fix: add SpamModeration to MemberProfile: when a User is created, a S…

…pamModeration object is automatically created for the associated MemberProfile

django/core/mixins.py

@@ -262,14 +262,19 @@ def _validate_content_object(self, instance):
             "spam_moderation",
             "is_marked_spam",
             "get_absolute_url",
-            "title",
         ]
         for field in required_fields:
             if not hasattr(instance, field):
                 raise ValueError(
                     f"instance {instance} does not have a {field} attribute"
                 )
 
+        # Ensure either 'title' () or 'username' (for MemberProfile) is present
+        if not (hasattr(instance, "title") or hasattr(instance, "username")):
+            raise ValueError(
+                f"instance {instance} must have either a 'title' or a 'username' attribute"
+            )
+
     @action(detail=True, methods=["post"], permission_classes=[ModeratorPermissions])
     def mark_spam(self, request, **kwargs):
         instance = self.get_object()

django/core/models.py

@@ -1,7 +1,7 @@
-from datetime import timedelta
-from enum import Enum
 import logging
 import pathlib
+from datetime import timedelta
+from enum import Enum
 
 from allauth.account.models import EmailAddress
 from django import forms
@@ -355,7 +355,7 @@ def find_users_with_email(self, candidate_email, exclude_user=None):
 
 @add_to_comses_permission_whitelist
 @register_snippet
-class MemberProfile(index.Indexed, ClusterableModel):
+class MemberProfile(index.Indexed, ModeratedContent, ClusterableModel):
     """
     Contains additional comses.net information, possibly linked to a CoMSES Member / site account
     """

django/core/views.py

@@ -8,64 +8,57 @@
 from django.contrib.auth.decorators import login_required
 from django.contrib.auth.mixins import LoginRequiredMixin
 from django.contrib.auth.models import User
-from django.core.files.images import ImageFile
 from django.core.exceptions import PermissionDenied
+from django.core.files.images import ImageFile
 from django.http import (
     Http404,
     HttpResponseBadRequest,
     HttpResponseRedirect,
     HttpResponseServerError,
 )
 from django.shortcuts import get_object_or_404, redirect, render
-from django.views.generic import DetailView, TemplateView, RedirectView
 from django.urls import reverse
-from rest_framework import (
-    viewsets,
-    generics,
-    parsers,
-    mixins,
-    filters,
-)
+from django.views.generic import DetailView, RedirectView, TemplateView
+from rest_framework import filters, generics, mixins, parsers, viewsets
+from rest_framework.decorators import action
 from rest_framework.exceptions import (
-    PermissionDenied as DrfPermissionDenied,
+    APIException,
     NotAuthenticated,
     NotFound,
-    APIException,
+    PermissionDenied as DrfPermissionDenied,
 )
-from rest_framework.decorators import action
 from rest_framework.filters import OrderingFilter
-from rest_framework.permissions import IsAuthenticated, AllowAny
+from rest_framework.permissions import AllowAny, IsAuthenticated
 from rest_framework.response import Response
 from rest_framework.views import APIView, exception_handler
 from taggit.models import Tag
 from wagtail.images.models import Image
 
 from library.models import Codebase
-from .models import Event, FollowUser, Job, MemberProfile
-from .serializers import (
-    EventSerializer,
-    JobSerializer,
-    MemberProfileSerializer,
-    RelatedMemberProfileSerializer,
-    TagSerializer,
-)
+from .discourse import build_discourse_url
 from .mixins import (
     CommonViewSetMixin,
     HtmlListModelMixin,
     HtmlRetrieveModelMixin,
     PermissionRequiredByHttpMethodMixin,
     SpamCatcherViewSetMixin,
 )
+from .models import Event, FollowUser, Job, MemberProfile
 from .pagination import SmallResultSetPagination
 from .permissions import ObjectPermissions, ViewRestrictedObjectPermissions
-from .discourse import build_discourse_url
+from .serializers import (
+    EventSerializer,
+    JobSerializer,
+    MemberProfileSerializer,
+    RelatedMemberProfileSerializer,
+    TagSerializer,
+)
+from .utils import parse_date, parse_datetime
 from .view_helpers import (
     add_user_retrieve_perms,
     get_search_queryset,
     retrieve_with_perms,
 )
-from .utils import parse_date, parse_datetime
-
 
 logger = logging.getLogger(__name__)
 
@@ -316,7 +309,9 @@ def filter_queryset(self, request, queryset, view):
         return get_search_queryset(qs, queryset, tags=tags)
 
 
-class MemberProfileViewSet(CommonViewSetMixin, HtmlNoDeleteViewSet):
+class MemberProfileViewSet(
+    SpamCatcherViewSetMixin, CommonViewSetMixin, HtmlNoDeleteViewSet
+):
     lookup_field = "user__pk"
     lookup_url_kwarg = "pk"
     queryset = MemberProfile.objects.public().with_tags()

django/curator/serializers.py

@@ -1,6 +1,6 @@
 from rest_framework import serializers
 
-from core.models import Event, Job, SpamModeration
+from core.models import Event, Job, MemberProfile, SpamModeration
 from library.models import Codebase
 
 
@@ -56,6 +56,23 @@ class Meta:
         ]
 
 
+class MinimalMemberProfileSerializer(serializers.ModelSerializer):
+    class Meta:
+        model = MemberProfile
+        fields = [
+            "id",
+            "username",
+            "name",
+            "email",
+            "bio",
+            "research_interests",
+            "affiliations_string",
+            "degrees",
+            "personal_url",
+            "professional_url",
+        ]
+
+
 class SpamUpdateSerializer(serializers.Serializer):
     id = serializers.IntegerField()
     is_spam = serializers.BooleanField()

django/curator/tests/test_llm_spam_moderation.py → django/curator/tests/test_llm_spam_moderation_api_endpoints.py

@@ -5,8 +5,9 @@
 from rest_framework import status
 from rest_framework.test import APIClient
 
-from core.models import Event, Job, SpamModeration
-from core.tests.base import BaseModelTestCase, EventFactory, JobFactory
+from core.models import Event, Job, MemberProfile, SpamModeration
+from core.tests.base import BaseModelTestCase, EventFactory, JobFactory, \
+    UserFactory
 from library.models import Codebase
 from library.tests.base import CodebaseFactory
 
@@ -42,14 +43,17 @@ def setUp(self):
             title="Test Codebase", description="Codebase Description"
         )
 
-        # Create SpamModeration objects
-        self.job_spam = SpamModeration.objects.create(
+        self.user_factory = UserFactory()
+        self.spammy_user = self.user_factory.create(username="scamlikely")
+
+        # Create SpamModeration objects (for MemberProfile the SpamModeration will be created automatically when user is created)
+        self.job_spam_moderation = SpamModeration.objects.create(
             content_object=self.job, status=SpamModeration.Status.SCHEDULED_FOR_CHECK
         )
-        self.event_spam = SpamModeration.objects.create(
+        self.event_spam_moderation = SpamModeration.objects.create(
             content_object=self.event, status=SpamModeration.Status.SCHEDULED_FOR_CHECK
         )
-        self.codebase_spam = SpamModeration.objects.create(
+        self.codebase_spam_moderation = SpamModeration.objects.create(
             content_object=self.codebase,
             status=SpamModeration.Status.SCHEDULED_FOR_CHECK,
         )
@@ -101,13 +105,16 @@ def test_get_latest_spam_batch(self):
         self.assertEqual(response.status_code, status.HTTP_200_OK)
 
         data = response.json()
-        self.assertEqual(len(data), 3)  # We expect 3 items in the batch
+        self.assertEqual(
+            len(data), 5
+        )  # We expect 5 items in the batch (Event, Job, Codebase, MemberProfile) + MemberProfile of the test_user from super().setUp()
 
         # Check if all content types are present
         content_types = [item["contentType"] for item in data]
         self.assertIn("job", content_types)
         self.assertIn("event", content_types)
         self.assertIn("codebase", content_types)
+        self.assertIn("memberprofile", content_types)
 
         # Check structure of a job item
         job_item = next(item for item in data if item["contentType"] == "job")
@@ -163,6 +170,40 @@ def test_update_spam_moderation_success(self):
         # Check if related content object was updated
         self.assertTrue(job.is_marked_spam)
 
+    def test_update_spam_moderation_success_memberprofile(self):
+        self.client.credentials(HTTP_X_API_KEY=self.api_key)
+
+        mp = MemberProfile.objects.get(id=self.spammy_user.member_profile.id)
+
+        data = {
+            "id": mp.spam_moderation.id,
+            "is_spam": True,
+            "spam_indicators": ["indicator1", "indicator2"],
+            "reasoning": "Test reasoning",
+            "confidence": 0.9,
+        }
+
+        response = self.client.post("/api/spam/update/", data, format="json")
+        self.assertEqual(response.status_code, status.HTTP_200_OK)
+
+        # Check if SpamModeration object was updated
+        mp.refresh_from_db()
+        self.assertIsNotNone(mp.spam_moderation)
+        self.assertEqual(mp.spam_moderation.status, SpamModeration.Status.SPAM_LIKELY)
+        self.assertTrue(mp.is_marked_spam)
+        self.assertEqual(mp.spam_moderation.detection_method, "LLM")
+        self.assertEqual(
+            mp.spam_moderation.detection_details["spam_indicators"],
+            ["indicator1", "indicator2"],
+        )
+        self.assertEqual(
+            mp.spam_moderation.detection_details["reasoning"], "Test reasoning"
+        )
+        self.assertEqual(mp.spam_moderation.detection_details["confidence"], 0.9)
+
+        # Check if related content object was updated
+        self.assertTrue(mp.is_marked_spam)
+
     def test_update_spam_moderation_not_spam(self):
         self.client.credentials(HTTP_X_API_KEY=self.api_key)
 
@@ -194,7 +235,7 @@ def test_update_spam_moderation_invalid_data(self):
         self.client.credentials(HTTP_X_API_KEY=self.api_key)
 
         data = {
-            "id": self.codebase_spam.id,
+            "id": self.codebase_spam_moderation.id,
             # Missing required 'is_spam' field
         }
 
@@ -205,7 +246,7 @@ def test_update_spam_moderation_partial_update(self):
         self.client.credentials(HTTP_X_API_KEY=self.api_key)
 
         data = {
-            "id": self.codebase_spam.id,
+            "id": self.codebase_spam_moderation.id,
             "is_spam": True,
             # Only providing partial data
         }

django/curator/views.py

@@ -25,6 +25,7 @@
     MinimalCodebaseSerializer,
     MinimalEventSerializer,
     MinimalJobSerializer,
+    MinimalMemberProfileSerializer,
     SpamModerationSerializer,
     SpamUpdateSerializer,
 )
@@ -105,6 +106,8 @@ def get_latest_spam_batch(request):
             content_serializer = MinimalEventSerializer(content_object)
         elif content_type == "codebase":
             content_serializer = MinimalCodebaseSerializer(content_object)
+        elif content_type == "memberprofile":
+            content_serializer = MinimalMemberProfileSerializer(content_object)
         else:
             continue
 

django/home/signals.py

@@ -1,15 +1,16 @@
 import logging
-import shortuuid
 
+import shortuuid
 from django.conf import settings
 from django.contrib.auth.models import User
+from django.contrib.contenttypes.models import ContentType
 from django.contrib.sites.models import Site
 from django.db.models.signals import post_save
 from django.dispatch import receiver
 from wagtail.models import Site as WagtailSite
 
 from core.discourse import create_discourse_user
-from core.models import MemberProfile, EXCLUDED_USERNAMES
+from core.models import EXCLUDED_USERNAMES, MemberProfile, SpamModeration
 
 logger = logging.getLogger(__name__)
 
@@ -34,6 +35,27 @@ def sync_discourse_user(user: User):
     return success
 
 
+def create_spam_moderation(mp: MemberProfile):
+    content_type = ContentType.objects.get_for_model(type(mp))
+    default_status = SpamModeration.Status.SCHEDULED_FOR_CHECK
+
+    default_spam_moderation = {
+        "status": default_status,
+        "detection_method": "",
+        "detection_details": "",
+    }
+
+    sm, created = SpamModeration.objects.update_or_create(
+        content_type=content_type,
+        object_id=mp.id,
+        defaults=default_spam_moderation,
+    )
+
+    # update the related object
+    mp.spam_moderation = sm
+    mp.save()
+
+
 @receiver(post_save, sender=User, dispatch_uid="member_profile_sync")
 def on_user_save(sender, instance: User, created, **kwargs):
     """
@@ -42,7 +64,9 @@ def on_user_save(sender, instance: User, created, **kwargs):
     if instance.username in EXCLUDED_USERNAMES:
         return
     if created:
-        sync_member_profile(instance)
+        mp = sync_member_profile(instance)
+        if mp:
+            create_spam_moderation(mp)
     if instance.email:
         # sync with discourse
         # to test discourse synchronization locally eliminate the DEPLOY_ENVIRONMENT check