knitr::opts_chunk$set( eval = FALSE, collapse = TRUE, comment = "#>" )
R's options are a useful tool for managing global settings. options
aims to
make them easy to configure and use, reduce boilerplate code and handle more
involved options handling so you don't have to.
While writing your package, we start by defining which options our package might use.
Define your options using the define_options()
shorthand. Interlace
descriptions and default values to define multiple options at once.
options::define_options( "This is an example of how a package author would document their internally used options. This option could make the package default to executing quietly.", quiet = TRUE, "Multiple options can be defined, providing default values if a global option or environment variable isn't set.", second_example = FALSE, "Default values are lazily evaluated, so you are free to use package functions without worrying about build-time evaluation order", lazy_example = fn_not_defined_until_later() )
When you want more control, you can use define_option()
to declare all aspects
of how your option behaves.
options::define_option( option = "concrete_example", default = TRUE, desc = paste0( "Or, if you prefer a more concrete constructor you can define each option ", "explicitly." ), option_name = "mypackage_concrete", # define custom option names envvar_name = "MYPACKAGE_CONCRETE", # and custom environment variable names envvar_fn = envvar_is_true() # and use helpers to handle envvar parsing )
As long as the options have been created as shown above, documenting your options is as easy as adding this small roxygen stub within your package.
#' @eval options::as_roxygen_docs() NULL
Which produces a ?mypackage::options
help page. Moreover, any options
page
will show up in an index when using ?options
to search for documentation,
making it easier to discover which packages have options for you to use.
mypackage Options Description: Internally used, package-specific options. All options will prioritize R options() values, and fall back to environment variables if undefined. If neither the option nor the environment variable is set, a default value is used. Options: quiet This is an example of how a package author would document their internally used options. This option could make the package default to executing quietly. default: TRUE option: mypackage.quiet envvar: R_MYPACKAGE_QUIET (raw) ...
When your options are used as default values to parameters, you can use the option documentation to populate your function parameter docs.
This is made simple when all of your parameters share the same names as your options.
#' @eval options::as_params() #' @name options_params #' NULL #' Count to Three #' #' @inheritParams option_params #' count_to_three <- function(quiet = opt("quiet")) { for (i in 1:3) if (!quiet) cat(i, "\n") }
In situations where you have identically named parameters where you don't want
to inherit the option documentation, you can provide their names to as_params
to use just a subset of options. You can also reassign documentation for an
option to a parameter of a different name.
#' Hello World! #' #' @eval options::as_params("silent" = "quiet") #' hello <- function(who, silent = opt("quiet")) { cat(paste0("Hello, ", who, "!"), "\n") }
When using define_option
you can set the option_name
and envvar_name
that
will be used directly.
But it can be tedious and typo-prone to write these out for each and every option. Instead, you might consider providing a function that sets the default format for your option and environment variable names.
For this, you can use set_option_name_fn
and set_envvar_name_fn
, which each
accept a function as an argument. This function accepts two arguments, a
package name and internal option name, which it uses to produce and return the
corresponding global option name or environment variable name.
options::set_option_name_fn(function(package, name) { tolower(paste0(package, ".", name)) }) options::set_envvar_name_fn(function(package, name) { gsub("[^A-Z0-9]", "_", toupper(paste0(package, "_", name))) })
Any scripts or data that you put into this service are public.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.