Re: New draft (#4) of SRFI 207: String-notated bytevectors
Daphne Preston-Kendal 07 Oct 2020 10:01 UTC
On 6 Oct 2020, at 17:36, John Cowan <xxxxxx@ccil.org> wrote:
> Okay. Are there any more issues before I request a last call?
One small thorny point I haven’t cleared up yet:
The formal syntax uses the R7RS-small ⟨inline hex escape⟩ rule to
define \x escapes, which allows any number of hex digits after the x.
The description of the notation, though, insists that there be no more
than two. Of course, anything more than \xFF; doesn't make sense anyway
— but the description as it stands doesn’t allow you to spell it
\x00FF;, for example.
I could either fix this inconsistency permissively (edit the
description to allow more hex digits as long as all but the rightmost
two are 0) or restrictively (edit the formal syntax to define a special
kind of inline hex escape only allowing). The former would be more
consistent with Scheme strings, which are allowed to spell characters
in ways like "\x000032E1;" notwithstanding that Unicode only goes up to
U+10FFFF. But I fear in the context of bytevectors it might just
confuse.
Daphne