I’ve used both CppUnit and boost::test for C++ unittesting. Generally I prefer boost::test, mainly because the auto-test macros minimise the effort to setup tests. But there’s one thing I really miss from CppUnit: the ability to register your own ‘protectors’, instances of which automatically wrap all the run tests. (Technically, you install a test ‘listener’, and that can wrap each test in a protector scope).
I’ve found these invaluable in the past for monitoring unittests for unexpected side effects (e.g checking code hasn’t changed the floating-point unit state flags). I can’t see any equivalent in the boost::test documentation, although BOOST_FIXTURE_TEST_CASE maybe comes closest.
Any suggestions for how to best achieve the same thing as CppUnit’s protectors in boost::test ?
(I haven’t really looked into boost::test‘s implementation yet, but if it’s anything like CppUnit it must use something very like protectors itself).
I’ve never used CppUnit, so not sure how protectors work. Are you looking for something that wraps individual tests, or the entire test suite?
For the former, you could use fixtures as you mention, but as I understand it, fixtures should be considered ‘outside’ the test. They set up whatever the test needs, and cleans it up afterwards. Any actual error-testing should be in the test itself, but can be easily implemented with RAII. Simply define a class which checks whatever you need in its destructor, and then create a local instance of it at the beginning of the test. Since it is constructed first, it gets destructed last, so it can easily check that the test hasn’t modified any unexpected state.
If you want it to check this after all the tests have executed, you probably want global fixtures