tests/testthat/test-args_to_list.R

In jeremyrcoyle/sl3: Pipelines for Machine Learning and Super Learning

context("test-args_to_list.R -- args_to_list")

my_function <- function(arg1 = "default_1", arg2 = "default_2", ...) {
  return(args_to_list())
}

test_that("basic args_to_list functionality works", {
  arglist <- my_function(arg2 = "specified_2", dot_arg = "test")
  expect_equal(arglist, list(arg1 = "default_1", arg2 = "specified_2", dot_arg = "test"))
})


test_that("list arguments work", {
  arglist <- my_function(arg3 = list(a = 2, b = list(c = 1, d = 3)))


  expect_equal(
    arglist,
    list(arg1 = "default_1", arg2 = "default_2", arg3 = list(
      a = 2, b = list(c = 1, d = 3)
    ))
  )
})


test_that("arguments are evaluated in the correct (calling) environment", {
  wrapper_function <- function() {
    a <- 1
    my_function(arg1 = a)
  }

  a <- 2
  arglist <- wrapper_function()
  expect_equal(arglist$arg1, 1)
})

test_that("argument defaults can depend on previous arguments", {
  my_function_defaults <- function(arg1 = 1, arg2 = arg1) {
    return(args_to_list())
  }

  arglist <- my_function_defaults()
  expect_equal(arglist, list(arg1 = 1, arg2 = 1))
})

test_that("NULL arguments work", {
  arglist <- my_function(arg1 = NULL)
  expect_equal(arglist, list(arg1 = NULL, arg2 = "default_2"))
})