I don't see what this function does that can't be done with `dplyr::na_if()`. Do we need it? I guess we could have a function called `replace_special()` that calls this on special characters?