R/add.config.R

In ProjectTemplate: Automates the Creation of New Statistical Analysis Projects

#' Add project specific config to the global config
#'
#' Enables project specific configuration to be added to the global config object.  The
#' allowable format is key value pairs which are appended to the end of the \code{config}
#' object, which is accessible from the global environment.
#'
#' Once defined, the value can be accessed from any \code{ProjectTemplate} script by
#' referencing \code{config$my_project_var}.
#'
#' @param ... A series of key-value pairs containing the configuration.  The key is the
#'            name that gets added to the config object. These can be overridden at load
#'            time through the \code{...} argument to \code{\link{load.project}}.
#' @param apply.override A boolean indicating whether overrides should be applied. This
#'   can be used to add a setting disregarding arguments to \code{load.project}
#'
#
#' @return NULL
#'
#' @export
#' @examples
#' library('ProjectTemplate')
#' \dontrun{
#' add.config(
#'     keep_bigdata=TRUE,     # Whether to keep the big data file in memory
#'     parse=7                # number of fields to parse
#' )
#'
#' if (config$keep_bigdata) ...
#' }
add.config <- function(..., apply.override = FALSE){
  project_config <- list(...)
  if (length(project_config) > 0 &&
        (is.null(names(project_config)) || any(names(project_config) == ""))) {
    stop('All options should be named')
  }

  config <- get('config', envir = .TargetEnv)
  if (apply.override) {
    project_config <- .apply.override.config(project_config, config$.override.config)
  }
  assign('config', c(config, project_config), envir = .TargetEnv)
}