non-ascii.r

clusters <- read.csv("clusters.csv", stringsAsFactors = FALSE)

x <- clusters$clean_text

non_ascii <- function(x) {
  any(charToRaw(x) > 0x7F)
}
bad <- x[vapply(x, non_ascii, logical(1), USE.NAMES = FALSE)]

get <- function(x) {
  rw <- charToRaw(x)
  rawToChar(rw[rw > 0x7F])
}
outside <- vapply(bad, get, character(1), USE.NAMES = FALSE)
iconv(outside, "UTF-8", "ASCII//translit", "byte")