Decoding and Encoding Consistency in Schema Design

I read from the schema docs, that decoding + encoding should give you back the same value. Is that why there's no "truncate" schema? Like "given a

chars length string, truncate it to

" thing? How should you describe such transformation instead?