Hi Evan,

Evan Silberman wrote on Thu, Feb 21, 2019 at 10:27:25AM -0800:

> less(1) replaces "nonprintable" unicode characters with a representation
> of their code point.  This determination is made using a table of code
> points

YIKES!

I really ought to clean that up and make it use standard functions
like iswprint(3).  Yes, that means we will restrict our implementation
of less(1) to platforms where wchar_t stores Unicode codepoints;
but OpenBSD will always do that and i don't think we care about
portability to platforms that store something in wchar_t which are
*not* code points.  Besides, following Illumos, we have forked
less anyway and maybe Illumos might even want to have that cleanup,
too.

Thank you very much for calling out the pigsty usr.bin/less/charset.c;
large parts of that file consist of nothing but very bad ideas.

Yours,
  Ingo

Reply via email to