HornDB docs
Guides for learning HornDB and reference for looking up exact behaviour.
HornDB is a graph database with a built-in OWL 2 RL reasoner. Load RDF, query it with SPARQL 1.1 over HTTP, and — once you turn reasoning on — get answers that include triples the reasoner derived, not only the ones you wrote.
What the server does today
- SPARQL 1.1 query and update at
/queryand/update, graded against the W3C SPARQL 1.1 evaluation suite on every run so behaviour cannot drift unnoticed. - Named graphs and datasets.
GRAPH(with a ground IRI or a variable),FROM/FROM NAMED, updates scoped to a named graph, and the SPARQL 1.1 Graph Store Protocol at/graphsfor reading and writing one whole graph over plain HTTP. Load quads from.nqand.trigfiles as well as triples from.ntand.ttl. - OWL 2 RL reasoning, off by default. Turn it on per view — your ontologies plus one data graph — and the triples it derives land in a separate inferred graph, so reading a data graph back returns exactly the quads you wrote to it. A view that derives a contradiction is flagged rather than allowed to poison the rest of the store.
- Operations. Layered configuration from
config.toml, environment variables, and flags, with a file watcher that applies edits to reloadable settings without a restart; Prometheus metrics at/metrics; liveness and readiness at/healthzand/readyz.
Where to go
These docs come in two kinds of page, so you always know what you’re reading.
- Guides — tutorials that walk you through a task end-to-end, how-to guides for a specific goal, and explanations of the design behind HornDB. Start with Get your first reasoned answer in five minutes if you are new here.
- Reference — exact syntax and behaviour: SPARQL functions and operators, configuration keys, and error codes. Look here when you already know what you want and need the precise details.
Looking for the project itself, rather than its docs? See horndb.io or the source on GitHub.