cookiecutter for R
The cookiecutter
package for R
makes it easy to create new projects and files based on templates. It can create directory structures and files, and customizes their contents based on predefined settings.
The package implements the core functionality of the excellent and much more flexible Python implementation, and complements the likewise fantastic usethis
package that automates development steps within a project.
Usage 🍪
Installation 📦
The package is available on CRAN, so that you can install it via
install.packages('cookiecutter')
Alternatively, if you like to live on the bleeding edge, you can install the development version:
devtools::install_github('felixhenninger/cookiecutter@main')
Making cookies 👩🍳
To get started, you’ll need a cookiecutter
template, and a place to extract it. For example, starting from a template directory you could run:
cookiecutter::bake(
'template_directory',
'output_directory'
)
If you’ve downloaded a template as a compressed archive or a directory, you can clone it locally.
cookiecutter::bake(
'template.zip',
'output_directory'
)
Crafting a template 🛠
In principle, any directory you already have can serve as a cookiecutter
template. Out of the box, if you use it as template, cookiecutter
will simply copy the directory for you. Likewise, if you create a zip
, .tar
or .tar.gz
archive of your directory, cookiecutter
will unpack it into the output location.
However, there are many more things you can do to make your template more user-friendly. Specifically, cookiecutter
will change the names and contents of files according to your specifications. The way this works is that you can insert a placeholder, {{ cookiecutter.my_variable }}
inside a file (or in its name or directory name), and cookiecutter
will try to replace this placeholder (curly braces and all) with the contents of my_variable
.
If you use the placeholder alone, your users will need to supply the values of any variables when they bake the template, like so (here, we set the variable topping
, which you could use in a file or file name via {{ cookiecutter.topping }}
):
cookiecutter::bake(
'template.zip',
'output_directory',
extra_context = list(
topping = 'sprinkles'
)
)
You can help your users figure out which variables are available by adding a cookiecutter.json
file to your template directory. If you do so, cookiecutter
will know about the variables it needs, and prompt users to fill in or to choose a value. It will provide a default, or a list of choices, respectively.
The format of this file is documented fully in the Python cookiecutter documentation, but to give you an impression of the options, here’s an example:
{
"cookie_eating_sound": "Om nom nom nom",
"cookie_local_term": ["cookie", "biscuit"]
}
Limitations ⚠️
The cookiecutter
package for R
aims to be compatible with its big sibling, but at present implements only a subset of their functionality.
Of course, few of these issues are insurmountable, and they are great places to contribute!
- We don’t (yet) provide the full extent of the Jinja templating language that
cookiecutter
uses. Specifically, at present, we don’t support logic in templates. - Naturally, we don’t support Python code in templates.
- We don’t currently support cloning templates directly from a git repository. You can still download the repository contents and clone from a directory or an archive.
- We don’t support
~/.cookiecutterrc
files. - We don’t support running scripts automatically after populating a template.
If you’re looking for the full power of the original, we wholeheartedly recommend using it — our clone is designed to be a stop-gap, not a replacement. If we can whet your appetite, that’s a win in our book!