Feature: Cross-library watch tracking for duplicate content (multi-language metadata)

[rfsbraz]

Problem

Users with multiple Plex libraries for the same content (e.g., different language metadata) face issues with watch tracking:

• Library A: "Movies (Portuguese)" - Portuguese titles/descriptions
• Library B: "Movies (English)" - English titles/descriptions
• Same physical content, different metadata language
• Users only watch in their preferred library
• Content appears unwatched in libraries users don't access → gets incorrectly deleted

Proposed Solution

Add automatic cross-library watch tracking that treats duplicate content as a single entity for watch history purposes.

Configuration Example

libraries:
  - name: "Movies (Portuguese)"
    radarr: "Radarr"
    last_watched_threshold: 180
    
    # Link to duplicate libraries
    duplicate_libraries:
      - "Movies (English)"
      - "Movies (French)"
    
    # Optional: control deletion behavior
    deletion_mode: "all_or_none"  # Delete from all linked libs or none
    # OR: "individual" - only delete from this library (default)
    
  - name: "Movies (English)"
    radarr: "Radarr"
    last_watched_threshold: 180
    duplicate_libraries:
      - "Movies (Portuguese)"
      - "Movies (French)"

Logic Flow

When evaluating if "The Matrix" should be deleted from "Movies (EN)":

1. Deleterr detects it's part of a duplicate library group
2. Automatically matches content across linked libraries using:
   • IMDB ID (preferred for movies)
   • TMDB ID (fallback for movies)
   • TVDB ID (for shows)
   • Title + Year (last resort if no IDs available)
3. Checks: "Was The Matrix watched in ANY linked library (PT/EN/FR)?"
4. If watched in any library → exclude from deletion in all linked libraries
5. If unwatched in all libraries → apply deletion rules

No manual matching configuration needed — Deleterr should use the best available identifier automatically.

Deletion Modes

all_or_none (recommended)

• If deleting, remove from ALL linked libraries
• Prevents orphaned copies
• Keeps libraries in sync

individual (default)

• Only delete from specific library
• Preserves content in other language libraries

Matching Strategy (Auto-Detected)

Deleterr should automatically use the best matching method:

1. Try IMDB ID first (most reliable, works across languages)
2. Fallback to TMDB ID (if IMDB not available)
3. Fallback to TVDB ID (for TV shows)
4. Last resort: Title + Year normalization (strip accents, normalize spacing)

No user configuration required — just specify which libraries are duplicates.

Use Cases

1. Multi-language households: Portuguese and English libraries for different family members
2. International content: Same shows with native vs. localized metadata
3. Shared Plex servers: Different user groups prefer different metadata languages

Benefits

• Language-agnostic watch tracking: Watch status syncs across metadata variants
• Prevents accidental deletions: Content watched in PT library won't be deleted from EN library
• Clean library management: Option to delete from all duplicates or keep them in sync

Implementation Notes

• Query Plex API for IMDB/TMDB/TVDB IDs from all linked libraries
• Build cross-library mapping at runtime
• Cache mappings to avoid repeated API calls
• Log warnings if content found in one library but not matched in linked libraries

Would appreciate feedback on the config structure and deletion mode defaults!