feat(crawler): support forced execution for manual crawl runs

backend/crawler/tasks.py

@@ -91,7 +91,7 @@ def extract_widgets_from_dict(d, found_ads):
             extract_widgets_from_dict(item, found_ads)
 
 @shared_task
-def run_crawl_pipeline(run_id):
+def run_crawl_pipeline(run_id, force=False):
     """
     Background Celery task that executes a crawling run.
     Parses Divar listing data, registers new Ads, and queues AI evaluations.
@@ -111,13 +111,14 @@ def run_crawl_pipeline(run_id):
         run.save()
         return
         
-    now = timezone.localtime(timezone.now()).time()
-    if not (task.start_hour <= now <= task.end_hour):
-        run.status = 'SUCCESS'
-        run.finished_at = timezone.now()
-        run.error_log = f"Skipped: current time {now.strftime('%H:%M')} is outside allowed window {task.start_hour.strftime('%H:%M')} to {task.end_hour.strftime('%H:%M')}"
-        run.save()
-        return
+    if not force:
+        now = timezone.localtime(timezone.now()).time()
+        if not (task.start_hour <= now <= task.end_hour):
+            run.status = 'SUCCESS'
+            run.finished_at = timezone.now()
+            run.error_log = f"Skipped: current time {now.strftime('%H:%M')} is outside allowed window {task.start_hour.strftime('%H:%M')} to {task.end_hour.strftime('%H:%M')}"
+            run.save()
+            return
 
     # Update run status to RUNNING
     run.status = 'RUNNING'

backend/crawler/views.py

@@ -25,7 +25,7 @@ class CrawlTaskViewSet(viewsets.ModelViewSet):
         try:
             from crawler.tasks import run_crawl_pipeline
             # run asynchronously
-            run_crawl_pipeline.delay(str(run.id))
+            run_crawl_pipeline.delay(str(run.id), force=True)
             is_queued = True
         except ImportError:
             # Fallback if tasks.py doesn't exist yet or run_crawl_pipeline is not defined