real-debrid-downloader/README.md
2026-03-09 04:48:58 +01:00

308 lines
13 KiB
Markdown

# Multi Debrid Downloader
Desktop downloader for Windows with package-based queue management, multi-provider fallback, automatic extraction, auto-rename, provider statistics, and built-in updates.
![Platform](https://img.shields.io/badge/platform-Windows%2010%2F11-0078D6)
![Electron](https://img.shields.io/badge/Electron-31.x-47848F)
![React](https://img.shields.io/badge/React-18.x-149ECA)
![TypeScript](https://img.shields.io/badge/TypeScript-5.x-3178C6)
![License](https://img.shields.io/badge/license-MIT-green)
## Why this tool?
- JDownloader-style workflow with packages, progress, extraction, history, and clean post-processing.
- Multiple debrid accounts in one app, including provider order, automatic fallback, and per-hoster routing.
- Built for large queues with session persistence, retries, reconnect handling, resume support, and integrity checks.
- Includes an in-app updater for releases published on `git.24-music.de`.
## Supported providers
- AllDebrid API
- AllDebrid Web via browser login
- BestDebrid API
- BestDebrid Web via cookie import
- Debrid-Link with multi-key support
- DDownload login
- 1fichier API
- LinkSnappy login
- Mega-Debrid API
- Mega-Debrid Web
- Real-Debrid
## Core features
### Queue and package handling
- Package-based queue with item status, retries, ETA, speed, provider, and account label.
- Start, pause, stop, cancel, reset, rename, and delete for packages and items.
- Ctrl+Click multi-select and bulk actions.
- Queue backup import/export as JSON.
- Context-menu export for selected packages or selected items as structured TXT re-import files.
- Duplicate handling when adding links: keep, skip, or overwrite.
- Optional start scheduling for a specific time.
- Session recovery after restart with optional auto-resume.
- Optional auto-sorting by progress.
### Link collection
- Paste links directly into the collector.
- Import `.txt` export files that preserve package names and optional per-file names.
- Clipboard watcher with automatic link detection.
- `.dlc` import via file picker and drag-and-drop.
- Drag-and-drop of plain links, `.txt` export files, and supported container files.
### Provider routing and fallback
- Configurable provider order with primary, secondary, and tertiary fallback.
- Optional automatic provider fallback on unrestrict/download failures.
- Per-hoster routing override, so specific hosters can always use a specific provider.
- Providers can be disabled without deleting stored account data.
- Daily traffic limits per provider.
- Debrid-Link per-key daily limits and per-key daily usage tracking.
### Accounts and provider tools
- Central Accounts view with account type, status, info, access data, and actions.
- BestDebrid cookie import directly from a Netscape cookies file.
- AllDebrid browser-login flow and in-app Rapidgator host status display.
- Debrid-Link multi-key management with optional detailed line-by-line key display.
- Debrid-Link API-key statistics popup with per-key Rapidgator traffic quota, link quota, reset, activate/deactivate, and click-to-copy masked keys.
- Reset button for stored account column widths in the Accounts table.
### Download engine
- Parallel downloads with resumable transfers when supported.
- Reconnect handling with configurable wait time.
- Circuit-breaker style cooldown and retry handling for provider issues.
- Global speed limit or per-download speed limit mode.
- Bandwidth schedules with time windows and speed caps.
- Live bandwidth chart and session statistics.
- Persistent all-time download counter.
### Extraction and post-processing
- Automatic extraction after download.
- Extraction can continue even when the session is stopped or after app restart.
- Hybrid download + extract workflow.
- Extraction backend using native tools by default, with JVM sidecar support available.
- Supports common archive formats including RAR, ZIP, and 7z.
- Nested extraction for archives found inside extracted output.
- Conflict handling: overwrite, skip, rename, or ask.
- Disk-space validation before extraction.
- Package-scoped password reuse for multi-archive sets.
- Optional cleanup of downloaded archives after extraction.
- Optional cleanup of link artifacts and sample files after extraction.
- Optional flat MKV collection folder after package completion.
### Auto-rename and media cleanup
- Auto-rename for extracted scene-style files based on folder/source naming.
- Multi-episode token parsing.
- Handles compact episode tokens like `s02e01` directly attached to the title.
- Optional skip of already extracted packages on start.
### Integrity, history, and backup
- Optional integrity verification with `CRC32`, `MD5`, and `SHA1`.
- Download history with package details, duration, size, provider, and target folder.
- Backup export/import for restoring app state.
- Persistent config, session, and history files in the Electron `userData` directory.
### UI and desktop integration
- Downloads, history, statistics, and settings tabs.
- Progress bars for packages and single items.
- Hoster/provider display showing both source and effective debrid account.
- Minimize-to-tray support.
- Dark/light theme setting.
- Long path support on Windows.
- Default startup window size of `1920x1080`.
## Installation
### Prebuilt releases
1. Download the latest installer or portable build from the releases page.
2. Start the app.
3. Add your provider credentials in `Settings > Accounts`.
Releases: [git.24-music.de Releases](https://git.24-music.de/Administrator/real-debrid-downloader/releases)
### Build from source
Requirements:
- Node.js `20+`
- npm
- Windows `10/11`
- Java Runtime `8+` for the optional JVM extraction backend
- Optional native extraction tools: 7-Zip / WinRAR / UnRAR
```bash
npm install
npm run dev
```
## NPM scripts
| Command | Description |
| --- | --- |
| `npm run dev` | Starts Vite, tsup watchers, and Electron in development mode |
| `npm run build` | Builds main and renderer bundles |
| `npm run start` | Starts the built app in production mode |
| `npm test` | Runs Vitest unit tests |
| `npm run self-check` | Runs integrated self-checks |
| `npm run release:win` | Builds Windows installer and portable EXE |
| `npm run release:gitea -- <version> [notes]` | Builds, tags, and uploads a release to `git.24-music.de` |
| `npm run release:forgejo -- <version> [notes]` | Alias for the same release workflow |
## Typical workflow
1. Add one or more provider accounts in `Settings > Accounts`.
2. Configure provider order, fallback, and optional hoster routing.
3. Paste links or import `.dlc` files.
4. Adjust package names, target folders, extraction, and cleanup settings if needed.
5. Start the queue and monitor downloads, extraction, and provider status.
6. Review history and statistics after completion.
## Link export format
Selected packages or items can be exported from the context menu as a structured text file. Re-importing that file restores the original package grouping, even if it only contains a subset of items from a larger package.
Example:
```txt
# rd-link-export: 1
# package: Dave Staffel 1
# file: Dave.S01E01.rar
https://example.com/e01
# file: Dave.S01E02.rar
https://example.com/e02
```
Supported import sources:
- collector text input
- `Datei importieren`
- drag-and-drop of `.txt` and `.json`
The optional `# file:` marker preserves the original item name so the imported subset can be rebuilt with the same package name and per-item filename hints.
## Project structure
- `src/main` - Electron main process, download engine, provider clients, updater, storage
- `src/preload` - secure IPC bridge
- `src/renderer` - React UI
- `src/shared` - shared types and IPC contracts
- `tests` - unit and integration-style tests
- `resources/extractor-jvm` - optional JVM extraction runtime
- `scripts` - release and build helpers
## Data and logs
Runtime files are stored in Electron's `userData` directory, including:
- `rd_downloader_config.json`
- `rd_session_state.json`
- `rd_history.json`
- `rd_downloader.log`
- `audit.log`
- `rename.log`
- `debug_ai_manifest.json`
- `trace.log`
- `trace_config.json`
- `session-logs/session_*.txt`
- `package-logs/package_*.txt`
- `item-logs/item_*.txt`
`audit.log`, `rename.log`, and `trace.log` are rotated automatically. The current file is kept plus one `.old` backup, and outdated backups are purged automatically.
### Remote debug server
For headless or server-style troubleshooting, the app can expose a small authenticated HTTP debug API with live status and log tails.
Enable it by creating these files in the same runtime folder that contains `rd_downloader.log`:
- `debug_token.txt`
Example: a long random token such as `rd-debug-please-change-me`
- `debug_port.txt`
Example: `9868`
- `debug_host.txt` (optional)
Default is `127.0.0.1`. Set `0.0.0.0` only if you really want remote access and protect it with firewall, VPN, or reverse proxy.
After startup, the app also writes `debug_ai_manifest.json` into the same runtime folder. This file is meant for support tooling and AI agents: it lists all available endpoints, the auth method, the related runtime files, and the one remaining external value the assistant may still need from you for remote access: the server IP or DNS name.
If you want extra support detail during a flaky or hard-to-reproduce issue, the app also maintains a `trace.log` plus `trace_config.json`. You can enable or disable the support trace from the app menu or remotely via the debug API. By default, the support trace now auto-disables again after 2 hours so it does not stay enabled forever by accident.
The app menu under `Hilfe` also includes a `Debug-Setup prüfen` action. It verifies the current host/port/token/AI-manifest/trace setup locally and now also reports free disk space, current support-log sizes, and an estimated support-bundle size.
Available endpoints after restart:
- `GET /health`
- `GET /meta`
- `GET /debug/setup`
- `GET /self-check`
- `GET /host/diagnostics`
- `GET /status`
- `GET /settings`
- `GET /accounts`
- `GET /stats`
- `GET /history?limit=50&status=completed`
- `GET /packages?package=Release&includeItems=1`
- `GET /items?status=downloading&package=Release`
- `GET /session?package=Release`
- `GET /log?lines=100&grep=keyword`
- `GET /logs/main?lines=100&grep=keyword`
- `GET /logs/audit?lines=100&grep=keyword`
- `GET /logs/rename?lines=100&grep=keyword`
- `GET /logs/trace?lines=100&grep=keyword`
- `GET /logs/session?lines=100&grep=keyword`
- `GET /logs/package?package=Release&lines=100&grep=keyword`
- `GET /logs/item?item=episode.part2.rar&lines=100&grep=keyword`
- `GET /trace/config?enable=1&note=support&durationMinutes=120`
- `GET /support/bundle`
- `GET /diagnostics?package=Release&lines=150`
Authentication works with either:
- header: `Authorization: Bearer <token>`
- query param: `?token=<token>`
Example from PowerShell:
```powershell
Invoke-RestMethod "http://SERVER:9868/diagnostics?token=YOUR_TOKEN&package=Release"
Invoke-RestMethod "http://SERVER:9868/settings?token=YOUR_TOKEN"
Invoke-RestMethod "http://SERVER:9868/accounts?token=YOUR_TOKEN"
Invoke-RestMethod "http://SERVER:9868/stats?token=YOUR_TOKEN"
Invoke-RestMethod "http://SERVER:9868/history?token=YOUR_TOKEN&limit=20"
Invoke-RestMethod "http://SERVER:9868/debug/setup?token=YOUR_TOKEN"
Invoke-RestMethod "http://SERVER:9868/self-check?token=YOUR_TOKEN"
Invoke-RestMethod "http://SERVER:9868/logs/audit?token=YOUR_TOKEN&lines=200"
Invoke-RestMethod "http://SERVER:9868/logs/rename?token=YOUR_TOKEN&lines=200"
Invoke-RestMethod "http://SERVER:9868/logs/trace?token=YOUR_TOKEN&lines=200"
Invoke-RestMethod "http://SERVER:9868/trace/config?token=YOUR_TOKEN&enable=1&note=support&durationMinutes=120"
Invoke-RestMethod "http://SERVER:9868/logs/package?token=YOUR_TOKEN&package=Release&lines=200"
Invoke-RestMethod "http://SERVER:9868/logs/item?token=YOUR_TOKEN&item=episode.part2.rar&lines=200"
Invoke-RestMethod "http://SERVER:9868/host/diagnostics?token=YOUR_TOKEN"
Invoke-WebRequest "http://SERVER:9868/support/bundle?token=YOUR_TOKEN" -OutFile ".\\rd-support-bundle.zip"
```
This makes it easy to share one URL plus token during support, so current package status, session state, history, redacted account/settings state, audit actions, rename/MKV move traces, trace data, package/session/item logs, host-side Windows crash hints, disk space, support-log volume, support-bundle size estimates, and even a full ZIP support bundle can be inspected remotely.
## Troubleshooting
- Provider does not work: verify credentials, enabled state, provider order, and daily limits.
- Debrid-Link quota looks wrong: open the API-key statistics popup and check the Rapidgator quota for the affected key.
- Extraction fails: verify passwords and installed extraction tools. The native backend is the default; JVM extraction is optional.
- Downloads stall: check active speed limits, bandwidth schedules, reconnect settings, and provider health.
- Accounts table looks misaligned on one machine: use `Spalten zuruecksetzen` in the Accounts view to clear the locally stored column widths.
## Changelog
Detailed release history is published on [git.24-music.de Releases](https://git.24-music.de/Administrator/real-debrid-downloader/releases).
## License
MIT - see `LICENSE`.