Scalar types: numbers, bool and char

Why so many number types?

Rust is a systems language, so it lets you choose exactly how much memory each number uses. A type is the kind of a value; it decides what the value can hold and what you can do with it. Rust is statically typed: every value's type is known when the program is compiled, so many mistakes are caught early.

Integers

A bit is a single 0 or 1. More bits hold bigger numbers. Signed types use one bit for the sign. usize matches your computer: 64 bits on a 64-bit machine. Rust uses usize for list positions and lengths, so you will see it often.

let population: u64 = 1_400_000_000;
let temperature: i8 = -5;
let byte: u8 = b'A';     // 65
let mask = 0xff;          // hexadecimal, i32 by default
println!("{population} {temperature} {byte} {mask}");

Integer overflow

Overflow means a result is too big (or too small) for its type, like 255 + 1 in a u8. In a debug build, Rust stops the program with a panic, an immediate crash with a message. In a release build, it wraps around silently (255 + 1 becomes 0), unless you turn overflow checks on. Neither is what you want for money or counters, so use the explicit methods:

let a: u8 = 250;
println!("{:?}", a.checked_add(10));    // None: would overflow
println!("{}", a.wrapping_add(10));     // 4: wraps around
println!("{}", a.saturating_add(10));   // 255: stops at the maximum

Floating-point numbers

f64 (the default) and f32 hold decimals. They are approximate, like most languages: 0.1 + 0.2 prints 0.30000000000000004. For money, store whole paise in an integer (Rs 12.50 as 1250) instead of a float.

Booleans and characters

bool is true or false. char is one Unicode character in single quotes, and it always takes 4 bytes, so it can hold any letter of any script or an emoji: 'a', '₹', 'क'. Text in double quotes is a string, not a char; strings come in module 4.

Converting with as

let price = 99.99_f64;
let whole = price as i32;      // 99: the decimal part is cut off
let big: i32 = 300;
let small = big as u8;         // 44: 300 does not fit, so it is truncated
println!("{whole} {small}");

Common mistakes

• Mixing types: an i32 plus a u8 does not compile. Convert one side first.
• Using f64 for money. Use integers of the smallest unit (paise).
• Using as to shrink numbers without checking. Prefer u8::try_from(x), which returns an error if it does not fit.
• Assuming overflow is handled the same in debug and release builds. Use checked_ or saturating_ when it matters.