CRAN status lifecycle R-CMD-check coverage


giftwrap takes shell commands and turns them into R functions. This enables R developers to immediately work with command lines tools like AWS CLI, Salesforce DX, Docker, git, and more.


If you have the AWS CLI installed and configured on your machine, you can install giftwrap and list your S3 buckets:


wrap_commands("aws s3 ls")

Or, if you have Docker installed on your machine, you can list your running containers.

This time, we’ll store our giftwrapped function in its own namespace that giftwrap creates. We’ll call the namespace gifts.

wrap_commands("docker ps", use_namespace = "gifts")

And we can add our S3 function to that same namespace. Notice how wrap commands can handle multiple commands.

wrap_commands(c("aws s3 ls", "docker ps"), use_namespace = "gifts")

The resulting giftwrapped functions can take any number of named or unnamed arguments, and will add those arguments to the command when the function is called. You can wrap any command available in your shell.


To enable a fast and standalone loading of commands, giftwrap employs the use of lexicons, and comes with several lexicons, which are accessed using the lexicon function, like lexicon("aws") or lexicon("docker").

The wrap_lexicon function takes a lexicon, accepts filtering for commands/subcommands, and has helpful options for where the resulting functions will live and what they will look like.

Let’s wrap the git lexicon.

             use_namespace = "git",
             commands = c("status", "reset"),
             drop_base = T)

commands and subcommands arguments accept regex, and drop_base removes the base ‘git’ from the R function. There is also an env argument to specify an environment, instead of using use_namespace.

Other Useful Features

Capture Output

You can capture the status, stdout, and stderr from your call to the shell using variable assignment.

output <- echo("hello world")

Current Lexicons

giftwrap currently comes with the following lexicons:

Making your own lexicon

Using the aws lexicon as an example, each lexicon contain columns for:

If you follow the format of an existing lexicon, you will likely be able to use wrap_lexicon with any command line tool of your choosing.

Note that all of the functionality in wrap_lexicon is identical to that of wrap_commands. wrap_lexicon just works with a dataframe, formatted as previously discussed. The hope is that wrap_lexicon will allow you to keep your command line commands organized, accessible, and reproducible.

Adding giftwrap functions to packages

If you are familiar with creating R packages, you may know you can specify actions to be taken when the package is loaded using the .onLoad function, in a file typically called zzz.R in the R folder of your package directory.

The following is a short code snippet you may place in zzz.R that allows you to load in giftwrapped functions from the aws lexicon into your package, accessible for your package users.

#' Generates functions on load
#' @importFrom giftwrap wrap_lexicon lexicon
.onLoad <- function(libname, pkgname) {
                           commands = "s3$|ec2$",
                           subcommands = "^ls$|^cp$|^describe-instances$",
                           use_namespace = "yourpackagenamehere",
                           drop_base = T)

Alternatively, if you only want your giftwrapped functions to be available to your package and not directly to the user, you can leverage environment caching instead.

#' Generates functions on load
#' @importFrom giftwrap wrap_lexicon lexicon
.onLoad <- function(libname, pkgname) {
    aws <- new.env()
                           commands = "s3$|ec2$",
                           subcommands = "^ls$|^cp$|^describe-instances$",
                           env = aws,
                           drop_base = T)
    assign("aws", aws, pos = parent.env(environment()))

Now your package can access a command like aws s3 ls with the syntax aws$s3_ls(), and you are free to develop on top of the giftwrapped function as you like.

gcloud in RStudio

After you have installed gcloud (and make sure to run when you do), follow these steps to ensure gcloud will work in RStudio on your local machine:

  1. In your terminal, run which gcloud
  2. From the result, something like /usr/local/google-cloud-sdk/bin/gcloud, copy up to bin. In this example, we’d copy /usr/local/google-cloud-sdk/bin
  3. In RStudio, run usethis::edit_r_environ()
  4. In the .Renviron file that opens, add gcloud to your PATH. Using our example, the line should read something like PATH=/usr/local/google-cloud-sdk/bin:$PATH
  5. Save your .Renviron file and restart R

Happy giftwrapping!