Numbers

WCL has fixed-width signed and unsigned integers plus two float widths. A bare integer literal defaults to i64; a suffix pins the exact type. Underscores may group digits for readability and are ignored. See Expressions for numeric promotion across mixed operands.

Literals

a = 42          // i64 (default)
b = 200u8       // unsigned 8-bit
c = 9_000i64    // underscores are ignored
d = 3.14f64     // float
e = -120i8      // signed
f = 0xFFu32     // hex
g = 0b1010_1100u8   // binary
h = 0o755u16    // octal

Types

The platform-default size (isize / usize) is the pointer width of the machine WCL runs on — 64-bit on most desktops/servers, 32-bit on smaller targets — so its exact width changes depending on the platform.

Numeric promotion

Arithmetic and comparison widen mixed numeric operands to a common type, so cross-width and integer/float mixing work without explicit casts.

a = 1 + 2.0        // i64 widened to f64 -> 3.0
b = 1u32 == 1i64   // true
c = 3.0 * 2u8      // 6.0

Number literals: radixes and width suffixes

An integer literal may be written in decimal, hexadecimal (0x), octal (0o), or binary (0b); floats accept scientific notation (1e6). A width suffix pins the exact numeric type, and underscores may group digits for readability (they are ignored).

a = 42          // i64 (default)
b = 200u8       // unsigned 8-bit
c = 9_000i64    // underscores are ignored
d = 3.14f64     // float
f = 0xFFu32     // hex
g = 0b1010_1100u8   // binary
h = 0o755u16    // octal