ctest is a unit test framework for software written in C.
Features:
- adding tests with minimal hassle (no manual adding to suites or testlists!)
- supports suites of tests
- supports setup() teardown() per test
- output format not messed up when tests fail, so easy to parse.
- displays elapsed time, so you can keep your tests fast
- uses coloring for easy error recognition
- only use coloring if output goes to terminal (not file/process)
- it's small (a little over 300 lines of code!)
- it's easy to integrate (only 1 header file)
- has SKIP option to skip certain test (no commenting test out anymore)
- Linux + OS/X + Windows MSVC support
CTEST(suite, test1) {
ASSERT_STR("foo", "foo");
}
CTEST(suite, test2) {
ASSERT_EQUAL(1, 2);
}
CTEST(suite, test_dbl) {
ASSERT_DBL_NEAR(0.0001, 0.00011);
ASSERT_DBL_NEAR_TOL(0.0001, 0.00011, 1e-5);
}
NO further typing is needed! ctest does the rest.
$ ./test
TEST 1/2 suite1:test1 [OK]
TEST 2/2 suite1:test2 [FAIL]
ERR: mytests.c:4 expected 1, got 2
RESULTS: 2 tests (1 ok, 1 failed, 0 skipped) ran in 1 ms
There can be one argument to: ./test . for example:
$ ./test timer
will run all tests from suites starting with 'timer'
NOTE: when piping output to a file/process, ctest will not color the output
A testcase with a setup()/teardown() is described below. An unsigned char buffer is malloc-ed before each test in the suite and freed afterwards.
CTEST_DATA(mytest) {
unsigned char* buffer;
};
NOTE: the mytest_data struct is available in setup/teardown/run functions as 'data'
CTEST_SETUP(mytest) {
data->buffer = (unsigned char*)malloc(1024);
}
CTEST_TEARDOWN(mytest) {
free(data->buffer);
}
NOTE: setup will be called before this test (and ony other test in the same suite)
NOTE: CTEST_LOG() can be used to log warnings consistent with the normal output format
CTEST2(mytest, test1) {
CTEST_LOG("%s() data=%p buffer=%p", __func__, data, data->buffer);
}
NOTE: teardown will be called after the test completes
NOTE: It's possible to only have a setup() or teardown()
Instead of commenting out a test (and subsequently never remembering to turn it back on, ctest allows skipping of tests. Skipped tests are still shown when running tests, but not run. To skip a test add _SKIP:
CTEST_SKIP(..) or CTEST2_SKIP(..)
The are some features that can be enabled/disabled at compile-time. Each can be enabled by enabling the #define before including ctest.h, see main.c.
#define CTEST_SEGFAULT
ctest will now catch segfaults and display them as error.
There are 2 features regarding colors:
#define CTEST_NO_COLORS
#define CTEST_COLOR_OK
The first one disables all color output (Note that color output will be disabled also when stdout is piped to file).
The CTEST_COLOR_OK will turn the [OK] messages green if enabled. Some users only want failing tests to draw attention and can leave this out then.
-
CTEST_ADD_TESTS_MANUALLY
If defined, test should be added manually using macros: CTEST_ADD(suite, test); // for CTEST CTEST_ADD2(suite, test) // for CTEST2
-
CTEST_NO_JMP
If longjmp() is not available, please define.