LDBC open-source GQL tools

By Alastair Green, Vice-chair LDBC, and author of the GQL Manifesto.

9 May 2024

The GQL standard was published in mid-April by ISO. See WG3 Convenor Keith Hare’s summary: ISO/IEC JTC 1 GQL Database Language

Linked Data Benchmark Council (LDBC) is releasing early-version open-source GQL grammar tools.

Open GQL Language Tools

There are three interlinked projects:

A commit of a new version of the grammar automatically rebuilds and deploys the Code Editor and the Railroad Diagrams.

Michael Burbidge (who leads our GQL Implementation Working Group), Damian Wileński, and Dominik Tomaszuk are responsible for making all this happen, so soon after the release of GQL. Wonderful work!

The tools are a work in progress, so expect evolution. Feel free to raise issues on Github.

The mission of the GQL Implementation Working Group is to create tooling and documentation to assist in and accelerate the implementation and adoption of GQL. See the working group charter for more information.

Code examples for Technical Reports

The Code Editor lets you create syntactically correct GQL examples. It is not connected to an implementation of the GQL spec, so type checking, variable scoping rules, etc., that are typically done by semantic analysis of the parse tree, are not enforced by the Code Editor.

I used Code Editor to create/check the code examples in my last post on LinkedIn, GQL in code, which links through to

GQL on one page: DDL, DML and GPML

It is also being used to help create a forthcoming LDBC Technical Report on GQL, which (unlike the spec) will be freely available to all, and will contain numerous examples.

Towards a GQL TCK

More is in the works: we have begun work on a Test Compatibility Kit, modelled on the openCypher TCK, but that’s a big job.

We need and welcome active contributions to all these community efforts.

If you are interested, please ping Michael at [email protected].

The Linked Data Benchmark Council

LDBC is a non-profit consortium of vendors, researchers and independent practitioners interested in graph data management.

LDBC defines benchmark standards for graph data workloads (using RDF, SQL, and property graph languages). It is a meeting point and working space for community efforts supportive of the GQL and SQL/PGQ property graph standards.

It supervises the audited execution of comparable benchmark runs which are reported with cost metrics, following the lead of TPC. Only audited results published by LDBC can be described as LDBC Benchmark(R) results.

LDBC is run by its 20+ organizational members including Oracle, Ant Group, Intel, Neo4j, TigerGraph, Fabarta, thatDot, Ontotext, ArangoDB, Relational AI, Stargraph, Nebula Graph, Sparksee, FORTH, Memgraph, Createlink, Alibaba DAMO Academy (Graphscope), Birkbeck University of London and AWS.

There are 70+ individual associate members (who join for free and support its working groups and task forces).

Recent and current initiatives include the Finance Benchmark, LDBC Extended GQL Schema (LEX), and GQL Implementation.

The G-CORE, PG-Keys, Graph Pattern Matching in GQL and SQL/PGQ and PG-Schema papers (all published at SIGMOD) directly reflect the work of LDBC participants on graph data languages, over the years.

LDBC is a Category C Liaison of ISO/IEC JTC1 SC32/WG3 (the SQL and GQL standards committee). Jan Hidders, Michael Burbidge and Alastair Green are LDBC’s representatives in WG3.

For all enquiries, including membership enquiries, please email [email protected].