feat(logging): implémenter la suppression des fichiers de log obsolètes et ajuster la configuration des logs
feat(settings): modifier la structure des paramètres de configuration et ajouter un niveau de log TRACE
feat(admin): fusionner les pages de configuration TMDB et Indexeurs, et mettre à jour les titres et descriptions
fix(system): corriger les messages d'alerte concernant le redémarrage des paramètres système
- add TMDB orphan matcher backend (orphan_matcher.py) and search client (tmdb_search.py)
- introduce Redis-backed run history tracker (history.py) for matcher runs
- extend TorrentItemModel with tmdb_match_attempted_at and index
- add Alembic migration to create tmdb_match_attempted_at and its index
- integrate new tmdb_orphan_matching job into scheduler; expose in UI
- add admin UI: tmdb_matcher.html, history/revert endpoints, and nav entry
- wire settings: scheduler_tmdb_match_interval_hours and batch size
- enhance title normalizer: strip possessive s and remove technical brackets/tokens
- update admin docs to reflect new matcher and deprecation of old cleanup
- update admin base navigation and scheduled tasks page accordingly
introduce initial torrent grouping: add torrent_groups table and
group_id on torrent_items. include migrations, ORM models, and DAOs.
expose admin/web endpoints to group by info_hash and by title/size.
propagate tmdb_id across grouped items. non-destructive and backwards
compatible with existing data
Introduce a comprehensive settings management system that allows runtime
configuration through a new admin interface. The system includes:
- Database-backed settings with Redis caching for performance
- Settings registry defining types, defaults, and categories
- New admin UI with navigation cards and dedicated sub-pages for each category
(general, cache, indexers, proxy, system, TMDB)
- Shared view logic for consistent GET/POST handling across sub-pages
- Automatic seeding of defaults and application of overrides at startup
- Validation and type coercion for setting values
The previous static configuration page has been replaced with a modular,
category-based interface that provides better organization and extensibility
for future configuration needs.
- Implemented a new HTML page for managing scheduled tasks in the admin panel.
- Added JavaScript functionality for auto-refreshing job statuses and triggering jobs manually.
- Enhanced job result display with relative time and duration formatting.
- Introduced a toast notification system for user feedback on job actions.
fix: update Redis cache key generation to include a prefix
- Modified the Redis cache key generation to prepend "search:" to the hashed key for better organization.
refactor: improve Yggflix API search functionality
- Added an optional parameter to the search_movie method to specify the search type.
- Implemented a fallback mechanism to broaden search if initial results are insufficient.
chore: enhance Yggflix service with improved result handling
- Introduced a threshold for retrying searches with a broader query type based on result ratios.
- Increased the limit for sorted results from 50 to 100.
feat: add maintenance endpoints for managing expired API and peer keys
- Created endpoints to disable expired API keys and peer keys in the admin panel.
- Added functionality to clean orphan torrent items based on a specified age.
feat: integrate scheduler into the application lifecycle
- Initialized a database cleanup scheduler during application startup.
- Ensured proper shutdown of the scheduler during application termination.
extend not_cached ttl from 20m to 3h to reduce re-checks
of not cached items
remove the file_index shortcut for series in Torbox
and rely on season/episode parsing
- Implemented a new admin page for managing TMDB mismatches, including listing, creating, and deleting mismatches.
- Added functionality to search for unmatched torrents with filters for language and type.
- Enhanced UI with improved filtering options and group toggling for better user experience.
- Introduced toast notifications for user actions such as creating mismatches and assigning TMDB IDs.
- Updated backend routes to handle mismatch creation, deletion, and assignment of TMDB IDs to selected torrents.
- Refactored title filtering logic in `filter_results.py` to utilize a new title matching module.
- Introduced normalization functions for search queries in `jackett_service.py` and `yggflix_service.py`.
- Added new endpoints for managing title and language normalization rules in the admin views.
- Implemented loading and reloading of title matching and language rules in the application lifecycle.
- Enhanced release group extraction and language detection in `parser_utils.py` using the new title matching module.
Refactor search_movie method to remove tmdb_id parameter and add offset support for pagination. Implement automatic fallback to search type when movie type returns empty results, improving result coverage. Add intelligent page 2 fetching in service layer based on title pre-filter ratio threshold to avoid unnecessary API calls.
Update aiohttp-socks dependency to support versions >=0.10.0 and add integer type conversion for tmdb_id field in TorrentItemModel to ensure proper data handling
Introduce a 4-phase search system where private indexers are grouped by priority:
- Phase 1 (priority_private): always queried first
- Phase 2 (intermediary_private): queried only if private results < minCachedResults
- Phase 3 (fallback_private): queried only if still insufficient
- Phase 4 (public): Yggflix relay position controlled by yggflixPriority toggle
Add rdMinCachedBeforeCheck setting to skip RealDebrid if prior services already found enough cached results, since RD API is slow (~3s) and rate-limited. Default to 3.
Non-blocking magnet cleanup in AllDebrid availability checks via background task. G3MINI now computes info_hash from .torrent files when API omits it. Public indexer results are excluded from final matches unless already cached by a debrid service.
Remove deprecated Sharewood tracker integration and add three new private trackers:
- ABN (Abnormal): keyword-based search with name filtering
- G3MINI (Gemini Tracker): TMDB/IMDB ID-based filtering
- TheOldSchool: TMDB/IMDB ID-based filtering
Includes new API clients, result parsers, and service layers. Updates frontend UI, settings configuration, and credential handling to support the new indexers with proper passkey management.
Implement POST /encode endpoint for encrypting user configuration tokens.
Uses Fernet encryption with automatic Base64 fallback when CONFIG_SECRET_KEY
is not set. Protected by signed CSRF token validation to prevent cross-site
attacks.
- Replace inline HTML templates with Jinja2 templates from static directory
- Add Redis client for rate limiting public registration endpoint
- Implement rate limiting of 5 requests per IP per hour
- Improve API key naming with date-based suffix for better traceability
- Add utility functions for IP hashing and rate limit enforcement
