Primitive Types

One thing that might be surprising about Storm is that primitive types are not treated specially by the system. Rather, they are implemented as any other type in the standard library. As far as Storm is concerned, the primitive types in the standard library are just opaque types with a set of operations that manipulate the opaque representation. That is, these types have no members, and therefore need to be manipulated through the members in the type. The only thing that is special about the primitive types is that they are treated specially during function calls (to follow the system's calling convention, to allow seamless calls to and from C++).

The standard library provides the following primitive types:

core.Bool
A type that may contain one of the values true and false.
core.Byte
An 8-bit unsigned integer.
core.Int
A 32-bit signed integer.
core.Nat
A 32-bit unsigned integer. Nat is short for "natural number".
core.Long
A 64-bit signed integer.
core.Word
A 64-bit unsigned integer.
core.Float
A 32-bit floating point number.
core.Double
A 64-bit floating point number.

All types have members that implement the expected comparison and arithmetic operators. The unsigned types also provide bitwise operations as operators. Types are converted between each other using members in the types. For example, the numeric types contain a member int that converts them to an integer.

The system also provides the following operations in addition to regular operators:

min(x, y), max(x, y) - minimum and maximum values
sqrt(x) - square root of floating point types
pow(base, exp) - exponential of floating point types
abs(x) - absolute value
clamp(value, min, max) - limit a value to a range

Finally, the types provides the following static operations:

min (e.g. Int:min) - provide the lowest representable value.
max (e.g. Long:max) - provide the highest representable value.

Note: Floating point types use min and max to represent the largest and smallest positive value that can be represented (i.e. min returns a positive but very small number). Floating point types also provide infinity to get the positive infinity value, as well as ceil and floor to round up and down respectively.

Previous: Standard Library

Next: Objects and Actors

Up: Standard Library