Merge pull request #666 from hubmapconsortium/Derek-Furst/limit-component-dataset

Derek furst/limit component dataset

Author: yuanzhou

src/app.py

@@ -3978,13 +3978,14 @@ def multiple_components():
     direct_ancestor_uuids = json_data_dict.get('direct_ancestor_uuids')
     if direct_ancestor_uuids is None or not isinstance(direct_ancestor_uuids, list) or len(direct_ancestor_uuids) !=1:
         bad_request_error(f"Required field 'direct_ancestor_uuids' must be a list. This list may only contain 1 item: a string representing the uuid of the direct ancestor")
-
     # validate existence of direct ancestors.
     for direct_ancestor_uuid in direct_ancestor_uuids:
         direct_ancestor_dict = query_target_entity(direct_ancestor_uuid, user_token)
         if direct_ancestor_dict.get('entity_type').lower() != "dataset":
             bad_request_error(f"Direct ancestor is of type: {direct_ancestor_dict.get('entity_type')}. Must be of type 'dataset'.")
-
+        dataset_has_component_children = app_neo4j_queries.dataset_has_component_children(neo4j_driver_instance, direct_ancestor_uuid)
+        if dataset_has_component_children:
+            bad_request_error(f"The dataset with uuid {direct_ancestor_uuid} already has component children dataset(s)")
     # validate that there is at least one component dataset
     if len(json_data_dict.get('datasets')) < 1:
         bad_request_error(f"'datasets' field must contain at leawst 1 dataset.")

src/app_neo4j_queries.py

@@ -97,6 +97,31 @@ def get_entities_by_type(neo4j_driver, entity_type, property_key = None):
 
     return results
 
+
+"""
+Determine if given dataset has componet children
+
+Parameters
+----------
+neo4j_driver : neo4j.Driver object
+    The neo4j database connection pool
+dataset_uuid : str
+    The uuid of the given dataset
+
+
+Returns
+-------
+boolean 
+"""
+def dataset_has_component_children(neo4j_driver, dataset_uuid):
+    query = (f"MATCH p=(ds1:Dataset)<-[:ACTIVITY_OUTPUT]-(a:Activity)<-[:ACTIVITY_INPUT]-(ds2:Dataset) "
+             f"WHERE ds2.uuid = '{dataset_uuid}' AND a.creation_action = 'Multi-Assay Split' "
+             f"RETURN (count(p) > 0) as {record_field_name}")
+
+    with neo4j_driver.session() as session:
+        result = session.run(query).value()
+    return result[0]
+
 """
 Retrieve the ancestor organ(s) of a given entity