Functions & Methods

In Vapour functions are declared much differently than they are in R, methods even more so. This is both to make it more readable and easier to parse, R has some oddities with how methods are declared.

The way functions and methods are declared is taken from Go.

info

In Vapour return is a keyword and is mandatory, this is what the language uses to know what is returned by functions.

Functions

Functions are declared with the func keyword, we indicate the types of each argument as well as the type the function returns.

Vapour

func add(x: int, y: int): int {
  return x + y
}

R

add = function(x, y) {
  return(x + y)
}

In R functions always return something there is no void, so use null instead.

Vapour

func add(x: char): null {
  print(x)
}

R

add = function(x) {
  return(x)
}

Anonymous Functions

Anonymous functions are declared as shown below, inspired from Javascript, unlike Javascript Vapour expects curly braces ({}).

Vapour

lapply(1..10, (x: int): int => {
  return x + 1
})

R

lapply(1:10, function(x) {
  return(x + 1)
})

Methods

You can define methods on your custom types.

Vapour

type car struct {
  int,
  name: char
}

func (c: char) print(name: char): car {
  c$name = 1
}

R

add = function(x) {
  return(x)
}

Class Decorator

You can use the decorator @class to customise the class assigned to your object if you are not happy with the above.

Vapour

@class(a, thing, on, wheels)
type vehicle: struct {
  bool,
  name: char
}

let train: vehicle = vehicle(
  false,
  name = "tchoo tchoo"
)

R

train = structure(
  FALSE,
  name = "tchoo tchoo",
  class = c("a", "thing", "on", "wheels")
)

Generic Decorator

You can use the decorator @generic to define a generic

Vapour

type user: object {
  id: int,
  name: char
}

@generic
func (u: any) set_id(id: int): any

func (u: user) set_id(id: int): user {
  u$id = id
  return u
}

R

set_id <- function(u, id) {
  UseMethod("set_id")
}

set_id.user <- function(u, id) {
  u$id = id
  return(user)
}