aatxe-bench

Rust authoring API + JSON emitter for aatxe — a statistical microbenchmark + regression-gate harness for TypeScript, Go, and Rust.

Quick start

[dev-dependencies]
aatxe-bench = "0.1"

Write a runner binary:

use aatxe_bench::{bench, keep, Suite};

fn main() {
    let mut suite = Suite::new("my-service");

    bench(&mut suite, "parse_phone", || {
        keep(parse_phone("+34 612 345 678"));
    });

    suite.emit_stdout();
}

The runner emits a single RunReport JSON on stdout. Wire it into aatxe run --lang rust and aatxe compares head vs. base, posts a sticky PR comment, and gates CI when a regression is statistically real.

Why a builder, not #[bench]?

The stable Rust toolchain does not ship #[bench]. Criterion is the standard alternative but introduces a heavy dependency tree and an HTML-report-centric flow. aatxe-bench instead exposes a small builder that integrates cleanly with cargo run --release — predictable, statically typed, and trivial to embed in CI.

Sampling matches the cross-language defaults:

• warmup iterations excluded from measurement;
• adaptive sampling that stops once the CV drops below Options::target_cv, the time budget expires, or Options::max_iterations is hit;
• automatic batch sizing for sub-µs operations.

Defeating dead-code elimination

use aatxe_bench::{black_box, keep};

bench(&mut suite, "parse_foo", || {
    keep(parse_foo("hello"));  // identity at the value level, defeats DCE
});

keep is named for parity with the TS (keep) and Go (Keep) SDKs. black_box is re-exported from std::hint for criterion-refugee muscle memory.

License

MIT.