What are Trap Representations in C, and How Can We Safely Access the Integer Representation of a Float?

Delving into Trap Representations in C: Understanding Undefined Behavior

In C programming, a "trap representation" refers to a bit pattern that falls within the memory space occupied by a particular data type but triggers undefined behavior if used as a value of that type. While not explicitly defined in C89, the term was introduced in C99 and serves as a catch-all for such bit patterns.

The existence of trap representations is not guaranteed for all data types. In fact, the standard explicitly states that unsigned char is the only type guaranteed not to have trap representations. However, it provides two hypothetical examples of trap representations, which are largely irrelevant in modern computing practices. Notably, a more practical example of a trap representation is a signaling NaN (Not-a-Number) in a floating-point type.

Regarding your code snippet, where a float variable is converted to an int pointer:

float f=3.5;
int *pi = (int*)&f;

Assuming sizeof(int) and sizeof(float) are equal, it's important to note that this code has undefined behavior due to pointer aliasing rules. It's not related to trap representations.

To correctly extract the integer representation of a float value, you can use the following code:

int extract_int(float f)
{
    union { int i; float f; } u;
    u.f = f;
    return u.i;
}

This code uses a union to interpret the float value as an integer, which provides unspecified behavior in C99. Although the precise integer value returned is not defined, it guarantees a valid integer, avoiding trap representations.

The above is the detailed content of What are Trap Representations in C, and How Can We Safely Access the Integer Representation of a Float?. For more information, please follow other related articles on the PHP Chinese website!