Initial commit: Phase 1 – Projektstruktur, DB-Schema, Core-API

- FastAPI-Backend mit vollständiger ABS v2.x API-Kompatibilität
- SQLAlchemy-Models: User, Library, LibraryItem, BookFile, Chapter,
  Podcast, PodcastEpisode, MediaProgress, Bookmark, PlaybackSession
- Auth: JWT-Login (/login, /logout, /api/authorize)
- Library + Items Endpoints inkl. camelCase ABS-Response-Format
- HLS-Streaming via FFmpeg (POST /api/items/:id/play, Session-Sync)
- Me/Progress Endpoints + Lesezeichen
- User-Management + Server-Settings (Admin)
- Library-Scanner (MP3/WAV Discovery, Hintergrund-Task)
- File Watcher (watchdog, 30s Debounce)
- Matching-Skelett (MusicBrainz, OpenLibrary, Google Books – Phase 5)
- Docker-Setup: backend (Python 3.12+FFmpeg), frontend (React/Vite),
  nginx Reverse-Proxy auf Port 3000
- setup.sh: Installiert Docker auf Debian/Ubuntu, richtet .env ein

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

backend/app/services/matching/musicbrainz.py

@@ -0,0 +1,40 @@
+"""MusicBrainz-Matching — Phase 5."""
+import httpx
+from .base import MatchResult
+
+MB_BASE = "https://musicbrainz.org/ws/2"
+HEADERS = {"User-Agent": "audiolib/1.0 (https://github.com/audiolib)"}
+
+
+async def search_musicbrainz(title: str, artist: str | None = None) -> list[MatchResult]:
+    query = f'release:"{title}"'
+    if artist:
+        query += f' AND artist:"{artist}"'
+    query += " AND format:Digital"
+
+    async with httpx.AsyncClient(headers=HEADERS, timeout=10) as client:
+        resp = await client.get(
+            f"{MB_BASE}/release",
+            params={"query": query, "fmt": "json", "limit": 5},
+        )
+        resp.raise_for_status()
+        data = resp.json()
+
+    results = []
+    for release in data.get("releases", []):
+        confidence = release.get("score", 0) / 100.0
+        artist_name = None
+        credits = release.get("artist-credit", [])
+        if credits:
+            artist_name = credits[0].get("name") or credits[0].get("artist", {}).get("name")
+
+        results.append(
+            MatchResult(
+                source="musicbrainz",
+                source_id=release.get("id", ""),
+                title=release.get("title", title),
+                author=artist_name,
+                confidence=confidence,
+            )
+        )
+    return results