restore annotation that got replaced by a subquery, restore output in scrub_validation

Author: lociii

CHANGELOG.md

@@ -12,6 +12,11 @@ and this project adheres to [Semantic Versioning](http://semver.org/spec/v2.0.0.
 - Nothing
 -->
 
+## [4.1.0] - 2025-03-04
+### Changed
+- Restore output in `scrub_validation` command
+- Move back to annotation, instead of Subquery for performance reasons
+
 ## [4.0.0] - 2025-02-19
 ### Breaking
 - Removed support for Python 3.8

django_scrubber/management/commands/scrub_data.py

@@ -7,6 +7,7 @@
 from django.contrib.sessions.models import Session
 from django.core.exceptions import FieldDoesNotExist
 from django.core.management.base import BaseCommand, CommandError
+from django.db.models import F
 from django.db.utils import DataError, IntegrityError
 
 from django_scrubber import settings_with_fallback
@@ -120,7 +121,9 @@ def _scrub_model(self, model_class, scrubber_apps_list, global_scrubbers):
         logger.info("Scrubbing %s with %s", model_class._meta.label, realized_scrubbers)
 
         try:
-            model_class.objects.update(**realized_scrubbers)
+            model_class.objects.annotate(
+                mod_pk=F("pk") % settings_with_fallback("SCRUBBER_ENTRIES_PER_PROVIDER"),
+            ).update(**realized_scrubbers)
         except IntegrityError as e:
             raise CommandError(
                 f"Integrity error while scrubbing {model_class} ({e}); maybe increase SCRUBBER_ENTRIES_PER_PROVIDER?",

django_scrubber/management/commands/scrub_validation.py

@@ -15,9 +15,16 @@ def handle(self, *args, **options):
 
         if len(non_scrubbed_field_list):
             for model_path, affected_field_list in non_scrubbed_field_list.items():
+                self.stdout.write(f"Model {model_path!r}:")
+
                 found_models += 1
                 for _field in affected_field_list:
+                    self.stdout.write(f"- {_field}")
                     found_fields += 1
 
+            self.stdout.write("")
             if found_models > 0:
+                self.stdout.write(f"{found_models} model(s) having {found_fields} unscrubbed field(s) detected.")
                 sys.exit(1)
+
+        self.stdout.write("No unscrubbed fields detected. Yeah!")

django_scrubber/scrubbers.py

@@ -6,7 +6,7 @@
 import faker
 from django.db import connections, router
 from django.db.models import Field, Func, OuterRef, Subquery
-from django.db.models.functions import Cast, Mod
+from django.db.models.functions import Cast
 from django.db.models.functions import Concat as DjangoConcat
 from django.db.utils import IntegrityError
 from django.utils.translation import get_language, to_locale
@@ -200,10 +200,7 @@ def __call__(self, field):
             Subquery(
                 FakeData.objects.filter(
                     provider=self.provider_key,
-                    provider_offset=Mod(
-                        OuterRef("pk"),
-                        Subquery(FakeData.objects.provider_count(OuterRef("provider"))),
-                    ),
+                    provider_offset=OuterRef("mod_pk"),
                 ).values("content")[:1],
             ),
             field,