working containerized build

2026-06-13 13:27:29 -07:00
parent 34690b3691
commit 530fd93a0b
12 changed files with 441 additions and 10 deletions
@@ -0,0 +1,136 @@
+"""HTTP server that serves the Guardian Daily Newscast MP3 index page and audio files."""
+
+import os
+import json
+import threading
+from http.server import HTTPServer, SimpleHTTPRequestHandler
+from socketserver import ThreadingMixIn
+from urllib.parse import unquote
+
+from loguru import logger
+
+
+class ThreadingHTTPServer(ThreadingMixIn, HTTPServer):
+    """HTTP server that handles each request in a new thread."""
+    daemon_threads = True
+
+
+class NewscastHandler(SimpleHTTPRequestHandler):
+    """Custom handler that serves a dynamic MP3 index at / and static files below."""
+
+    data_dir: str = "/app/data"
+
+    def do_GET(self):
+        if self.path == "/" or self.path == "":
+            self.serve_index()
+        else:
+            # Strip any leading / to get a relative path
+            rel_path = self.path.lstrip("/")
+            # Prevent directory traversal
+            if ".." in rel_path or rel_path.startswith("/"):
+                self.send_error(403, "Forbidden")
+                return
+            super().do_GET()
+
+    def serve_index(self):
+        """Serve the dynamic MP3 listing page."""
+        audio_dir = os.path.join(self.data_dir, "audio")
+        scripts_file = os.path.join(self.data_dir, "scripts.json")
+
+        # Load script metadata for timestamps
+        scripts_map = {}
+        if os.path.exists(scripts_file):
+            try:
+                with open(scripts_file, "r", encoding="utf-8") as f:
+                    scripts = json.load(f)
+                    for item in scripts:
+                        scripts_map[item["headline"]] = item
+            except (json.JSONDecodeError, KeyError):
+                pass
+
+        mp3_files = []
+        if os.path.isdir(audio_dir):
+            mp3_files = sorted(
+                f for f in os.listdir(audio_dir) if f.lower().endswith(".mp3")
+            )
+
+        html = _generate_index(mp3_files, scripts_map)
+        self.send_response(200)
+        self.send_header("Content-Type", "text/html; charset=utf-8")
+        self.send_header("Content-Length", str(len(html.encode("utf-8"))))
+        self.end_headers()
+        self.wfile.write(html.encode("utf-8"))
+
+    def log_message(self, format, *args):
+        """Suppress per-request logs to avoid noisy output."""
+        pass
+
+
+def _generate_index(mp3_files, scripts_map):
+    """Generate the HTML index page for the MP3 newscast."""
+    audio_entries = ""
+    for f in mp3_files:
+        safe_name = f.replace('"', "&quot;")
+        audio_entries += f"""
+        <div class="episode">
+          <audio controls preload="none">
+            <source src="/{f}" type="audio/mpeg">
+            Your browser does not support the audio element.
+          </audio>
+          <div class="episode-meta">
+            <span class="episode-title">{safe_name}</span>
+            <a href="/{f}" class="download-btn" download>Download</a>
+          </div>
+        </div>"""
+
+    if not mp3_files:
+        audio_entries = '<p class="empty">No newscast files available yet. The pipeline will populate this directory.</p>'
+
+    return f"""<!DOCTYPE html>
+<html lang="en">
+<head>
+  <meta charset="UTF-8">
+  <meta name="viewport" content="width=device-width, initial-scale=1.0">
+  <title>Guardian Daily Newscast</title>
+  <style>
+    * {{ margin: 0; padding: 0; box-sizing: border-box; }}
+    body {{ font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, sans-serif; background: #0f1117; color: #e1e4e8; min-height: 100vh; }}
+    .container {{ max-width: 720px; margin: 0 auto; padding: 2rem 1rem; }}
+    h1 {{ font-size: 1.8rem; margin-bottom: 0.25rem; color: #fff; }}
+    .subtitle {{ color: #8b949e; margin-bottom: 2rem; font-size: 0.95rem; }}
+    .episode {{ background: #161b22; border: 1px solid #30363d; border-radius: 8px; padding: 1rem; margin-bottom: 0.75rem; }}
+    audio {{ width: 100%; margin-bottom: 0.5rem; border-radius: 4px; }}
+    .episode-meta {{ display: flex; align-items: center; justify-content: space-between; gap: 1rem; }}
+    .episode-title {{ color: #c9d1d9; font-size: 0.9rem; word-break: break-word; }}
+    .download-btn {{ background: #238636; color: #fff; padding: 0.35rem 0.75rem; border-radius: 6px; text-decoration: none; font-size: 0.85rem; white-space: nowrap; }}
+    .download-btn:hover {{ background: #2ea043; }}
+    .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; }}
+  </style>
+  <script>
+    function refresh() {{ location.reload(); }}
+    setInterval(refresh, 300000);
+  </script>
+</head>
+<body>
+  <div class="container">
+    <h1>Guardian Daily Newscast</h1>
+    <p class="subtitle">Daily US news — spoken by AI</p>
+    <button class="refresh-btn" onclick="refresh()">Refresh</button>
+    {audio_entries}
+  </div>
+</body>
+</html>"""
+
+
+def run_server(port: int, data_dir: str) -> None:
+    """Start the HTTP server in a background thread and block."""
+    os.chdir("/")  # Serve from root (paths are absolute)
+    NewscastHandler.data_dir = data_dir
+    handler = lambda *a, **kw: NewscastHandler(
+        *a, directory=data_dir, **kw
+    )
+    server = ThreadingHTTPServer(("0.0.0.0", port), handler)
+    logger.info(f"Web server listening on port {port}")
+    server.serve_forever()