Getting started

Corium currently builds with Rust 1.85 or newer. From the repository root:

cargo build --workspace
cargo test --workspace

Start a local transactor in one terminal:

cargo run -p corium-cli -- transactor \
  --data-dir ./corium-data \
  --listen 127.0.0.1:4334 \
  --metrics-listen 127.0.0.1:9464

The transactor keeps its blobs, root pointers, and transaction logs in one of the selected stores: fs (the default, under --data-dir), mem (fully in-memory and ephemeral — handy for demos and tests), postgres (build with --features postgres and pass --postgres-url), or turso (an embeddable-SQLite database; build with --features turso and pass --turso-path). postgres and turso store their transaction logs natively alongside blobs and roots; fs uses versioned log files and mem uses an in-process registry.

cargo run -p corium-cli -- transactor --store mem --data-dir ./corium-data
cargo run -p corium-cli --features postgres -- \
  transactor --store postgres --postgres-url "$DATABASE_URL" \
  --data-dir ./corium-data
cargo run -p corium-cli --features turso -- \
  transactor --store turso --data-dir ./corium-data

Create a schema file named schema.toml:

schema-version = 1

[[entity]]
name = "person"

[entity.attributes]
name = { type = "string", unique = "identity", index = true }
age = "long"

The entity block is schema-authoring sugar: it supplies the person group for the person/name and person/age attributes, but does not install or enforce an entity type. Flat [[attribute]] declarations and the original Datomic-style EDN schema format are also supported. See TOML schemas.

Create the database and inspect it:

cargo run -p corium-cli -- db create people --schema schema.toml
cargo run -p corium-cli -- db list
cargo run -p corium-cli -- db stats people
cargo run -p corium-cli -- console people

The console accepts EDN Datalog directly:

[:find ?name ?age
 :where [?e :person/name ?name]
        [?e :person/age ?age]]

Transactions can be submitted from the Rust peer API or from the corium.api/transact Clojurust binding. See Clojurust integration for the boundary API and Operations for production process flags, backup, metrics, and recovery.

For a fuller worked example — a MusicBrainz schema, a streaming data loader, a Clojurust query REPL, and one-command scripts for every store type — see examples/musicbrainz.