Measuring Data Sizes

Measuring Data Sizes

You may remember in Hour 4, "Data Types and Names in C," I mentioned that each data type has its own size. Depending on the operating system and the C compiler you're using, the size of a data type varies. For example, on most UNIX workstations, an integer is 32 bits long, while most C compilers only support 16-bit integers on a DOS-based machine.

So, how do you know the size of a data type on your machine? The answer is that you can measure the data type size by using the sizeof operator provided by C.

The general form of the sizeof operator is

sizeof (expression)

Here expression is the data type or variable whose size is measured by the sizeof operator. The value of the size is returned, in units of bytes, by the sizeof operator. For instance, if an integer is 16 bits long, the value returned by the sizeof operator will be 2 (bytes). (Note that 8 bits are equal to 1 byte.)

The parentheses are optional in the general form of the operator. If the expression is not a C keyword for a data type, the parentheses can be discarded.

For instance, the following statement

size = sizeof(int);

measures the size of the int data type and returns the number of bytes required by the data type to an int variable size.

The program in Listing 8.1 finds the sizes of the char, int, float, and double data types on my machine.
TYPE
Listing 8.1. Using the sizeof operator.

1:  /* 08L01.c: Using the sizeof operator */
2:  #include <stdio.h>
3:
4:  main()
5:  {
6:     char   ch = ` `;
7:     int    int_num = 0;
8:     float  flt_num = 0.0f;
9:     double dbl_num = 0.0;
10:
11:    printf("The size of char is: %d-byte\n", sizeof(char));
12:      printf("The size of ch is: %d-byte\n", sizeof ch );
13:        printf("The size of int is: %d-byte\n", sizeof(int));
14:          printf("The size of int_num is: %d-byte\n", sizeof int_num);
15:            printf("The size of float is: %d-byte\n", sizeof(float));
16:          printf("The size of flt_num is: %d-byte\n", sizeof flt_num);
17:        printf("The size of double is: %d-byte\n", sizeof(double));
18:      printf("The size of dbl_num is: %d-byte\n", sizeof dbl_num);
19:    return 0;
20: }

    OUTPUT
    After this program is compiled and linked, an executable file, 08L01.exe, is created. The following is the output printed on the screen after the executable is run from a DOS prompt on my machine:

    C:\app> 08L01
    The size of char is: 1-byte
    The size of ch is: 1-byte
    The size of int is: 2-byte
    The size of int_num is: 2-byte
    The size of float is: 4-byte
    The size of flt_num is: 4-byte
    The size of double is: 8-byte
    The size of dbl_num is: 8-byte
    C:\app>

    ANALYSIS
    Line 2 in Listing 8.1 includes the header file stdio.h for the printf() function used in the statements inside the main() function body. Lines 6_9 declare a char variable (ch), an int variable (int_num), a float variable (flt_num), and a double variable (dbl_num), respectively. Also, these four variables are initialized. Note that in line 8, the initial value to flt_num is suffixed with f to specify float. (As you learned in Hour 4, you can use f or F to specify the float type for a floating-point number.)

Lines 11 and 12 display the size of the char data type, as well as the char variable ch. Note that the sizeof operator is used in both line 11 and line 12 to obtain the number of bytes the char data type or the variable ch can have. Because the variable ch is not a keyword in C, the parentheses are discarded for the sizeof operator in line 12.

The first two lines in the output are printed out by executing the two statements in line 11 and 12, respectively. From the output, you see that the size of the char data type is 1 byte long, which is the same as the size of the variable ch. This is not surprising because the variable ch is declared as the char variable.

Likewise, lines 13 and 14 print out the sizes of the int data type and the int variable int_num by using the sizeof operator. You see that the size of each is 2 bytes.

Also, by using the sizeof operator, lines 15_18 give the sizes of the float data type, the float variable flt_num, the double data type, and the double variable dbl_num, respectively. The results in the output section show that the float data type and the variable flt_num have the same size (4 bytes). The double data type and the variable dbl_num are both 8 bytes long.

From the output you see that char uses 1 byte, while an int uses 2 bytes. Accordingly, while variables of type char have been used to store integers, they cannot store integers of the same range as a variable of type int can.