- Allow manual cross-season reassignment without merging entries
- Require validated title matches for automatic merges
- Share title normalization across AniList and stats flows
- queue a telemetry write behind subtitle lines so it lands past the first batch
- assert imm_lifetime_anime picks up linesSeen/activeMs/cards only after the full queue drains
- Extract the mergeAnime/moveVideoToAnime queue-drain test into immersion-tracker-write-queue.test.ts
- Factor shared setup (seedTwoEntries, queueSubtitleLines, countLinesForAnime) into helpers and add a moveVideoToAnime coverage case
- Replace single flushNow() with drainWriteQueue loop so forced telemetry appended after a full batch isn't left unwritten before merge/move/rebuild summaries recompute
- Add dialog a11y to AnimeMergeDialog/LibraryEntryPicker: aria-modal, labelled headings, alert roles for errors, labelled search input, close button labels
- flush pending telemetry before merge/move so in-progress session time isn't dropped from lifetime totals
- repoint subtitle lines by video_id instead of anime_id so lines recorded before the async title parse assigns a link aren't stranded
- stop absorbing metadata into the target when moving an episode out of an emptied entry (a move isn't a same-show claim)
- return 404 only for missing episode/target, not storage failures; return 404 when a merge folds nothing
- keep AnimeMergeDialog open while a merge is in flight instead of letting dismiss race the request
- surface library load failures in LibraryEntryPicker instead of showing an empty list
- Add multi-select "Merge Selected" flow to fold duplicate library cards into one, preserving sessions, mined cards, and watch time
- Add per-episode "move to another entry" action for reassigning stray episodes, pruning the source entry when emptied
- Auto-merge library entries that resolve to the same AniList id when their parsed seasons are compatible
- Add mergeAnime/moveVideoToAnime service methods, stats-server routes, and HTTP contract types
The query test suite opened each temp database with SQLite's defaults
(rollback journal, synchronous=FULL) while the runtime opens them via
applyPragmas (WAL, synchronous=NORMAL). Every statement therefore ran as
its own fsync-ing transaction.
That is unnoticeable on a fast local disk but dominates on CI: the file
took 40.6s there versus 0.5s locally, and the 106-insert case
'getVocabularyStats pages past hidden rows' crossed the 5s per-test
timeout and failed the quality gate.
Open test databases through a helper that applies the same pragmas as
production, so the tests exercise the runtime's actual SQLite
configuration instead of a slower one.
- Add allowReadingOnlyMatch:false to kanji token known-word lookups so 渓谷/けいこく no longer matches a mined 警告/けいこく card via reading-only index
- Greedy name pre-pass yields when a strictly longer generic word starts at the same position (空 no longer splits 空気; ties still go to the name)
- 確かに and やはり no longer excluded from frequency/JLPT highlighting and vocab stats
- Stop-word list now covers only interjections, pronouns, and grammar fragments
sudacode
