fix: set is_marked_spam to True for SpamModeration.Status SPAM or SPAM_LIKELY

- fix tests
- add asdf & direnv to .gitignore and .dockerignore

Author: asuworks

.dockerignore

@@ -1,3 +1,6 @@
+.direnv/
+.tool-versions
+.envrc
 .git
 .yarn/cache
 .yarn/install-state.gz

.gitignore

@@ -180,3 +180,7 @@ vignettes/*.pdf
 
 # End of https://www.toptal.com/developers/gitignore/api/r
 
+# asdf & direnv
+.direnv/
+.tool-versions
+.envrc

django/core/mixins.py

@@ -249,11 +249,11 @@ class SpamCatcherViewSetMixin:
 
     def perform_create(self, serializer: serializers.Serializer):
         super().perform_create(serializer)
-        self.handle_spam_detection(serializer)
+        self.create_spam_moderation_object(serializer)
 
     def perform_update(self, serializer):
         super().perform_update(serializer)
-        self.handle_spam_detection(serializer)
+        self.create_spam_moderation_object(serializer)
 
     def _validate_content_object(self, instance):
         # make sure that the instance has a spam_moderation attribute as well as the
@@ -294,10 +294,10 @@ def mark_spam(self, request, **kwargs):
             spam_moderation.save()
         return redirect(instance.get_list_url())
 
-    def handle_spam_detection(self, serializer: serializers.Serializer):
+    def create_spam_moderation_object(self, serializer: serializers.Serializer):
         try:
             self._validate_content_object(serializer.instance)
-            self._record_spam(
+            self._create_spam_moderation_object(
                 serializer.instance,
                 (
                     serializer.context["spam_context"]
@@ -308,24 +308,25 @@ def handle_spam_detection(self, serializer: serializers.Serializer):
         except ValueError as e:
             logger.warning("Cannot flag %s as spam: %s", serializer.instance, e)
 
-    def _record_spam(self, instance, spam_context: dict = None):
+    def _create_spam_moderation_object(self, instance, spam_context: dict = None):
         content_type = ContentType.objects.get_for_model(type(instance))
+        default_status = (
+            SpamModeration.Status.SPAM_LIKELY
+            if spam_context
+            else SpamModeration.Status.SCHEDULED_FOR_CHECK
+        )
+        default_spam_moderation = {
+            "status": default_status,
+            "detection_method": (
+                spam_context.get("detection_method", "") if spam_context else ""
+            ),
+            "detection_details": (
+                spam_context.get("detection_details", "") if spam_context else ""
+            ),
+        }
 
-        # SpamModeration updates the content instance on save
-        spam_moderation, created = SpamModeration.objects.get_or_create(
+        SpamModeration.objects.update_or_create(
             content_type=content_type,
             object_id=instance.id,
-            defaults={
-                "status": SpamModeration.Status.SCHEDULED_FOR_CHECK,
-                "detection_method": (
-                    spam_context["detection_method"] if spam_context else ""
-                ),
-                "detection_details": (
-                    spam_context["detection_details"] if spam_context else ""
-                ),
-            },
+            defaults=default_spam_moderation,
         )
-
-        if not created:
-            spam_moderation.status = SpamModeration.Status.SCHEDULED_FOR_CHECK
-            spam_moderation.save()

django/core/models.py

@@ -134,16 +134,15 @@ def deploy_environment(self):
 
 class SpamModeration(models.Model):
     class Status(models.TextChoices):
-        UNREVIEWED = "unreviewed", _("Unreviewed")
         SPAM = "spam", _("Confirmed spam")
         NOT_SPAM = "not_spam", _("Confirmed not spam")
         SCHEDULED_FOR_CHECK = "scheduled_for_check", _("Scheduled for check by LLM")
-        SPAM_LIKELY = "spam_likely", _("Marked spam by LLM")
-        NOT_SPAM_LIKELY = "not_spam_likely", _("Marked as not spam by LLM")
+        SPAM_LIKELY = "spam_likely", _("Automatically marked as spam")
+        NOT_SPAM_LIKELY = "not_spam_likely", _("Automatically marked as not spam")
 
     status = models.CharField(
         choices=Status.choices,
-        default=Status.UNREVIEWED,
+        default=Status.SCHEDULED_FOR_CHECK,
         max_length=32,
     )
     content_type = models.ForeignKey(ContentType, on_delete=models.CASCADE)
@@ -195,7 +194,10 @@ def update_related_object(self):
         related_object = self.content_object
         if hasattr(related_object, "is_marked_spam"):
             related_object.spam_moderation = self
-            related_object.is_marked_spam = self.status == self.Status.SPAM
+            related_object.is_marked_spam = self.status in {
+                self.Status.SPAM,
+                self.Status.SPAM_LIKELY,
+            }
             related_object.save()
 
     def __str__(self):

django/core/settings/defaults.py

@@ -527,7 +527,9 @@ def is_test(self):
 DISCOURSE_API_KEY = read_secret("discourse_api_key", "unconfigured")
 DISCOURSE_API_USERNAME = os.getenv("DISCOURSE_API_USERNAME", "unconfigured")
 
-LLM_SPAM_CHECK_API_KEY = read_secret("llm_spam_check_api_key", "unconfigured")
+LLM_SPAM_CHECK_API_KEY = (
+    read_secret("llm_spam_check_api_key", "unconfigured") or "unconfigured"
+)
 
 # https://docs.djangoproject.com/en/4.2/ref/settings/#templates
 TEMPLATES = [

django/core/tests/test_views.py

@@ -162,9 +162,13 @@ def test_event_creation_with_honeypot_spam(self):
         )
         self.assertResponseCreated(response)
         event = Event.objects.get(title=data["title"])
-        self.assertTrue(event.is_marked_spam)
+
         self.assertIsNotNone(event.spam_moderation)
+        self.assertEqual(
+            event.spam_moderation.status, SpamModeration.Status.SPAM_LIKELY
+        )
         self.assertEqual(event.spam_moderation.detection_method, "honeypot")
+        self.assertTrue(event.is_marked_spam)
 
     def test_job_creation_with_timer_spam(self):
         # FIXME: should incorporate how long a typical request takes to resolve
@@ -179,9 +183,11 @@ def test_job_creation_with_timer_spam(self):
         )
         self.assertResponseCreated(response)
         job = Job.objects.get(title=data["title"])
-        self.assertTrue(job.is_marked_spam)
+
         self.assertIsNotNone(job.spam_moderation)
+        self.assertEqual(job.spam_moderation.status, SpamModeration.Status.SPAM_LIKELY)
         self.assertEqual(job.spam_moderation.detection_method, "form_submit_time")
+        self.assertTrue(job.is_marked_spam)
 
     def test_mark_spam(self):
         data = self.event_factory.get_request_data()
@@ -193,18 +199,29 @@ def test_mark_spam(self):
             format="json",
         )
         event = Event.objects.get(title=data["title"])
+        self.assertIsNotNone(event.spam_moderation)
+        self.assertEqual(
+            event.spam_moderation.status, SpamModeration.Status.SCHEDULED_FOR_CHECK
+        )
+        # by default, all created objects will have is_marked_spam = False unless spam_moderation.status is explicitly SPAM or SPAM_LIKELY
         self.assertFalse(event.is_marked_spam)
-        self.assertIsNone(event.spam_moderation)
+
         response = self.client.post(
             reverse("core:event-mark-spam", kwargs={"pk": event.id}),
             data,
             HTTP_ACCEPT="application/json",
             format="json",
         )
+
         event.refresh_from_db()
-        # non-moderators cannot mark content as spam
+        # non-moderators cannot mark content as spam (set status to SPAM)
         self.assertEquals(response.status_code, 403)
+        self.assertIsNotNone(event.spam_moderation)
+        self.assertEqual(
+            event.spam_moderation.status, SpamModeration.Status.SCHEDULED_FOR_CHECK
+        )
         self.assertFalse(event.is_marked_spam)
+
         # check moderator
         self.client.login(
             username=self.moderator.username, password=self.user_factory.password
@@ -217,12 +234,17 @@ def test_mark_spam(self):
             format="json",
         )
         event.refresh_from_db()
-        self.assertTrue(event.is_marked_spam)
         self.assertIsNotNone(event.spam_moderation)
+        self.assertEqual(event.spam_moderation.status, SpamModeration.Status.SPAM)
+        self.assertTrue(event.is_marked_spam)
+
         event.mark_not_spam(self.moderator)
         event.refresh_from_db()
-        self.assertFalse(event.is_marked_spam)
+
         self.assertIsNotNone(event.spam_moderation)
+        self.assertEqual(event.spam_moderation.status, SpamModeration.Status.NOT_SPAM)
+        self.assertFalse(event.is_marked_spam)
+
         # check superuser
         self.client.login(
             username=self.superuser.username, password=self.user_factory.password
@@ -234,9 +256,10 @@ def test_mark_spam(self):
             format="json",
         )
         event.refresh_from_db()
-        self.assertTrue(event.is_marked_spam)
+
         self.assertIsNotNone(event.spam_moderation)
         self.assertEqual(event.spam_moderation.status, SpamModeration.Status.SPAM)
+        self.assertTrue(event.is_marked_spam)
 
     def test_event_creation_without_spam(self):
         data = self.event_factory.get_request_data()
@@ -248,8 +271,12 @@ def test_event_creation_without_spam(self):
         )
         self.assertResponseCreated(response)
         event = Event.objects.get(title=data["title"])
+
+        self.assertIsNotNone(event.spam_moderation)
+        self.assertEqual(
+            event.spam_moderation.status, SpamModeration.Status.SCHEDULED_FOR_CHECK
+        )
         self.assertFalse(event.is_marked_spam)
-        self.assertIsNone(event.spam_moderation)
 
     def test_job_update_with_spam(self):
         data = self.job_factory.get_request_data()
@@ -261,8 +288,13 @@ def test_job_update_with_spam(self):
         )
         self.assertResponseCreated(response)
         job = Job.objects.get(title=data["title"])
+
+        self.assertIsNotNone(job.spam_moderation)
+        self.assertEqual(
+            job.spam_moderation.status, SpamModeration.Status.SCHEDULED_FOR_CHECK
+        )
         self.assertFalse(job.is_marked_spam)
-        self.assertIsNone(job.spam_moderation)
+
         data = self.job_factory.get_request_data(
             honeypot_value="spammy content",
             elapsed_time=settings.SPAM_LIKELY_SECONDS_THRESHOLD + 1,
@@ -274,9 +306,11 @@ def test_job_update_with_spam(self):
             format="json",
         )
         job.refresh_from_db()
-        self.assertTrue(job.is_marked_spam)
+
         self.assertIsNotNone(job.spam_moderation)
+        self.assertEqual(job.spam_moderation.status, SpamModeration.Status.SPAM_LIKELY)
         self.assertEqual(job.spam_moderation.detection_method, "honeypot")
+        self.assertTrue(job.is_marked_spam)
 
     def test_exclude_spam_from_public_views(self):
         data = self.event_factory.get_request_data(honeypot_value="spammy content")

django/curator/tests/test_llm_spam_moderation.py

@@ -21,6 +21,7 @@ def setUp(self):
 
         self.job_factory = JobFactory(submitter=self.user)
         self.event_factory = EventFactory(submitter=self.user)
+        self.codebase_factory = CodebaseFactory(submitter=self.user)
 
         today = timezone.now()
 
@@ -38,8 +39,7 @@ def setUp(self):
             title="Test Event",
         )
 
-        codebase_factory = CodebaseFactory(submitter=self.user)
-        self.codebase = codebase_factory.create(
+        self.codebase = self.codebase_factory.create(
             title="Test Codebase", description="Codebase Description"
         )