# Music Orchestrator

Search → preview on YouTube → download as tagged MP3, organized for PlexAmp.

## Production deploy (Ubuntu server, local network)

Copy the project to your Ubuntu machine, then:

```bash
bash deploy.sh
```

The script installs all dependencies, builds the frontend, configures nginx and
a systemd service, and starts everything. Re-run it at any time to apply updates.

**Access:** `http://<server-ip>` from any device on your local network.

### After deploy

```bash
# Stream live logs
journalctl -u music-orchestrator@<your-user> -f

# Restart the API
sudo systemctl restart music-orchestrator@<your-user>

# Change music library path
nano /opt/music-orchestrator/backend/.env
sudo systemctl restart music-orchestrator@<your-user>
```

---

## Local development

**Requirements:** Python 3.11+, Node 18+, ffmpeg

```bash
# Backend
cd backend
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt
cp .env.example .env   # set MUSIC_DIR
uvicorn main:app --reload --port 8000
```

```bash
# Frontend (separate terminal)
cd frontend
npm install
npm run dev
```

Open `http://localhost:5173` — Vite proxies `/api` to the backend automatically.

---

## Music library structure

Downloads are saved as:

```
Music/
└── Queen/
    └── A Night at the Opera/
        └── 01 - Bohemian Rhapsody.mp3
```

Set `MUSIC_DIR` in `backend/.env` to your Plex library folder.

## Architecture notes

- **1 uvicorn worker only** — download job state is held in memory. Multiple
  workers would each have their own job store, breaking status polling.
- nginx serves the React build as static files and reverse-proxies `/api/*`
  to the FastAPI backend on `127.0.0.1:8000`.
- Metadata (title, artist, album, artwork) comes from the iTunes Search API.
  The YouTube video is only used for audio.