Adds a new job type that refreshes metadata from external providers for
all series already linked via approved external_metadata_links. Tracks
and displays per-field diffs (series and book level), respects locked
fields, and provides a detailed change report in the job detail page.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add a new metadata row to the dashboard with three cards:
- Series metadata coverage (linked vs unlinked donut)
- Provider breakdown (donut by provider)
- Book metadata quality (summary and ISBN fill rates)
Includes API changes (stats.rs), frontend types, and FR/EN translations.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Bedetheque uses "Légendaires (Les) - Résistance" while local series
names are "Les légendaires - Résistance". Add normalize_title() that
strips leading articles and articles in parentheses before comparing,
so these forms correctly produce a 100% confidence match.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
When multiple provider results exist but the best has 100% confidence,
compare local book count with external total_volumes. If they match,
treat it as an auto-match and link+sync series and book metadata
automatically instead of requiring manual review.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add `metadata_provider` query param to series API endpoints (linked/unlinked/specific provider)
- Return `metadata_provider` field in series response
- Add metadata filter dropdown on series page with all provider options
- Show small provider icon badge on linked series cards
- LiveSearchForm now wraps filters on two rows when needed
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add metadata_batch job type with background processing via tokio::spawn
- Auto-apply metadata only when single result at 100% confidence
- Support primary + fallback provider per library, "none" to opt out
- Add batch report/results API endpoints and job detail UI
- Add series_status and has_missing filters to both series listing pages
- Add GET /series/statuses endpoint for dynamic filter options
- Normalize series_metadata status values (migration 0036)
- Hide ComicVine provider tab when no API key configured
- Translate entire backoffice UI from English to French
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add series status concept (ongoing/ended/hiatus/cancelled/upcoming)
with normalization across all providers
- Add status field to series_metadata table (migration 0033)
- AniList: use chapters as fallback for volume count on ongoing series,
add books_message when both volumes and chapters are null
- Bedetheque: extract description from meta tag, genres, parution status,
origin/language; rewrite book parsing with itemprop microdata for
clean ISBN, dates, page counts, covers; filter placeholder authors
- Add comprehensive e2e provider tests with field coverage reporting
- Wire status into EditSeriesForm, MetadataSearchModal, and series page
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
When books have no volume number, use their 1-based position in the
backoffice sort order (volume ASC NULLS LAST, natural title sort) as
effective volume for matching against external provider books.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Pre-fetch all local books in one query instead of N queries per external
book. Match by volume number first, then bidirectional title containment
(external in local OR local in external). Track matched IDs to prevent
double-matching.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add a complete metadata synchronization system allowing users to search
and sync series/book metadata from external providers (Google Books,
Open Library, ComicVine, AniList, Bédéthèque). Each library can use a
different provider. Matching requires manual approval with detailed sync
reports showing what was updated or skipped (locked fields protection).
Key changes:
- DB migrations: external_metadata_links, external_book_metadata tables,
library metadata_provider column, locked_fields, total_volumes, book
metadata fields (summary, isbn, publish_date)
- Rust API: MetadataProvider trait + 5 provider implementations,
7 metadata endpoints (search, match, approve, reject, links, missing,
delete), sync report system, provider language preference support
- Backoffice: MetadataSearchModal, ProviderIcon, SafeHtml components,
settings UI for provider/language config, enriched book detail page,
edit forms with locked fields support, API proxy routes
- OpenAPI/Swagger documentation for all new endpoints and schemas
Closes#3
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace sccache with a two-stage build strategy: dummy source files cache
dependency compilation in a separate layer, and --mount=type=cache for
Cargo registry/git/target directories. Source-only changes now skip
full dependency recompilation.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Fix Meilisearch indexing to use authors[] array instead of scalar author field
- Join series_metadata to include series-level authors in search documents
- Configure searchable attributes (title, authors, series) in Meilisearch
- Convert EditSeriesForm and EditBookForm from inline forms to modals
- Add tabbed navigation (General / Integrations) to Settings page
- Add Force Search Resync button (POST /settings/search/resync)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Adds Komga sync feature to import read status from a Komga server.
Books are matched by title (case-insensitive) with series+title primary
match and title-only fallback. Sync reports are persisted with matched,
newly marked, and unmatched book lists. UI shows check icon for newly
marked books, sorted to top. Credentials (URL+username) are saved
between sessions. Uses HashSet for O(1) lookups to handle large libraries.
Closes#2
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The total_authors stat now combines distinct values from the legacy
author column and the new authors array column.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Nouveaux endpoints PATCH /books/:id et PATCH /libraries/:id/series/:name pour éditer les métadonnées
- GET /libraries/:id/series/:name/metadata pour récupérer les métadonnées de série
- Ajout du champ `authors` (Vec<String>) sur les structs Book/BookDetails
- 3 migrations : table series_metadata, colonne authors sur series_metadata et books
- Composants EditBookForm et EditSeriesForm dans le backoffice
- Routes API Next.js correspondantes
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Remplace REGEXP_REPLACE(title, '[0-9]+', '', 'g') par
REGEXP_REPLACE(title, '[0-9].*$', '') pour n'extraire que le préfixe
avant le premier chiffre.
L'ancienne regex supprimait TOUS les chiffres du titre entier, y compris
dans les sous-titres. Quand chaque volume a un sous-titre différent
(ex: "Deadpool marvel deluxe 3 - Je suis ton homme"), la partie texte
restante variait → l'ordre alphabétique des sous-titres prenait le dessus.
Avec le nouveau préfixe "deadpool marvel deluxe " identique pour tous,
le numéro de volume (2ème clé) dicte correctement l'ordre.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Ajout de `b.volume NULLS LAST` comme première clé de tri dans list_books
et dans tous les ROW_NUMBER() OVER (...) des CTEs series, pour corriger
l'ordre des volumes dont les titres varient en format (ex: "Round" vs "R")
- Suppression de l'ancienne extract_page publique et de ses 4 helpers
(extract_cbz_page_n, extract_cbz_page_n_streaming, extract_cbr_page_n,
extract_pdf_page_n) remplacés par la nouvelle implémentation avec cache
- Suppression de archive_index_cache dans AppState (remplacé par le cache
statique CBZ_INDEX_CACHE dans parsers), import StdMutex nettoyé
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Chaque cold render ré-énumérait toutes les entrées ZIP/RAR pour construire
la liste triée des images. Maintenant la liste est mise en cache dans l'AppState
(LruCache<String, Arc<Vec<String>>>, std::sync::Mutex pour accès spawn_blocking).
Nouvelles fonctions dans parsers :
- list_archive_images(path, format) -> Vec<String>
- extract_image_by_name(path, format, name) -> Vec<u8>
Mesures avant/après (cache disque froid, n=20) :
- CBZ cold : 43ms → 11.9ms (-73%)
- CBR cold : 46ms → 11.0ms (-76%)
- Warm/concurrent : identique
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Add sort=latest option to GET /books and GET /series API endpoints,
and expose a Sort select in the backoffice books and series pages.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add GET /stats API endpoint with collection overview, reading status,
format/library breakdowns, top series, and monthly additions.
Replace static home page with interactive dashboard featuring donut
charts, bar charts, and progress bars. Use distinct colors for series
(warning/yellow) across nav, page titles, and quick links.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- API: add POST /series/mark-read to batch mark all books in a series
- API: add GET /series cross-library endpoint with search, library and status filters
- API: add library_id to SeriesItem response
- Backoffice: mark book as read/unread button on book detail page
- Backoffice: mark series as read/unread button on series cards
- Backoffice: new /series top-level page with search and filters
- Backoffice: new /libraries/[id]/series/[name] series detail page
- Backoffice: opacity on fully read books and series cards
- Backoffice: live search with debounce on books and series pages
- Backoffice: reading status filter on books and series pages
- Fix $2 -> $1 parameter binding in mark-series-read SQL
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Two new read routes for the home screen:
- /series/ongoing: partially read series sorted by last activity
- /books/ongoing: next unread book per ongoing series
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add POST /admin/tokens/{id}/delete endpoint that permanently removes
a token from the database (only if already revoked). Add delete button
in backoffice UI for revoked tokens.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Pages: mode Original (zero-transcoding), ETag/304, cache index CBZ,
préfetch next 2 pages, filtre Triangle par défaut
- Thumbnails: DCT scaling JPEG via jpeg-decoder (decode 7x plus rapide),
img.thumbnail() pour resize, support format Original, fix JPEG RGBA8
- API fallback thumbnail: OutputFormat::Original + DCT scaling au lieu
de WebP full-decode, retour (bytes, content_type) dynamique
- Watcher: remplacement notify par poll léger sans inotify/fd,
skip poll quand job actif, snapshots en mémoire
- Jobs: mutex exclusif corrigé (tous statuts actifs, tous types exclusifs)
- Robustesse: suppression fs::canonicalize (problèmes fd Docker),
list_folders avec erreurs explicites, has_children default true
- Backoffice: FormRow items-start pour alignement inputs avec helper text,
labels settings clarifiés
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Parsers: raw ZIP reader (flate2) contournant la validation CRC32 des
Unicode extra fields (0x7075) qui bloquait certains CBZ
- Parsers: nouvelle API publique extract_page() pour extraire une page
par index depuis CBZ/CBR/PDF avec fallbacks automatiques
- API: suppression du code d'extraction dupliqué, délégation à parsers::extract_page()
- API: retrait des dépendances directes zip/unrar/pdfium-render/natord
- Indexer: nettoyage Meili systématique à chaque sync (au lieu de ~10%)
avec pagination pour supporter les grosses collections — corrige les
doublons dans la recherche
- Indexer: retrait de la dépendance rand (plus utilisée)
- Backoffice: popin jobs rendue via createPortal avec positionnement
dynamique — corrige le débordement desktop et le header cassé en mobile
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Migration 0020 : colonne format sur books, backfill depuis book_files
- batch.rs / scanner.rs : l'indexer écrit le format dans books
- books.rs : format dans BookItem + filtre ?format= dans list_books
- perf_pages.sh : benchmarks par format CBZ/CBR/PDF
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Une seule entrée illisible dans le central directory ne doit pas bloquer
l'analyse de tout le livre. Le count et la première page lisible sont
retournés même si certaines entrées sont endommagées.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- zip 8.x résout nativement les extra fields NTFS (source du bug EOCD)
- notify 8.x améliore le support inotify Linux
- lopdf 0.39 contient des correctifs de parsing PDF
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Les ZIP créés par des outils Windows (version 6.3) contiennent des extra
fields NTFS (tag 0x000A) qui font échouer ZipArchive::new() avec "Could
not find EOCD". Ajout d'un fallback via read_zipfile_from_stream qui lit
les local file headers sans dépendre du central directory.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Même correctif que dans le parsers/indexer : un .cbr qui est en réalité
un ZIP (et vice-versa) retourne maintenant la bonne page au lieu d'un 500.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- cancel_job: ajouter 'extracting_pages' aux statuts annulables
- cleanup_stale_jobs: couvrir 'extracting_pages' et 'generating_thumbnails' au redémarrage
- analyzer: ne pas régénérer le thumbnail si déjà existant (skip sub-phase B)
- analyzer: supprimer les dotfiles macOS (._*) encore en DB
- SSE backoffice: réduire le spam de logs en cas d'API injoignable
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Added a new job status 'extracting_pages' to represent the first sub-phase of thumbnail generation.
- Updated the database schema to include a timestamp for when thumbnail generation starts.
- Enhanced job progress components to handle the new status, including UI updates for displaying progress and status labels.
- Refactored job-related logic to accommodate the two-phase process: extracting pages and generating thumbnails.
- Adjusted SQL queries and job detail responses to include the new fields and statuses.
This change improves the clarity of job processing states and enhances user feedback during the thumbnail generation process.
CBR: extract_cbr_page extrayait TOUT le CBR sur disque pour lire une
seule page. Reécrit avec le crate unrar : listing en mémoire + extraction
ciblée de la page demandée uniquement. Zéro subprocess, zéro temp dir.
PDF: render_pdf_page utilisait pdftoppm subprocess + temp dir. Reécrit
avec pdfium-render in-process. Zéro subprocess, zéro temp dir.
CBZ: sort naturel (natord) pour l'ordre des pages.
Dockerfile API: retire unar et poppler-utils, ajoute libpdfium.so.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Déplace les migrations du service `migrate` séparé vers un entrypoint.sh
- L'API exécute `sqlx migrate run` au démarrage avant de lancer le binaire
- Gestion de la rétrocompatibilité : détecte un schéma pre-sqlx et crée
une baseline `_sqlx_migrations` pour éviter les conflits sur les instances existantes
- Installe sqlx-cli dans le builder, copie le binaire et les migrations dans l'image finale
- Supprime le service `migrate` du docker-compose.yml ; l'indexer dépend maintenant
du healthcheck de l'API (qui garantit que les migrations sont appliquées)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- meili.rs: corrige la désérialisation de la réponse paginée de
Meilisearch (attendait Vec<Value>, l'API retourne {results:[...]}) —
la suppression des documents obsolètes ne s'exécutait jamais, laissant
d'anciens UUIDs qui généraient des 404 sur les thumbnails
- books.rs: fallback sur render_book_page_1 si le fichier thumbnail
n'est plus accessible sur le disque (au lieu de 500)
- pages.rs: retourne 404 au lieu de 500 quand le fichier CBZ est absent
- search.rs + api.ts + BookCard: ajout série hits, statut lecture,
pagination OFFSET, filtre reading_status, et placeholder onError
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
API:
- Remplace cursor par page (1-indexé) + OFFSET sur GET /books et GET /libraries/:id/series
- BooksPage et SeriesPage retournent total, page, limit
- GET /libraries/:id/series supporte ?q pour filtrer par nom (ILIKE)
Backoffice:
- Remplace CursorPagination par OffsetPagination sur les 3 pages de liste
- Adapte fetchBooks et fetchSeries (cursor → page)
- Met à jour les types BooksPageDto, SeriesPageDto, SeriesDto, BookDto
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- fix(auth): parse_prefix supporte les préfixes de token contenant '_'
- feat: GET /books expose reading_status, reading_current_page, reading_last_read_at
- feat: GET /books accepte ?reading_status=unread,reading (CSV multi-valeur)
- feat: SeriesItem expose books_read_count pour dériver le statut de lecture
- feat: GET /libraries/:id/series accepte ?reading_status=unread,reading
- feat: BooksPage et SeriesPage exposent total (count matchant les filtres)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- API : nouvelle table book_reading_progress (migration 0016) et module
reading_progress.rs avec GET/PATCH /books/:id/progress (token read)
- API : GET /books/:id enrichi avec reading_status, reading_current_page,
reading_last_read_at via LEFT JOIN
- Backoffice : badge de statut (Non lu / En cours · p.N / Lu) sur la page
de détail et overlay sur les BookCards
- OpenSpec : change reading-progress avec proposal/design/specs/tasks
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Ajoute migration 0015 : colonne phase2_started_at sur index_jobs
- Indexer : renseigne phase2_started_at lors du passage à generating_thumbnails
- API : expose phase2_started_at et book_id dans IndexJobDetailResponse
- Page détail : timeline avec durée de chaque phase (Discovery / Thumbnails)
- Page détail : banners contextuels (success/failed/cancelled) avec résumé en une ligne
- Page détail : description textuelle du type de job, durée dans l'overview
- Page détail : stats normalisées selon le type (index vs thumbnail-only)
- JobRow : affiche le type via JobTypeBadge (cohérence visuelle)
- Badge : labels lisibles pour tous les types de jobs
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Ajout de DynamicSettings dans AppState (Arc<RwLock>) chargé depuis la DB
- rate_limit_per_second, timeout_seconds : plus hardcodés, lus depuis settings
- image_processing (format, quality, filter, max_width) : appliqués comme
valeurs par défaut sur les requêtes de pages (overridables via query params)
- cache.directory : lu depuis settings au lieu de la variable d'env
- update_setting recharge immédiatement le DynamicSettings en mémoire
pour les clés limits, image_processing et cache (sans redémarrage)
- parse_filter() : mapping lanczos3/triangle/nearest → FilterType
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Ajoute la possibilité de convertir un livre CBR en CBZ depuis le backoffice.
La conversion est sécurisée : le CBR original n'est supprimé qu'après vérification
du CBZ généré et mise à jour de la base de données.
- parsers: nouvelle fn `convert_cbr_to_cbz` (unar extract → zip pack → vérification → rename atomique)
- api: `POST /books/:id/convert` crée un job `cbr_to_cbz` (vérifie format CBR, détecte collision)
- indexer: nouveau `converter.rs` dispatché depuis `job.rs`
- backoffice: bouton "Convert to CBZ" sur la page détail (visible si CBR), label dans JobRow
- migrations: colonne `book_id` sur `index_jobs` + type `cbr_to_cbz` dans le check constraint
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Add missing utoipa annotations:
- GET /books/{id}/thumbnail
- GET/POST /settings, /settings/{key}
- POST /settings/cache/clear
- GET /settings/cache/stats, /settings/thumbnail/stats
Add 'settings' tag and register all new schemas.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Phase 1 (discovery): walkdir + filename-only metadata, zero archive I/O.
Books are visible immediately in the UI while Phase 2 runs in background.
Phase 2 (analysis): open each archive once via analyze_book() to extract
page_count and first page bytes, then generate WebP thumbnail directly in
the indexer — removing the HTTP roundtrip to the API checkup endpoint.
- Add parse_metadata_fast() (infallible, no archive I/O)
- Add analyze_book() returning (page_count, first_page_bytes) in one pass
- Add looks_like_image() magic bytes check for unrar p stdout validation
- Add lsar fallback in list_cbr_images() for UTF-16BE encoded filenames
- Add directory_mtimes table to skip unchanged dirs on incremental scans
- Add analyzer.rs: generate_thumbnail, analyze_library_books, regenerate_thumbnails
- Remove run_checkup() from API; indexer handles thumbnail jobs directly
- Remove api_base_url/api_bootstrap_token from IndexerConfig and AppState
- Add unar + poppler-utils to indexer Dockerfile
- Fix smoke.sh: wait for job completion, check thumbnail_url field
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Add CLAUDE.md at root and AGENTS.md in apps/api, apps/indexer,
apps/backoffice, crates/parsers with module-specific guidelines
- Unify all service ports to 70XX (no more internal/external split):
API 7080, Indexer 7081, Backoffice 7082
- Update docker-compose.yml, Dockerfiles, config.rs defaults,
.env.example, backoffice routes, bench.sh, smoke.sh
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Change all instances of AppState to reference the new state module across multiple files for consistency.
- Clean up imports in auth, books, index_jobs, libraries, pages, search, settings, thumbnails, and tokens modules.
- Simplify main.rs by removing unused code and organizing middleware and route handlers under the new handlers module.
- Extend thumbnail regeneration logic to support full rebuilds, allowing for the deletion of orphaned thumbnails.
- Implement database updates to clear thumbnail paths for books during regeneration and full rebuild processes.
- Improve logging to provide detailed insights on the number of deleted thumbnails and cleared database entries.
- Refactor code for better organization and clarity in handling thumbnail files.
- Introduce dynamic loading of concurrent render limits from the database for both page rendering and thumbnail generation.
- Update API to utilize the loaded concurrency settings, defaulting to 8 for page renders and 4 for thumbnails.
- Modify front-end settings page to reflect changes in concurrency limits and provide user guidance on their impact.
- Ensure that changes to limits require a server restart to take effect, with clear messaging in the UI.
