SAL

SAL, the Semantic Accessibility Layer CLI, turns RDF source files into versioned data products. It is designed for projects where the source data lives in Git, the built output is an Apache Iceberg table, and published artifacts need to preserve the link back to the exact source commit.

How SAL fits together

Initialize a project

Start with Getting Started to create the .sal workspace inside a Git repository.

Build RDF

Use sal build to validate Turtle or JSON-LD and write triples into Iceberg.

Inspect data

Query the local Iceberg table with sal query, which runs DuckDB inside sal.

Publish output

Publish to OCI registries or object stores with the workflows in Publishing Data.

Understand the layout

Read Data Layout for how .sal/data stores generated Iceberg files.

Core workflow

1. Run sal init in a Git repository with a configured remote.
2. Add RDF source files to the project.
3. Run sal validate while editing to catch invalid or undefined terms.
4. Commit source changes, then run sal build to create or update the local data product.
5. Use sal query to inspect the built table.
6. Publish the result with sal push or sal upload.

For command flags and examples, see the command reference.