R/regexplain.R

In regexTestR: A 'Shiny' App to Test Regular Expressions in an R Environment

split_vec = function(vec, splits) {
  #' Spit at a specified list of indices
  #'
  #' @param vec vector to split
  #' @param splits vector of split locations
  #'
  #' @return list of vectors
  #' @keywords internal
  #' @noRd

  split(vec, cumsum(seq_along(vec) %in% (splits)))
}


regexplain = function(regx) {
  #' Create an 'explanation' of an inputted regular expression
  #'
  #' @details Provided by rick.measham.id.au
  #'
  #' @param regx string containing regular expression to explain
  #'
  #' @return a dataframe with columns c("regex", "explanation")
  #' @keywords internal
  #' @noRd

  base_url = "http://rick.measham.id.au/paste/explain?regex="

  regx_url = stringr::str_replace_all(
    utils::URLencode(regx), stringr::fixed("+"), "%2B"
  )
  regx_url = stringr::str_replace_all(
    regx_url, stringr::fixed("[[:blank:]]"), "[:blank:]"
  )
  regx_url = stringr::str_replace_all(
    regx_url, stringr::fixed("[%5E[:blank:]]"), "[%5E:blank:]"
  )

  full_url = paste0(base_url, regx_url)

  page_html = xml2::read_html(full_url)
  regexplain_tbl_txt = rvest::html_text(rvest::html_node(page_html, "pre"))
  regexplain_tbl_txt = stringr::str_split(trimws(regexplain_tbl_txt), "\n")[[1]]

  row_breaker = "--------------------------------------------------------------------------------"

  regexplain_tbl_txt = trimws(
    regexplain_tbl_txt[regexplain_tbl_txt != row_breaker]
  )

  split_rows = stringr::str_split(regexplain_tbl_txt, stringr::regex("\\s{2,}"))

  headers = split_rows[[1]]
  rows = split_rows[-1]

  good_inds = which(vapply(rows, length, numeric(1)) != 1)

  clean_row_list = lapply(split_vec(rows, good_inds), unlist)
  clean_row_list = lapply(clean_row_list, function(x) {
    trimws(c(x[1], paste(x[-1], collapse = " ")))
  })
  names(clean_row_list) = paste0("node_", seq_along(clean_row_list))

  explain_df = data.table::as.data.table(do.call("rbind", clean_row_list))
  data.table::setnames(explain_df, c("regex", "explanation"))

  if (
    stringr::str_detect(regx, stringr::fixed("[[:blank:]]")) &
    any(stringr::str_detect(explain_df$regex, stringr::fixed("[:blank:]")))
  ) {
    blank_inds = which(explain_df$regex == "[:blank:]")
    explain_df$regex[blank_inds] = "[[:blank:]]"
    explain_df$explanation[blank_inds] = "any character of: blank characters (space and tab, and possibly other locale-dependent characters such as non-breaking space)"
  }

  if (
    stringr::str_detect(regx, stringr::fixed("[^[:blank:]]")) &
    any(stringr::str_detect(explain_df$regex, stringr::fixed("[^:blank:]")))
  ) {
    blank_inds = which(explain_df$regex == "[^:blank:]")
    explain_df$regex[blank_inds] = "[^[:blank:]]"
    explain_df$explanation[blank_inds] = "any character except: blank characters (space and tab, and possibly other locale-dependent characters such as non-breaking space)"
  }

  explain_df
}