developers.md

Developers

Rust

The Rust API is broken down into 2 crates

• ./crates/order_book: Simulated order-book and stock-market implementations.
• ./crates/step_sim: A discrete event simulation library built around the order-book library.

The crates are organised under a single workspace so cargo commands can be run from the repo root.

The rust CI workflow expects the following checks to pass

cargo test
cargo fmt --all -- --check
cargo clippy -- -Dwarnings
cargo doc --no-deps

and the examples to successfully run

cargo run --example random_agents
cargo run --example order_book
cargo run --example multi_asset

Python

The Python API consists of

• ./rust: A Rust-Python that provides Python classes forwarding functionality from the Rust simulation libraries.
• ./src: Python library wrapping the Rust interface with additional simulation utilities.
• ./docs: Sphinx documentation.
• ./examples: Python simulation examples.

Bourse uses hatch for dependency management and maturin to build the combined Python-Rust package.

Development commands can be run from a virtual environment using hatch

• hatch run dev:develop: Build and install the Python library
• hatch run dev:lint: Run Python linting
• hatch run dev:test: Run Python tests
• hatch run dev:bench: Run Python benchmarks
• hatch run dev:build: Build the Python library
• hatch run dev:examples: Run the Python examples

Documentation can be built and tested using hatch run docs:build and hatch run docs:test respectively.