Async job-based HTTP wrapper around the musicfetch binary, dockerized for the
Lidarr stack, X-API-Key auth, Siri-friendly human messages, port via
MUSICFETCH_PORT (default 6769).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Accept free-form queries (artist/album/title and combos) instead of strict
"Artist - Track". Search Lidarr and YouTube Music concurrently, present a
unified rich picker with bold keyword matching, and act on the chosen hit.
- Normalize results via a Hit dataclass across both sources
- Lidarr: /api/v1/search (album+artist) with album/artist lookup fallback
- YouTube: ytmusicapi for accurate metadata + music.youtube.com URLs,
yt-dlp scrape fallback; tag overrides from the chosen hit
- Lidarr album pick adds artist+album monitored, runs interactive release
search, and falls through to top YouTube hit when no indexer release exists
- argparse CLI: -n/--noninteractive, -s/--ytsearch, -d/--dry-run,
-q/--quality, --limit, --lidarr-only/--yt-only, -o/--root, --search-all
- Config via LIDARR_URL / LIDARR_API_KEY / MUSICFETCH_ROOT env vars
- Update README; add .gitignore for __pycache__
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
--debug for debug mode
--search-all will search all music sources as soon as the artist is added to Lidarr. The default behavior is to not search all sources.
