R/regex-methods.R

In rebus.base: Core Functionality for the 'rebus' Package

#' Convert or test for regex objects
#'
#' \code{as.regex} gives objects the class \code{"regex"}. \code{is.regex}
#' tests for objects of class \code{"regex"}.
#' @param x An object to test or convert.
#' @return \code{as.regex} returns the inputs object, with class
#' \code{c("regex", "character")}.
#' \code{is.regex} returns \code{TRUE} when the input inherits from class
#' \code{"regex"} and \code{FALSE} otherwise.
#' @examples
#' x <- as.regex("month.abb")
#' is.regex(x)
#' @export
as.regex <- function(x)
{
  class(x) <- c("regex", "character")
  x
}

#' @rdname as.regex
#' @export
is.regex <- function(x)
{
  inherits(x, "regex")
}

#' Create a regex
#'
#' Creates a regex object.
#' @param ... Passed to \code{paste0}.
#' @return An object of class \code{regex}.
#' @note This works like \code{paste0}, but the returns value has class
#' \code{c("regex", "character")}.
#' @seealso \code{\link[base]{paste0}}
#' as.regex(month.abb)
#' regex(letters[1:5], "?")
#' @export
regex <- function(...)
{
  as.regex(paste0(...))
}

#' Print or format regex objects
#'
#' Prints/formats objects of class \code{regex}.
#' @param x A regex object.
#' @param encode_string If \code{TRUE}, the regex is encoded with
#' \code{\link[base]{encodeString}}. This means that backslashes are doubled,
#' compared to the default of \code{FALSE}.
#' @param ... Passed from other format methods.  Currently ignored.
#' @return \code{format.regex} returns a character vector. \code{print.regex}
#' is invoked for the side effect of printing the regex object.
#' @examples
#' group(1:5)
#' lookahead(1:5)
#' @export
format.regex <- function(x, ...)
{
  paste0("<regex> ", x)
}

#' @rdname format.regex
#' @export
print.regex <- function(x, encode_string = FALSE, ...)
{
  x2 <- if(encode_string)
  {
    as.regex(encodeString(x))
  } else x
  cat(format(x2), sep = "\n")
  invisible(x)
}