Switching to metube for downloads

app/scanner.py

@@ -0,0 +1,92 @@
+
+
+import os
+import time
+import signal
+import sys
+from pathlib import Path
+from redis import Redis
+from rq import Queue
+
+# Config via env (matches docker-compose)
+LIB = Path(os.getenv("LIBRARY_ROOT", "/library"))
+TRN = Path(os.getenv("TRANSCRIPT_ROOT", "/transcripts"))
+REDIS_URL = os.getenv("REDIS_URL", "redis://redis:6379/0")
+SCAN_INTERVAL = int(os.getenv("SCAN_INTERVAL", "30"))  # seconds
+
+# Media types to track
+MEDIA_EXT = {
+    ".mp3", ".m4a", ".mp4", ".mkv", ".wav", ".flac", ".webm", ".ogg", ".opus"
+}
+
+# In-memory seen set to avoid re-enqueueing during a single run
+_seen: set[str] = set()
+
+
+def already_transcribed(p: Path) -> bool:
+    """Heuristic: if <basename>.json exists in transcripts, consider it done."""
+    base_json = TRN / f"{p.stem}.json"
+    return base_json.exists()
+
+
+def iter_media_files(root: Path):
+    for path in root.rglob("*"):
+        if not path.is_file():
+            continue
+        if path.suffix.lower() in MEDIA_EXT:
+            yield path
+
+
+def enqueue_new_files():
+    q = Queue(connection=Redis.from_url(REDIS_URL))
+
+    # Ensure target dirs exist
+    TRN.mkdir(parents=True, exist_ok=True)
+    LIB.mkdir(parents=True, exist_ok=True)
+
+    new_jobs = 0
+    for p in iter_media_files(LIB):
+        key = str(p.resolve())
+        if key in _seen:
+            continue
+        if already_transcribed(p):
+            _seen.add(key)
+            continue
+        # Enqueue the worker to process this local file
+        q.enqueue("worker.handle_local_file", key)
+        _seen.add(key)
+        new_jobs += 1
+    return new_jobs
+
+
+_shutdown = False
+
+
+def _handle_sig(sig, frame):
+    global _shutdown
+    _shutdown = True
+
+
+def main():
+    signal.signal(signal.SIGINT, _handle_sig)
+    signal.signal(signal.SIGTERM, _handle_sig)
+
+    print(f"[scanner] Watching {LIB} → transcripts in {TRN}; interval={SCAN_INTERVAL}s", flush=True)
+    while not _shutdown:
+        try:
+            jobs = enqueue_new_files()
+            if jobs:
+                print(f"[scanner] Enqueued {jobs} new file(s)", flush=True)
+        except Exception as e:
+            print(f"[scanner] Error: {e}", file=sys.stderr, flush=True)
+        # Sleep between passes
+        for _ in range(SCAN_INTERVAL):
+            if _shutdown:
+                break
+            time.sleep(1)
+
+    print("[scanner] Shutting down", flush=True)
+
+
+if __name__ == "__main__":
+    main()

app/worker.py

@@ -231,6 +231,32 @@ def publish_to_openwebui(paths):
     except Exception as e:
         log({"status": "owui_error", "error": str(e)})
 
+from pathlib import Path
+
+def handle_local_file(path_str: str):
+    """Transcribe & index a local media file that already exists in /library.
+    Safe to call repeatedly; it skips if transcript JSON already exists.
+    """
+    try:
+        p = Path(path_str)
+        if not p.exists():
+            log({"url": path_str, "status": "error", "error": "file_not_found"})
+            return
+        title = p.stem
+        base_json = TRN / f"{title}.json"
+        if base_json.exists():
+            log({"url": path_str, "status": "skip", "reason": "already_transcribed"})
+            return
+        info = {"url": path_str, "status": "transcribing", "title": title, "uploader": p.parent.name, "date": "", "path": str(p)}
+        log(info)
+        base = transcribe(p)
+        index_meili(base.with_suffix(".json"))
+        publish_to_openwebui([base.with_suffix(".txt")])
+        log({**info, **{"status": "done"}})
+    except Exception as e:
+        log({"url": path_str, "status": "error", "error": str(e)})
+        raise
+
 def handle_web(url: str):
     info = {"url": url, "status":"web-downloading", "title":"", "uploader":"", "date":"", "path":""}
     log(info)
@@ -244,6 +270,12 @@ def handle_web(url: str):
 
 def handle_url(url: str):
     try:
+        # If a local file path (or file:// URL) is provided, process it directly
+        if url.startswith("file://"):
+            return handle_local_file(url[7:])
+        if url.startswith("/") and Path(url).exists():
+            return handle_local_file(url)
+
         if not is_media_url(url):
             handle_web(url)
             return

docker-compose.yml

@@ -63,3 +63,41 @@ services:
     volumes:
       - ${REDIS_DATA_HOST_DIR:-./data/redis}:/data
     restart: unless-stopped
+
+  metube:
+    image: alexta69/metube:latest
+    container_name: metube
+    ports:
+      - "8081:8081"
+    environment:
+      - PUID=1000
+      - PGID=1000
+      - TZ=Europe/Prague
+      - DOWNLOAD_DIR=/downloads
+      - OUTPUT_TEMPLATE=%(uploader)s/%(upload_date)s - %(title)s.%(ext)s
+      # Optional: pass a cookies file to bypass consent/age walls
+      # - COOKIE_FILE=/config/cookies.txt
+      # Optional: yt-dlp options (JSON). Example enables Android client fallback
+      # - YTDL_OPTIONS={"extractor_args":{"youtube":{"player_client":"android"}}}
+    volumes:
+      - ${LIBRARY_HOST_DIR:-./library}:/downloads
+      # Optional cookies file on host → /config/cookies.txt inside container
+      # - /mnt/secure/cookies.txt:/config/cookies.txt:ro
+    restart: unless-stopped
+
+  podx-scanner:
+    build: ./app
+    container_name: podx-scanner
+    command: ["python", "scanner.py"]
+    env_file: [.env]
+    environment:
+      MEILI_URL: http://meili:7700
+      REDIS_URL: redis://redis:6379/0
+      LIBRARY_ROOT: /library
+      TRANSCRIPT_ROOT: /transcripts
+      SCAN_INTERVAL: 30
+    volumes:
+      - ${LIBRARY_HOST_DIR:-./library}:/library
+      - ${TRANSCRIPTS_HOST_DIR:-./transcripts}:/transcripts
+    depends_on: [redis]
+    restart: unless-stopped