Merge pull request #11756 from adlius/handle-validation-remove-context

Remove non-required fields from payload @context; add tests

Author: adlius

osf/models/cedar_metadata.py

@@ -12,6 +12,9 @@
     'pav:createdBy',
     'pav:lastUpdatedOn',
     'oslc:modifiedBy',
+    'schema:isBasedOn',
+    'schema:name',
+    'schema:description',
 })
 
 
@@ -66,9 +69,22 @@ def clean(self):
                     f for f in schema['required']
                     if f not in CEDAR_SERVER_GENERATED_FIELDS
                 ]
-            metadata = {k: v for k, v in self.metadata.items() if v != {}}
+            # Workaround for a cedar-embeddable-editor bug that injects extra keys into
+            # @context; strip it down to only the schema-required keys before saving.
+            context = self.metadata.get('@context')
+            if isinstance(context, dict):
+                required_context = schema.get('properties', {}).get('@context', {}).get('required')
+                if isinstance(required_context, list):
+                    filtered_context = {
+                        k: v for k, v in context.items()
+                        if k in required_context
+                    }
+                    if filtered_context:
+                        self.metadata['@context'] = filtered_context
+                    else:
+                        self.metadata.pop('@context', None)
             try:
-                jsonschema_validate(metadata, schema)
+                jsonschema_validate(self.metadata, schema)
             except JsonSchemaValidationError as e:
                 raise ValidationError(
                     f'CEDAR metadata does not validate against template "{self.template.schema_name}": {e.message}'

osf_tests/test_cedar_metadata.py

@@ -0,0 +1,42 @@
+import pytest
+
+from osf.models import CedarMetadataRecord, CedarMetadataTemplate
+from osf_tests.factories import AuthUserFactory, ProjectFactory
+
+
+@pytest.mark.django_db
+class TestCedarMetadataRecordClean:
+
+    def test_clean_filters_context_items_not_in_schema_required(self):
+        template = CedarMetadataTemplate.objects.create(
+            schema_name='cedar_test_schema',
+            cedar_id='cedar_test_id',
+            template_version=1,
+            template={
+                'type': 'object',
+                'properties': {
+                    '@context': {
+                        'type': 'object',
+                        'required': ['foo', 'bar'],
+                        'properties': {
+                            'foo': {'type': 'string'},
+                            'bar': {'type': 'string'},
+                        },
+                        'additionalProperties': False,
+                    },
+                },
+                'required': ['@context'],
+            },
+            active=True,
+        )
+        user = AuthUserFactory()
+        project = ProjectFactory(creator=user)
+        record = CedarMetadataRecord(
+            guid=project.guids.first(),
+            template=template,
+            metadata={'@context': {'foo': 'value1', 'bar': 'value2', 'extra': 'drop'}},
+            is_published=True,
+        )
+
+        record.clean()  # should not raise after filtering @context items
+        assert record.metadata == {'@context': {'foo': 'value1', 'bar': 'value2'}}