Remove duplicate neo4j func and add upload to delete_cache()

Author: yuanzhou

src/app.py

@@ -4514,25 +4514,29 @@ def require_json(request):
 """
 def delete_cache(id):
     if MEMCACHED_MODE and memcached_client_instance and MEMCACHED_PREFIX:
+        # First delete the target entity cache
         entity_dict = query_target_entity(id, get_internal_token())
         entity_uuid = entity_dict['uuid']
 
+        # If the target entity is Sample (`direct_ancestor`) or Dataset/Publication (`direct_ancestors`)
         # Delete the cache of all the direct descendants (children)
-        # Otherwise they'll have old cached data for the `direct_ancestor` (Sample) `direct_ancestors` (Dataset/Publication) fields
-        children_uuids = schema_neo4j_queries.get_children(neo4j_driver_instance, entity_uuid , 'uuid')
+        child_uuids = schema_neo4j_queries.get_children(neo4j_driver_instance, entity_uuid , 'uuid')
 
         # If the target entity is Collection, delete the cache for each of its associated 
-        # Datasets (via [:IN_COLLECTION] relationship) and Publications (via [:USES_DATA] relationship)
-        collection_associated_uuids = schema_neo4j_queries.get_collection_associated_entities(neo4j_driver_instance, entity_uuid , 'uuid')
+        # Datasets and Publications (via [:IN_COLLECTION] relationship) as well as just Publications (via [:USES_DATA] relationship)
+        collection_dataset_uuids = schema_neo4j_queries.get_collection_associated_datasets(neo4j_driver_instance, entity_uuid , 'uuid')
 
-        # Also find the associated Collections if the target entity is Datasets or Publication
-        # `get_dataset_collections()` applies to both Dataset and Publication (subclass of Dataset)
+        # If the target entity is Upload, delete the cache for each of its associated Datasets (via [:IN_UPLOAD] relationship)
+        upload_dataset_uuids = schema_neo4j_queries.get_upload_datasets(neo4j_driver_instance, entity_uuid , 'uuid')
+
+        # If the target entity is Datasets/Publication, delete the associated Collections cache, Upload cache
         collection_uuids = schema_neo4j_queries.get_dataset_collections(neo4j_driver_instance, entity_uuid , 'uuid')
-        pubication_associated_collection_dict = schema_neo4j_queries.get_publication_associated_collection(neo4j_driver_instance, entity_uuid)
+        collection_dict = schema_neo4j_queries.get_publication_associated_collection(neo4j_driver_instance, entity_uuid)
+        upload_dict = schema_neo4j_queries.get_dataset_upload(neo4j_driver_instance, entity_uuid)
 
         # We only use uuid in the cache key acorss all the cache types
         cache_keys = []
-        for uuid in ([entity_uuid] + children_uuids + collection_associated_uuids + collection_uuids + [pubication_associated_collection_dict['uuid']]):
+        for uuid in ([entity_uuid] + child_uuids + collection_dataset_uuids + upload_dataset_uuids + collection_uuids + [collection_dict['uuid']] + [upload_dict['uuid']]):
             cache_keys.append(f'{MEMCACHED_PREFIX}_neo4j_{uuid}')
             cache_keys.append(f'{MEMCACHED_PREFIX}_complete_{uuid}')
             cache_keys.append(f'{MEMCACHED_PREFIX}_normalized_{uuid}')

src/schema/schema_neo4j_queries.py

@@ -691,7 +691,7 @@ def get_next_revision_uuid(neo4j_driver, uuid):
 
 
 """
-Get a list of associated dataset and publication uuids for a given collection
+Get a list of associated Datasets and Publications (subclass of Dataset) uuids for a given collection
 
 Parameters
 ----------
@@ -707,7 +707,7 @@ def get_next_revision_uuid(neo4j_driver, uuid):
 list
     A list of datasets and publications
 """
-def get_collection_associated_entities(neo4j_driver, uuid, property_key = None):
+def get_collection_associated_datasets(neo4j_driver, uuid, property_key = None):
     results = []
 
     if property_key:
@@ -719,7 +719,7 @@ def get_collection_associated_entities(neo4j_driver, uuid, property_key = None):
                  f"WHERE c.uuid = '{uuid}' "
                  f"RETURN apoc.coll.toSet(COLLECT(e)) AS {record_field_name}")
 
-    logger.info("======get_collection_associated_entities() query======")
+    logger.info("======get_collection_associated_datasets() query======")
     logger.info(query)
 
     with neo4j_driver.session() as session:
@@ -735,50 +735,6 @@ def get_collection_associated_entities(neo4j_driver, uuid, property_key = None):
 
     return results
 
-"""
-Get a list of associated dataset and publication uuids for a given collection
-
-Parameters
-----------
-neo4j_driver : neo4j.Driver object
-    The neo4j database connection pool
-uuid : str
-    The uuid of collection
-property_key : str
-    A target property key for result filtering
-
-Returns
--------
-list
-    A list of datasets and publications
-"""
-def get_collection_associated_entities(neo4j_driver, uuid, property_key = None):
-    results = []
-
-    if property_key:
-        query = (f"MATCH (e:Entity)-[:IN_COLLECTION|:USES_DATA]->(c:Collection) "
-                 f"WHERE c.uuid = '{uuid}' "
-                 f"RETURN apoc.coll.toSet(COLLECT(e.{property_key})) AS {record_field_name}")
-    else:
-        query = (f"MATCH (e:Entity)-[:IN_COLLECTION|:USES_DATA]->(c:Collection) "
-                 f"WHERE c.uuid = '{uuid}' "
-                 f"RETURN apoc.coll.toSet(COLLECT(e)) AS {record_field_name}")
-
-    logger.info("======get_collection_associated_entities() query======")
-    logger.info(query)
-
-    with neo4j_driver.session() as session:
-        record = session.read_transaction(execute_readonly_tx, query)
-
-        if record and record[record_field_name]:
-            if property_key:
-                # Just return the list of property values from each entity node
-                results = record[record_field_name]
-            else:
-                # Convert the list of nodes to a list of dicts
-                results = nodes_to_dicts(record[record_field_name])
-
-    return results
 
 """
 Get a list of associated collection uuids for a given dataset or publication (subclass)
@@ -870,15 +826,13 @@ def get_publication_associated_collection(neo4j_driver, uuid):
     The neo4j database connection pool
 uuid : str
     The uuid of dataset
-property_key : str
-    A target property key for result filtering
 
 Returns
 -------
 dict
     A Upload dict
 """
-def get_dataset_upload(neo4j_driver, uuid, property_key = None):
+def get_dataset_upload(neo4j_driver, uuid):
     result = {}
 
     query = (f"MATCH (e:Entity)-[:IN_UPLOAD]->(s:Upload) "
@@ -1111,18 +1065,27 @@ def unlink_datasets_from_upload(neo4j_driver, upload_uuid, dataset_uuids_list):
     The neo4j database connection pool
 uuid : str
     The uuid of Upload
+property_key : str
+    A target property key for result filtering
 
 Returns
 -------
 list
     The list containing associated dataset dicts
 """
-def get_upload_datasets(neo4j_driver, uuid):
+def get_upload_datasets(neo4j_driver, uuid, property_key = None):
     results = []
 
-    query = (f"MATCH (e:Dataset)-[:IN_UPLOAD]->(s:Upload) "
-             f"WHERE s.uuid = '{uuid}' "
-             f"RETURN apoc.coll.toSet(COLLECT(e)) AS {record_field_name}")
+    if property_key:
+        query = (f"MATCH (e:Dataset)-[:IN_UPLOAD]->(s:Upload) "
+                 f"WHERE s.uuid = '{uuid}' "
+                 # COLLECT() returns a list
+                 # apoc.coll.toSet() reruns a set containing unique nodes
+                 f"RETURN apoc.coll.toSet(COLLECT(e.{property_key})) AS {record_field_name}")
+    else:
+        query = (f"MATCH (e:Dataset)-[:IN_UPLOAD]->(s:Upload) "
+                 f"WHERE s.uuid = '{uuid}' "
+                 f"RETURN apoc.coll.toSet(COLLECT(e)) AS {record_field_name}")
 
     logger.info("======get_upload_datasets() query======")
     logger.info(query)
@@ -1131,8 +1094,12 @@ def get_upload_datasets(neo4j_driver, uuid):
         record = session.read_transaction(execute_readonly_tx, query)
 
         if record and record[record_field_name]:
-            # Convert the list of nodes to a list of dicts
-            results = nodes_to_dicts(record[record_field_name])
+            if property_key:
+                # Just return the list of property values from each entity node
+                results = record[record_field_name]
+            else:
+                # Convert the list of nodes to a list of dicts
+                results = nodes_to_dicts(record[record_field_name])
 
     return results