3.5 Bytes and Byte Strings

A byte is an exact integer between 0 and 255, inclusive. The byte? predicate recognizes numbers that represent bytes.


  > (byte? 0)


  > (byte? 256)


A byte string is similar to a string – see Strings (Unicode) – but its content is a sequence of bytes instead of characters. Byte strings can be used in applications that process pure ASCII instead of Unicode text. The printed form of a byte string supports such uses in particular, because a byte string prints like the ASCII decoding of the byte string, but prefixed with a #. Unprintable ASCII characters or non-ASCII bytes in the byte string are written with octal notation.

+Reading Strings in Reference: Racket documents the fine points of the syntax of byte strings.


  > #"Apple"


  > (bytes-ref #"Apple" 0)


  > (make-bytes 3 65)


  > (define b (make-bytes 2 0))
  > b


  > (bytes-set! b 0 1)
  > (bytes-set! b 1 255)
  > b


The display form of a byte string writes its raw bytes to the current output port (see Input and Output). Technically, display of a normal (i.e,. character) string prints the UTF-8 encoding of the string to the current output port, since output is ultimately defined in terms of bytes; display of a byte string, however, writes the raw bytes with no encoding. Along the same lines, when this documentation shows output, it technically shows the UTF-8-decoded form of the output.


  > (display #"Apple")


  > (display "\316\273")  ; same as "λ"


  > (display #"\316\273") ; UTF-8 encoding of λ


For explicitly converting between strings and byte strings, Racket supports three kinds of encodings directly: UTF-8, Latin-1, and the current locale’s encoding. General facilities for byte-to-byte conversions (especially to and from UTF-8) fill the gap to support arbitrary string encodings.


  > (bytes->string/utf-8 #"\316\273")


  > (bytes->string/latin-1 #"\316\273")


  > (parameterize ([current-locale "C"])  ; C locale supports ASCII,
      (bytes->string/locale #"\316\273")) ; only, so...

  bytes->string/locale: byte string is not a valid encoding

  for the current locale: #"\316\273"

  > (let ([cvt (bytes-open-converter "cp1253" ; Greek code page
          [dest (make-bytes 2)])
      (bytes-convert cvt #"\353" 0 1 dest)
      (bytes-close-converter cvt)
      (bytes->string/utf-8 dest))


+Byte Strings in Reference: Racket provides more on byte strings and byte-string procedures.