This screenshot shows what it looks like when run twice. Only upper case A through Z are encoded / decoded. All other characters pass through unchanged. To decode, use the negative of the offset that was entered during encoding. e.g. if you encode with 7, use -7 to decode.