Previous: 2.5 More C Statements
Up: 2 Basic Concepts
Next: 2.7 Summary
Previous Page: 2.5.6 More Complex Loop Constructs - Nested Loops
Next Page: 2.7 Summary

2.6 Common Errors

In this section we list some common problems and programming errors that beginners often make. We also suggest steps to avoid these pitfalls.

1. Program logic is incorrect. This could be due to an incorrect understanding of the problem statement or improper algorithm design. To check what the program is doing, manually trace the program and use debug statements. Introduce enough debug statements to narrow down the code in which there is an error. Once an error is localized to a critical point in the code or perhaps to one or two statements, it is easier to find the error. Critical points in the code include before a loop starts, at the start of a loop, at the end of a loop and so forth.
2. Variables are used before they are initialized. This often results in garbage values occurring in the output of results. For example:

   int x, y;
   
        x = x * y;

   There is no compiler error, x and y have unknown, garbage values. Be sure to initialize all variables.
3. The assignment operator, =, is used when an ``equal to'' operator, ==, is meant, e.g.:

   while (x = y)
             ...
   
        if (x = y)
             printf("x is equal to y\n");

   There will be no compiler error since any valid expression is allowed as an if or while condition. The expression is True if non-zero is assigned, and False if zero is assigned. Always double check conditions to see that a correct equality operator, ==, is used.
4. Object names are passed, instead of addresses of objects, in function calls to scanf():

   scanf("%d", n);          /* should be &n */

   Again this is not a compile time error; the compiler will assume the value of n is the address of an integer object and will attempt to store a value in it. This often results in a run time addressing error. Make sure the passed arguments in scanf() calls are addresses of the objects where data is to be stored.
5. Loop variables are not initialized:

   while (i < n)
             ...

   i is garbage; the while expression is evaluated with unknown results.
6. Loop variables are not updated:

   i = 0;
        while (i < n) {
             ...
        }

   i is unchanged within the loop; it is always 0. The result is an infinite loop.
7. Loop conditions are in error. Suppose, a loop is to be executed ten times:

   n = 10;
        i = 0;
        while (i <= n) {
             ...
             i = i + 1;
        }

   (i <= n) will be True for i = 0, 1, ..., 10, i.e. 11 times. The loop is executed one more time than required. Loop expressions should be examined for values of loop variables at the boundaries. Suppose n is zero; should the loop be executed? Suppose it is 1, suppose it is 10, etc.
8. User types in numbers incorrectly. This will be explained more fully in Chapter . Consider the loop:

   while (x != 0) {
             ...
             scanf("%d", &x);
        }

   Suppose a user types: 23r. An integer is read by scanf() until a non-digit is reached, in this case, until r is reached. The first integer read will be 23. However, the next time scanf() is executed it will be unable to read an integer since the first non-white space character is a non-digit. The loop will be an infinite loop.
9. Expressions should use consistent data types. If necessary, use a cast operator to convert one data type to another.

   int sum, count;
        float avg;
   
        avg = sum / count;

   Suppose sum is 30 and count is 7. The operation sum / count will be the integer value of 30 / 7, i.e. 4; the fractional part is truncated. The result 4 is assigned to a float variable avg as 4.0. If a floating point value is desired for the ratio of sum / count, then cast the integers to float:

   avg = (float) sum / (float) count;

   Now, the expression evaluates to 30.0 / 7.0 whose result is a floating point value 4.285 assigned to avg