library-with-tests

library-with-tests

A single Cabin package with a library target and two test targets that exercise it. This is the example to read for cabin test.

A Cabin test target is an ordinary executable: it passes when its main() returns 0 and fails otherwise. There is no framework, macro, or attribute to learn. cabin test builds every type = "test" target, runs each one, and reports a per-target ... ok / ... FAILED line plus a summary. Tests run in a deterministic order - by package name, then target name - so calc_test always runs before parity_test.

Both tests depend on the calc library through deps = ["calc"] and include its public header through calc's include-dirs = ["include"].

Run the tests

cd examples/library-with-tests
cabin test

Expected output (the Compiling ... build line goes to stderr and is omitted here):

running 2 tests
test library-with-tests:calc_test ... ok
test library-with-tests:parity_test ... ok

test result: ok. 2 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s

To run a single test target, name it with --test:

cabin test --test calc_test

A passing test is silent on stdout. The check(...) helper in tests/*.cc only writes (to stderr) when an assertion fails, which also makes that target exit non-zero so cabin test reports it as FAILED (exit N) and the command exits non-zero.

This package has no executable target, so cabin run does not apply here. A plain cabin build compiles only the calc library; the test targets are dev-only, so cabin test is what builds and runs the two test binaries.