GSQL Language Overview

GSQL™ is TigerGraph's comprehensive environment for designing graph schemas, loading data, and performing advanced data analysis through queries. It is a powerful, SQL-like language designed specifically for the complexities of graph data.

Unified Query Engine

TigerGraph's execution engine is highly versatile, supporting multiple query languages within the same environment:

• GSQL: The native, high-performance graph query language.
• OpenCypher: Support for popular Cypher-style pattern matching.
• GQL: Integration with the emerging Graph Query Language (ISO standard).

Core Capabilities

1. Data Definition (DDL)

Use GSQL to define your graph's blueprint (the schema). This includes creating vertex types (entities), edge types (relationships), and the graph containers themselves.

2. Data Loading (GDSL)

GSQL features a unique and powerful loading sub-language. You can define "Loading Jobs" that map structured data (like CSVs or JSON) into the graph schema with high parallelism.

3. Graph Querying

The GSQL Query Language allows for deep, multi-hop exploration and analysis. It excels at:

• Parameterized Queries: Write reusable logic that accepts inputs.
• Accumulators: A unique GSQL feature for high-performance distributed computations.
• Vector Search: Integrated support for hybrid graph+vector searches for AI/LLM applications.

Key Terminology

• Vertex: A data entity (often called a "node").
• Edge: A connection between two vertices.
• Graph: A defined collection of vertex and edge types forming a specific database domain.
• Catalog: The internal dictionary storing all definitions (vertices, edges, queries, jobs).