FAQ¶
Q: No test detected why?¶
If no test was ran or your test cases were not detected, in most cases it because your compiler did not define the __BASE_FILE__ macro. If you are using the Visual studio IDE you should define the macro in Properties -> C/C++ -> Preprocessor -> Preprocessor Definition as __BASE_FILE__=”%(Filename)%(Extension)”.
Or you can define the macro at compile time as argument to your compiler using the macro option. e.g. in gcc
gcc -D__BASE_FILE__=\"/the/path/to/yout/testfile.c\" testfile.c -I.
You can also define the __BASE_FILE__ at the beginning of your test file with the absolute path to the test file. E.g for the test file test.c:
#define __BASE_FILE__ "/path/to/test.c"
#include <exotic/cester.h>
CESTER_TEST(test1, test_instance,
cester_assert_equal(NULL, NULL);
)
Alternatively the test cases should be manually registered in the main method, you will have to disable cester main function by defining the macro CESTER_NO_MAIN.
#include <exotic/cester.h>
#define CESTER_NO_MAIN
CESTER_TEST(test1, test_instance,
cester_assert_equal(NULL, NULL);
)
CESTER_BODY(
int main(int argc, char** argv) {
CESTER_REGISTER_TEST(test1);
return CESTER_RUN_ALL_TESTS(argc, argv);
}
)
Visit this link Manual test registration for more detail on manual test registration.
Q: What is the best practice for test detection¶
Include the guide for __BASE_FILE__
#ifndef __BASE_FILE__
#define CESTER_NO_MAIN
#endif
#include <exotic/cester.h>
//your test cases here
#ifndef __BASE_FILE__
CESTER_BODY(
int main(int argc, char** argv) {
// register your test cases here
return CESTER_RUN_ALL_TESTS(argc, argv);
}
)
#endif
This ensure that if the directive __BASE_FILE__ is not defined by the compiler the test case will be registered manually.
Q: Why is test_instance->arg value not changing¶
In isolated tests the test_instance object is not preserved between the test cases, each test case has a new copy of the test_instance object the same applied to CESTER_BEFORE_EACH and CESTER_AFTER_EACH. changes made in each test cases, the CESTER_BEFORE_EACH and CESTER_AFTER_EACH does not reflect in each test case as each test case is ran on a new process, the solution is to declare shared datas in CESTER_BEFORE_ALL and CESTER_AFTER_ALL functions. Or add the –cester-noisolation to run the tests on a single process this way all data are shared. argv and argc values of the test_instance is shared between each test process in isolated test only arg data is not shared.
Note
Using the –cester-noisolation cester will rely on signal to recover from critical crash.
The example below recreates the problem and provides a solution.
#include <exotic/cester.h>
CESTER_BEFORE_EACH(test_instance, test_name, index,
if (test_name == "test_one") {
test_instance->arg = "Hello World";
}
)
CESTER_TEST(test_one, test_instance,
cester_assert_equal(test_instance->arg, "Hello World");
)
CESTER_TEST(test_two, test_instance,
cester_assert_equal(test_instance->arg, "Yahoo");
)
CESTER_AFTER_EACH(test_instance, test_name, index,
test_instance->arg = "Yahoo";
)
Compiling and running test on seperate process.
gcc ./crash_test.c -I. -o out
./out
+ (0.01) test_one
- (0.00) test_two
AssertionError crash_test.c:14: in test case 'test_two' => expected "Yahoo", found test_instance->arg
Ran 2 test(s) in 0.01 Seconds
Synthesis: FAILURE Tested: 2 | Passing: 1 | Failing: 1 | Skipped: 0
The test above wil fail because for test case test_two because the test_instance->arg value will still be “Hello World” as the CESTER_AFTER_EACH executed for test_one will be exclusive to test_one process only. The solution is to applly the option –cester-noisolation to the test to use a single process for all the tests.
gcc ./crash_test.c -I. -o out
./out --cester-noisolation
+ (0.00) test_one
+ (0.00) test_two
Ran 2 test(s) in 0.00 Seconds
Synthesis: SUCCESS Tested: 2 | Passing: 2 | Failing: 0 | Skipped: 0