fixes for the service layer. Updated readme.

README.md

@@ -215,18 +215,24 @@ docker tag guardian-newscast truenas.local:30095/guardian-newscast:latest
 docker push truenas.local:30095/guardian-newscast:latest
 ```
 
-### Deploying on TrueNAS
+#### Cross-platform builds
+
+For Intel (x86_64):
 
 ```bash
-docker pull truenas.local:30095/guardian-newscast:latest
-docker run -d \
-  --name guardian-newscast \
-  --network host \
-  -v /path/to/data:/app/data \
-  -e OLLAMA_URL=http://127.0.0.1:11434/api/generate \
-  -e PORT=30095 \
-  --restart unless-stopped \
-  truenas.local:30095/guardian-newscast:latest
+docker buildx build --platform linux/amd64 -t guardian-newscast:amd64 .
+```
+
+For ARM (aarch64 / Apple Silicon):
+
+```bash
+docker buildx build --platform linux/arm64 -t guardian-newscast:arm64 .
+```
+
+To build and push both variants as a single multi-arch manifest:
+
+```bash
+docker buildx build --platform linux/amd64,linux/arm64 -t truenas.local:30095/guardian-newscast:latest --push .
 ```
 
 ### Docker Compose
@@ -242,14 +248,10 @@ docker compose --profile full up -d
 ### GPU Acceleration
 
 The default Dockerfile uses `python:3.11-slim` (Debian), which has glibc — fully compatible with NVIDIA CUDA.
-
-To run with GPU acceleration on TrueNAS:
-1. Install the [NVIDIA Container Toolkit](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html).
-2. Build with CUDA-enabled torch:
+1. Build with CUDA-enabled torch:
    ```bash
    docker build --target gpu -t guardian-newscast:gpu .
    ```
    (Add a `FROM python:3.11-slim AS gpu` stage with `pip install torch --index-url https://download.pytorch.org/whl/cu121`)
 3. Run with `--gpus all`.
-4. Minimum **4 GB VRAM** recommended for Ollama + Kokoro in the same container.
-
+4. Minimum **4 GB VRAM** recommended for Ollama + Kokoro in the same container.

pipeline.py

@@ -37,8 +37,21 @@ from server.app import run_server as start_server
 logger.add("pipeline.log", rotation="5 MB", retention="7 days")
 
 
+def clear_json_data() -> None:
+    """Remove stale JSON files at the start of each pipeline cycle."""
+    if not os.path.isdir(DATA_DIR):
+        return
+    for fname in ("output.json", "scripts.json"):
+        fpath = os.path.join(DATA_DIR, fname)
+        if os.path.exists(fpath):
+            os.remove(fpath)
+            logger.info(f"Cleared {fpath}")
+    logger.info("Cleared JSON data at cycle start")
+
+
 async def run_pipeline() -> bool:
     """Run all three pipeline stages sequentially. Returns True on success."""
+    clear_json_data()
     logger.info("=" * 60)
     logger.info("Stage 1: Scraping Guardian articles...")
     try:

server/app.py

@@ -1,5 +1,6 @@
 """HTTP server that serves the Guardian Daily Newscast MP3 index page and audio files."""
 
+import glob
 import os
 import json
 import threading
@@ -65,6 +66,41 @@ class NewscastHandler(SimpleHTTPRequestHandler):
         """Suppress per-request logs to avoid noisy output."""
         pass
 
+    def do_POST(self):
+        if self.path == "/clear":
+            self.clear_audio()
+        else:
+            self.send_error(405, "Method Not Allowed")
+
+    def clear_audio(self):
+        """Remove all MP3 files from the audio directory."""
+        if ".." in self.path:
+            self.send_error(403, "Forbidden")
+            return
+        audio_dir = os.path.join(self.data_dir, "audio")
+        if not os.path.isdir(audio_dir):
+            self._json_response({"status": "ok", "cleared": 0})
+            return
+        mp3_paths = glob.glob(os.path.join(audio_dir, "*.mp3"))
+        count = 0
+        for path in mp3_paths:
+            try:
+                os.remove(path)
+                count += 1
+            except OSError:
+                logger.exception(f"Failed to remove {path}")
+        logger.info(f"Cleared {count} audio files.")
+        self._json_response({"status": "ok", "cleared": count})
+
+    def _json_response(self, data):
+        """Send a JSON response."""
+        body = json.dumps(data).encode("utf-8")
+        self.send_response(200)
+        self.send_header("Content-Type", "application/json")
+        self.send_header("Content-Length", str(len(body)))
+        self.end_headers()
+        self.wfile.write(body)
+
 
 def _generate_index(mp3_files, scripts_map):
     """Generate the HTML index page for the MP3 newscast."""
@@ -107,9 +143,12 @@ def _generate_index(mp3_files, scripts_map):
     .empty {{ text-align: center; color: #8b949e; padding: 3rem 0; }}
     .refresh-btn {{ background: #21262d; color: #c9d1d9; border: 1px solid #30363d; padding: 0.5rem 1rem; border-radius: 6px; cursor: pointer; font-size: 0.85rem; margin-bottom: 1.5rem; }}
     .refresh-btn:hover {{ background: #30363d; }}
+    .clear-btn {{ background: #490202; color: #f85149; border: 1px solid #f85149; padding: 0.5rem 1rem; border-radius: 6px; cursor: pointer; font-size: 0.85rem; margin-bottom: 1.5rem; }}
+    .clear-btn:hover {{ background: #da3633; color: #fff; }}
   </style>
   <script>
     function refresh() {{ location.reload(); }}
+    function clearAudio() {{ fetch('/clear', {{ method: 'POST' }}).then(() => location.reload()); }}
     setInterval(refresh, 300000);
   </script>
 </head>
@@ -118,6 +157,7 @@ def _generate_index(mp3_files, scripts_map):
     <h1>Guardian Daily Newscast</h1>
     <p class="subtitle">Daily US news — spoken by AI</p>
     <button class="refresh-btn" onclick="refresh()">Refresh</button>
+    <button class="clear-btn" onclick="clearAudio()">Clear Audio</button>
     {audio_entries}
   </div>
 </body>