Prevent using elements without zones in ML APIs

arkindex/documents/serializers/ml.py

@@ -66,8 +66,12 @@ class TranscriptionCreateSerializer(serializers.Serializer):
     """
     Allows for insertion of new transcriptions and zones
     """
-    element = serializers.PrimaryKeyRelatedField(queryset=Element.objects.all())
-    source = serializers.PrimaryKeyRelatedField(queryset=DataSource.objects.filter(type=MLToolType.Recognizer))
+    element = serializers.PrimaryKeyRelatedField(
+        queryset=Element.objects.filter(zone__isnull=False),
+    )
+    source = serializers.PrimaryKeyRelatedField(
+        queryset=DataSource.objects.filter(type=MLToolType.Recognizer),
+    )
     polygon = serializers.ListField(
         child=serializers.ListField(
             child=serializers.IntegerField(),
@@ -85,6 +89,7 @@ class TranscriptionCreateSerializer(serializers.Serializer):
         if not self.context.get('request'):  # May be None when generating an OpenAPI schema or using from a REPL
             return
         self.fields['element'].queryset = Element.objects.filter(
+            zone__isnull=False,
             corpus__in=Corpus.objects.writable(self.context['request'].user),
         )
 
@@ -113,7 +118,9 @@ class TranscriptionsSerializer(serializers.Serializer):
     Allows for insertion of new transcriptions and zones
     in Bulk (uses serializer above) on a common parent
     """
-    parent = serializers.PrimaryKeyRelatedField(queryset=Element.objects.all())
+    parent = serializers.PrimaryKeyRelatedField(
+        queryset=Element.objects.filter(zone__isnull=False),
+    )
     recognizer = MLToolField(MLToolType.Recognizer)
     transcriptions = TranscriptionBulkSerializer(many=True, allow_empty=False)
 
@@ -122,6 +129,7 @@ class TranscriptionsSerializer(serializers.Serializer):
         if not self.context.get('request'):  # May be None when generating an OpenAPI schema or using from a REPL
             return
         self.fields['parent'].queryset = Element.objects.filter(
+            zone__isnull=False,
             corpus__in=Corpus.objects.writable(self.context['request'].user),
         )
 
@@ -139,7 +147,9 @@ class ClassificationsSerializer(serializers.Serializer):
     """
     Insert N classifications on a single element from a single ML tool
     """
-    parent = serializers.PrimaryKeyRelatedField(queryset=Page.objects.all())
+    parent = serializers.PrimaryKeyRelatedField(
+        queryset=Page.objects.filter(zone__isnull=False),
+    )
     classifier = MLToolField(MLToolType.Classifier)
     classifications = ClassificationBulkSerializer(many=True, allow_empty=False)
 
@@ -148,5 +158,6 @@ class ClassificationsSerializer(serializers.Serializer):
         if not self.context.get('request'):  # May be None when generating an OpenAPI schema or using from a REPL
             return
         self.fields['parent'].queryset = Page.objects.filter(
+            zone__isnull=False,
             corpus__in=Corpus.objects.writable(self.context['request'].user),
         )

arkindex/documents/tests/test_transcription_create.py

@@ -16,6 +16,7 @@ class TestTranscriptionCreate(FixtureAPITestCase):
     def setUpTestData(cls):
         super().setUpTestData()
         cls.page = Page.objects.get(corpus=cls.corpus, zone__image__path='img1')
+        cls.vol = cls.corpus.elements.get(name='Volume 1')
         cls.src = DataSource.objects.get(slug='test')
 
     def test_require_login(self):
@@ -136,7 +137,6 @@ class TestTranscriptionCreate(FixtureAPITestCase):
         self.client.force_login(self.user)
         response = self.client.post(reverse('api:transcription-bulk'), format='json', data={
             "parent": str(self.page.id),
-            "image": str(self.page.zone.image.id),
             "recognizer": self.src.slug,
             "transcriptions": [
                 {
@@ -167,3 +167,28 @@ class TestTranscriptionCreate(FixtureAPITestCase):
 
         # Indexer called
         self.assertEqual(indexer.return_value.run_index.call_count, 2)
+
+    @patch('arkindex.project.serializer_fields.MLTool.get')
+    def test_bulk_transcription_no_zone(self, ml_get_mock):
+        self.assertIsNone(self.vol.zone)
+
+        ml_get_mock.return_value.type = self.src.type
+        ml_get_mock.return_value.slug = self.src.slug
+        ml_get_mock.return_value.version = self.src.revision
+        self.src.internal = True
+        self.src.save()
+
+        self.client.force_login(self.user)
+        response = self.client.post(reverse('api:transcription-bulk'), format='json', data={
+            "parent": str(self.vol.id),
+            "recognizer": self.src.slug,
+            "transcriptions": [
+                {
+                    "type": "word",
+                    "polygon": [(0, 0), (100, 0), (100, 100), (0, 100), (0, 0)],
+                    "text": "NEKUDOTAYIM",
+                    "score": 0.83,
+                },
+            ]
+        })
+        self.assertEqual(response.status_code, status.HTTP_400_BAD_REQUEST)