Represent integer as binary in C

Question

To practice my C, I've written a small function in C to display an integer as an 8-bit binary number. Please let me know how to improve this function, and other C-style coding conventions. Thanks!

/**
 * Converts an integer to binary.
 * 
 * @param num Integer to convert.
 * 
 * @return c-string.
**/
const char* to_binary(int num) {

    static char binary[10];
    binary[0] = '0';
    binary[1] = 'b';

    int value = 128;
    int number = num;

    for (int i = 2; i < 10; i++) {
        if (number - value >= 0) {
            binary[i] = '1';
            number = number - value;
        } else {
            binary[i] = '0';
        }
        value = value / 2;
    }

    return binary;

}

And here's how you call it:

int main() {

    const char* result = to_binary(146);

    printf("%s\n", result);

}

Answer 1

• More specific types are available for an eight bit number, such as uint8_t
• static keyword is dangerous (when used as an output buffer) and results in weird behaviour if your method is called twice
• The arithmetic can be replaced with a more simple & (bitwise AND)

void to_binary(uint8_t x, char *output) {
    *output++ = '0';
    *output++ = 'b';
    for (int mask = 1<<CHAR_BIT-1; mask!=0 ; mask>>=1) {
        *output++ = x & mask ? '1':'0'; 
    }
    *output = 0; // null terminate the string
}

Answer 2

Bug

Code attempts to form a string. Yet it lacks a terminating null character and space for it.

// static char binary[10];
static char binary[11];

As a static, it is initialized to all zeros. With increased size, an explicit setting of the null character is not needed.

OP's code undefined behavior perhaps "works" as a zero may exist just past binary[10].

Loop Simplification

Iterate right (most significant) to left (least significant).

for (int i = 10; --i >= 2; ) {
  binary[i] = '0' + (num & 1);
  num /= 2;
}

Unclear functionality

When num > 255 code prints all '1'.
When num < 0 code prints all '0'.

Code should 1) not use an int argument or 2) state functionality when num is outside the [0-255] range.

Alternative

I'd avoid the use of static and provide a buffer to the function. Example.