- Add utils/cache.py with invalidate_cache_patterns and smart_cached_view
- Support cache bypass on Cache-Control: no-cache headers and query params
- Add automatic signals (post_save, post_delete, m2m_changed) for library, video, podcast, article, calendar, and hadis
- Hook cache invalidation in custom admin viewset actions
- video: remove Q(playlist_items__video__id__in=...) to strictly filter by bookmarked playlist ID
- podcast: ensure playlist bookmark filter targets playlist IDs and accepts both PODCAST_PLAYLIST and PODCAST
- security/correctness: return empty queryset if is_bookmark is requested without authentication across video, podcast, library, and article views
- Add request-level user bookmark memoization in BookmarkStatusSerializer to eliminate redundant queries across all lists
- Optimize BookReferenceSyncView and BookReference.authors caching (reduced queries from 185 to 10)
- Add multi-level prefetch_related for references and images in HadisMainListView (reduced queries from 109 to 6)
- Add incremental sync support (updated_after/since) in HadisSyncView
- Add select_related for language in BookListView (reduced queries from 35 to 18)
- Annotate episode count and cache effective_thumbnail in podcast and video playlist views (reduced queries from 22/17 to 4)
- Prefetch categories with annotated article counts in ArticleListAPIView (reduced queries from 41 to 3)
- Scale Gunicorn workers/threads to 4w/4t (16 concurrency) with gthread in docker-compose.prod.yml
- Add Redis caching (2h TTL) to categories, collections, and pinned collections in video, library, article, podcast, and dobodbi_calendar (6h TTL)
- Implement conditional_cached_view (30m TTL) for list endpoints (books, podcasts, articles, videos, playlists) that caches public queries and bypasses caching when is_bookmark=true or is_bookmarked=true
- Remove bookmark/user count queries from pinned collection list views to avoid user-specific cache pollution
- Prevent user bookmark/rate leakage into public catalog cache in BookListView via AnonymousUser context wrapper
- Fix video bookmark filtering bug in VideoPlaylistListAPIView and VideoListAPIView to accept both is_bookmark and is_bookmarked, and query both VIDEO_PLAYLIST and VIDEO bookmarks
- Add assign_random_videos_to_playlists management command
- Create reusable watermark utility in utils/watermark.py supporting PIL, UploadedFile, bytes, and DRF serializer fields
- Default position set to bottom-left with responsive 32% width scaling and alpha blending
- Auto-resolve SVG watermark to high-resolution PNG
- Integrate automatic watermarking into Article thumbnails, Hadith/Book Reference images, Corrections, Original Texts, and Interpretations
- Add OptionalPagination to Dovodi Category and Collection viewsets (podcast, video, article)
- Support status query filtering (status=true/false) in category and collection viewsets
- Sort categories and collections by newest first (-created_at, -id) across podcast, video, article, and library admin viewsets
- Make slug optional in AdminArticle serializers with auto-generation fallback
- Support unicode and URL-encoded slugs with case-insensitive lookup in ArticleDetailAPIView
- Count only active books (status=True) in library Category.books_count and CategoryListView
- Introduced `SimilarArticleSerializer` to provide a structured representation of similar articles, including categories and thumbnail.
- Updated `ArticleDetailSerializer` to include a new `similar_articles` field, which fetches articles from the same categories, excluding the current article.
- Implemented a method to retrieve similar articles, improving the article detail view with relevant content suggestions.
- Introduced the `EmbeddingSession` model to track the status and progress of embedding tasks, including fields for status, progress, processed items, and error messages.
- Created `EmbeddingSessionAdmin` to manage embedding sessions in the admin interface, featuring custom display methods for status badges and progress bars.
- Implemented background processing to trigger a FastAPI agent upon creating a new embedding session, enhancing the system's responsiveness and user experience.
- Added migrations to establish the new model in the database.
- Introduced a `share_link` property in the `Article`, `Hadis`, `Book`, `Podcast`, `Video`, and `Transmitter` models to generate shareable links based on their slugs.
- Updated corresponding serializers to include the `share_link` field as read-only.
- Added a management command to refresh share links for existing Hadis models, ensuring all instances have updated slugs and share links.
- Adjusted the `settings` to define `DOVODI_DOMAIN` for consistent link generation across models.
- Changed URL configurations in article, course, and transaction apps to use re_path for improved regex matching of slugs.
- This update enhances flexibility in slug handling across various endpoints, ensuring consistency in URL patterns.
- Added new models for `ContentPart` and `TextSection` to support structured article content, allowing for bilingual Arabic text and translations.
- Updated serializers to include nested text sections within content parts, enhancing the API response structure.
- Introduced a management command to seed article data with Russian content about Imams and Prophets, including categories, collections, and structured content sections.
- Enhanced the admin interface to manage text sections within content parts effectively.
- Created a sample API response for articles to demonstrate the new structure.
- Updated `ArticleCollectionAdminBase` and `PodcastCollectionAdminBase` to include a new `get_display_position` method for better visual distinction between pinned and regular collections.
- Modified verbose names for `PinnedArticleCollection` and `MiddleArticleCollection` to clarify their purpose in the admin interface.
- Added a new README for podcast collections to provide guidance on managing collections and highlight key differences between pinned and regular collections.
- Updated settings to improve sidebar navigation for article and podcast collections, ensuring clear access to both collection types.
H.shokhmgar86
