Langarr

⚠️ AI-Generated Code - Review before production use. Test with DRY_RUN=true first. Backup your databases.

Automatic quality profile management for Radarr & Sonarr based on original language

Perfect for multilingual libraries where you want original audio for some languages (English, German) but dubbed audio for foreign content (Korean, Japanese, French, etc.).

Quick Start

What you need:

• Radarr v3+ and/or Sonarr v3+ (v4 recommended)
• Recyclarr already running in your docker compose stack
• Docker & Docker Compose

What it does:

1. Creates two quality profiles: Original Preferred and Dub Preferred
2. Automatically assigns profiles based on original language
3. Tags foreign content with prefer-dub for easy filtering
4. (Optional) Tags media based on actual audio tracks present in files

Installation

1. Add Langarr Service

Using pre-built image (recommended):

Choose your version:

• latest - Stable releases (recommended for production)
• v1.0.0 - Specific version (pin for stability)
• develop - Development builds (testing/bleeding edge)

services:
  langarr:
    image: ghcr.io/qudiqudi/langarr:latest  # or :v1.0.0 or :develop
    container_name: langarr
    ports:
      - "5678:5678"  # For webhook support
    environment:
      - RADARR_URL=${RADARR_URL}
      - RADARR_API_KEY=${RADARR_API_KEY}
      - SONARR_URL=${SONARR_URL}
      - SONARR_API_KEY=${SONARR_API_KEY}
      - OVERSEERR_URL=${OVERSEERR_URL}  # Optional
      - OVERSEERR_API_KEY=${OVERSEERR_API_KEY}  # Optional
    volumes:
      - ./langarr-config:/config:ro
    networks:
      - your_network_name
    restart: unless-stopped

Building from source:

git clone https://github.com/qudiqudi/langarr.git
cd langarr

Then use build: ./langarr/language-tagger instead of image: in docker-compose.yml.

2. Add Recyclarr Profiles

New to Recyclarr? Copy the full example:

cp recyclarr/recyclarr-full-example.yml /path/to/recyclarr/config/recyclarr.yml

Existing config? See recyclarr/README.md to merge profiles.

Sync profiles:

docker exec recyclarr recyclarr sync

Verify in Radarr/Sonarr → Settings → Profiles

3. Configure Languages

Create ./langarr-config/config.yml:

radarr:
  main:
    enabled: true
    original_languages:
      - en  # English
      - de  # German
    original_profile: Original Preferred
    dub_profile: Dub Preferred

sonarr:
  main:
    enabled: true
    original_languages:
      - en
      - de
    original_profile: Original Preferred
    dub_profile: Dub Preferred

webhook:
  enabled: true
  port: 5678
  auth_token: "your-secure-random-token-here"  # REQUIRED for security

4. Start Service

docker compose up -d langarr
docker logs -f langarr

Configuration

Environment Variables

# Required (usually already defined)
RADARR_URL=http://radarr:7878
RADARR_API_KEY=your-api-key
SONARR_URL=http://sonarr:8989
SONARR_API_KEY=your-api-key

# Optional: Testing
DRY_RUN=true  # Preview changes without applying

Multiple Instances

For 4K/Anime instances, add to config:

radarr:
  main:
    enabled: true
    # ...
  4k:  # Matches RADARR_4K_URL env var
    enabled: true
    original_languages:
      - en

Monitored Items Only

To only process monitored items (skip unmonitored):

radarr:
  main:
    only_monitored: true  # Default: false

Overseerr/Seerr Integration (Optional)

Automatically set correct profiles on requests before they reach Radarr/Sonarr.

Setup

1. Add environment variables:

OVERSEERR_URL=http://overseerr:5055
OVERSEERR_API_KEY=your-api-key

2. Enable in config.yml:

overseerr:
  main:
    enabled: true
    radarr_servers:
      0: main  # Overseerr server ID 0 → langarr radarr.main
    sonarr_servers:
      0: main

Finding Server IDs:

• Settings → Services → click server → check URL
• Example: /settings/services/radarr/0 → ID is 0
• Note: Seerr uses 0-based IDs (starts at 0, not 1)

Webhook Support (Recommended for Auto-Approve)

For instant profile updates with auto-approve enabled:

1. Enable webhook in config.yml:

webhook:
  enabled: true
  port: 5678
  auth_token: "your-secure-random-token-here"  # REQUIRED for security

2. Expose port in docker-compose.yml:

services:
  langarr:
    ports:
      - "5678:5678"

3. Configure in Seerr/Overseerr:

• Settings → Notifications → Webhook
• URL: http://langarr:5678/webhook
• Authorization Header: your-secure-random-token-here (must match config.yml auth_token)
  • Note: Enter just the token value, not X-Auth-Token: or Bearer
  • The server accepts: Authorization, Bearer <token>, or X-Auth-Token formats
• Enable: Media Auto Approved + Media Pending

How it works:

User requests → Webhook fires → Langarr updates profile in Radarr
→ Triggers search → Downloads with correct quality ✅

Without webhook: Profiles updated every 24 hours (still works, just slower)

How It Works

Profile Assignment Logic

• Original language matches config (en, de) → Original Preferred
• Foreign language (ko, ja, fr) → Dub Preferred + prefer-dub tag

Workflow

1. Recyclarr syncs quality profiles with custom formats
2. Langarr assigns profiles based on original language
3. Runs every 24 hours to tag new content
4. (Optional) Webhook for instant updates on new requests

Audio Track Tagging (Optional)

Tag media based on actual audio tracks in downloaded files, not just original language metadata.

Use case: Tag all movies/shows that have a German audio track, regardless of original language. Useful for filtering in Plex/Jellyfin.

Setup

Add audio_tags to your instance config in config.yml:

schedule:
  interval_hours: 24
  audio_scan_interval_hours: 24  # Optional: separate interval for audio scanning

radarr:
  main:
    # ...existing config...
    audio_tags:              # Add this section
      - language: de         # ISO code or full name
        tag_name: german-audio
      - language: en
        tag_name: english-audio

sonarr:
  main:
    # ...existing config...
    audio_tags:
      - language: de
        tag_name: german-audio

How It Works

1. Fetches mediaInfo.audioLanguages from each downloaded file
2. Falls back to languages field (parsed from release name) if mediaInfo is empty
3. Parses languages (handles "English / German" format)
4. Adds tags when language is detected
5. Removes tags if language no longer present (file replaced)

For Sonarr: Tags the series only if all episodes have the audio track.

Difference from Profile Assignment

Feature	Profile Assignment	Audio Tagging
Data source	TMDB original language	Actual file mediaInfo
When	Before/during download	After download
Purpose	Quality profile selection	Filtering/organization

Troubleshooting

Test with Dry-Run

docker exec langarr env DRY_RUN=true RUN_MODE=once python3 /app/arr-language-tagger.py

Common Issues

Profiles not appearing?

docker exec recyclarr recyclarr sync
# Check Radarr/Sonarr → Settings → Profiles

Items not being tagged?

docker logs langarr
# Verify profile names match exactly

Webhook not working?

docker logs -f langarr
# Check for "Received webhook" messages
# Verify port 5678 is exposed

Language detection issues?

• Sonarr v4+ required for full language support
• Check if content has language metadata in Radarr/Sonarr

Adapting for Other Languages

Using TRaSH Guides Templates

1. Find your language: TRaSH Guides

   • French, Spanish, Italian, Portuguese, Japanese, etc.

2. Update recyclarr template:

include:
  - template: radarr-custom-formats-hd-bluray-web-french  # Replace 'german'

3. Replace custom format IDs: Follow language-specific IDs from TRaSH Guides

4. Update language codes:

original_languages:
  - en
  - fr  # Your language

Share your config! Submit a PR to help others.

Versioning

Langarr follows Semantic Versioning:

• Production: Use latest tag or pin to specific version v1.0.0
• Testing: Use develop tag for bleeding-edge features
• Releases: Tagged as v1.0.0, v1.1.0, etc.

Docker image tags:

ghcr.io/qudiqudi/langarr:latest   # Latest stable release
ghcr.io/qudiqudi/langarr:v1.0.0   # Specific version
ghcr.io/qudiqudi/langarr:develop  # Development builds

Credits

• Recyclarr - Quality profile management
• TRaSH Guides - Custom formats
• Built with AI assistance (Claude Code)

License

MIT