fixed bug with identities from neo4j being 0 and therefore not distin…

…guishable from None under simple if identity: clauses. Switched everything to if identity is None

rel2graph/core/factories/matcher.py

@@ -97,5 +97,6 @@ def match(self, resource: Resource) -> List[Node]:
 
             # Convert to nodes
             match_list = [Node.from_dict(r['labels'], r['properties'], identity=r["identity"]) for r in match_list]
+
             return match_list
 

rel2graph/neo4j/graph_elements.py

@@ -228,6 +228,7 @@ def __db_merge__(self, tx, primary_label=None, primary_key=None):
                     p_key = node.__primarykey__
                 else:
                     p_key = primary_key
+
                 # Add node to the node dictionary
                 key = (p_label, p_key, frozenset(node.labels))
                 node_dict.setdefault(key, []).append(node)
@@ -365,7 +366,7 @@ def __getstate__(self):
 
     def __hash__(self):
         try:
-            if self._identity:
+            if self._identity is not None:
                 return hash(self._identity)
         except:
             pass
@@ -424,7 +425,7 @@ def from_dict(labels: List[str], properties: dict, primary_key: str = None, prim
             node.set_primary_key(primary_key)
         if primary_label:
             node.set_primary_label(primary_label)
-        if identity:
+        if identity is not None:
             node.identity = identity
         return node
 
@@ -455,6 +456,8 @@ def __repr__(self):
         args = list(self.labels)
         kwargs = OrderedDict()
         d = dict(self)
+        if self.identity is None:
+            d["identity"] = self.identity
         for key in sorted(d):
             if is_safe_key(key):
                 args.append("%s=%r" % (key, d[key]))
@@ -517,7 +520,7 @@ def from_dict(start_node: Node, end_node: Node, type: str, properties: dict, pri
         relationship = Relationship(start_node, type, end_node, **properties)
         if primary_key:
             relationship.set_primary_key(primary_key)
-        if identity:
+        if identity is not None:
             relationship.identity = identity
         return relationship
 

tests/integration/helpers.py

@@ -44,13 +44,13 @@ def auth():
     return neo4j.basic_auth("neo4j", "password")
 
 def delete_all(session):
-    session.run("MATCH (n) DETACH DELETE n")
+    session.run("MATCH (n) DETACH DELETE n;")
 
 def num_nodes(session):
-    return session.run("MATCH (n) RETURN COUNT(n) AS num_nodes").single()["num_nodes"]
+    return session.run("MATCH (n) RETURN COUNT(DISTINCT n) AS num_nodes").single()["num_nodes"]
 
 def num_relationships(session):
-    return session.run("MATCH ()-[r]->() RETURN COUNT(r) AS num_relations").single()["num_relations"]
+    return session.run("MATCH ()-[r]->() RETURN COUNT(DISTINCT r) AS num_relations").single()["num_relations"]
 
 def get_nodes(session, labels=[]):
     if not isinstance(labels, list) and not isinstance(labels, tuple):

tests/integration/test_merge_relationships.py → tests/integration/test_merge.py

@@ -2,7 +2,7 @@
 # -*- coding: utf-8 -*-
 
 """
-Integration tests for testing the MERGE_RELATIONSHIPS wrapper.
+Integration tests for testing the merging and MERGE_RELATIONSHIPS wrapper.
 
 authors: Julian Minder
 """
@@ -58,7 +58,20 @@ def test_standart_same_resource(config, session, uri, auth):
 
 
 @pytest.mark.parametrize("config",[(1,True), (1, False) ,(5, False)])
-def test_merge(config, session, uri, auth):
+def test_merge_nodesasdf(config, session, uri, auth):
+    schema = """
+    ENTITY("Entity"):
+        NODE("Entity") node:
+            + id = INT(Entity.id)
+    """
+    entities = pd.DataFrame({"id": [1,2, 1, 2]})
+    iterator = IteratorIterator([PandasDataFrameIterator(entities, "Entity")])
+    converter = Converter(schema, iterator, uri, auth, serialize=config[1], num_workers=config[0])    
+    converter()
+    assert num_nodes(session) == 2
+
+@pytest.mark.parametrize("config",[(1,True), (1, False) ,(5, False)])
+def test_merge_relationships(config, session, uri, auth):
     schema = """
     ENTITY("Entity"):
         NODE("Entity") node:
@@ -70,12 +83,12 @@ def test_merge(config, session, uri, auth):
     entities = pd.DataFrame({"id": [1,2]})
     relations = pd.DataFrame({"source_id": [1,1], "target_id": [2,2]})
     iterator = IteratorIterator([PandasDataFrameIterator(entities, "Entity"), PandasDataFrameIterator(relations, "Relation")])
-    converter = Converter(schema, iterator, uri, auth, serialize=config[1], num_workers=config[0])
+    converter = Converter(schema, iterator, uri, auth, serialize=config[1], num_workers=config[0], batch_size=1)    
     converter()
     assert num_relationships(session) == 1
 
-@pytest.mark.parametrize("config",[(1,True), (1, False) ,(5, False)])
-def test_merge_same_resource(config, session, uri, auth):
+@pytest.mark.parametrize("config",[(1,True)])
+def test_merge_relationships_same_resource(config, session, uri, auth):
     schema = """
     ENTITY("Entity"):
         NODE("Entity") node:
@@ -88,6 +101,15 @@ def test_merge_same_resource(config, session, uri, auth):
     entities = pd.DataFrame({"id": [1,2]})
     relations = pd.DataFrame({"source_id": [1], "target_id": [2]})
     iterator = IteratorIterator([PandasDataFrameIterator(entities, "Entity"), PandasDataFrameIterator(relations, "Relation")])
-    converter = Converter(schema, iterator, uri, auth, serialize=config[1], num_workers=config[0])
+    converter = Converter(schema, iterator, uri, auth, serialize=config[1], num_workers=config[0], batch_size=1)
     converter()
+    n = num_relationships(session)
+    if n != 1:
+        with session.begin_transaction() as tx:
+            for record in tx.run("MATCH (a) RETURN a.id, id(a)"):
+                print(record)
+            for record in tx.run("MATCH (a)-[r]->(b) RETURN r, a.id, b.id, id(r), id(a), id(b)"):
+                print(record)
+        print("n", n)
+        exit()
     assert num_relationships(session) == 1