Numbers

Types

There is no implicit truthy conversion of numbers to bool; conditions must be bool (use n != 0).

Aliases for generic parameters live in the prelude: Int, Float, SignedInt, UnsignedInt.

Literals

d = 1_000_000        # underscores are ignored
h = 0xFF             # hexadecimal
b = 0b1010           # binary
o = 0o17             # octal
f = 1.5              # f64
g = 1.5f32           # f32 suffix

An integer literal takes the type its context requires, so x: i64 = 1 and f(1u8) both work.

Arithmetic

print(7 / 2)      # 3   (integer division)
print(7.0 / 2.0)  # 3.5
print(7 % 3)      # 1
print(2 ** 10)    # 1024

For integer operands both / and // perform truncating division — the fractional part is discarded and the sign follows the dividend:

print(7 / 2)     # 3
print(-7 / 2)    # -3   (truncated toward zero, not floored)
print(7 // 2)    # 3

If you need a fractional result or Python's floor semantics, convert to a float first (f64(a) / f64(b)).

Mixed types

When the two operands differ, the result is the wider type; any float operand makes the result floating-point:

x = 1 + 2.5      # f64
y: i64 = 3
z = y + 1        # i64 (the literal takes i64)

Bitwise and shifts

print(0b1100 & 0b1010)   # 8
print(0b1100 | 0b1010)   # 14
print(0b1100 ^ 0b1010)   # 6
print(~0)                # -1
print(1 << 4)            # 16
print(256 >> 2)          # 64

Bitwise operators require integer operands. ~, <<, >> work on any integer width; the result keeps the operand's width.

Comparison

== != < <= > >= compare numbers by value and yield bool. Mixing signed and unsigned operands follows the wider type.

Casting

n: i32 = 300
big = i64(n)             # widening
small = n as u8          # truncating, via `as`
f = f64(n)
back = i32(f)            # truncates toward zero
print(i32(len(xs)))      # very common: len() is i64, indexing wants i32

as and T(value) are equivalent. Narrowing keeps the low bits.

Prelude helpers

print(abs(-3))           # 3
print(min(1, 2), max(1, 2))
print(clamp(15, 0, 10))  # 10
print(sign(-7))          # -1

For hexadecimal and other bases in output, use formatting.md (f"{n:#x}") or the prelude hex() helper.