Skip to content

arkady69/neo4j-gremlin-bolt-master

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

neo4j-gremlin-bolt

This project allows the use of the Apache Tinkerpop Java API with the neo4j server using the BOLT protocol.

Build status

Build Status Coverage Status Maven Central Contributors

Requirements

  • Java 8.
  • Maven 3.0.0 or newer.

Usage

Add the Neo4j Apache Tinkerpop implementation to your project:

Maven

    <dependency>
        <groupId>com.steelbridgelabs.oss</groupId>
        <artifactId>neo4j-gremlin-bolt</artifactId>
        <version>{version}</version>
    </dependency>

*Please check the Maven Central for the latest version available.

License

neo4j-gremlin-bolt and it's modules are licensed under the Apache License v 2.0.

Features

  • Apache Tinkerpop 3.x Online Transactional Processing Graph Systems (OLTP) support.
  • neo4j implementation on top of the BOLT protocol.
  • Support for Graph partitioning, out of the box implementation for All labels and Any label partitions.

Graph API

Element ID providers

The library supports an open architecture for element ID generation for new Vertices and Edges. The following element ID providers are supported out of the box:

Neo4J native id() support, see Neo4JNativeElementIdProvider for more information.

    // create id provider
    Neo4JElementIdProvider<?> provider = new Neo4JNativeElementIdProvider();

Pros:

  • IDs are stored as java.lang.Long instances.
  • Fewer database hits on MATCH statements since index lookups are not required at the time of locating an entity by id: MATCH (n:Label) WHERE ID(n) = {id} RETURN n

Cons:

  • CREATE statements will run slower since the entity id must be retrieved from the database after insertion: CREATE (n:label{field1: value, ..., fieldN: valueN}) RETURN ID(n)
  • Entity IDs in Neo4J are not guaranteed to be the same after a database restart/upgrade. Storing links to Neo4J entities outside the database based on IDs could become invalid after a database restart/upgrade.

Database sequence support, see DatabaseSequenceElementIdProvider for more information.

    // create id provider
    Neo4JElementIdProvider<?> provider = new DatabaseSequenceElementIdProvider(driver);

Pros:

  • IDs are stored as java.lang.Long instances.
  • CREATE statements will run faster since there is no need to retrieve the entity after an insert operation: CREATE (n:label{id: 1, field1: value, ..., fieldN: valueN})
  • Entity IDs are guaranteed to be the same after a database restart/upgrade since they are stored as property values.

Cons:

  • A unique index is required for each one of the Labels used in your model.
  • More database hits on MATCH statements since an index lookup is required in order to locate an entity by id: MATCH (n:Label) WHERE n.id = {id} RETURN n

Custom providers, by implementing the Neo4JElementIdProvider interface.

Connecting to the database

    // create driver instance
    Driver driver = GraphDatabase.driver("bolt://localhost", AuthTokens.basic("neo4j", "neo4j"));
  • Create element id provider instances, see providers for more information.
    // create id provider instances
    vertexIdProvider = ...
    edgeIdProvider = ...
    // create graph instance
    try (Graph graph = new Neo4JGraph(driver, vertexIdProvider, edgeIdProvider)) {
        
    }

Working with transactions

    // create graph instance
    try (Graph graph = new Neo4JGraph(driver, vertexIdProvider, edgeIdProvider)) {
        // begin transaction
        try (Transaction transaction = graph.tx()) {
            // use Graph API to create, update and delete Vertices and Edges
            
            // commit transaction
            transaction.commit();
        }
    }

Enabling Neo4J profiler

  • Set logger INFO level to the package: com.steelbridgelabs.oss.neo4j.structure.summary

  • Enable profiler to the Graph instance.

    // create graph instance
    try (Neo4JGraph graph = new Neo4JGraph(driver, vertexIdProvider, edgeIdProvider)) {
        // enable profiler
        graph.setProfilerEnabled(true);
        
    }

The library will prefix CYPHER statements with the PROFILE clause dumping the output into the log file, example:

2016-08-26 23:19:42.226  INFO 98760 --- [-f6753a03391b-1] c.s.o.n.s.summary.ResultSummaryLogger    : Profile for CYPHER statement: Statement{text='PROFILE MATCH (n:Person{id: {id}})-[r:HAS_ADDRESS]->(m) RETURN n, r, m', parameters={id: 1306984}}

+----------------------+----------------+------+---------+-----------+
| Operator             + Estimated Rows + Rows + DB Hits + Variables |
+----------------------+----------------+------+---------+-----------+
| +ProduceResults      |              0 |    1 |       0 | m, n, r   |
| |                    +----------------+------+---------+-----------+
| +Expand(All)         |              0 |    1 |       2 | m, n, r   |
| |                    +----------------+------+---------+-----------+
| +Filter              |              0 |    1 |       1 | n         |
| |                    +----------------+------+---------+-----------+
| +NodeUniqueIndexSeek |              0 |    1 |       2 | n         |
+----------------------+----------------+------+---------+-----------+

Working with Vertices and Edges

Create a Vertex

Create a new Vertex in the current graph call the Graph.addVertex() method.

  // create a vertex in current graph
  Vertex vertex = graph.addVertex();

Create a new Vertex in the current graph with property values:

  // create a vertex in current graph with property values
  Vertex vertex = graph.addVertex("name", "John", "age", 50);

Create a new Vertex in the current graph with a Label:

  // create a vertex in current graph with label
  Vertex vertex1 = graph.addVertex("Person");
  // create another vertex in current graph with label
  Vertex vertex2 = graph.addVertex(T.label, "Company");

Building the library

To compile the code and run all the unit tests:

mvn clean install

To run the Tinkerpop integration tests you need a running instance of the neo4j server. The easiest way to get one up and running is by using the official neo4j docker image:

docker run -d --name neo4j -p 7687:7687 -e NEO4J_AUTH=neo4j/neo4j123 neo4j:3.2-enterprise

And then execute the integration tests by running the following command:

mvn test -Pintegration-test

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published