feat: populate important legacy data fields and unset MSSDK ones

Keeping both model fields around during the transition will prevent the
resulting documents from being committed into the database, as they will
exceed the 16MB BSON document limit of MongoDB.

The test data field is highly custom to the pipeline and ingrained in
the test code. It is a flat list as compared to the nicely structured
one in the MSSDK. We missed to include its cloning in the initial
post-model validation function.

Author: schivmeister

src/ted_sws/core/model/transform.py

@@ -27,6 +27,7 @@
 from mapping_suite_sdk.core.models.file_asset import (
     RMLMappingFileAsset,
     VocabularyMappingFileAsset,
+    TestDataFileAsset,
 )
 from mapping_suite_sdk.mapping_package_v2.models.mapping_package_v2_metadata import (
     MappingPackageV2Metadata,
@@ -369,3 +370,23 @@ def _populate_legacy_from_mssdk(self) -> None:
                 resources=resources,
                 rml_mapping_rules=rml_rules
             )
+            # Clear MSSDK v2 technical and vocabulary suites after populating
+            self.technical_mapping_suite = None
+            self.vocabulary_mapping_suite = None
+
+        # Populate transformation_test_data from test_data_suites
+        if (self.transformation_test_data is None or not getattr(self.transformation_test_data, 'test_data', None)) and self.test_data_suites:
+            all_files = []
+            for suite in self.test_data_suites:
+                all_files.extend([
+                    FileResource(
+                        file_name=file.path.name,
+                        file_content=file.content,
+                        original_name=file.path.name
+                    ) for file in suite.files if isinstance(file, TestDataFileAsset)
+                ])
+            self.transformation_test_data = TransformationTestData(
+                test_data=all_files
+            )
+            # Clear MSSDK v2 test_data_suites after populating
+            self.test_data_suites = []