Contributing

Thanks for your interest in wickra-backtest — the streaming-native backtester for the Wickra indicator library.

Development

cargo build --workspace
cargo test --workspace --all-features
cargo fmt --all
cargo clippy --workspace --all-targets --all-features -- -D warnings
cargo deny check

Run a backtest locally:

cargo run --bin wkbt -- run --data examples/sample.csv --spec examples/ema-cross.json

Before opening a PR

• cargo fmt --all and cargo clippy … -D warnings must be clean (CI enforces this on three operating systems and the MSRV).
• Add tests. The engine is correctness-critical — prefer a hand-computed expectation (see engine.rs tests) over a smoke test.
• One logical change per PR; a clear, imperative commit message.

Design rules

• Strategies are data, not code. The StrategySpec is JSON so the same strategy runs identically across every Wickra language binding. Keep the DSL small and serialisable.
• No look-ahead bias. Signals are decided on a bar's close and fill on the next bar's open; stop/target/trailing levels fill intrabar. Any change to the fill model must preserve this.
• The engine is feed-agnostic. It consumes a bar stream; loaders and (later) live feeds live outside the core so backtest and live share one engine.

Indicators

The registry (registry.rs) wraps wickra-core indicators behind a uniform EvalIndicator. New indicators are added there (and, eventually, generated from the Wickra manifest). Multi-output indicators expose named fields referenced as "name.field".

License

By contributing you agree that your contributions are licensed under the project's dual MIT OR Apache-2.0 license.