Szathmáry László honlapja @ DEIK | Nim2 / Procedures and functions

proc

In Nim, a proc can be either a procedure (no return value) or a function (has a return value).

(1) procedure [no return value]
#include <stdio.h> void hello() { puts("hello world"); } int main() { hello(); return 0; }	proc hello() = echo "hello world" hello()

(2) function [return a value]
#include <stdio.h> int twice(const int n) { return 2 * n; } int main() { int value = twice(8); printf("%d\n", value); // 16 return 0; }	proc twice(n: int): int = return 2 * n let value = twice(8) echo value # 16

In Nim, formal parameters are in read-only mode. So inside twice(), the value of n cannot be changed. This Nim code is equivalent to the C code on the left side.

func

A func is a pure function.

What is a pure function? A function that has no side effect (doesn't change its environment) and for the same input it always returns the same output.

In the example above, twice() is actually a pure function, thus it could also be defined like this:

func twice(n: int): int =
return 2 * n

Short form

If the body is just 1 line, you can write the body after the '=' sign:

proc hello() = echo "hello world"

If the subroutine has no input parameters, you can omit the parentheses:

proc hello = echo "hello world"

_{Note: I don't really like this last short form.}