Skip to content

testing.md

Latest commit

 

History

History
162 lines (110 loc) · 3.98 KB

testing.md

File metadata and controls

162 lines (110 loc) · 3.98 KB

Testing

Rusty Wire uses three test layers:

  • Rust tests via cargo test (unit + integration)
  • scripts/test-itu-region-bands.sh for region-band regression checks
  • scripts/test-multi-optima.sh for non-resonant multi-optima regression sweeps
  • scripts/test-nec-calibration.sh for conductor-calibration regression checks

All of the above can be run together:

./scripts/test-all.sh

Primary Command

cargo test

Useful variants:

cargo test -- --nocapture
cargo test cli_integration

Integration Coverage

Integration tests live in tests/cli_integration.rs and validate real binary behavior, including:

  • no-argument help output
  • invalid input validation (for example velocity and mixed-unit window flags)
  • region-aware band listing
  • transformer recommendation resolution
  • export path behavior for single and multi-format runs

Regression Scripts

  • no-argument invocation prints help
  • mixed meter/feet constraints return a validation error
  • invalid velocity values return a validation error
  • --list-bands --region 2 shows region-specific output
  • recommended transformer defaults resolve correctly for non-resonant runs and EFHW mode
  • multiple export formats ignore a custom --output path and use default names
  • single-format export respects the requested --output path

These tests are intentionally high-level so that clap parsing and the real CLI flow are both covered.

Future Validation Idea

  • Add an advise-mode cross-tool validation script that calls fnec-rust for the top-ranked wire + transformer candidates.
  • Record per-candidate sustainability status (for example: validated, warning, rejected) based on agreed efficiency/thermal thresholds.
  • Use this as a regression guard when optimizer scoring weights or practical-limit coefficients are changed.

Script: Multi-Optima Sweep

Run:

./scripts/test-multi-optima.sh

Purpose:

  • builds the project
  • sweeps band selections, velocity factors, and wire-length windows
  • stops at the first case where multiple non-resonant optima are found

This is not a unit test. It is an empirical regression script used to confirm that the optimization logic still produces multiple-optima cases under realistic parameter sweeps.

Environment variables:

  • BIN: path to the binary to execute, default target/debug/rusty-wire
  • SWEEP_OUT: path to the temporary sweep output file, default /tmp/rw_sweep_result.txt

The sweep uses CLI band names/ranges (for example 40m, 20m,17m, 40m-10m) to match the current --bands parser behavior.

Script: ITU Region Band Regression

Run:

./scripts/test-itu-region-bands.sh

Checks listed ranges for Regions 1, 2, and 3.

Script: NEC Calibration Regression

Run:

./scripts/test-nec-calibration.sh

Checks:

  • template dataset fit stays at documented values (k = 0.011542, RMSE = 0.000000)
  • parser accepts blank lines and # comment lines in calibration CSV input
  • malformed rows are rejected with non-zero exit

Multi-optima sweep

./scripts/test-multi-optima.sh

Builds and sweeps parameter combinations, then exits at the first confirmed multi-optima case.

Environment variables for the sweep script:

  • BIN default: target/debug/rusty-wire
  • SWEEP_OUT default: /tmp/sweep_out.txt

Recommended Workflow

Full suite (format + compile + unit + all regression scripts):

./scripts/test-all.sh

Default pre-push verification (Rust only):

cargo fmt
cargo check
cargo test

On version bump (before tagging):

cargo sbom
cargo sbom-cdx
git add sbom/

For the full release flow (including manual real-terminal screenshot refresh), see docs/release-checklist.md.

When changing region or band behavior:

cargo test
./scripts/test-itu-region-bands.sh

When changing non-resonant optimization behavior:

cargo test
./scripts/test-multi-optima.sh

When changing conductor-diameter correction constants:

cargo test
./scripts/test-nec-calibration.sh