feat: implement background scheduler for job scraping with Gunicorn support

tests/test_scheduler.py

@@ -39,6 +39,24 @@ class TestScheduler:
         from web.craigslist import schedule
         assert schedule is not None
 
+    @patch('web.craigslist.threading.Thread')
+    def test_start_scheduler_in_background_idempotent(self, mock_thread):
+        """Ensure background scheduler starts only once per process."""
+        import web.craigslist as craigslist
+
+        thread_instance = MagicMock()
+        thread_instance.is_alive.return_value = True
+        mock_thread.return_value = thread_instance
+
+        craigslist._scheduler_thread = None
+
+        first = craigslist.start_scheduler_in_background()
+        second = craigslist.start_scheduler_in_background()
+
+        assert first is second
+        mock_thread.assert_called_once()
+        thread_instance.start.assert_called_once()
+
     @patch('web.craigslist.db_get_all_job_urls')
     @patch('web.craigslist.seed_regions_keywords_from_listings')
     @patch('web.craigslist.get_all_regions')