Why Doesn't C Throw an Exception for Integer Divide-by-Zero?

Catching Divide-by-Zero Exceptions

Despite attempts to catch an exception using a try/catch block, the following code inexplicably bypasses the exception when dividing by zero:

int i = 0;

cin >> i; // input validation not present

try {
    i = 5/i;
}
catch (std::logic_error e) {

    cerr << e.what();
}

Exception Handling in C

Contrary to expectations, integer division by zero is not a built-in exception in standard C . This is also true for floating-point division/remainder by zero, although in this case, specific rational values like NaN/Inf may result.

The list of exceptions defined in the ISO C 20 standard includes:

logic_error
domain_error
invalid_argument
length_error
out_of_range
runtime_error
range_error
overflow_error
underflow_error

While arguments could be made for using overflow_error or domain_error to indicate a divide-by-zero, the standard explicitly states that the behavior is undefined in such cases.

Custom Exception Handling

If a divide-by-zero exception is desired, it can be implemented manually. The following example uses the intDivEx function to divide integers with exception handling:

inline int intDivEx (int numerator, int denominator) {
    if (denominator == 0)
        throw std::overflow_error("Divide by zero exception");
    return numerator / denominator;
}

The output of using this function demonstrates the exception being thrown and caught, leaving the original variable untouched in the case of a divide-by-zero:

i = intDivEx(10, 0); // throws exception

i = intDivEx(10, 2); // returns 5

The above is the detailed content of Why Doesn't C Throw an Exception for Integer Divide-by-Zero?. For more information, please follow other related articles on the PHP Chinese website!