Ranged tests with parameters

A very useful feature of the TestApe framework, is its ability to execute the same test with a given list of parameters, e.g. a range between 0 and 255, or a set of strings e.g. "", NULL, "SAMPLE" that is known to be of importance for the unit being tested. This also gives the posibility to vary the test slightly, in order to achieve the final increase in coverage. Of course, this can be done by copying existing tests; changing a parameter here and there and implementing new mocks. However - it is much easier to use the possibility to execute a ranged test with parameters using the macro PARAMETER, PARAMETER_RANGE, and PARAMETER_SET.

These macros will prepare one or more values for the next executing test. Suppose you want to test myfunc using a bool input parameter with values true and false, then you would prepare the test using PARAMETER(true); PARAMETER(false); and execute the test as using EXECUTE(myfunc)

Some more examples are shown in the the code below. In the sample directory, sample15 has a more complete working example.^2.1

  void interesting_integers(void) {
    PARAMETER(0);
    PARAMETER(-1);
    PARAMETER_RANGE(128,255);
    PARAMETER("some input string");
    PARAMETER("");
    PARAMETER(NULL);
    EXECUTE(interesting_parameters);
    }

This would generate the following output

 Executing test interesting_parameters (step 1 of 133 using value 0)
 ....
 PASSED test interesting_parameters
 Executing test interesting_parameters (step 2 of 133 using value -1)
 ....
 PASSED test interesting_parameters

 ....

 Executing test interesting_integers (step 130 of 133 using value 255)
 ....
 PASSED test interesting_parameters
 Executing test interesting_integers (step 131 of 133 using value "some input string")
 ....
 PASSED test interesting_parameters
 Executing test interesting_integers (step 132 of 133 using value "")
 ....
 PASSED test interesting_parameters
 Executing test interesting_integers (step 132 of 133 using value NULL)
 ....
 PASSED test interesting_parameters

When testing a value in a range from one to 10000, the same test is literally executed 10000 times. This can take some time and generate huge logfiles. As an option, the amount of values tested, and/or the output generated, can be limited from the command line when invoking the test executable. See section [*]