DEFINITION MODULE Utf8;
(* Strict RFC3629 UTF-8 codec, shared by the front end and the text
   output libraries.  Codepoints are UCHAR scalars. *)

CONST
  MaxUTF8Length = 4;
  MaxCodePoint  = 1114111;

PROCEDURE Decode (s : ARRAY OF CHAR; lo : CARDINAL;
                  VAR cp : UCHAR; VAR next : CARDINAL) : BOOLEAN;
(* Decode one sequence starting at s[lo]: cp := the codepoint, next :=
   the index just past it.  FALSE on malformed or truncated input
   (overlong, surrogate, > MaxCodePoint, bad continuation byte). *)

PROCEDURE Encode (cp : UCHAR; VAR buf : ARRAY OF CHAR) : CARDINAL;
(* Encode cp as UTF-8 into buf[0..]; returns the byte count (0 when cp
   is not a Unicode scalar value).  buf needs MaxUTF8Length slots. *)

END Utf8.