fix(parsers): close stale source jobs

src/apps/core/services.py

@@ -6,12 +6,14 @@ They are easily testable and can manage transactions.
 """
 
 import logging
+from datetime import timedelta
 from typing import Any, Generic, TypeVar
 
 import django
 from apps.core.exceptions import NotFoundError
 from django.db import models, transaction
-from django.db.models import QuerySet
+from django.db.models import Q, QuerySet
+from django.utils import timezone
 
 logger = logging.getLogger(__name__)
 
@@ -678,3 +680,37 @@ class BackgroundJobService(BaseReadOnlyService):
             .delete()
         )
         return deleted
+
+    @classmethod
+    def mark_stale_active_jobs_failed(
+        cls,
+        *,
+        max_age_minutes: int,
+        task_names: set[str] | None = None,
+        meta_sources: set[str] | None = None,
+    ) -> int:
+        """Mark old active jobs as failed after worker restarts or hard kills."""
+        from apps.core.models import JobStatus
+
+        cutoff = timezone.now() - timedelta(minutes=max_age_minutes)
+        queryset = cls.get_queryset().filter(
+            status__in=[JobStatus.PENDING, JobStatus.STARTED, JobStatus.RETRY],
+            updated_at__lt=cutoff,
+        )
+        if task_names:
+            queryset = queryset.filter(task_name__in=task_names)
+        if meta_sources:
+            source_filter = Q()
+            for source in meta_sources:
+                source_filter |= Q(meta__source=source)
+            queryset = queryset.filter(source_filter)
+
+        stale_message = (
+            "Stale background job was marked failed after "
+            f"{max_age_minutes} minutes without progress."
+        )
+        updated = 0
+        for job in queryset.order_by("created_at"):
+            job.fail(error=stale_message)
+            updated += 1
+        return updated