Users were incorrectly redirected to /login during the initial loading phase
before the loading was actually complete. Added loading_started and
loading_complete flags to properly track the initialization state so
the setup redirect middleware knows when it's safe to redirect.
_execute_nfo_scan() was importing get_anime_service from anime_service.py
which is a factory requiring series_app argument. Changed to import from
dependencies.py which handles series_app internally and provides proper
dependency injection with caching.
- Add /api/setup/unresolved/done endpoint to mark phase complete
- NFO scan now runs after series sync during initialization
- Middleware redirects to /login after setup complete (was /loading)
- Done button allows skipping folder resolution with redirect to NFO scan phase
- Add check for existing series by key in SetupService.run to skip duplicates
- Fix Path construction in initialization_service.py cleanup function
- Update unit tests to mock get_by_key and get_series_app
When the search provider returns a link like 'shinobi-no-ittoki' instead of
'/anime/stream/shinobi-no-ittoki', the key was not being extracted and all
folders were marked as unresolved.
Now handles both link formats:
- URL format: '/anime/stream/key' -> extract key
- Direct format: 'key' -> use as-is
Also added debug logging for both resolution paths to aid troubleshooting.
When SetupService cannot auto-resolve a provider key for an anime folder,
the folder is now tracked in the new 'unresolved_folders' table instead of
being silently skipped. Users can then resolve these via the new API:
- GET /api/setup/unresolved - list unresolved folders with search suggestions
- POST /api/setup/unresolved/{folder}/resolve - provide key to resolve folder
The SetupService.run() now:
- Tracks unresolved folders instead of skipping them
- Re-creates AnimeSeries for previously unresolved folders that are now resolved
- Includes unresolved count in logs
New files:
- src/server/api/setup_endpoints.py - API endpoints for unresolved management
- tests/unit/test_unresolved_folder_service.py - service and model tests
Modified:
- src/server/database/models.py - add UnresolvedFolder model
- src/server/database/service.py - add UnresolvedFolderService
- src/server/services/setup_service.py - track unresolved folders
- src/server/fastapi_app.py - include setup router
- Add _normalize_title() to strip anime suffixes (TV, OVA, Movie, etc.)
- Add _titles_match() using SequenceMatcher for similarity (threshold 0.85)
- Replace exact string match with fuzzy match to fix skipped folders
- Add debug logging for title mismatches and multiple results
- Set LOG_LEVEL=DEBUG in docker-compose.yml
- Fix _resolve_key_via_search to use 'title' instead of 'name'
- Extract key from 'link' field URL (e.g., /anime/stream/naruto -> naruto)
- Skip folders with unresolved keys instead of crashing with 'Series key cannot be empty'
- Update tests to use correct field names (title/link)
SetupService.run() now checks each anime folder for tvshow.nfo,
logo.png, and poster/fanart images instead of using hardcoded
defaults. Provider key resolution via search is unchanged.
Extract SetupService class from initialization_service to handle:
- Scan data/ folder subdirectories
- Extract title and year from folder names (YYYY pattern)
- Create AnimeSeries records in database
- Resolve provider keys via search (single exact match)
Updates _scan_folders_to_database() to delegate to SetupService.run().
Adds comprehensive unit tests for SetupService.
- Add _compute_folder_name helper that deduplicates year (handles cases like 'Name (2023)' not becoming 'Name (2023) (2023)')
- Create anime folder on disk when adding series (not just DB + memory)
- Add rename_folder_if_needed to auto-rename existing folders without year
- Fetch year from aniworld_provider and include in folder as 'Name (YYYY)'
Closes: anime folders now include release year when available from provider
- Add nfo_scan_after_rescan config option (default: true)
- Implement year caching in AniworldLoader and EnhancedAniWorldLoader
- Make get_year abstract method in base provider
- Run NFO validation/creation after scheduled rescan completes
- Add _YearDict cache to avoid re-extracting year from HTML
Delete folder_rename_service.py. Stub out get_duplicate_folders API to return
empty response. Update folder_scan_service and tests to skip rename step.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
- Add _scan_folders_to_database() - iterates anime_directory subdirs
- Extract title/year from folder names via (YYYY) pattern
- Resolve provider key via search when single match found
- Create AnimeSeries records for new folders only
- Add corresponding unit tests
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
- Move src/core/ → src/server/
- Split SerieList.py (531 lines) and series.py (414 lines) into src/server/database/
- Add database/models.py for SQLAlchemy models
- Update all test imports to reflect new structure
- Remove deprecated test files (test_serie_class.py, test_serie_folder_with_year.py)
RescanService was thin wrapper. Its logic (rescan, auto-download, folder
scan, WebSocket broadcasts) moved into SchedulerService as private methods.
RescanService and its module deleted.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
- Extract rescan logic into new RescanService (src/server/services/rescan_service.py)
- SchedulerService now only handles APScheduler cron scheduling
- Move scheduler sub-services (folder_rename, folder_scan, key_resolution) to scheduler/ folder
- Keep RescanOrchestrator as backward-compatible alias
- Update all imports across api/, server/, and test files
Update_database_paths and duplicate folder cleanup were using get_by_key()
(provider key lookup) instead of get_by_folder() when operating on folder names.
This caused orphaned DB records when removing duplicate folders like 'Hells Paradise'
that mapped to an already-existing 'Hell\'s Paradise (2023)'.
folder_rename_service depends on clean NFO files but repair tasks
were fire-and-forget. Now collect all repair tasks and await them
with asyncio.gather before validate_and_rename_series_folders runs.
Also update tests that mock asyncio.create_task to also mock
asyncio.gather since perform_nfo_repair_scan now awaits tasks.
- Add key_resolution_service.py to resolve provider keys for folders without key/data files
- Search anime provider and match folder names (case-insensitive, exact match required)
- Only save to DB if exactly one match found; otherwise skip
- Add comprehensive unit tests (28 tests)
- Integrate into scheduler_service after nfo_repair scan
- Update ARCHITECTURE.md documentation
- Add is_valid_key check in SerieScanner._read_data_from_file() to prevent
passing invalid keys to Serie constructor (caused ValueError)
- Improve error message for key generation failures
- Add warning log before removing duplicate source folders in rename service
Add configurable folder rename patterns via settings with anime_folder_rename_regex and custom_pattern options. Integrate into SerieScanner and SeriesApp for consistent episode organization.
- SerieScanner: Remove key file fallback, keep data file fallback
- SystemSettings: Add legacy_key_cleanup_completed flag
- initialization_service: Add cleanup task to remove key files from folders with DB entries
- Tests updated to reflect key file removal from legacy path
Key files caused duplicate key errors on folder rename. DB is now sole source of truth.
- Add DuplicateFolderGroup and DuplicateFoldersResponse Pydantic models
- Add /duplicate-folders GET endpoint for listing pre-existing duplicates
- Add _scan_for_pre_existing_duplicates() function for NFO-based detection
- Add _try_merge_duplicate_group() for auto-merging empty/symlink-only duplicates
- Integrate duplicate detection into validate_and_rename_series_folders workflow
- Skip rename for flagged duplicates to prevent data loss during merge
- Add _cleanup_orphaned_folder() to delete/move old folder contents after rename
- Empty folders: delete directly via rmdir()
- Non-empty folders: move contents to new path, then delete old folder
- Handle PermissionError and OSError gracefully with logging
- Add dry_run parameter to preview changes without applying them
- Add --dry-run support to validate_and_rename_series_folders()
- Add unit tests for _cleanup_orphaned_folder and dry-run mode
- All 66 related tests pass
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
SchedulerConfig.__init__ maps legacy auto_download/folder_scan keys to the
primary auto_download_after_rescan/folder_scan_enabled fields. However,
model_dump() was including auto_download=null and folder_scan=null in
serialised output. When this was written to config.json and reloaded,
those keys were present (albeit null), so the alias mapping was skipped
and the primary fields retained default False values instead of the
configured True values.
Fix:
- Override SchedulerConfig.model_dump() to drop None-valued alias fields
before returning the serialised dict.
- ConfigService.save_config() re-serialises the scheduler field through
its overridden model_dump() so the fix applies when writing to disk.
Tests added:
- test_roundtrip_excludes_none_alias_fields: verifies model_dump omits
null auto_download/folder_scan keys.
- test_save_and_load_scheduler_flags_roundtrip: end-to-end roundtrip
through ConfigService confirms raw JSON and loaded values match.
Pre-existing failure in test_core_error_handler.py is unrelated.
- SerieScanner: generate key from folder when no key/data files exist
- Handle edge cases: non-Latin characters, special symbols in folder names
- anime_service: expose loading_status and loading_error fields
- Update tests to match new fallback behavior
- Add on_warning event system with subscribe/unsubscribe methods
- Change duplicate key handling from error to warning
- Fire on_warning event when duplicate series detected
- Include metadata: key, duplicate_folder, existing_folder
Scheduler used a separate SQLite file (scheduler.db) only to persist one
cron job. This was originally required because APScheduler's
SQLAlchemyJobStore is sync-only, creating an async/sync driver conflict
when accessing the same file.
The job is rebuilt from config.json on every startup regardless
(replace_existing=True), so the persisted state only served misfire
detection. Moved misfire detection into the app layer by querying
system_settings.last_scan_timestamp on startup: if the last scan is
>23h but <25h ago, an immediate rescan is triggered.
Change summary:
- Remove SQLAlchemyJobStore; use default MemoryJobStore instead
- Add _check_missed_run() that reads last_scan_timestamp from aniworld.db
- Update docs/DEVELOPMENT.md scheduler troubleshooting section
- Update the scheduler unit test that verified SQLAlchemyJobStore
When episodes are downloaded successfully, the in-memory Serie.episodeDict
is updated, but the deprecated data file was not being synced. This caused
UI to show episodes as missing when already downloaded.
Changes:
- Update data file in _remove_episode_from_memory when download completes
- DB is authoritative; data file is optional backup (deprecated)
- Gracefully skip update if data file doesn't exist
New integration tests for episode download sync:
- Verify episode removed from missing list after download
- Verify in-memory cache updated after download
- Verify data file updated after download (when it exists)
- Verify downloads work without data file
- Rename function to reflect its legacy status
- Add deprecation warning log on execution
- Update all callers (initialization_service, api/config, fastapi_app)
- Update tests to use new name
- Add deprecation notice to DEVELOPMENT.md
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
- Add migration_legacy_files_completed flag to SystemSettings model
- Create legacy_file_migration service to migrate series from key/data files
- Integrate legacy migration into initialization_service startup flow
- Add integration tests for legacy file migration
- Update DATABASE.md documentation with migration details
- Fix various test and service issues (nfo_repair, tmdb_client, download_service)
- Add test_database_schema unit tests
Add ensure_started() to SchedulerService as idempotent entry point.
Start scheduler in auth setup run_initialization() after NFO scan.
Sync anime_directory and start scheduler in config update endpoint.
Add unit and endpoint tests for ensure_started() behavior.
- Switch scheduler_service from structlog to standard logging for consistency
- Add detailed lifecycle logging in SchedulerService (start, stop, rescan)
- Add debug logging in fastapi_app scheduler initialization
- Fix test_add_series_episodes to mock EpisodeService.get_by_series
Use the database as the authoritative source for missing-episode lists so
that episodes marked is_downloaded=True are never shown as missing, even
when the in-memory state is stale.
Key changes:
- EpisodeService.get_by_series() gains only_missing flag
- AnimeService uses DB-backed episodeDict and preserves downloaded episodes
during sync, skipping them when adding/removing missing episodes
- DownloadService broadcasts series_updated after marking an episode downloaded
so the frontend reflects the change immediately
- Frontend filters out series with zero missing episodes client-side and
fixes renderSeries to respect the active filter
- Unit tests updated to assert the broadcast is sent
- Add missing year field when building series list in anime_service
- Add _create_missing_nfo to generate minimal NFO for series without one
- Update perform_nfo_repair_scan to detect and create missing NFOs
- Add semaphore-protected async creation with TMDB rate limiting
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
- New migration script: populate year from folder (YYYY) pattern
- SerieScanner: refactor year extraction logic
- anime_service: pass year when syncing from data files
Retry count and queue status were in-memory only and lost on
restart, so failed downloads could not be safely resumed and
permanently-failed episodes silently blocked re-queueing via the
episode-id unique index.
- Add status + retry_count columns to DownloadQueueItem
- Replace unique(episode_id) with unique(episode_id, status) so
permanently_failed rows do not block new pending entries
- Add PERMANENTLY_FAILED to DownloadStatus enum
- Persist retry_count on each failure; mark permanently_failed once
max_retries reached
- QueueRepository reads status/retry_count from DB instead of
defaulting to PENDING/0
- Stop double-incrementing retry_count in retry_failed_items;
increment only happens in _process_download on failure
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Change _remove_episode_from_missing_list to set is_downloaded=True
and populate file_path via EpisodeService.mark_downloaded, instead of
deleting the Episode row. Preserves download history so queries can
distinguish series with downloaded episodes from completely unwatched
series.
- Pass serie_folder to construct file_path
- Look up series_id via AnimeSeriesService.get_by_key
- Update tests to mock mark_downloaded path
- Document episode lifecycle in docs/DEVELOPMENT.md
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
- Fix race condition: next_run_time only available after scheduler.start()
- Handle None gracefully in logging
- Add debug logging to _perform_rescan and _run_rescan_job
- Document scheduler troubleshooting in DEVELOPMENT.md
