Remove synthetic file names from TDR bundle manifest entries

DataBiosphere · Aug 29, 2024 · e25a26b · e25a26b
1 parent 922ba3d
commit e25a26b
Show file tree

Hide file tree

Showing 3 changed files with 52 additions and 54 deletions.
diff --git a/src/azul/plugins/repository/tdr_hca/__init__.py b/src/azul/plugins/repository/tdr_hca/__init__.py
@@ -184,14 +184,13 @@ def canning_qualifier(cls) -> str:
 
     def add_entity(self,
                    *,
-                   entity_key: str,
                    entity: EntityReference,
                    row: BigQueryRow,
                    is_stitched: bool
                    ) -> None:
         if is_stitched:
             self.stitched.add(entity.entity_id)
-        self._add_manifest_entry(name=entity_key,
+        self._add_manifest_entry(name='',
                                  uuid=entity.entity_id,
                                  version=TDRPlugin.format_version(row['version']),
                                  size=row['content_size'],
@@ -356,11 +355,10 @@ def _emulate_bundle(self, bundle_fqid: TDRBundleFQID) -> TDRHCABundle:
                     rows = future.result()
                     pk_column = entity_type + '_id'
                     rows.sort(key=itemgetter(pk_column))
-                    for i, row in enumerate(rows):
+                    for row in rows:
                         entity = EntityReference(entity_id=row[pk_column], entity_type=entity_type)
                         is_stitched = entity not in root_entities
-                        bundle.add_entity(entity_key=f'{entity_type}_{i}.json',
-                                          entity=entity,
+                        bundle.add_entity(entity=entity,
                                           row=row,
                                           is_stitched=is_stitched)
                 else:

diff --git a/test/indexer/data/1b6d8348-d6e9-406a-aa6a-7ee886e52bf9.tdr.hca.json b/test/indexer/data/1b6d8348-d6e9-406a-aa6a-7ee886e52bf9.tdr.hca.json