The entire Complement API now no longer uses t *testing.T, instead opting for ct.TestLike which is an interface. In practice, this changes nothing. However, this enables Complement to be used in many more places where you do not have a t *testing.T object to use e.g benchmarks, scripts. This is of particular use for Complement-Crypto which has to run parts of the test as a standalone binary, and therefore has no t *testing.T to use.
The entire Complement API now uses ct.Fatalf and ct.Errorf for highlighting test failures in red. This should make it significantly easier to skim for the test failure message.
Add Deployment.ContainerID(TestLike, hsName) string to allow tests to interact with the container beyond the abilities of the Complement API e.g log extraction, memory use, CPU use.
t *testing.T, instead opting forct.TestLikewhich is an interface. In practice, this changes nothing. However, this enables Complement to be used in many more places where you do not have at *testing.Tobject to use e.g benchmarks, scripts. This is of particular use for Complement-Crypto which has to run parts of the test as a standalone binary, and therefore has not *testing.Tto use.ct.Fatalfandct.Errorffor highlighting test failures in red. This should make it significantly easier to skim for the test failure message.Deployment.ContainerID(TestLike, hsName) stringto allow tests to interact with the container beyond the abilities of the Complement API e.g log extraction, memory use, CPU use.