Align limit tests with direct snapshot depth

archivebox/services/runner.py

@@ -1197,9 +1197,11 @@ class CrawlRunner:
         snapshot = Snapshot.objects.select_related("crawl", "crawl__created_by").filter(id=snapshot_id).first()
         if snapshot is None or snapshot.status == Snapshot.StatusChoices.SEALED:
             return
-        if snapshot.status == Snapshot.StatusChoices.STARTED:
-            snapshot.sm.seal()
-            return
+        # Limit stops are runner-owned cancellation decisions, not normal
+        # "all ArchiveResults finished" lifecycle seals. Updating the row
+        # directly avoids racing the state machine's in-memory state while
+        # concurrent snapshot tasks are stopping because the crawl-wide limit
+        # has already been reached.
         snapshot.update_and_requeue(
             status=Snapshot.StatusChoices.SEALED,
             retry_at=None,

archivebox/tests/test_config_DELETE_AFTER.py

@@ -231,8 +231,9 @@ def test_delete_after_real_add_page_and_rest_create_paths(client):
     from archivebox.services.runner import run_due_crawl
 
     assert run_due_crawl(ui_crawl, lock_seconds=10)
-    ui_snapshot = ui_crawl.snapshot_set.get(url="https://example.com/delete-after-ui")
+    ui_snapshot = ui_crawl.snapshot_set.get(url="archivebox://internal")
     assert ui_snapshot.delete_at is not None
+    assert ui_snapshot.output_dir.joinpath("staticfile", "stdin.txt").read_text() == "https://example.com/delete-after-ui"
 
     from archivebox.api.auth import get_or_create_api_token
 

archivebox/tests/test_config_MAX_limits.py

@@ -287,9 +287,10 @@ def test_recursive_crawl_respects_max_urls(tmp_path, initialized_archive, recurs
             depth: Snapshot.objects.filter(depth=depth).count() for depth in set(Snapshot.objects.values_list("depth", flat=True))
         }
 
-    assert crawl == (2, 4)
+    assert crawl == (3, 4)
     assert len(snapshot_rows) == 4
-    assert depth_counts.get(0, 0) == 1
-    assert depth_counts.get(1, 0) == 3
-    assert depth_counts.get(2, 0) == 0
-    assert set(recursive_test_site["child_urls"]).issubset({url for url, depth, _parent in snapshot_rows if depth == 1})
+    assert depth_counts.get(0, 0) == 0
+    assert depth_counts.get(1, 0) == 1
+    assert depth_counts.get(2, 0) == 3
+    assert depth_counts.get(3, 0) == 0
+    assert set(recursive_test_site["child_urls"]).issubset({url for url, depth, _parent in snapshot_rows if depth == 2})