Copy an environment or a function

Source: R/copy_env.R

By default, R does not create a copy of an environment when it is assigned to a new variable. If a is an environment, b <- a just makes b point to the same environment as a. This means that any changes that happen to b also happen to a (or rather, the environment that they point to). Sometimes, this is not preferred and it would be better to explicitly make a copy of the environment.

The copy_env() function creates a new environment with the same parent as env, and all the objects in env are copied (in their usual way). The only exception to this is if env contains an environment object, in which case, it will also be copied using copy_env(). This also means that the only changes occurring is the re-direction of pointers and so copy_env() is surprisingly memory-light as it does not actually copy any large objects.

The copy_fun() function similarly creates a copy of a function, including that applying copy_env() to that function's environment. This is particularly useful for fseq which rely heavily on their function environment

copy_env(env)

copy_fun(fun)

Arguments

env

environment to be copied
fun

function to be copied

Examples

a <- new.env()
a$x <- 1
a$y <- "hello"
a$z <- data.frame(x = runif(20))

b <- copy_env(a)
identical(a$z, b$z)

#> [1] TRUE
rlang::env_label(a) == rlang::env_label(b)

#> [1] FALSE

f1 <- . %>%
  magrittr::add(2) %>%
  magrittr::multiply_by(2)
f2 <- f1
f3 <- copy_fun(f1)

f1_env_label <- rlang::env_label(rlang::get_env(f1))
f2_env_label <- rlang::env_label(rlang::get_env(f2))
f3_env_label <- rlang::env_label(rlang::get_env(f3))
f1_env_label == f2_env_label

#> [1] TRUE
f1_env_label == f3_env_label

#> [1] FALSE