Using pure functions

Pure functions are the second technique that we recommend to prevent hard-to-reproduce bugs. Pure functions can be thought of as an extension of the avoid side-effects principle. The definition of a pure function is a function where the following is true:

• No changes are caused outside of the function (no side-effects)
• The return value does not depend on anything but the function parameters

Here are some examples of pure functions:

fn p0() {}

fn p1() -> u64 {
   444
}

fn p2(x: u64) -> u64 {
   x * 444
}

fn p3(x: u64, y: u64) -> u64 {
   x * 444 + y
}

fn main()
{
   p0();
   p1();
   p2(3);
   p3(3,4);
}

Here are some examples of impure functions:

use std::cell::Cell;

static mut blah: u64 = 3;
fn ip0() {
   unsafe {
      blah = 444;
   }
}

fn ip1(c: &Cell<u64>) {
   c.set(333);
}

fn main()
{
   ip0();
   let r = Cell::new(3);
   ip1(&r);
   ip1(&r);
}

Rust does...