Virtuoso-to-Neo4j Handler

A custom Neo4j procedure that allows SPARQL-within-SQL queries to be returned in Neo4j over JDBC, with support for virtuoso.jdbc4.VirtuosoExtendedString and virtuoso.jdbc4.VirtuosoRdfBox classes.

Quick Installation Guide

1. Download the best matching .JAR file from the Releases Section and the Virtuoso JDBC Driver into your Neo4j installation's plugins directory

2. Open your neo4j.conf file and add openlink.* to your dbms.security.procedures.unrestricted parameter value.

3. Restart your Neo4j instance.

4. Test that the custom procedure has successfully installed by running:

SHOW PROCEDURES YIELD name
WHERE name STARTS WITH 'openlink'
RETURN name;

5. If successful, you can now use openlink.virtuoso_jdbc_connect()

Building and Installation Guide for Developers

1. Clone the git repo using git clone
2. Build the .JAR using mvn clean package
3. Copy the .JAR and included JDBC Driver to your Neo4j installation's plugins directory.
4. Open your neo4j.conf file and add openlink.* to your dbms.security.procedures.unrestricted parameter value.
5. Restart your Neo4j instance.
6. Test that the custom procedure has successfully installed by running:

SHOW PROCEDURES YIELD name
WHERE name STARTS WITH 'openlink'
RETURN name;

7. If successful, you can now use openlink.virtuoso_jdbc_connect()

Usage

The Procedure uses the following parameters

• jdbc_url: Your JDBC URL string or variable
• query: Your SQL or SPARQL-within-SQL query
• read/write (default = null)
  • 'r' = read
  • 'rw' = read/write
  • It is recommended to apply correct privileges on your Virtuoso instance rather than fully depending on this parameter.

JDBC URL Format

A basic Virtuoso JDBC URL uses the following format:

jdbc:virtuoso://:<Port#>/DATABASE=/UID=/PWD=/

• Virtuoso JDBC Driver Documentation

Cypher Merge Example

Returned SQL or SPARQL-within-SQL Query Results can be further queried and manipulated using Cypher

WITH 'jdbc:virtuoso://localhost:1111/UID={username}/PWD={password}/CHARSET=UTF-8' AS url
CALL openlink.virtuoso_jdbc_connect(
  url,
  'SPARQL SELECT ?person1 ?person2  WHERE { SERVICE <https://linkeddata.uriburner.com/sparql/>{ SELECT * FROM <urn:analytics> WHERE {?person1 foaf:knows ?person2} }}',
  'r'
) YIELD value

UNWIND value AS row
WITH row['person1'] AS person1, row['person2'] AS person2
MERGE (p1:Person {id: person1})
MERGE (p2:Person {id: person2})
MERGE (p1)-[:KNOWS]->(p2);