Session directories are now created only when the first MP3 file is saved, not at session creation. Added EnsureSessionDirectoryExists to ISessionService and SessionService to handle this logic. Updated VideoService to call this method before saving files. Cleanup now logs when no directory exists for expired sessions.
Introduce automatic ZIP file creation and management for session-based downloads.
- Server: Add endpoints to serve and provide info about session ZIPs; VideoService now updates ZIP after each MP3 download with user-friendly filenames.
- Client: Add "Download All in ZIP" button with loading state and ZIP info; trigger browser download via JS interop; responsive and styled UI.
- API: Extend service interfaces for ZIP download/info; add SessionZipInfo model.
- Docs: Update SESSION_MANAGEMENT.md with ZIP workflow, endpoints, and UX details.
- Robust error handling and logging throughout.
Organize downloaded files into per-job subdirectories under "downloads" instead of a flat structure. Update DownloadController and VideoService to read, write, and delete files within these job-specific directories. Add helper methods for directory management and improve logging. Update download/delete endpoints and result URLs to reflect the new structure. Clean up empty job directories after file deletion.
Adopt GUID-based {jobId}_{title}.ext filename convention for all downloads, ensuring unique, human-readable, and traceable files. Update DownloadController to use precise jobId-based file matching and generate user-friendly download names. Pass jobId to VideoService for consistent naming; add robust title sanitization. Remove fallback file search logic for predictability. Update README and API docs to reflect new format and workflow. Improve logging and error messages for clarity.
- Integrate yt-dlp.exe for real YouTube audio download/conversion
- Serve actual MP3 files from wwwroot/downloads via API and static files
- Report real-time progress from yt-dlp to frontend
- Enhance DownloadController: file serving, metadata, deletion endpoints
- Update VideoController to track and return real file info/URLs
- Enable static file serving in Program.cs; update API info/logging
- Update Blazor frontend to use real download URLs and remove old JS logic
- Add README_UPDATED.md with new workflow and troubleshooting
- Improve error handling, logging, and file cleanup throughout backend
- Update .gitignore for downloads and yt-dlp.exe
Refactored backend to extract actual YouTube title and thumbnail from HTML, removing simulated metadata fields (description, author, duration, views, upload date, tags). Simplified UI to display only title and thumbnail, and enhanced visuals with animated musical notes background. Removed related formatting and display logic for unused metadata.
- Add Health, Video, and Download controllers with endpoints for health checks, video details, processing, status, results, and simulated downloads
- Implement IVideoService for YouTube URL validation and mock video info
- Add API models for responses and job status tracking
- Configure DI, CORS (BlazorPolicy), OpenAPI, and root metadata endpoint in Program.cs
- Add README with usage, troubleshooting, and development notes
- Minor project file cleanup and encoding fix
Michal Vrba