community[patch]: Add neo4j timeout and value sanitization option (#16138)

The timeout function comes in handy when you want to kill longrunning
queries.
The value sanitization removes all lists that are larger than 128
elements. The idea here is to remove embedding properties from results.

libs/community/tests/integration_tests/graphs/test_neo4j.py

@@ -69,3 +69,22 @@ def test_cypher_return_correct_schema() -> None:
         sorted(relationships, key=lambda x: x["output"]["end"])
         == expected_relationships
     )
+
+
+def test_neo4j_timeout() -> None:
+    """Test that neo4j uses the timeout correctly."""
+    url = os.environ.get("NEO4J_URI")
+    username = os.environ.get("NEO4J_USERNAME")
+    password = os.environ.get("NEO4J_PASSWORD")
+    assert url is not None
+    assert username is not None
+    assert password is not None
+
+    graph = Neo4jGraph(url=url, username=username, password=password, timeout=0.1)
+    try:
+        graph.query("UNWIND range(0,100000,1) AS i MERGE (:Foo {id:i})")
+    except Exception as e:
+        assert (
+            e.code
+            == "Neo.ClientError.Transaction.TransactionTimedOutClientConfiguration"
+        )

libs/community/tests/unit_tests/graphs/test_neo4j_graph.py

@@ -0,0 +1,32 @@
+from langchain_community.graphs.neo4j_graph import value_sanitize
+
+
+def test_value_sanitize_with_small_list():
+    small_list = list(range(15))  # list size > LIST_LIMIT
+    input_dict = {"key1": "value1", "small_list": small_list}
+    expected_output = {"key1": "value1", "small_list": small_list}
+    assert value_sanitize(input_dict) == expected_output
+
+
+def test_value_sanitize_with_oversized_list():
+    oversized_list = list(range(150))  # list size > LIST_LIMIT
+    input_dict = {"key1": "value1", "oversized_list": oversized_list}
+    expected_output = {
+        "key1": "value1"
+        # oversized_list should not be included
+    }
+    assert value_sanitize(input_dict) == expected_output
+
+
+def test_value_sanitize_with_nested_oversized_list():
+    oversized_list = list(range(150))  # list size > LIST_LIMIT
+    input_dict = {"key1": "value1", "oversized_list": {"key": oversized_list}}
+    expected_output = {"key1": "value1", "oversized_list": {}}
+    assert value_sanitize(input_dict) == expected_output
+
+
+def test_value_sanitize_with_dict_in_list():
+    oversized_list = list(range(150))  # list size > LIST_LIMIT
+    input_dict = {"key1": "value1", "oversized_list": [1, 2, {"key": oversized_list}]}
+    expected_output = {"key1": "value1", "oversized_list": [1, 2, {}]}
+    assert value_sanitize(input_dict) == expected_output