Styjun

Consider the following snippet:

int i = 5 / 0;

This gives compiler error CS0020: Division by constant zero, which is fine. However, the next snippet:

int i = 10;
i = i / 0;

Compiles just fine.

Does someone know why? I see no reason why the compiler allows an integer variable to be divided by a zero integer constant.

In the general case, the compiler has no reason to disallow division by zero (or any other number) at runtime.

Your first example, though, is a compile time constant, i.e. it's calculated by the compiler and replaced with the result of the evaluation. This is what the compiler's complaining about as it rightly doesn't know what integer value to put in place of 5/0.

It is simply the difference between the compile time check and the runtime check. While

i = i / 0;

throws

System.DivideByZeroException: 'Attempted to divide by zero.'

the

i = 5 / 0;

gives a compilation error

Error CS0020 Division by constant zero.

The value of 5 / 0 is evaluated at compile time, this is what makes the difference between receiving an error and getting an exception.

EDIT:

Given the comments to the answers here, I will take some time to write an assumption of why the compiler works this way. The compiler was never intended to prevent division by zero of a variable. If you want to do it, you are allowed to, and you could work in some sneaky ways to get your program dividing by zero. What was the compiler supposed to do is just making some optimization when it can, by computing it advance some calculus and replace them with the resulting value. When it sees the division by zero, it just gives you an error as it can not compute it.

Because of the compiler converts int i = 5 / 0; to a constant value , so it must calculate the value during compile time, which gives it an error.

But the code

int i = 10;
i = i / 0;

is formally correct. It throws an exception and it is your decision and your intention to generate such an exception.

same as everyone saying here.
This line here is compile time and since compiler doesn't know the answer of this.It is giving error

int i = 5 / 0;

where as this line is a run time check and its just a general mathematical calculation so compiler no reason to throw this as an error

 int i = 10;
 i = i / 0;