define_questions_from.R

#' Define questions from a data frame
#'
#' Each row in the text data frame is interpreted as a question. We only have to define
#' the columns that we are going to use, the rest of the columns are taken by default.
#'
#' For answers where a vector is required, "<|>" is used as a separator of the vector
#' elements.
#'
#' @param qc A `question_category` object.
#' @param df A data frame.
#'
#' @return A `question_category`.
#'
#' @family question definition
#'
#' @examples
#'
#' file <- system.file("extdata", "questions.csv", package = "moodef")
#' df <- read_question_csv(file = file)
#'
#' qc <-
#'   question_category(category = 'Initial test', adapt_images = TRUE) |>
#'   define_questions_from_data_frame(df)
#'
#' @export
define_questions_from_data_frame <- function(qc, df)
  UseMethod("define_questions_from_data_frame")

#' @rdname define_questions_from_data_frame
#' @export
define_questions_from_data_frame.question_category <- function(qc, df) {
  attributes <- names(df)
  df[, attributes] <- data.frame(lapply(df[, attributes], as.character), stringsAsFactors = FALSE)
  if (nrow(df) == 1) {
    df[, attributes] <-
      tibble::as_tibble(as.list(apply(df[, attributes, drop = FALSE], 2, function(x)
        tidyr::replace_na(x, ''))))
  } else {
    df[, attributes] <-
      apply(df[, attributes, drop = FALSE], 2, function(x)
        tidyr::replace_na(x, ''))
  }
  attributes <- snakecase::to_snake_case(attributes)
  names(df) <- attributes
  for (opcional in c('type', 'image', 'image_alt')) {
    if (!(opcional %in% attributes)) {
      df[, opcional] <- ''
    }
  }
  rest <- setdiff(attributes, c("type", "question", "image", "image_alt", "answer"))
  for (i in 1:nrow(df)) {
    text <- paste0(
      'define_question(qc, type = "',
      df[i, 'type'],
      '", question = "',
      df[i, 'question'],
      '", image = "',
      df[i, 'image'],
      '", image_alt = "',
      df[i, 'image_alt'],
      '", answer = ',
      string_to_string_vector(df[i, 'answer'][[1]])
    )
    j <- 0
    for (r in rest) {
      if (df[i, r][[1]] != '') {
        j <- j + 1
        text <- paste0(text,
                       ", a_", j, " = ", string_to_string_vector(df[i, r][[1]]))
      }
    }
    text <- paste0(text, ")")
    qc <- eval(parse(text = text))
  }
  qc
}



#' Define questions from a csv file
#'
#' Each row in the text file is interpreted as a question. We only have to define
#' the columns that we are going to use, the rest of the columns are taken by default.
#'
#' For answers where a vector is required, "<|>" is used as a separator of the vector
#' elements.
#'
#' @param qc A `question_category` object.
#' @param file A string, name of a text file.
#' @param sep Column separator character.
#'
#' @return A `question_category`.
#'
#' @family question definition
#'
#' @examples
#'
#' file <- system.file("extdata", "questions.csv", package = "moodef")
#' qc <-
#'   question_category(category = 'Initial test', adapt_images = TRUE) |>
#'   define_questions_from_csv(file = file)
#'
#' @export
define_questions_from_csv <- function(qc, file, sep)
  UseMethod("define_questions_from_csv")

#' @rdname define_questions_from_csv
#' @export
define_questions_from_csv.question_category <- function(qc, file, sep = ',') {
  df <- readr::read_delim(
    file,
    delim = sep,
    col_types = readr::cols(.default = readr::col_character())
  )
  define_questions_from_data_frame(qc, df)
}




#' Define questions from a Excel file
#'
#' Each row in the Excel file is interpreted as a question. We only have to define
#' the columns that we are going to use, the rest of the columns are taken by default.
#'
#' In addition to the file, we can indicate the sheet by its name or index. If we
#' do not indicate anything, it considers the first sheet.
#'
#' For answers where a vector is required, "<|>" is used as a separator of the vector
#' elements.
#'
#' @param qc A `question_category` object.
#' @param file A string, name of an Excel file.
#' @param sheet_index A number, sheet index in the workbook.
#' @param sheet_name A string, sheet name.
#'
#' @return A `question_category`.
#'
#' @family question definition
#'
#' @examples
#'
#' \donttest{
#' file <- system.file("extdata", "questions.xlsx", package = "moodef")
#' qc <-
#'   question_category(category = 'Initial test', adapt_images = TRUE) |>
#'   define_questions_from_excel(file = file)
#' }
#'
#' @export
define_questions_from_excel <- function(qc,
                                        file,
                                        sheet_index,
                                        sheet_name)
  UseMethod("define_questions_from_excel")

#' @rdname define_questions_from_excel
#' @export
define_questions_from_excel.question_category <- function(qc,
                                                          file,
                                                          sheet_index = NULL,
                                                          sheet_name = NULL) {
  if (is.null(sheet_index) & is.null(sheet_name)) {
    sheet_name <- readxl::excel_sheets(file)
  } else if (is.null(sheet_name)) {
    sheet_name <- readxl::excel_sheets(file)[sheet_index]
  }
  sheet_name <- sheet_name[1]
  df <- suppressMessages(
    readxl::read_excel(
      file,
      sheet = sheet_name,
      col_names = TRUE,
      col_types = "text",
      trim_ws = TRUE
    )
  )
  define_questions_from_data_frame(qc, df)
}