Integer types

Integers represent whole numbers. Pop provides four signed widths and four unsigned widths.

Type	Minimum	Maximum
Int8	-128	127
Int16	-32,768	32,767
Int32	-2,147,483,648	2,147,483,647
Int64 / Int	-9,223,372,036,854,775,808	9,223,372,036,854,775,807
UInt8 / Byte	0	255
UInt16	0	65,535
UInt32	0	4,294,967,295
UInt64	0	18,446,744,073,709,551,615

Use Int for ordinary whole-number work. Choose a fixed width when a file format, protocol, native interface, or memory layout requires it. Unsigned types are appropriate when the represented format is explicitly unsigned; they are not a substitute for validating that a value is nonnegative.

Writing integer literals

Integer literals use decimal digits. Underscores may separate groups for readability:

local population = 8_100_000_000
local mask: UInt8 = 255

The underscore does not change the value. This release does not implement hexadecimal, octal, or binary literal prefixes.

A negative value is written by applying unary - to a literal or another signed expression:

local temperature = -12

Unsigned values do not support unary negation.

Checked arithmetic

Pop checks +, -, and * against the range of their exact type. If a result cannot be represented, execution traps:

local maximum: UInt8 = 255
local overflow = maximum + 1 -- traps

Division and remainder trap when the divisor is zero. Signed division also traps when dividing the minimum value by -1, because the positive result is one larger than the type can represent.

No arithmetic promotion is performed. Both operands must already have the same type:

local left: Int32 = 10
local right: Int64 = 20
local total = left + right -- type error

General numeric conversion functions are not exposed by the Standard library in 0.1.0-rc.2, so choose compatible types at the boundaries of your program.