Bleeparr 2.0 is a profanity-cleaning tool for TV shows and movies that integrates with Sonarr and Radarr media management systems. The application automatically detects and censors profanity in videos by muting bad words based on subtitles and AI speech recognition.

• Modern Web Interface: Clean, responsive UI built with React
• Sonarr Integration: Connect to your Sonarr instance to process TV shows
• Radarr Integration: Connect to your Radarr instance to process movies (optional)
• Automatic Processing: Monitor Sonarr/Radarr for new downloads and process them automatically
• Manual Processing: Manually trigger processing for specific episodes or entire series
• Smart Profanity Detection: Uses multi-layered detection strategy:
  • Subtitle scanning for obvious profanity
  • Whisper AI transcription (small + medium fallback) for audio verification
  • Final fallback to mute entire subtitle segments if needed
• Configurable Settings: Adjust detection sensitivity, output formats, and more
• Dashboard: View processing statistics, queue status, and system health
• Dockerized: Easy deployment with Docker Compose

• Docker and Docker Compose
• Sonarr and/or Radarr setup with API access
• A list of words to censor in swears.txt

1. Clone the repository:

   git clone https://github.com/davidpeele/bleeparr.git
   cd bleeparr

2. Create or modify the swears.txt file with the words you want to censor, one per line.

3. Edit the docker-compose.yml file to set your Sonarr/Radarr URLs and API keys.

4. Start the application:

   docker compose up -d

5. Access the web interface at http://localhost:5050

You can configure Bleeparr by editing the environment variables in docker-compose.yml:

• SONARR_URL: URL to your Sonarr instance (e.g., http://localhost:8989)
• SONARR_API_KEY: Your Sonarr API key
• RADARR_URL: URL to your Radarr instance (optional)
• RADARR_API_KEY: Your Radarr API key (optional)
• SWEARS_FILE: Path to your swear words file
• OUTPUT_PREFIX: Prefix for cleaned files (default: "clean_")
• OUTPUT_DIRECTORY: Directory for cleaned files (default: same as input)
• ENABLE_AUTO_PROCESSING: Set to 1 to enable automatic processing
• POLL_INTERVAL_SECONDS: How often to check for new downloads (default: 300)
• BOOST_DB: Audio boost level in dB for improved detection (default: 6)
• PRE_BUFFER: Pre-mute buffer in milliseconds (default: 100)
• POST_BUFFER: Post-mute buffer in milliseconds (default: 100)
• BLEEPTOOL: Passes to run - options: S-M-FSM, S-FSM, S-M, S (default: S-M-FSM)

The web interface allows you to:

1. Dashboard: View system status, processing statistics, and recent activities
2. TV Shows: Manage which shows are filtered and manually trigger processing
3. Settings: Configure Sonarr/Radarr connections and profanity cleaning options

• backend/: FastAPI backend
  • api/: API routes and core functionality
  • db.py: Database management
  • tasks.py: Background tasks and polling
• frontend/: React frontend
  • src/components/: React components
  • src/assets/: Static assets
• bleeparr-1.1.py: Core profanity cleaning engine
• Dockerfile: Container definition
• docker-compose.yml: Container orchestration

For development, you can run the frontend and backend separately:

cd backend
pip install -r ../requirements.txt
uvicorn main:app --reload --host 0.0.0.0 --port 5050

cd frontend
npm install
npm run dev

• Database Issues: If the database becomes corrupted, you can delete the data/bleeparr.db file and restart the application.
• Connection Problems: Ensure your Sonarr/Radarr URLs and API keys are correct.
• Processing Failures: Check that the swears.txt file exists and contains valid entries.
• Container Access: If using network_mode: host, ensure that the ports are not already in use.

Contributions are welcome! Please feel free to submit a Pull Request.

This project is open-source and available under the MIT License.

Bleeparr uses:

• Faster-Whisper
• Subliminal
• FFmpeg
• FastAPI
• React