectionTasteMatch:
return f.fetchRecommendationSection(ctx, s, libraryID, libraryIDs, userID, profileID, filter)
case SectionHiddenGems:
return f.fetchHiddenGems(ctx, s, libraryID, libraryIDs, userID, profileID, filter)
case SectionForgottenFavorites:
return f.fetchForgottenFavorites(ctx, s, libraryID, libraryIDs, userID, profileID, filter)
case SectionEditorialSpotlight:
return f.fetchEditorialSpotlight(ctx, s, libraryID, libraryIDs, filter)
case SectionProfileActivityFeed:
return f.fetchProfileActivityFeed(ctx, s, libraryID, libraryIDs, profileID, filter)
default:
// Profile-scoped types (continue_watching, watchlist, favorites)
// will be wired later when user store integration is added.
return nil, 0, fmt.Errorf("unsupported section type: %s", s.SectionType)
}
}
func (f *Fetcher) fetchCollection(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, userID int, profileID string, filter catalog.AccessFilter) ([]*models.MediaItem, int, error) {
cfg := ParseCollectionConfig(s.Config)
// User collection path: resolve from the user's personal store.
if userCollID := strings.TrimSpace(cfg.UserCollectionID); userCollID != "" {
return f.fetchUserCollection(ctx, s, libraryID, libraryIDs, userID, profileID, filter, userCollID)
}
// Library collection path (existing behaviour).
if f.CollectionRepo == nil {
return nil, 0, fmt.Errorf("collection sections require a collection repository")
}
if strings.TrimSpace(cfg.LibraryCollectionID) == "" {
return []*models.MediaItem{}, 0, nil
}
if _, err := f.CollectionRepo.GetByID(ctx, cfg.LibraryCollectionID); err != nil {
return nil, 0, fmt.Errorf("loading library collection: %w", err)
}
collectionItems, err := f.CollectionRepo.ListItems(ctx, cfg.LibraryCollectionID)
if err != nil {
return nil, 0, fmt.Errorf("listing collection items: %w", err)
}
if len(collectionItems) == 0 {
return []*models.MediaItem{}, 0, nil
}
limit := s.ItemLimit
if limit <= 0 || limit > len(collectionItems) {
limit = len(collectionItems)
}
contentIDs := make([]string, 0, limit)
for _, item := range collectionItems[:limit] {
contentIDs = append(contentIDs, item.MediaItemID)
}
items, err := f.fetchItemsByContentIDs(ctx, contentIDs, libraryID, libraryIDs, filter)
if err != nil {
return nil, 0, err
}
itemByID := make(map[string]*models.MediaItem, len(items))
for _, item := range items {
itemByID[item.ContentID] = item
}
orderedItems := make([]*models.MediaItem, 0, len(contentIDs))
for _, contentID := range contentIDs {
item, ok := itemByID[contentID]
if !ok {
continue
}
orderedItems = append(orderedItems, item)
}
return orderedItems, len(orderedItems), nil
}
// fetchUserCollection resolves a personal (profile-scoped) user collection.
// For smart collections it delegates to fetchFiltered; for manual ones it
// fetches the stored item list and looks up the media items by content ID.
func (f *Fetcher) fetchUserCollection(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, userID int, profileID string, filter catalog.AccessFilter, collectionID string) ([]*models.MediaItem, int, error) {
if f.StoreProvider == nil || userID <= 0 || profileID == "" {
return []*models.MediaItem{}, 0, nil
}
store, err := f.StoreProvider.ForUser(ctx, userID)
if err != nil {
return nil, 0, fmt.Errorf("getting user store for collection: %w", err)
}
collection, err := store.GetCollection(ctx, collectionID)
if err != nil {
return nil, 0, fmt.Errorf("loading user collection: %w", err)
}
// Verify the requesting profile can access this collection.
if collection.ProfileID != profileID && collection.CreatorProfileID != profileID {
canAccess := false
for _, allowed := range collection.AllowedProfileIDs {
if allowed == profileID {
canAccess = true
break
}
}
if !canAccess {
return []*models.MediaItem{}, 0, nil
}
}
// Smart / live-query collection: parse the query definition and use the
// filtered fetch path, mirroring resolveUserCollectionSource in catalog_resolver.
isSmartType := strings.EqualFold(strings.TrimSpace(collection.CollectionType), "smart")
if isSmartType {
var qd catalog.QueryDefinition
if err := json.Unmarshal([]byte(collection.QueryDefinition), &qd); err != nil {
return nil, 0, fmt.Errorf("parsing user collection query_definition: %w", err)
}
if strings.TrimSpace(collection.DisplayQueryDefinition) == "" {
qd = qd.Normalize()
// Build a synthetic resolved section with the query definition as config.
cfgBytes, _ := json.Marshal(qd)
synth := ResolvedSection{
ID: s.ID,
SectionType: SectionCustomFilter,
Title: s.Title,
Featured: s.Featured,
ItemLimit: s.ItemLimit,
Config: cfgBytes,
Position: s.Position,
}
return f.fetchFiltered(ctx, synth, libraryID, libraryIDs, filter)
}
qd = applySectionLibraryScopeToQuery(qd.Normalize(), libraryID, libraryIDs)
qd = catalog.ApplySmartCollectionItemLimit(qd)
limit := catalog.DefaultSmartCollectionItemLimit
if qd.Limit != nil && *qd.Limit > 0 {
limit = *qd.Limit
}
executor := &catalog.QueryExecutor{Pool: f.pool}
items, _, err := executor.Preview(ctx, qd, filter, limit)
if err != nil {
return nil, 0, err
}
displayAccess := filter
displayAccess.UserID = userID
displayAccess.ProfileID = profileID
items, err = catalog.FilterCollectionItemsByDisplayQuery(ctx, f.pool, items, collection.DisplayQueryDefinition, displayAccess)
if err != nil {
return nil, 0, err
}
items, total := limitUserCollectionSectionItems(items, s.ItemLimit)
return items, total, nil
}
// Exact collection: fetch stored items.
collectionItems, err := store.ListCollectionItems(ctx, collectionID)
if err != nil {
return nil, 0, fmt.Errorf("listing user collection items: %w", err)
}
if len(collectionItems) == 0 {
return []*models.MediaItem{}, 0, nil
}
contentIDs := make([]string, 0, len(collectionItems))
for _, item := range collectionItems {
contentIDs = append(contentIDs, item.MediaItemID)
}
items, err := f.fetchItemsByContentIDs(ctx, contentIDs, libraryID, libraryIDs, filter)
if err != nil {
return nil, 0, err
}
itemByID := make(map[string]*models.MediaItem, len(items))
for _, item := range items {
itemByID[item.ContentID] = item
}
orderedItems := make([]*models.MediaItem, 0, len(contentIDs))
for _, contentID := range contentIDs {
item, ok := itemByID[contentID]
if !ok {
continue
}
orderedItems = append(orderedItems, item)
}
displayAccess := filter
displayAccess.UserID = userID
displayAccess.ProfileID = profileID
orderedItems, err = catalog.FilterCollectionItemsByDisplayQuery(ctx, f.pool, orderedItems, collection.DisplayQueryDefinition, displayAccess)
if err != nil {
return nil, 0, err
}
orderedItems, total := limitUserCollectionSectionItems(orderedItems, s.ItemLimit)
return orderedItems, total, nil
}
func limitUserCollectionSectionItems(items []*models.MediaItem, limit int) ([]*models.MediaItem, int) {
total := len(items)
if limit > 0 && limit < len(items) {
return items[:limit], total
}
return items, total
}
func (f *Fetcher) fetchRecommendationSection(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, userID int, profileID string, filter catalog.AccessFilter) ([]*models.MediaItem, int, error) {
if f.RecommendationReader == nil {
return []*models.MediaItem{}, 0, nil // graceful degradation
}
cfg := parseRecommendationSectionConfig(s.Config)
var scoredItems []recommendations.ScoredItem
switch s.SectionType {
case SectionRecommendedForYou:
row, err := f.RecommendationReader.GetForYouMain(ctx, userID, profileID, s.ItemLimit, filter)
if err != nil || row == nil {
return []*models.MediaItem{}, 0, err
}
scoredItems = row.Items
case SectionBecauseYouWatched:
items, err := f.RecommendationReader.GetBecauseYouWatched(ctx, userID, profileID, cfg.SourceItemID, s.ItemLimit, filter)
if err != nil {
return []*models.MediaItem{}, 0, err
}
scoredItems = items
case SectionSimilarUsersLiked:
items, err := f.RecommendationReader.GetSimilarUsersLiked(ctx, userID, profileID, s.ItemLimit, filter)
if err != nil {
return []*models.MediaItem{}, 0, err
}
scoredItems = items
case SectionTasteMatch:
if strings.TrimSpace(cfg.Genre) == "" {
return []*models.MediaItem{}, 0, nil
}
row, err := f.RecommendationReader.GetTasteMatchRow(ctx, userID, profileID, cfg.Genre, s.ItemLimit, filter)
if err != nil || row == nil {
return []*models.MediaItem{}, 0, err
}
scoredItems = row.Items
default:
return []*models.MediaItem{}, 0, nil
}
if len(scoredItems) == 0 {
return []*models.MediaItem{}, 0, nil
}
// Resolve item IDs to full MediaItem objects
itemIDs := make([]string, len(scoredItems))
for i, item := range scoredItems {
itemIDs[i] = item.MediaItemID
}
mediaItems, err := f.fetchItemsByContentIDs(ctx, itemIDs, libraryID, libraryIDs, filter)
if err != nil {
return nil, 0, err
}
orderedItems := orderMediaItems(mediaItems, itemIDs)
return orderedItems, len(orderedItems), nil
}
func (f *Fetcher) fetchHiddenGems(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, userID int, profileID string, filter catalog.AccessFilter) ([]*models.MediaItem, int, error) {
var p recipes.HiddenGemsParams
if len(s.Config) > 0 {
_ = json.Unmarshal(s.Config, &p)
}
if p.MinRating == 0 {
p.MinRating = 7.5
}
// Require a valid user/profile to exclude watched items.
if userID <= 0 || profileID == "" {
return []*models.MediaItem{}, 0, nil
}
repo := catalog.NewDiscoveryRepository(f.pool)
items, err := repo.ListUnplayedHighRated(ctx, catalog.UnplayedFilter{
MinRating: p.MinRating,
Limit: s.ItemLimit,
UserID: userID,
ProfileID: profileID,
Filter: filter,
})
if err != nil {
return nil, 0, err
}
return items, len(items), nil
}
func (f *Fetcher) fetchCriticallyAcclaimed(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]*models.MediaItem, int, error) {
var p recipes.CriticallyAcclaimedParams
if len(s.Config) > 0 {
_ = json.Unmarshal(s.Config, &p)
}
if p.MinScore == 0 {
p.MinScore = 8.0
}
repo := catalog.NewDiscoveryRepository(f.pool)
items, err := repo.ListByRatingThreshold(ctx, catalog.RatingFilter{
Min: p.MinScore,
Limit: s.ItemLimit,
LibraryID: libraryID,
Filter: filter,
})
if err != nil {
return nil, 0, err
}
return items, len(items), nil
}
func (f *Fetcher) fetchAwardWinners(_ context.Context, s ResolvedSection) ([]*models.MediaItem, int, error) {
// TODO(awards-data): wire up once award metadata exists
return []*models.MediaItem{}, 0, nil
}
func (f *Fetcher) fetchForgottenFavorites(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, userID int, profileID string, filter catalog.AccessFilter) ([]*models.MediaItem, int, error) {
var p recipes.ForgottenFavoritesParams
if len(s.Config) > 0 {
_ = json.Unmarshal(s.Config, &p)
}
if p.LookbackDays <= 0 {
p.LookbackDays = 365
}
// Require a valid user/profile to check watch history.
if userID <= 0 || profileID == "" {
return []*models.MediaItem{}, 0, nil
}
repo := catalog.NewDiscoveryRepository(f.pool)
items, err := repo.ListForgottenFavorites(ctx, catalog.ForgottenFavoritesFilter{
LookbackDays: p.LookbackDays,
Limit: s.ItemLimit,
UserID: userID,
ProfileID: profileID,
Filter: filter,
})
if err != nil {
return nil, 0, err
}
return items, len(items), nil
}
// fetchFormatShowcase returns items that have at least one media file matching the
// requested stream format. Filtering is done against media_files columns:
// - 4k: resolution IN ('4k','uhd','2160p')
// - dolby_vision: EXISTS a video_tracks element with a non-empty dolby_vision field
// - hdr: hdr = true
//
// When no format is supplied all items with any of the above formats are returned.
func (f *Fetcher) fetchFormatShowcase(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]*models.MediaItem, int, error) {
var p recipes.FormatShowcaseParams
if len(s.Config) > 0 {
_ = json.Unmarshal(s.Config, &p)
}
// Build the per-format EXISTS subquery condition against media_files.
var formatCond string
switch p.Format {
case "4k":
formatCond = `EXISTS (
SELECT 1 FROM media_files mf
WHERE (mf.content_id = mi.content_id OR mf.episode_id = mi.content_id)
AND mf.missing_since IS NULL
AND LOWER(mf.resolution) IN ('4k','uhd','2160p')
)`
case "dolby_vision":
formatCond = `EXISTS (
SELECT 1 FROM media_files mf
WHERE (mf.content_id = mi.content_id OR mf.episode_id = mi.content_id)
AND mf.missing_since IS NULL
AND EXISTS (
SELECT 1 FROM jsonb_array_elements(COALESCE(mf.video_tracks, '[]'::jsonb)) AS vt
WHERE vt->>'dolby_vision' IS NOT NULL AND vt->>'dolby_vision' != ''
)
)`
case "hdr":
formatCond = `EXISTS (
SELECT 1 FROM media_files mf
WHERE (mf.content_id = mi.content_id OR mf.episode_id = mi.content_id)
AND mf.missing_since IS NULL
AND mf.hdr = true
)`
default:
// No format specified: return items matching any of the above formats.
formatCond = `EXISTS (
SELECT 1 FROM media_files mf
WHERE (mf.content_id = mi.content_id OR mf.episode_id = mi.content_id)
AND mf.missing_since IS NULL
AND (
LOWER(mf.resolution) IN ('4k','uhd','2160p')
OR mf.hdr = true
OR EXISTS (
SELECT 1 FROM jsonb_array_elements(COALESCE(mf.video_tracks, '[]'::jsonb)) AS vt
WHERE vt->>'dolby_vision' IS NOT NULL AND vt->>'dolby_vision' != ''
)
)
)`
}
var conditions []string
var args []any
argIdx := 1
conditions = append(conditions, formatCond)
fromClause, libConditions, libArgs, newArgIdx := buildLibraryScope(libraryID, libraryIDs, nil, filter.DisabledLibraryIDs, argIdx)
conditions = append(conditions, libConditions...)
args = append(args, libArgs...)
argIdx = newArgIdx
catalog.ApplySectionAccessFilter("mi", filter, &conditions, &args, &argIdx)
conditions = append(conditions, catalog.MangaChapterExclusionWhere("mi"))
whereClause := "WHERE " + strings.Join(conditions, " AND ")
limit := s.ItemLimit
if limit <= 0 {
limit = 20
}
query := fmt.Sprintf(
`SELECT %s FROM %s %s ORDER BY mi.rating_imdb DESC NULLS LAST, mi.content_id ASC LIMIT $%d`,
itemColumns("mi"), fromClause, whereClause, argIdx,
)
args = append(args, limit)
rows, err := f.pool.Query(ctx, query, args...)
if err != nil {
return nil, 0, fmt.Errorf("fetching format showcase: %w", err)
}
defer rows.Close()
items, err := scanMediaItems(rows)
if err != nil {
return nil, 0, err
}
return items, len(items), nil
}
func (f *Fetcher) fetchAdminCuratedList(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]*models.MediaItem, int, error) {
var p recipes.AdminCuratedListParams
if len(s.Config) > 0 {
_ = json.Unmarshal(s.Config, &p)
}
if len(p.ItemIDs) == 0 {
return []*models.MediaItem{}, 0, nil
}
// Build conditions: content_id IN curated set + library scope + access filter.
var conditions []string
var args []any
argIdx := 1
conditions = append(conditions, fmt.Sprintf("mi.content_id = ANY($%d)", argIdx))
args = append(args, p.ItemIDs)
argIdx++
fromClause, libConditions, libArgs, newArgIdx := buildLibraryScope(libraryID, libraryIDs, nil, filter.DisabledLibraryIDs, argIdx)
conditions = append(conditions, libConditions...)
args = append(args, libArgs...)
argIdx = newArgIdx
catalog.ApplySectionAccessFilter("mi", filter, &conditions, &args, &argIdx)
whereClause := "WHERE " + strings.Join(conditions, " AND ")
// Order by position in the curated array. Reuse the item-IDs parameter
// index (1) for array_position; do NOT pass the slice twice.
query := fmt.Sprintf(
`SELECT %s FROM %s %s ORDER BY array_position($1::text[], mi.content_id)`,
itemColumns("mi"), fromClause, whereClause,
)
rows, err := f.pool.Query(ctx, query, args...)
if err != nil {
return nil, 0, fmt.Errorf("fetching admin curated list: %w", err)
}
defer rows.Close()
items, err := scanMediaItems(rows)
if err != nil {
return nil, 0, err
}
return items, len(items), nil
}
// fetchEditorialSpotlight returns items for an editorial_spotlight section.
// When auto_rotate is true it selects a subject from a candidate list using a
// deterministic ISO-week hash; otherwise it uses the configured subject directly.
func (f *Fetcher) fetchEditorialSpotlight(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]*models.MediaItem, int, error) {
items, total, _, err := f.fetchEditorialSpotlightWithTitle(ctx, s, libraryID, libraryIDs, filter)
return items, total, err
}
func (f *Fetcher) fetchEditorialSpotlightWithTitle(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]*models.MediaItem, int, string, error) {
var p recipes.EditorialSpotlightParams
if len(s.Config) > 0 {
_ = json.Unmarshal(s.Config, &p)
}
if p.SubjectType == "" {
return []*models.MediaItem{}, 0, "", nil
}
subject := p.Subject
if p.AutoRotate {
cands, err := f.cachedEditorialCandidates(ctx, p.SubjectType, libraryID, libraryIDs, filter, editorialCandidateCacheTTL, f.editorialCandidates)
if err != nil {
return nil, 0, "", fmt.Errorf("editorial_spotlight candidates: %w", err)
}
if len(cands) == 0 {
return []*models.MediaItem{}, 0, "", nil
}
days := editorialCadenceDays(p.RotationCadence)
// Build a value-based rotation key. Using `%v` on the *int would print
// the pointer address, which changes on every process restart and
// would break the deterministic-rotation contract.
libKey := "all"
if libraryID != nil {
libKey = strconv.Itoa(*libraryID)
}
key := fmt.Sprintf("%s|%s", p.SubjectType, libKey)
idx := recipes.RotationIndex(f.now(), key, len(cands), days)
subject = cands[idx]
}
if subject == "" {
return []*models.MediaItem{}, 0, "", nil
}
// Build a QueryDefinition with the subject filter and run it via the executor.
def, err := editorialBuildQueryDef(p.SubjectType, subject)
if err != nil {
return nil, 0, "", fmt.Errorf("editorial_spotlight build query: %w", err)
}
switch {
case libraryID != nil:
def.LibraryIDs = []int{*libraryID}
case libraryIDs != nil:
if len(def.LibraryIDs) == 0 {
def.LibraryIDs = append([]int(nil), libraryIDs...)
} else {
def.LibraryIDs = intersectLibraryIDs(def.LibraryIDs, libraryIDs)
}
}
limit := s.ItemLimit
if limit <= 0 {
limit = 20
}
def.Limit = &limit
executor := &catalog.QueryExecutor{Pool: f.pool}
items, _, _, err := executor.PreviewPage(ctx, def, filter, limit, 0, false)
if err != nil {
return nil, 0, "", fmt.Errorf("editorial_spotlight query: %w", err)
}
return items, len(items), editorialSpotlightDisplayTitle(s.Title, subject), nil
}
func editorialSpotlightDisplayTitle(baseTitle, subject string) string {
baseTitle = strings.TrimSpace(baseTitle)
subject = strings.TrimSpace(subject)
if subject == "" {
return baseTitle
}
if baseTitle == "" || strings.EqualFold(baseTitle, subject) {
return subject
}
suffix := " - " + subject
if strings.HasSuffix(baseTitle, suffix) {
return baseTitle
}
return baseTitle + suffix
}
// editorialCadenceDays converts a RotationCadence to a number of days for RotationIndex.
func editorialCadenceDays(c recipes.RotationCadence) int {
switch c {
case recipes.CadenceDaily:
return 1
case recipes.CadenceMonthly:
return 30
default: // weekly or empty
return 7
}
}
// editorialBuildQueryDef builds a QueryDefinition for the given subject type and subject value.
func editorialBuildQueryDef(subjectType, subject string) (catalog.QueryDefinition, error) {
var rule catalog.QueryRule
switch subjectType {
case "director":
rule = catalog.QueryRule{Field: "director", Op: "is", Value: subject}
case "actor":
rule = catalog.QueryRule{Field: "actor", Op: "is", Value: subject}
case "studio":
rule = catalog.QueryRule{Field: "studio", Op: "is", Value: subject}
case "era":
startYear, endYear, err := eraToYearRange(subject)
if err != nil {
return catalog.QueryDefinition{}, err
}
// release_date between uses string dates ("YYYY-01-01", "YYYY-12-31").
startISO := fmt.Sprintf("%d-01-01", startYear)
endISO := fmt.Sprintf("%d-12-31", endYear)
rule = catalog.QueryRule{Field: "release_date", Op: "between", Value: []any{startISO, endISO}}
default:
return catalog.QueryDefinition{}, fmt.Errorf("editorial_spotlight: unsupported subject_type %q", subjectType)
}
def := catalog.QueryDefinition{
Match: "all",
Groups: []catalog.QueryGroup{
{Match: "all", Rules: []catalog.QueryRule{rule}},
},
}
return def.Normalize(), nil
}
// eraToYearRange converts an era label like "1980s" to its start and end years.
func eraToYearRange(era string) (int, int, error) {
switch era {
case "1970s":
return 1970, 1979, nil
case "1980s":
return 1980, 1989, nil
case "1990s":
return 1990, 1999, nil
case "2000s":
return 2000, 2009, nil
case "2010s":
return 2010, 2019, nil
case "2020s":
return 2020, 2029, nil
default:
return 0, 0, fmt.Errorf("editorial_spotlight: unknown era %q", era)
}
}
// editorialCandidates returns the top subject names for auto-rotation, applying
// library scope and access filter so results are scoped to the user's libraries.
func (f *Fetcher) editorialCandidates(ctx context.Context, subjectType string, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]string, error) {
switch subjectType {
case "director":
return f.topPersonCandidates(ctx, 2 /* PersonKindDirector */, libraryID, libraryIDs, filter)
case "actor":
return f.topPersonCandidates(ctx, 1 /* PersonKindActor */, libraryID, libraryIDs, filter)
case "studio":
return f.topStudioCandidates(ctx, libraryID, libraryIDs, filter)
case "era":
// Auto-rotate excludes the current decade — incomplete data + stylistic skew.
// "2020s" is still accepted as a manually-pinned subject via eraToYearRange.
return []string{"1970s", "1980s", "1990s", "2000s", "2010s"}, nil
case "franchise":
// TODO: franchise data not yet available; auto-rotate returns empty.
return nil, nil
default:
return nil, fmt.Errorf("editorial_spotlight: unsupported subject_type %q", subjectType)
}
}
const editorialCandidateLimit = 50
// topPersonCandidates returns the top N person names by item count for the given kind.
func (f *Fetcher) topPersonCandidates(ctx context.Context, kind int, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]string, error) {
var conditions []string
var args []any
argIdx := 1
fromClause, libConditions, libArgs, newArgIdx := buildLibraryScope(libraryID, libraryIDs, nil, filter.DisabledLibraryIDs, argIdx)
conditions = append(conditions, libConditions...)
args = append(args, libArgs...)
argIdx = newArgIdx
catalog.ApplySectionAccessFilter("mi", filter, &conditions, &args, &argIdx)
whereClause := "WHERE 1=1"
if len(conditions) > 0 {
whereClause += " AND " + strings.Join(conditions, " AND ")
}
args = append(args, kind, editorialCandidateLimit)
// Weighted billing: SUM(1/(sort_order+1)) favors top-billed credits
// (TMDB convention: 0 = lead). HAVING MIN(sort_order) <= 5 excludes
// people who only ever appear in deep supporting roles.
query := fmt.Sprintf(`
SELECT p.name
FROM people p
JOIN item_people ip ON ip.person_id = p.id AND ip.kind = $%d
JOIN %s ON mi.content_id = ip.content_id
%s
GROUP BY p.name
HAVING MIN(ip.sort_order) <= 5
ORDER BY SUM(1.0 / (ip.sort_order + 1)) DESC, COUNT(*) DESC
LIMIT $%d
`, argIdx, fromClause, whereClause, argIdx+1)
rows, err := f.pool.Query(ctx, query, args...)
if err != nil {
return nil, fmt.Errorf("querying top persons (kind=%d): %w", kind, err)
}
defer rows.Close()
var names []string
for rows.Next() {
var name string
if err := rows.Scan(&name); err != nil {
return nil, fmt.Errorf("scanning person name: %w", err)
}
names = append(names, name)
}
return names, rows.Err()
}
// topStudioCandidates returns the top N studio names by item count.
func (f *Fetcher) topStudioCandidates(ctx context.Context, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]string, error) {
var conditions []string
var args []any
argIdx := 1
fromClause, libConditions, libArgs, newArgIdx := buildLibraryScope(libraryID, libraryIDs, nil, filter.DisabledLibraryIDs, argIdx)
conditions = append(conditions, libConditions...)
args = append(args, libArgs...)
argIdx = newArgIdx
catalog.ApplySectionAccessFilter("mi", filter, &conditions, &args, &argIdx)
whereClause := ""
if len(conditions) > 0 {
whereClause = "AND " + strings.Join(conditions, " AND ")
}
args = append(args, editorialCandidateLimit)
query := fmt.Sprintf(`
SELECT studio
FROM (
SELECT unnest(mi.studios) AS studio
FROM %s
WHERE mi.studios IS NOT NULL
%s
) sub
GROUP BY studio
ORDER BY COUNT(*) DESC
LIMIT $%d
`, fromClause, whereClause, argIdx)
rows, err := f.pool.Query(ctx, query, args...)
if err != nil {
return nil, fmt.Errorf("querying top studios: %w", err)
}
defer rows.Close()
var studios []string
for rows.Next() {
var studio string
if err := rows.Scan(&studio); err != nil {
return nil, fmt.Errorf("scanning studio name: %w", err)
}
studios = append(studios, studio)
}
return studios, rows.Err()
}
type recommendationSectionConfig struct {
SourceItemID string `json:"source_item_id"`
Genre string `json:"genre"`
}
func parseRecommendationSectionConfig(config json.RawMessage) recommendationSectionConfig {
var cfg recommendationSectionConfig
if len(config) == 0 {
return cfg
}
_ = json.Unmarshal(config, &cfg)
return cfg
}
func orderMediaItems(items []*models.MediaItem, orderedIDs []string) []*models.MediaItem {
if len(items) <= 1 {
return items
}
byID := make(map[string]*models.MediaItem, len(items))
for _, item := range items {
byID[item.ContentID] = item
}
ordered := make([]*models.MediaItem, 0, len(items))
for _, id := range orderedIDs {
item, ok := byID[id]
if !ok {
continue
}
ordered = append(ordered, item)
}
return ordered
}
func (f *Fetcher) fetchRecentlyAdded(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]*models.MediaItem, int, error) {
query, args := buildRecentlyAddedQuery(s, libraryID, libraryIDs, filter)
rows, err := f.pool.Query(ctx, query, args...)
if err != nil {
return nil, 0, err
}
defer rows.Close()
items, err := scanMediaItems(rows)
return items, len(items), err
}
type sectionQuery struct {
sql string
args []any
}
func buildRecentlyAddedQuery(s ResolvedSection, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) (string, []any) {
cfgFilters := ParseConfigFilters(s.Config)
// Backwards compat: support legacy "types" config field
var legacyCfg struct {
Types []string `json:"types"`
}
if len(s.Config) > 0 {
_ = json.Unmarshal(s.Config, &legacyCfg)
}
if cfgFilters.FilterType == "" && len(legacyCfg.Types) > 0 {
cfgFilters.FilterType = legacyCfg.Types[0]
}
if query, ok := buildRecentlyAddedSingleLibraryQuery(s, cfgFilters, libraryID, libraryIDs, filter); ok {
return query.sql, query.args
}
var conditions []string
var args []any
argIdx := 1
applyConfigTypeFilter("mi", cfgFilters.FilterType, &conditions, &args, &argIdx)
fromClause, libConditions, libArgs, newArgIdx := buildLibraryScope(libraryID, libraryIDs, cfgFilters.LibraryIDs(), filter.DisabledLibraryIDs, argIdx)
conditions = append(conditions, libConditions...)
args = append(args, libArgs...)
argIdx = newArgIdx
catalog.ApplySectionAccessFilter("mi", filter, &conditions, &args, &argIdx)
conditions = append(conditions, catalog.MangaChapterExclusionWhere("mi"))
whereClause := ""
if len(conditions) > 0 {
whereClause = "WHERE " + strings.Join(conditions, " AND ")
}
query := fmt.Sprintf(
`SELECT %s FROM %s %s ORDER BY mi.created_at DESC, mi.content_id ASC LIMIT $%d`,
itemColumnsLatestMangaPoster("mi"), fromClause, whereClause, argIdx,
)
args = append(args, s.ItemLimit)
return query, args
}
func buildRecentlyAddedSingleLibraryQuery(s ResolvedSection, cfgFilters SectionConfigFilters, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) (sectionQuery, bool) {
configLibraryIDs := cfgFilters.LibraryIDs()
var singleLibraryID int
switch {
case libraryID != nil:
singleLibraryID = *libraryID
case len(configLibraryIDs) == 1:
singleLibraryID = configLibraryIDs[0]
default:
return sectionQuery{}, false
}
if singleLibraryID <= 0 {
return sectionQuery{}, false
}
var conditions []string
var args []any
argIdx := 1
conditions = append(conditions, fmt.Sprintf("mil.media_folder_id = $%d", argIdx))
args = append(args, singleLibraryID)
argIdx++
if libraryIDs != nil {
if len(libraryIDs) == 0 {
conditions = append(conditions, "1 = 0")
} else {
placeholders := make([]string, len(libraryIDs))
for i, id := range libraryIDs {
placeholders[i] = fmt.Sprintf("$%d", argIdx)
args = append(args, id)
argIdx++
}
conditions = append(conditions, fmt.Sprintf("mil.media_folder_id IN (%s)", strings.Join(placeholders, ", ")))
}
}
if len(filter.DisabledLibraryIDs) > 0 {
placeholders := make([]string, len(filter.DisabledLibraryIDs))
for i, id := range filter.DisabledLibraryIDs {
placeholders[i] = fmt.Sprintf("$%d", argIdx)
args = append(args, id)
argIdx++
}
conditions = append(conditions, fmt.Sprintf("mil.media_folder_id NOT IN (%s)", strings.Join(placeholders, ", ")))
}
applyConfigTypeFilter("mi", cfgFilters.FilterType, &conditions, &args, &argIdx)
catalog.ApplySectionAccessFilter("mi", filter, &conditions, &args, &argIdx)
conditions = append(conditions, catalog.MangaChapterExclusionWhere("mi"))
whereClause := "WHERE " + strings.Join(conditions, " AND ")
query := fmt.Sprintf(
`SELECT %s FROM media_item_libraries mil JOIN media_items mi ON mi.content_id = mil.content_id %s ORDER BY mil.first_seen_at DESC, mil.content_id ASC LIMIT $%d`,
itemColumnsLatestMangaPoster("mi"), whereClause, argIdx,
)
args = append(args, s.ItemLimit)
return sectionQuery{sql: query, args: args}, true
}
func buildRecentlyReleasedQuery(s ResolvedSection, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) (string, []any) {
cfgFilters := ParseConfigFilters(s.Config)
var conditions []string
var args []any
argIdx := 1
applyConfigTypeFilter("mi", cfgFilters.FilterType, &conditions, &args, &argIdx)
fromClause, libConditions, libArgs, newArgIdx := buildLibraryScope(libraryID, libraryIDs, cfgFilters.LibraryIDs(), filter.DisabledLibraryIDs, argIdx)
conditions = append(conditions, libConditions...)
args = append(args, libArgs...)
argIdx = newArgIdx
catalog.ApplySectionAccessFilter("mi", filter, &conditions, &args, &argIdx)
conditions = append(conditions, catalog.MangaChapterExclusionWhere("mi"))
whereClause := ""
if len(conditions) > 0 {
whereClause = "WHERE " + strings.Join(conditions, " AND ")
}
query := fmt.Sprintf(
`SELECT %s FROM %s %s ORDER BY mi.year DESC, mi.created_at DESC LIMIT $%d`,
itemColumnsLatestMangaPoster("mi"), fromClause, whereClause, argIdx,
)
args = append(args, s.ItemLimit)
return query, args
}
func (f *Fetcher) fetchRecentlyReleased(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]*models.MediaItem, int, error) {
query, args := buildRecentlyReleasedQuery(s, libraryID, libraryIDs, filter)
rows, err := f.pool.Query(ctx, query, args...)
if err != nil {
return nil, 0, err
}
defer rows.Close()
items, err := scanMediaItems(rows)
return items, len(items), err
}
func (f *Fetcher) fetchFiltered(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]*models.MediaItem, int, error) {
def, err := ParseQueryDefinition(s.Config)
if err != nil {
return nil, 0, fmt.Errorf("parsing query definition: %w", err)
}
def = applySectionLibraryScopeToQuery(def, libraryID, libraryIDs)
if s.ItemLimit > 0 {
limit := s.ItemLimit
def.Limit = &limit
}
executor := &catalog.QueryExecutor{Pool: f.pool}
items, _, hasMore, err := executor.PreviewPage(ctx, def, filter, s.ItemLimit, 0, false)
if err != nil {
return nil, 0, err
}
total := len(items)
if hasMore && s.ItemLimit > 0 && total <= s.ItemLimit {
total = s.ItemLimit + 1
}
return items, total, nil
}
func applySectionLibraryScopeToQuery(def catalog.QueryDefinition, libraryID *int, libraryIDs []int) catalog.QueryDefinition {
switch {
case libraryID != nil:
def.LibraryIDs = []int{*libraryID}
case libraryIDs != nil:
if len(def.LibraryIDs) == 0 {
def.LibraryIDs = append([]int(nil), libraryIDs...)
} else {
def.LibraryIDs = intersectLibraryIDs(def.LibraryIDs, libraryIDs)
}
}
return def
}
func buildRandomQuery(s ResolvedSection, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) (string, []any, int) {
cfgFilters := ParseConfigFilters(s.Config)
var conditions []string
var args []any
argIdx := 1
applyConfigTypeFilter("mi", cfgFilters.FilterType, &conditions, &args, &argIdx)
fromClause, libConditions, libArgs, newArgIdx := buildLibraryScope(libraryID, libraryIDs, cfgFilters.LibraryIDs(), filter.DisabledLibraryIDs, argIdx)
conditions = append(conditions, libConditions...)
args = append(args, libArgs...)
argIdx = newArgIdx
catalog.ApplySectionAccessFilter("mi", filter, &conditions, &args, &argIdx)
conditions = append(conditions, catalog.MangaChapterExclusionWhere("mi"))
whereClause := ""
if len(conditions) > 0 {
whereClause = "WHERE " + strings.Join(conditions, " AND ")
}
limit := s.ItemLimit
if limit <= 0 {
limit = 20
}
queryLimit := limit + 1
query := fmt.Sprintf(
`SELECT candidate.content_id
FROM (
SELECT DISTINCT mi.content_id
FROM %s
%s
) candidate
ORDER BY RANDOM()
LIMIT $%d`,
fromClause, whereClause, argIdx,
)
args = append(args, queryLimit)
return query, args, limit
}
func (f *Fetcher) fetchRandom(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]*models.MediaItem, int, error) {
query, args, limit := buildRandomQuery(s, libraryID, libraryIDs, filter)
rows, err := f.pool.Query(ctx, query, args...)
if err != nil {
return nil, 0, err
}
defer rows.Close()
var contentIDs []string
for rows.Next() {
var contentID string
if err := rows.Scan(&contentID); err != nil {
return nil, 0, fmt.Errorf("scanning random section content id: %w", err)
}
contentIDs = append(contentIDs, contentID)
}
if err := rows.Err(); err != nil {
return nil, 0, err
}
hasMore := len(contentIDs) > limit
if hasMore {
contentIDs = contentIDs[:limit]
}
items, err := f.fetchItemsByContentIDs(ctx, contentIDs, libraryID, libraryIDs, filter)
if err != nil {
return nil, 0, err
}
ordered := orderMediaItems(items, contentIDs)
total := len(ordered)
if hasMore && total <= limit {
total = limit + 1
}
return ordered, total, nil
}
func (f *Fetcher) fetchItemsByContentIDs(ctx context.Context, contentIDs []string, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]*models.MediaItem, error) {
if len(contentIDs) == 0 {
return []*models.MediaItem{}, nil
}
var conditions []string
var args []any
argIdx := 1
placeholders := make([]string, len(contentIDs))
for i, contentID := range contentIDs {
placeholders[i] = fmt.Sprintf("$%d", argIdx)
args = append(args, contentID)
argIdx++
}
conditions = append(conditions, fmt.Sprintf("mi.content_id IN (%s)", strings.Join(placeholders, ", ")))
effectiveLibraryIDs := effectiveFetchLibraryIDs(libraryIDs, filter)
fromClause, libConditions, libArgs, newArgIdx := buildLibraryScope(libraryID, effectiveLibraryIDs, nil, filter.DisabledLibraryIDs, argIdx)
conditions = append(conditions, libConditions...)
args = append(args, libArgs...)
argIdx = newArgIdx
catalog.ApplySectionAccessFilter("mi", filter, &conditions, &args, &argIdx)
query := fmt.Sprintf(
`SELECT %s FROM %s WHERE %s`,
itemColumns("mi"), fromClause, strings.Join(conditions, " AND "),
)
rows, err := f.pool.Query(ctx, query, args...)
if err != nil {
return nil, err
}
defer rows.Close()
return scanMediaItems(rows)
}
func (f *Fetcher) fetchEpisodeTargetsByContentIDs(ctx context.Context, contentIDs []string, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]*models.MediaItem, map[string]SectionItemMeta, error) {
if len(contentIDs) == 0 {
return []*models.MediaItem{}, map[string]SectionItemMeta{}, nil
}
var conditions []string
var args []any
argIdx := 1
placeholders := make([]string, len(contentIDs))
for i, contentID := range contentIDs {
placeholders[i] = fmt.Sprintf("$%d", argIdx)
args = append(args, contentID)
argIdx++
}
conditions = append(conditions, fmt.Sprintf("e.content_id IN (%s)", strings.Join(placeholders, ", ")))
effectiveLibraryIDs := effectiveFetchLibraryIDs(libraryIDs, filter)
fromClause := "episodes e JOIN media_items si ON e.series_id = si.content_id LEFT JOIN seasons s ON s.content_id = e.season_id"
if libraryID != nil || effectiveLibraryIDs != nil {
fromClause += " JOIN media_item_libraries mil ON si.content_id = mil.content_id"
}
if libraryID != nil {
conditions = append(conditions, fmt.Sprintf("mil.media_folder_id = $%d", argIdx))
args = append(args, *libraryID)
argIdx++
}
if effectiveLibraryIDs != nil {
if len(effectiveLibraryIDs) == 0 {
return []*models.MediaItem{}, map[string]SectionItemMeta{}, nil
}
placeholders = make([]string, len(effectiveLibraryIDs))
for i, id := range effectiveLibraryIDs {
placeholders[i] = fmt.Sprintf("$%d", argIdx)
args = append(args, id)
argIdx++
}
conditions = append(conditions, fmt.Sprintf("mil.media_folder_id IN (%s)", strings.Join(placeholders, ", ")))
}
catalog.ApplySectionAccessFilter("si", filter, &conditions, &args, &argIdx)
query := fmt.Sprintf(`
SELECT
e.content_id,
e.series_id,
e.title,
e.overview,
e.runtime,
e.rating_imdb,
COALESCE(NULLIF(s.poster_path, ''), NULLIF(si.poster_path, ''), NULLIF(e.still_path, ''), '') AS poster_path,
COALESCE(NULLIF(s.poster_thumbhash, ''), NULLIF(si.poster_thumbhash, ''), NULLIF(e.still_thumbhash, ''), '') AS poster_thumbhash,
e.season_number,
e.episode_number,
e.air_date,
si.title,
si.genres,
si.content_rating,
COALESCE(NULLIF(e.still_path, ''), NULLIF(si.backdrop_path, ''), '') AS backdrop_path,
COALESCE(NULLIF(e.still_thumbhash, ''), NULLIF(si.backdrop_thumbhash, ''), '') AS backdrop_thumbhash,
si.logo_path,
si.status
FROM %s
WHERE %s
`, fromClause, strings.Join(conditions, " AND "))
rows, err := f.pool.Query(ctx, query, args...)
if err != nil {
return nil, nil, err
}
defer rows.Close()
items := []*models.MediaItem{}
itemMeta := map[string]SectionItemMeta{}
for rows.Next() {
var (
item models.MediaItem
seriesID string
seasonNumber int
episodeNumber int
seriesTitle string
airDate *time.Time
)
item.Type = "episode"
err := rows.Scan(
&item.ContentID,
&seriesID,
&item.Title,
&item.Overview,
&item.Runtime,
&item.RatingIMDB,
&item.PosterPath,
&item.PosterThumbhash,
&seasonNumber,
&episodeNumber,
&airDate,
&seriesTitle,
&item.Genres,
&item.ContentRating,
&item.BackdropPath,
&item.BackdropThumbhash,
&item.LogoPath,
&item.Status,
)
if err != nil {
return nil, nil, fmt.Errorf("scanning episode section item: %w", err)
}
items = append(items, &item)
itemMeta[item.ContentID] = SectionItemMeta{
SeriesID: &seriesID,
SeriesTitle: seriesTitle,
SeasonNumber: &seasonNumber,
EpisodeNumber: &episodeNumber,
Badges: recentSeasonPremiereBadges(seasonNumber, episodeNumber, airDate),
}
}
if err := rows.Err(); err != nil {
return nil, nil, err
}
return items, itemMeta, nil
}
func effectiveFetchLibraryIDs(libraryIDs []int, filter catalog.AccessFilter) []int {
if libraryIDs != nil {
return libraryIDs
}
if filter.AllowedLibraryIDs != nil {
return filter.AllowedLibraryIDs
}
return nil
}
func recentSeasonPremiereBadges(seasonNumber, episodeNumber int, airDate *time.Time) []string {
if seasonNumber <= 1 || episodeNumber != 1 || airDate == nil {
return nil
}
nowDate := utcDay(time.Now().UTC())
premiereDate := utcDay(airDate.UTC())
windowStart := nowDate.AddDate(0, 0, -(recentSeasonPremiereBadgeWindowDays - 1))
if premiereDate.Before(windowStart) || premiereDate.After(nowDate) {
return nil
}
return []string{"season_premiere"}
}
func utcDay(t time.Time) time.Time {
return time.Date(t.Year(), t.Month(), t.Day(), 0, 0, 0, 0, time.UTC)
}
// itemColumnsList returns the alias-prefixed SELECT columns matching
// scanMediaItems, in scan order. Mirrors catalog.browseItemColumns. The shared
// source of truth for itemColumns and its manga-poster-override variant.
func itemColumnsList(alias string) []string {
cols := []string{
"content_id", "type", "title", "sort_title", "original_title", "year", "genres",
"content_rating", "runtime", "overview", "tagline",
"rating_imdb", "rating_tmdb", "rating_rt_critic", "rating_rt_audience",
"imdb_id", "tmdb_id", "tvdb_id",
"poster_path", "poster_thumbhash", "backdrop_path", "backdrop_thumbhash", "logo_path",
"metadata_s3_path", "metadata_etag", "season_count",
"studios", "networks", "countries", "release_date::text", "first_air_date", "last_air_date",
"show_status",
"matched_at", "status", "created_at", "updated_at",
}
prefixed := make([]string, len(cols))
for i, c := range cols {
prefixed[i] = alias + "." + c
}
return prefixed
}
// itemColumns returns the SELECT column list matching scanMediaItems.
// Mirrors catalog.browseItemColumns.
func itemColumns(alias string) string {
return strings.Join(itemColumnsList(alias), ", ")
}
// itemColumnsLatestMangaPoster returns the same SELECT column list as
// itemColumns (identical order and aliases, so scanMediaItems is unchanged) but
// overrides poster_path/poster_thumbhash for type='manga' SERIES rows: a manga
// series card shows the cover of its latest-added volume/chapter (the linked
// manga_chapters row with the greatest created_at) instead of the AniList series
// cover, falling back to the series' own poster when no chapter cover exists.
//
// Strictly gated on mi.type = 'manga' so movies/TV/audiobooks/ebooks keep their
// own poster exactly. Used only by the recently-added / recently-released
// section builders; all other queries keep itemColumns.
func itemColumnsLatestMangaPoster(alias string) string {
cols := itemColumnsList(alias)
for i, c := range cols {
switch c {
case alias + ".poster_path":
cols[i] = mangaLatestVolumePosterExpr(alias, "poster_path")
case alias + ".poster_thumbhash":
cols[i] = mangaLatestVolumePosterExpr(alias, "poster_thumbhash")
}
}
return strings.Join(cols, ", ")
}
// mangaLatestVolumePosterExpr emits the manga-gated CASE override for a single
// poster column, aliased back to the original column name so the scan order and
// column set are unchanged.
func mangaLatestVolumePosterExpr(alias, col string) string {
// NULLIF(...,'') on each operand: poster columns default to '' (empty
// string), not NULL, so a plain COALESCE would surface a cover-less latest
// chapter's empty poster instead of falling back to the series' own cover.
// Mirrors the episode poster expressions above; trailing '' keeps the THEN
// branch non-NULL.
return "CASE WHEN " + alias + ".type = 'manga' THEN COALESCE(NULLIF((" +
"SELECT c." + col + " FROM media_items c " +
"JOIN manga_chapters mc ON mc.chapter_content_id = c.content_id " +
"WHERE mc.series_content_id = " + alias + ".content_id " +
"ORDER BY c.created_at DESC, c.content_id DESC LIMIT 1), ''), " +
"NULLIF(" + alias + "." + col + ", ''), '') ELSE " + alias + "." + col + " END AS " + col
}
// scanMediaItems scans rows into MediaItem slices. Must match itemColumns order.
func scanMediaItems(rows pgx.Rows) ([]*models.MediaItem, error) {
var items []*models.MediaItem
for rows.Next() {
var item models.MediaItem
err := rows.Scan(
&item.ContentID, &item.Type, &item.Title, &item.SortTitle, &item.OriginalTitle,
&item.Year, &item.Genres, &item.ContentRating, &item.Runtime, &item.Overview, &item.Tagline,
&item.RatingIMDB, &item.RatingTMDB, &item.RatingRTCritic, &item.RatingRTAudience,
&item.ImdbID, &item.TmdbID, &item.TvdbID,
&item.PosterPath, &item.PosterThumbhash, &item.BackdropPath, &item.BackdropThumbhash, &item.LogoPath,
&item.MetadataS3Path, &item.MetadataEtag, &item.SeasonCount,
&item.Studios, &item.Networks, &item.Countries, &item.ReleaseDate, &item.FirstAirDate, &item.LastAirDate,
&item.ShowStatus,
&item.MatchedAt, &item.Status, &item.CreatedAt, &item.UpdatedAt,
)
if err != nil {
return nil, fmt.Errorf("scanning item: %w", err)
}
items = append(items, &item)
}
return items, rows.Err()
}
// trendingDiscoverEntry is a provider-agnostic external trending result.
type trendingDiscoverEntry struct {
tmdbID string
imdbID string
tvdbID string
mediaType string // "movie" | "tv"
}
// newTrendingEntry normalizes a provider entry into a trendingDiscoverEntry,
// stringifying the numeric external IDs and dropping any that are unset (<= 0).
func newTrendingEntry(tmdbID, tvdbID int, imdbID, mediaType string) trendingDiscoverEntry {
e := trendingDiscoverEntry{imdbID: imdbID, mediaType: mediaType}
if tmdbID > 0 {
e.tmdbID = strconv.Itoa(tmdbID)
}
if tvdbID > 0 {
e.tvdbID = strconv.Itoa(tvdbID)
}
return e
}
// fetchTrendingDiscover surfaces external global trending (TMDB or Trakt),
// mixing movies + series, matched to titles in the viewer's enabled libraries.
// The external fetch + ID resolution are cached briefly so it does not hit the
// upstream API on every home-page load.
func (f *Fetcher) fetchTrendingDiscover(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]*models.MediaItem, int, error) {
var p recipes.TrendingDiscoverParams
if len(s.Config) > 0 {
_ = json.Unmarshal(s.Config, &p)
}
source, window := canonicalTrendingKey(p.Source, p.Window)
limit := s.ItemLimit
if limit <= 0 {
limit = 20
}
orderedIDs, err := f.loadTrendingDiscoverContentIDs(ctx, source, window)
if err != nil {
return nil, 0, err
}
if len(orderedIDs) == 0 {
return []*models.MediaItem{}, 0, nil
}
items, err := f.fetchItemsByContentIDs(ctx, orderedIDs, libraryID, libraryIDs, filter)
if err != nil {
return nil, 0, err
}
// Re-order to trending rank (fetchItemsByContentIDs returns DB order) and
// truncate to the section's display limit.
ordered := orderMediaItems(items, orderedIDs)
if len(ordered) > limit {
ordered = ordered[:limit]
}
return ordered, len(ordered), nil
}
// loadTrendingDiscoverContentIDs returns the persisted, catalog-resolved content
// IDs for the canonical (source, window). It reads only the snapshot table; the
// upstream fetch happens out-of-band in TrendingRefresher. Returns nil when no
// snapshot reader is configured or no snapshot exists yet.
func (f *Fetcher) loadTrendingDiscoverContentIDs(ctx context.Context, source, window string) ([]string, error) {
if f.TrendingSnapshots == nil {
return nil, nil
}
snap, found, err := f.TrendingSnapshots.Get(ctx, source, window)
if err != nil {
return nil, err
}
if !found {
return nil, nil
}
return snap.ContentIDs, nil
}
// orderedTrendingContentIDs maps trending entries to library content IDs in
// trending order, preferring TVDB (series) > TMDB > IMDb, de-duplicated.
func orderedTrendingContentIDs(entries []trendingDiscoverEntry, movieLookup, seriesLookup *catalog.ExternalIDLookup) []string {
seen := make(map[string]struct{}, len(entries))
out := make([]string, 0, len(entries))
for _, e := range entries {
// Only movies and series are resolvable; skip anything else (TMDB
// trending/all also returns media_type "person").
var lookup *catalog.ExternalIDLookup
isSeries := e.mediaType == "tv"
switch e.mediaType {
case "movie":
lookup = movieLookup
case "tv":
lookup = seriesLookup
default:
continue
}
if lookup == nil {
continue
}
var id string
if isSeries && e.tvdbID != "" {
id = lookup.ByTVDB[e.tvdbID]
}
if id == "" && e.tmdbID != "" {
id = lookup.ByTMDB[e.tmdbID]
}
if id == "" && e.imdbID != "" {
id = lookup.ByIMDb[e.imdbID]
}
if id == "" {
continue
}
if _, dup := seen[id]; dup {
continue
}
seen[id] = struct{}{}
out = append(out, id)
}
return out
}
func (f *Fetcher) fetchTrending(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]*models.MediaItem, int, error) {
var p recipes.TrendingParams
if len(s.Config) > 0 {
_ = json.Unmarshal(s.Config, &p)
}
interval := "7 days"
switch p.Window {
case "24h":
interval = "24 hours"
case "7d", "":
interval = "7 days"
case "30d":
interval = "30 days"
}
var conditions []string
var args []any
argIdx := 1
fromClause, libConditions, libArgs, newArgIdx := buildLibraryScope(libraryID, libraryIDs, nil, filter.DisabledLibraryIDs, argIdx)
conditions = append(conditions, libConditions...)
args = append(args, libArgs...)
argIdx = newArgIdx
catalog.ApplySectionAccessFilter("mi", filter, &conditions, &args, &argIdx)
conditions = append(conditions, catalog.MangaChapterExclusionWhere("mi"))
conditions = append(conditions, fmt.Sprintf("uwh.watched_at > NOW() - $%d::interval", argIdx))
args = append(args, interval)
argIdx++
limit := s.ItemLimit
if limit <= 0 {
limit = 20
}
whereClause := "WHERE " + strings.Join(conditions, " AND ")
query := fmt.Sprintf(
`SELECT %s FROM %s JOIN user_watch_history uwh ON uwh.media_item_id = mi.content_id %s GROUP BY mi.content_id ORDER BY COUNT(DISTINCT uwh.profile_id) DESC, COUNT(*) DESC LIMIT $%d`,
itemColumns("mi"), fromClause, whereClause, argIdx,
)
args = append(args, limit)
rows, err := f.pool.Query(ctx, query, args...)
if err != nil {
return nil, 0, fmt.Errorf("fetching trending: %w", err)
}
defer rows.Close()
items, err := scanMediaItems(rows)
if err != nil {
return nil, 0, err
}
return items, len(items), nil
}
func (f *Fetcher) fetchProfileActivityFeed(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, profileID string, filter catalog.AccessFilter) ([]*models.MediaItem, int, error) {
var p recipes.ProfileActivityFeedParams
if len(s.Config) > 0 {
_ = json.Unmarshal(s.Config, &p)
}
target := p.ProfileID
// Household mode (target == "") leaks all history when caller is unauthenticated.
if target == "" && profileID == "" {
return []*models.MediaItem{}, 0, nil
}
var args []any
argIdx := 1
// CTE deduplicates per media_item_id and keeps the most-recent watched_at,
// so each item appears once and is ordered by latest-watch DESC.
var cteCond string
var cteWindow string
if target == "" {
cteCond = fmt.Sprintf("profile_id <> $%d", argIdx)
args = append(args, profileID)
argIdx++
cteWindow = "INTERVAL '7 days'"
} else {
cteCond = fmt.Sprintf("profile_id = $%d", argIdx)
args = append(args, target)
argIdx++
cteWindow = "INTERVAL '30 days'"
}
var conditions []string
fromClause, libConditions, libArgs, newArgIdx := buildLibraryScope(libraryID, libraryIDs, nil, filter.DisabledLibraryIDs, argIdx)
conditions = append(conditions, libConditions...)
args = append(args, libArgs...)
argIdx = newArgIdx
catalog.ApplySectionAccessFilter("mi", filter, &conditions, &args, &argIdx)
limit := s.ItemLimit
if limit <= 0 {
limit = 20
}
whereClause := ""
if len(conditions) > 0 {
whereClause = "WHERE " + strings.Join(conditions, " AND ")
}
query := fmt.Sprintf(
`WITH most_recent AS (
SELECT media_item_id, MAX(watched_at) AS latest
FROM user_watch_history
WHERE %s AND watched_at > NOW() - %s
GROUP BY media_item_id
)
SELECT %s
FROM %s
JOIN most_recent mr ON mr.media_item_id = mi.content_id
%s
ORDER BY mr.latest DESC
LIMIT $%d`,
cteCond, cteWindow,
itemColumns("mi"), fromClause, whereClause, argIdx,
)
args = append(args, limit)
rows, err := f.pool.Query(ctx, query, args...)
if err != nil {
return nil, 0, fmt.Errorf("fetching profile activity feed: %w", err)
}
defer rows.Close()
items, err := scanMediaItems(rows)
if err != nil {
return nil, 0, err
}
return items, len(items), nil
}
func (f *Fetcher) fetchNewToLibrary(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]*models.MediaItem, int, error) {
var p recipes.NewToLibraryParams
if len(s.Config) > 0 {
_ = json.Unmarshal(s.Config, &p)
}
days := p.LookbackDays
if days <= 0 {
days = 30
}
var conditions []string
var args []any
argIdx := 1
fromClause, libConditions, libArgs, newArgIdx := buildLibraryScope(libraryID, libraryIDs, nil, filter.DisabledLibraryIDs, argIdx)
conditions = append(conditions, libConditions...)
args = append(args, libArgs...)
argIdx = newArgIdx
catalog.ApplySectionAccessFilter("mi", filter, &conditions, &args, &argIdx)
conditions = append(conditions, catalog.MangaChapterExclusionWhere("mi"))
conditions = append(conditions, fmt.Sprintf("mi.created_at > NOW() - ($%d || ' days')::interval", argIdx))
args = append(args, days)
argIdx++
limit := s.ItemLimit
if limit <= 0 {
limit = 20
}
whereClause := "WHERE " + strings.Join(conditions, " AND ")
query := fmt.Sprintf(
`SELECT %s FROM %s %s ORDER BY mi.created_at DESC LIMIT $%d`,
itemColumns("mi"), fromClause, whereClause, argIdx,
)
args = append(args, limit)
rows, err := f.pool.Query(ctx, query, args...)
if err != nil {
return nil, 0, fmt.Errorf("fetching new to library: %w", err)
}
defer rows.Close()
items, err := scanMediaItems(rows)
if err != nil {
return nil, 0, err
}
return items, len(items), nil
}
func (f *Fetcher) fetchMostWatched(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]*models.MediaItem, int, error) {
var p recipes.MostWatchedParams
if len(s.Config) > 0 {
_ = json.Unmarshal(s.Config, &p)
}
interval := "7 days"
if p.Window == "month" {
interval = "30 days"
}
var conditions []string
var args []any
argIdx := 1
fromClause, libConditions, libArgs, newArgIdx := buildLibraryScope(libraryID, libraryIDs, nil, filter.DisabledLibraryIDs, argIdx)
conditions = append(conditions, libConditions...)
args = append(args, libArgs...)
argIdx = newArgIdx
catalog.ApplySectionAccessFilter("mi", filter, &conditions, &args, &argIdx)
conditions = append(conditions, catalog.MangaChapterExclusionWhere("mi"))
conditions = append(conditions, fmt.Sprintf("uwh.watched_at > NOW() - $%d::interval", argIdx))
args = append(args, interval)
argIdx++
limit := s.ItemLimit
if limit <= 0 {
limit = 20
}
whereClause := "WHERE " + strings.Join(conditions, " AND ")
query := fmt.Sprintf(
`SELECT %s FROM %s JOIN user_watch_history uwh ON uwh.media_item_id = mi.content_id %s GROUP BY mi.content_id ORDER BY COUNT(*) DESC LIMIT $%d`,
itemColumns("mi"), fromClause, whereClause, argIdx,
)
args = append(args, limit)
rows, err := f.pool.Query(ctx, query, args...)
if err != nil {
return nil, 0, fmt.Errorf("fetching most watched: %w", err)
}
defer rows.Close()
items, err := scanMediaItems(rows)
if err != nil {
return nil, 0, err
}
return items, len(items), nil
}
func buildLibraryScope(libraryID *int, libraryIDs []int, configLibraryIDs []int, disabledLibraryIDs []int, argIdx int) (string, []string, []any, int) {
fromClause := "media_items mi"
var conditions []string
var args []any
needsJoin := libraryID != nil || libraryIDs != nil || len(configLibraryIDs) > 0 || len(disabledLibraryIDs) > 0
if needsJoin {
fromClause = "media_items mi JOIN media_item_libraries mil ON mi.content_id = mil.content_id"
}
if libraryID != nil {
conditions = append(conditions, fmt.Sprintf("mil.media_folder_id = $%d", argIdx))
args = append(args, *libraryID)
argIdx++
}
if len(configLibraryIDs) > 0 && libraryID == nil {
placeholders := make([]string, len(configLibraryIDs))
for i, id := range configLibraryIDs {
placeholders[i] = fmt.Sprintf("$%d", argIdx)
args = append(args, id)
argIdx++
}
conditions = append(conditions, fmt.Sprintf("mil.media_folder_id IN (%s)", strings.Join(placeholders, ", ")))
}
if libraryIDs != nil {
if len(libraryIDs) == 0 {
conditions = append(conditions, "1 = 0")
return fromClause, conditions, args, argIdx
}
placeholders := make([]string, len(libraryIDs))
for i, id := range libraryIDs {
placeholders[i] = fmt.Sprintf("$%d", argIdx)
args = append(args, id)
argIdx++
}
conditions = append(conditions, fmt.Sprintf("mil.media_folder_id IN (%s)", strings.Join(placeholders, ", ")))
}
if len(disabledLibraryIDs) > 0 {
placeholders := make([]string, len(disabledLibraryIDs))
for i, id := range disabledLibraryIDs {
placeholders[i] = fmt.Sprintf("$%d", argIdx)
args = append(args, id)
argIdx++
}
conditions = append(conditions, fmt.Sprintf("mil.media_folder_id NOT IN (%s)", strings.Join(placeholders, ", ")))
}
return fromClause, conditions, args, argIdx
}
func fetchSortClause(sort, order string) string {
dir := "DESC"
if order == "asc" {
dir = "ASC"
}
switch sort {
case "rating":
return fmt.Sprintf("ORDER BY mi.rating_imdb %s NULLS LAST", dir)
case "year":
return fmt.Sprintf("ORDER BY mi.year %s", dir)
case "title":
return fmt.Sprintf("ORDER BY mi.sort_title %s", dir)
default:
return fmt.Sprintf("ORDER BY mi.created_at %s", dir)
}
}
func intersectLibraryIDs(a, b []int) []int {
if len(a) == 0 || len(b) == 0 {
return nil
}
allowed := make(map[int]struct{}, len(b))
for _, value := range b {
allowed[value] = struct{}{}
}
var result []int
seen := make(map[int]struct{})
for _, value := range a {
if _, ok := allowed[value]; !ok {
continue
}
if _, ok := seen[value]; ok {
continue
}
seen[value] = struct{}{}
result = append(result, value)
}
return result
}
// applyConfigTypeFilter adds a WHERE condition for the config's filter_type.
func applyConfigTypeFilter(alias string, filterType string, conditions *[]string, args *[]any, argIdx *int) {
if filterType == "" {
return
}
*conditions = append(*conditions, fmt.Sprintf("%s.type = $%d", alias, *argIdx))
*args = append(*args, filterType)
*argIdx++
}
func matchesContinueWatchingFilter(filterType string, itemType string) bool {
switch filterType {
case "movie", "series", "episode":
return ContinueTypeMatchesItem(ContinueTypeWatching, itemType)
case "audiobook":
return ContinueTypeMatchesItem(ContinueTypeListening, itemType)
case "ebook":
return ContinueTypeMatchesItem(ContinueTypeReading, itemType)
default:
return true
}
}
// fetchSeasonalThemed returns items for a seasonal_themed section.
//
// Multi-theme mode (EnabledThemes set): picks the highest-priority enabled
// theme whose predicate matches now and queries items for that theme. Off-
// season (no enabled theme matches) the section is empty and gets suppressed
// by the API dispatcher.
//
// Legacy single-theme mode (Theme set, EnabledThemes empty): mode controls
// suppression — "auto" hides off-season, "pinned" always renders.
func (f *Fetcher) fetchSeasonalThemed(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]*models.MediaItem, int, error) {
var p recipes.SeasonalThemedParams
if len(s.Config) > 0 {
_ = json.Unmarshal(s.Config, &p)
}
now := time.Now()
if f.Clock != nil {
now = f.Clock.Now()
}
// Resolve which theme to use. Multi-theme mode wins when populated.
var theme string
switch {
case len(p.EnabledThemes) > 0:
theme = recipes.ActiveSeasonalTheme(p.EnabledThemes, now)
if theme == "" {
// No enabled theme is currently in season — hide the section.
return []*models.MediaItem{}, 0, nil
}
case p.Theme != "":
// Legacy single-theme path with mode-based suppression.
pred, ok := recipes.SeasonalPredicates[p.Theme]
if !ok {
return []*models.MediaItem{}, 0, nil
}
mode := p.Mode
if mode == "" {
mode = "auto"
}
if mode == "auto" && !pred(now) {
return []*models.MediaItem{}, 0, nil
}
theme = p.Theme
default:
return []*models.MediaItem{}, 0, nil
}
def, hasQuery := seasonalQueryDef(theme)
if !hasQuery {
// Theme exists but has no executable genre query yet (christmas,
// st_patricks, thanksgiving need keyword/tag column support).
// Return empty until that data lands.
return []*models.MediaItem{}, 0, nil
}
// Apply library scope the same way fetchFiltered / fetchEditorialSpotlight do.
switch {
case libraryID != nil:
def.LibraryIDs = []int{*libraryID}
case libraryIDs != nil:
if len(def.LibraryIDs) == 0 {
def.LibraryIDs = append([]int(nil), libraryIDs...)
} else {
def.LibraryIDs = intersectLibraryIDs(def.LibraryIDs, libraryIDs)
}
}
limit := s.ItemLimit
if limit <= 0 {
limit = 20
}
def.Limit = &limit
executor := &catalog.QueryExecutor{Pool: f.pool}
items, _, _, err := executor.PreviewPage(ctx, def, filter, limit, 0, false)
if err != nil {
return nil, 0, fmt.Errorf("seasonal_themed query: %w", err)
}
return items, len(items), nil
}
// seasonalQueryDef returns a catalog.QueryDefinition for the given theme and
// whether an executable query exists. Themes that require a keyword/tag column
// (christmas, st_patricks, thanksgiving) return false until that data lands.
func seasonalQueryDef(theme string) (catalog.QueryDefinition, bool) {
newDef := func(rules ...catalog.QueryRule) catalog.QueryDefinition {
return catalog.QueryDefinition{
Match: "all",
Groups: []catalog.QueryGroup{
{Match: "any", Rules: rules},
},
}
}
rule := func(field, op, value string) catalog.QueryRule {
return catalog.QueryRule{Field: field, Op: op, Value: value}
}
switch theme {
case "halloween":
return newDef(
rule("genre", "contains", "Horror"),
rule("genre", "contains", "Thriller"),
), true
case "valentines":
return newDef(
rule("genre", "contains", "Romance"),
), true
case "summer", "summer_blockbuster":
return newDef(
rule("genre", "contains", "Action"),
rule("genre", "contains", "Adventure"),
), true
case "saturday_morning":
return newDef(
rule("genre", "contains", "Animation"),
rule("genre", "contains", "Family"),
), true
case "christmas", "st_patricks", "thanksgiving":
// TODO: needs keyword/tag column to filter by theme-specific keywords.
return catalog.QueryDefinition{}, false
default:
return catalog.QueryDefinition{}, false
}
}
// fetchMoodCollection returns items for a mood_collection section.
func (f *Fetcher) fetchMoodCollection(ctx context.Context, s ResolvedSection, libraryID *int, libraryIDs []int, filter catalog.AccessFilter) ([]*models.MediaItem, int, error) {
var p recipes.MoodCollectionParams
if len(s.Config) > 0 {
_ = json.Unmarshal(s.Config, &p)
}
info, ok := recipes.MoodByKey(p.Mood)
if !ok {
return []*models.MediaItem{}, 0, nil
}
// Build genre rules (OR'd together).
genreRules := make([]catalog.QueryRule, 0, len(info.GenresAny))
for _, genre := range info.GenresAny {
genreRules = append(genreRules, catalog.QueryRule{Field: "genre", Op: "contains", Value: genre})
}
def := catalog.QueryDefinition{
Match: "all",
Groups: []catalog.QueryGroup{
{Match: "any", Rules: genreRules},
{Match: "all", Rules: []catalog.QueryRule{
{Field: "rating_imdb", Op: "gte", Value: info.MinRating},
}},
},
}
switch {
case libraryID != nil:
def.LibraryIDs = []int{*libraryID}
case libraryIDs != nil:
if len(def.LibraryIDs) == 0 {
def.LibraryIDs = append([]int(nil), libraryIDs...)
} else {
def.LibraryIDs = intersectLibraryIDs(def.LibraryIDs, libraryIDs)
}
}
limit := s.ItemLimit
if limit <= 0 {
limit = 20
}
def.Limit = &limit
executor := &catalog.QueryExecutor{Pool: f.pool}
items, _, _, err := executor.PreviewPage(ctx, def, filter, limit, 0, false)
if err != nil {
return nil, 0, fmt.Errorf("mood_collection query: %w", err)
}
return items, len(items), nil
}
// mangaChapterSeriesMetaQuery resolves the owning manga series for chapter
// items (type='ebook' rows linked via manga_chapters) appearing on section
// cards, so continue-reading surfaces can show the series instead of the
// chapter's raw file title.
const mangaChapterSeriesMetaQuery = `
SELECT mc.chapter_content_id, mc.series_content_id, si.title
FROM manga_chapters mc
JOIN media_items si ON si.content_id = mc.series_content_id
WHERE mc.chapter_content_id = ANY($1)
`
// FetchMangaChapterSeriesMeta returns series linkage keyed by chapter content
// id. IDs that are not manga chapters simply have no entry.
func (f *Fetcher) FetchMangaChapterSeriesMeta(ctx context.Context, ids []string) (map[string]SectionItemMeta, error) {
meta := make(map[string]SectionItemMeta, len(ids))
if f == nil || f.pool == nil || len(ids) == 0 {
return meta, nil
}
rows, err := f.pool.Query(ctx, mangaChapterSeriesMetaQuery, ids)
if err != nil {
return nil, err
}
defer rows.Close()
for rows.Next() {
var chapterID, seriesID, seriesTitle string
if err := rows.Scan(&chapterID, &seriesID, &seriesTitle); err != nil {
return nil, err
}
id := seriesID
meta[chapterID] = SectionItemMeta{SeriesID: &id, SeriesTitle: seriesTitle}
}
return meta, rows.Err()
}
// applyMangaChapterSeriesMeta resolves the owning manga series for any chapter
// (ebook) items in the set and merges SeriesID/SeriesTitle into the existing
// itemMeta, preserving the progress fields already populated there. Lets the
// shared series-collapse group multiple in-progress chapters of one manga.
func (f *Fetcher) applyMangaChapterSeriesMeta(ctx context.Context, items []*models.MediaItem, itemMeta map[string]SectionItemMeta) {
ids := make([]string, 0, len(items))
for _, item := range items {
if item != nil && item.Type == "ebook" && strings.TrimSpace(item.ContentID) != "" {
ids = append(ids, item.ContentID)
}
}
if len(ids) == 0 {
return
}
seriesMeta, err := f.FetchMangaChapterSeriesMeta(ctx, ids)
if err != nil {
slog.Warn("continue-reading: manga series linkage lookup failed", "error", err)
return
}
for chapterID, sm := range seriesMeta {
m := itemMeta[chapterID]
m.SeriesID = sm.SeriesID
m.SeriesTitle = sm.SeriesTitle
itemMeta[chapterID] = m
}
}