pub fn encoded_word_encode<O, R: AsRef<str>>(input: R, out: &mut O)where O: EncodedWordWriter,
for now this only supports utf8/ascii input, as we have to know where we can split