OntoCode

A planned VS Code extension for ontology-as-code — powered by a Rust backend.

OntoCode aims to become a full ontology engineering workbench inside VS Code: browse classes and properties, edit OWL/RDF, run queries, validate in CI, review semantic diffs in pull requests, and work the way modern software teams already work with Git and editors.

Build, query, validate, refactor, reason over, and document OWL/RDF ontologies directly in VS Code.

Status: Early development. The VS Code extension is not shipped yet. This repository currently contains OntoIndex — the Rust engine that will power the extension — plus planning specs for the full product.

Two-layer architecture

OntoCode is designed as two products that ship together:

┌─────────────────────────────────────┐
│  OntoCode (planned)                 │
│  VS Code extension + UI panels      │
└─────────────────┬───────────────────┘
                  │ Language Server
┌─────────────────▼───────────────────┐
│  OntoIndex (v0.1.0)                 │
│  Rust index, catalog, query, CLI    │
└─────────────────┬───────────────────┘
                  │ Oxigraph / RDF parsers
┌─────────────────▼───────────────────┐
│  Your ontology repo                 │
│  .ttl .owl .rdf .jsonld …           │
└─────────────────────────────────────┘

OntoIndex is useful on its own today (CLI, CI, local analysis). The extension will call into the same engine via a language server rather than reimplementing ontology logic in TypeScript.

Why OntoCode?

Protégé is strong for traditional ontology editing, but most engineering teams live in Git, pull requests, and VS Code. OntoCode is being built for that workflow:

• Git-native semantic diffs and review
• CI-friendly validation (ontoindex validate)
• Editor-native navigation and refactoring
• SQL-like and SPARQL querying over a workspace index
• Local-first indexing — no upload by default

Long-term goal: routine ontology work in VS Code without opening Protégé.

What's in v0.1.0 (OntoIndex foundation)

This release delivers the Rust backend described in the v0.1 roadmap:

• Workspace scanner — recursive discovery, .gitignore support, content hashing
• RDF/OWL parsing — Turtle, RDF/XML, OWL, JSON-LD, N-Triples, N-Quads, TriG via Oxigraph
• Semantic catalog — ontologies, classes, properties, individuals, annotations, axioms, namespaces, imports
• SQL-like queries — SELECT, FROM, WHERE, projections, CSV/JSON export
• SPARQL — query indexed triples directly
• CLI — ontoindex index, query, sparql, validate, inspect

Exit criterion (works today):

cargo run -- query ./fixtures "SELECT * FROM classes"

Quick start

# Build
cargo build --release

# Index and inspect a workspace
cargo run -- inspect fixtures

# Query classes
cargo run -- query fixtures "SELECT * FROM classes"

# Filter results
cargo run -- query fixtures "SELECT short_name, labels FROM classes WHERE short_name = 'Person'"

# SPARQL
cargo run -- sparql fixtures "SELECT ?s ?p ?o WHERE { ?s ?p ?o } LIMIT 5"

# Validate (non-zero exit on parse errors — CI-friendly)
cargo run -- validate fixtures

# JSON output
cargo run -- query fixtures "SELECT * FROM classes" --format json

Install the CLI from crates.io (binary name: ontoindex):

cargo install ontoindex-cli
ontoindex query ./fixtures "SELECT * FROM classes"

Or build from source after cloning this repository.

Planned VS Code experience (not yet built)

Specs and wireframes live in ontocode_ontoindex_docs/. Upcoming OntoCode UI includes:

• Ontology Explorer sidebar (classes, properties, individuals)
• Entity inspector with jump-to-source
• Inline diagnostics and validation
• Class/property/individual authoring
• SPARQL and SQL query panels
• Reasoner integration and graph visualization
• Semantic Git diff viewer

The extension will be a thin TypeScript shell over ontoindex-lsp and the OntoIndex crates — not a second ontology stack.

Roadmap

See ROADMAP.md and PLAN.md for the full product plan.

Repository layout

crates/
├── ontoindex-core      # types, workspace scanner
├── ontoindex-parser    # RDF parsing and entity extraction
├── ontoindex-catalog   # index builder and semantic catalog
├── ontoindex-query     # SQL-like and SPARQL engines
└── ontoindex-cli       # `ontoindex` binary
fixtures/               # sample ontology for tests
ontocode_ontoindex_docs/  # specs, ADRs, wireframes, backlog
tests/                  # integration and golden snapshot tests

Virtual tables

Development

cargo test --workspace
cargo fmt --all
cargo clippy --all-targets -- -D warnings

Update golden snapshots:

ONTOINDEX_UPDATE_GOLDEN=1 cargo test golden_classes

Releasing

Published crates (v0.1.0):

Push a tag matching [workspace.package].version in Cargo.toml (e.g. v0.1.0):

git tag v0.1.0
git push origin v0.1.0

The release workflow verifies packages, runs tests, publishes workspace crates to crates.io in dependency order, and creates a GitHub Release with the ontoindex Linux binary. Requires the CARGO_REGISTRY_TOKEN repository secret.

See CHANGELOG.md for release notes.

License

MIT OR Apache-2.0