AutoTest

##Introduction This simple tool can be used for generating testing part for C functions, which means you can just write the function and the testing cases. I spent one day to create this simple version 1.0 because I am practising algorithms recently, and writing testing code is so boring and repetitive, and I hope it will be also convenient for you if you would like to have a try.

##Usage Up to now, one line "//FUNC" must be right before the C function to indicate the tested function, and the "//IN" to indicate the input of the function, which can't be more than one line and we use "//RET" to indicate the return value.The C function must be standard format, which means the return type in a separate line.

All the symbols are listed below:

• //FUNC: the start of a function
• //IN: one input case
• //RET: one return value

For example:

//FUNC
int
sum(int A[], int N)
{
	/*
		//IN
		A[3] = {1, 2, 3}; N = 3;
		//RET
		6
		//IN
		A[2] = {1, 2}; N = 2;
		//RET
		3
	*/
	int i, total = 0;
	
	for(i = 0; i < N; i++)
		total += A[i];
	
	return total;
}

Notes

• Type. The input needs to be written like C code, but don't need to add the data type, which will be added automatically.
• Name. The name of the variables can be arbitrary and it's ok be the same in different test cases, which will be renamed automatically.
• Order. The order of the input must be the same with the arguments declared in the function prototype.
• Indicator. The "//FUNC" is not necessary for every function and you just need add this label for the function you want to test.Even if you add the "//FUNC", but you don't write the test cases, the test code of this function will not be generated.

##Commands Under the Linux with Python, if you have a source file named file.c ,you could use this tool by typing this command:

python Tester.py file.c

Then a new file called test_file.c will be created, which contains the function and the test code.

If you don't want to create a new file, you can use the below command:

python Tester.py file.c -nk

Then the file.c will contains the testCode.

By default, the new test file is just add a prefix "test_" before your file. So if you want to rename this file, you can add one more argument:

python Tester.py file.c dd newfile.c

Noted that if you rename the file containing the test code, the third argument which indicates whether to keep original file can be arbitrary.

No matter whether we generate a new file, the source file containing test code will then be executed by gcc, thus checking all the test and showing the result.

##Examples If we have a C source file sum.c

//FUNC
int
sum(int A[], int N)
{
	/*
		//IN
		A[3] = {1, 2, 3}; N = 3;
		//RET
		6
		//IN
		A[2] = {1, 2}; N = 2;
		//RET
		3
	*/
	int i, total = 0;
	
	for(i = 0; i < N; i++)
		total += A[i];
	
	return total;
}

Then this tool will generate test file like below:

#include <stdio.h>
#include <stdlib.h>

//FUNC
int
sum(int A[], int N)
{
	/*
		//IN
		A[3] = {1, 2, 3}; N = 3;
		//RET
		6
		//IN
		A[2] = {1, 2}; N = 2;
		//RET
		3
	*/
	int i, total = 0;
	
	for(i = 0; i < N; i++)
		total += A[i];
	
	return total;
}

void
test_sum()
{
	int A0[3] = {1, 2, 3};
	int N0 = 3;
	
	int A1[2] = {1, 2};
	int N1 = 2;
	
	if(6 != sum(A0, N0) )
	{
		printf("sum: Case 0 fail.\n");
		exit(1);
	}

	if(3 != sum(A1, N1) )
	{
		printf("sum: Case 1 fail.\n");
		exit(1);
	}

}

int
main()
{
	test_sum();

	return 0;
}

The output is shown below:

gcc compiles successfully.
test_sum runs successfully, and pass all the tests.

If we modify the second test case to make it wrong, the output will be like this (Noted that the index starts from zero):

gcc compiles successfully.
sum: Case 1 fail.
[Error]: test_sum can not pass all the tests.