fix: start archive scheduler in background thread when serving

run_archiver.py

@@ -11,7 +11,9 @@ import argparse
 import atexit
 import logging
 import os
+import signal
 import sys
+import threading
 import time
 from pathlib import Path
 
@@ -166,18 +168,42 @@ def run_scheduler(interval_minutes: int, logger: logging.Logger, verbose: bool =
         sys.exit(1)
 
 
-def run_web_server(host: str, port: int, logger: logging.Logger, verbose: bool = False) -> None:
+def run_web_server(host: str, port: int, logger: logging.Logger, verbose: bool = False, interval_minutes: int = None) -> None:
-    """Run Flask web server.
+    """Run Flask web server with optional background scheduler.
     
     Args:
         host: Host to bind to
         port: Port to bind to
         logger: Logger instance
         verbose: If True, enable verbose logging
+        interval_minutes: If set, start background scheduler at this interval
     """
     logger.info("Starting web server")
     logger.info("=" * 60)
     
+    _scheduler_stopped = [False]
+
+    def _start_scheduler():
+        try:
+            init_db_rss()
+            initialize_storage()
+            start_scheduler(interval_minutes)
+            logger.info("Background scheduler started (every %d min)", interval_minutes)
+            logger.info("Running initial archive...")
+            scheduled_archive(logger)
+            logger.info("Initial archive complete")
+        except Exception as e:
+            logger.error("Scheduler thread failed: %s", str(e))
+        finally:
+            _scheduler_stopped[0] = True
+
+    def _shutdown_scheduler(signum=None, frame=None):
+        if not _scheduler_stopped[0]:
+            logger.info("Stopping scheduler...")
+            stop_scheduler()
+            logger.info("Scheduler stopped")
+        sys.exit(0)
+
     try:
         if not (ARCHIVE_DIR / 'cache.db').exists():
             logger.info("Database not found, initializing...")
@@ -186,6 +212,13 @@ def run_web_server(host: str, port: int, logger: logging.Logger, verbose: bool =
         if not check_singlefile_available():
             logger.warning("SingleFile CLI not available. Some features may not work.")
         
+        if interval_minutes:
+            logger.info("Starting background archive scheduler (interval: %d min)...", interval_minutes)
+            t = threading.Thread(target=_start_scheduler, daemon=True)
+            t.start()
+            signal.signal(signal.SIGINT, _shutdown_scheduler)
+            signal.signal(signal.SIGTERM, _shutdown_scheduler)
+        
         logger.info("Web server starting on %s:%d", host, port)
         logger.info("=" * 60)
         
@@ -230,8 +263,8 @@ Examples:
     parser.add_argument(
         '--interval',
         type=int,
-        default=60,
+        default=None,
-        help='Run background scheduler with specified interval (minutes, default: 60)'
+        help='Run background scheduler with specified interval (minutes). When used with --serve, runs in background thread (default: 60 min). Standalone blocks.'
     )
     
     parser.add_argument(
@@ -272,7 +305,8 @@ Examples:
         sys.exit(0 if success else 1)
 
     elif args.serve:
-        run_web_server(args.host, args.port, logger, args.verbose)
+        interval = args.interval if args.interval else 60
+        run_web_server(args.host, args.port, logger, args.verbose, interval_minutes=interval)
 
     elif args.interval:
         run_scheduler(args.interval, logger, args.verbose)