Vijay Marupudi schreef op vr 21-01-2022 om 20:21 [-0500]:
> +SCM_DEFINE (scm_utf8_range_to_string, "utf8->string",
> +            1, 2, 0,
> +            (SCM utf, SCM start, SCM end),
> +            "Return a newly allocate string that contains from the
> UTF-8-"
> +            "encoded contents of bytevector @var{utf}.")

This is incorrect, since the nul character is encoded even though UTF-
proper does not allow encoding the nul character -- UTF-8 with an
encoding of the nul character is sometimes called ‘modified UTF-8’.

The distinction is sometimes relevant, e.g. the GNS specifications asks
for labels to be encoded in UTF-8, and according to the spec writers,
that implied that nul characters are forbidden.

As such, I cannot rely on 'utf8->string' to verify that there aren't
any nul characters.

Greetings,
Maxime.