read_documentation is ambiguous

[Tomeriko96]

The name read_documentation() is, in hindsight, quite ambiguous.

• [x] Create a new function called read_import_definitions that follows the same pattern as read_documentation.

[Tomeriko96]

We could resort to the following function instead:

#' read_import_definitions
#'
#' A function that reads an import definitions file.
#'
#' @param filename The filename of the import definitions file.
#' @param metadata_location Location of the import definitions files,
#' if NULL system variables will be used.
#' @param ... Additional parameters to be passed to the file reading function.
#'
#' @return A data frame containing the import definitions.
#' @export
read_import_definitions <- function(filename = "", metadata_location = NULL, ...) {
  if (is.null(metadata_location)) {
    if (!Sys.getenv("METADATA_IMPORT_DEFINITIONS_DIR") == "") {
      metadata_location <- Sys.getenv("METADATA_IMPORT_DEFINITIONS_DIR")
    } else {
      stop("System variable for metadata_location is missing.")
    }
  }

  file_path <- file.path(metadata_location, filename)

  # Determine the file reading function based on the file extension
  file_extension <- tools::file_ext(filename)
  if (file_extension == "csv") {
    return(utils::read.csv(file_path, stringsAsFactors = FALSE, ...))
  } else if (file_extension == "txt" || file_extension == "dat") {
    # Assume tab-separated file
    return(utils::read.delim(file_path, stringsAsFactors = FALSE, ...))
  } else {
    stop("Unsupported file format. Only CSV, TXT, and DAT files are supported.")
  }
}

• Naming: Clearly uses import definitions files
• System variable: A clear name to differentiate from other metadata/documentation

[Tomeriko96]

@CorneeldH Do you agree with this function?

[CorneeldH]

Yes, agreed 👍 Do you also want to supersede read_documenation, with read_data_dictionary or read_dictionary or something else less generic than read_documenation?

https://lifecycle.r-lib.org/articles/stages.html

[Tomeriko96]

That sounds reasonable enough:

We will create (pending the naming)

• read_import_definitions
• read_data_dictionary

As you mention, the goal then is to supersede read_documentation, in favor of more specific functions. These have their own system variables to differentiate between them.

[Tomeriko96]

read_data_dictionary has been added to the dev version:

#' read_data_dictionary
#'
#' A function that reads a data dictionary file.
#'
#' @param filename The filename of the data dictionary file.
#' @param metadata_location Location of the data dictionary files,
#' if NULL system variables will be used.
#' @param ... Additional parameters to be passed to the file reading function.
#'
#' @return A data frame containing the data dictionary.
#' @export
read_data_dictionary <- function(filename = "", metadata_location = NULL, ...) {
  if (is.null(metadata_location)) {
    if (!Sys.getenv("METADATA_DICTIONARY_DIR") == "") {
      metadata_location <- Sys.getenv("METADATA_DICTIONARY_DIR")
    } else {
      stop("System variable for metadata_location is missing.")
    }
  }

  file_path <- file.path(metadata_location, filename)

  # Determine the file reading function based on the file extension
  file_extension <- tools::file_ext(filename)
  if (file_extension == "csv") {
    return(utils::read.csv(file_path, stringsAsFactors = FALSE, ...))
  } else if (file_extension == "txt" || file_extension == "dat") {
    # Assume tab-separated file
    return(utils::read.delim(file_path, stringsAsFactors = FALSE, ...))
  } else {
    stop("Unsupported file format. Only CSV, TXT, and DAT files are supported.")
  }
}

It relies on system variable: METADATA_DICTIONARY_DIR. We can always choose a different name for it afterwards