Overview
Overview of the data sources supported by the `hugr` query engine and how to configure them.
DuckDB
DuckDB data source allows you to connect to a DuckDB database file, which can be accessed by multiple users. It can be placed in the file system or in object storage like S3 or GCS.
PostgreSQL
PostgreSQL data source configuration and usage in the Hugr engine. Includes query push-down, range types, PostGIS, and TimescaleDB support.
MySQL
MySQL data source configuration and usage in the Hugr engine. Includes query push-down and JSON support.
SQL Server / Azure SQL
SQL Server and Azure SQL data source configuration in the Hugr engine. Includes SQL Server authentication (mssql:// protocol) and Azure EntraID service principal authentication (azure:// protocol) for Fabric Warehouse and Analytical Endpoints.
HTTP RESTful API
Learn how to configure HTTP RESTful APIs as data sources in the `hugr` GraphQL engine to enable integration with external services.
Extension Data Source
Extensions are a special type of data source that allow you to extend existing data objects with cross-source subquery and function call fields.
Embeddings and Semantic Search
Add embeddings models and perform semantic search queries.
DuckLake
DuckLake data source provides lakehouse capabilities with snapshot-based versioning, time-travel queries, and support for various metadata backends (DuckDB, PostgreSQL).
Iceberg
Apache Iceberg data source provides access to Iceberg catalogs with table discovery, time-travel queries, and standard DML operations via DuckDB's iceberg extension.
Hugr App
The hugr-app data source type connects hugr to external applications via the DuckDB Airport extension (Arrow Flight gRPC protocol).
LLM (AI Models)
Connect LLM providers (OpenAI, Anthropic, Gemini) for text generation and tool calling via GraphQL.
Redis
Redis key-value store data source for caching, counters, and shared state.
Airport
Attach remote DuckDB catalogs via the Airport extension (Arrow Flight gRPC).