Make function calls explicit

This function takes a block of code and seeks to make all function calls 'explicit' through the use of the double-colon operator ::. This function is bound to the RStudio addin "Make function calls explicit". See examples for usage.

Usage

add_double_colons(
  code = NULL,
  use_packages = current_packages(),
  ignore_functions = imported_functions()
)

Arguments

code

Code to transform. Either a character vector or NULL, in which case any highlighted code (in RStudio) will be used.

use_packages

A character vector of package names. The order is important here - see examples for details.

ignore_functions

Functions to ignore when applying the transformation

Value

The transformed code as a character string

Details

This function behaves differently depending on the context.

• Not package development: If the current context is not package development, then it will make function calls explicit using the currently attached packages (i.e. the ones attached by calls to library()).

• Package development: If it detects that the current context is package development it will make function calls explicit using packages in the 'Imports' field of the package DESCRIPTION. If the package being developed imports any packages in their entirety (i.e if Import pkg appears in the NAMESPACE file), calls to functions from these packages will be left unchanged. See current_packages() for more information.

Examples

code <- "
  cars <- as_tibble(mtcars)
  cars %>%
    filter(mpg > 20) %>%
    summarise(across(everything(), n_distinct))
"

# Code will be transformed to use the double-colon operator, but notice
# that `n_distinct` is not transformed as it is not followed by `()`
cat(add_double_colons(code, "dplyr"))
#> 
#>   cars <- dplyr::as_tibble(mtcars)
#>   cars %>%
#>     dplyr::filter(mpg > 20) %>%
#>     dplyr::summarise(dplyr::across(dplyr::everything(), n_distinct))

# You can specify functions that shouldn't be transformed:
cat(add_double_colons(code, "dplyr", ignore_functions = "across"))
#> 
#>   cars <- dplyr::as_tibble(mtcars)
#>   cars %>%
#>     dplyr::filter(mpg > 20) %>%
#>     dplyr::summarise(across(dplyr::everything(), n_distinct))

# Beware namespace conflicts! The following are not the same, mimicking
# the effects of reordering calls to `library()`:
cat(add_double_colons(code, c("dplyr", "stats")))
#> 
#>   cars <- dplyr::as_tibble(mtcars)
#>   cars %>%
#>     dplyr::filter(mpg > 20) %>%
#>     dplyr::summarise(dplyr::across(dplyr::everything(), n_distinct))

cat(add_double_colons(code, c("stats", "dplyr")))
#> 
#>   cars <- dplyr::as_tibble(mtcars)
#>   cars %>%
#>     stats::filter(mpg > 20) %>%
#>     dplyr::summarise(dplyr::across(dplyr::everything(), n_distinct))