rontolisp:octets-to-string

(rontolisp:octets-to-string octets)

Decodes a packed (unsigned-byte 8) vector as the UTF-8 text its bytes spell, returning a fresh string. rontolisp:string-to-octets is the encoder that reverses it.

The decode is total and lenient: octets is never rejected. A byte that leads no valid sequence, and a sequence the vector's end cuts short, both decode to their own byte value as a one-character result rather than signaling -- the same rule an HTTP body carried as raw bytes is decoded by, so a body decodes the same whichever transport carried it. An overlong encoding and a UTF-8-encoded surrogate are not rejected either: each decodes to the code point its bits assemble, since a rontolisp character admits any code point in 0 to #x10FFFF, surrogates included.

The second example is a 3-byte sequence (€) cut off after its first continuation byte: each of the two remaining bytes decodes to its own value rather than being dropped or signaled.