Function to check C-C++ code

Added a function which compiles C and C++ code submitted by the student.

1) If compilation is successful, then the submitted code is tested using
   test-cases.

2) To test the function written by the student, a C++ file calls the
   function and passes the argument to the function. Then the function
   checks for the expected return value.

3) If the return value is as expected, then a different set of arguments
   are passed, and the output is checked.

4) If for all set of arguments the output is as expected then the student
   code is graded correct else the error is displayed to the student.

Changed the way the code is graded.

Previously, the algorithm checked the student code for all test-cases.
If all the test-cases were satisfied, the last-line
of the program was reached and printed "All Correct". So at
any point if a test-case fails, the last line is not reached as
the program was terminate. When the string "All Correct" was
found in the output, the code was graded as RIGHT else WRONG.

This is not a proper way for code checking, as the student code *may*
contain a print statement with the string('All Correct'), and thus
can get program RIGHT even though it is WRONG.

So now the student code is tested as follows:

1) The code checks for all test-cases.
2) If all test-cases are satisfied then it returns 0.
3) If any one of the test-case fails, the program is terminated and will return 1.

4) Now depending on the return status(0 or 1), it will grade the code.
a) if 0 then RIGHT
b) if 1 then WRONG
This ensures, no manipulation from student side.

1 files changed, 10 insertions, 0 deletions

diff --git a/testapp/docs/array_sum.cpp b/testapp/docs/array_sum.cpp
new file mode 100644
index 0000000..9b82d3b
--- /dev/null
+++ b/testapp/docs/array_sum.cpp
@@ -0,0 +1,10 @@
+int array_sum(int a[])
+{
+ int i=0, sum=0;
+ for(i=0;i<5;i++)
+ {
+   sum = sum + a[i];
+ }
+ return sum;
+}
+