Re: [bug-gawk] v4.1.3 (run on OSX 10.11.3): potential gsub() bug

[Aharon Robbins]

Hi.

Thanks for the notes.  The current code base should not dump core,
although I see that with stock 4.1.3.

The questions raised are messy.  I don't have good answers.  I think
that if you use [...] with real UTF-8 encoded characters as the
start and end point of the ranges, things will work OK.  But I'm not
sure.

There is no intent to support things like \x10f7ff. If such a thing
works it's by accident and it won't last; the master branch was changed
to accept no more than two hex digits after \x.

I am not in a rush to add things like \uXXXX to gawk.

For now, you are probably best off avoiding things like [\x80-\xFF] in
Unicode locales.  Or using LC_ALL=C.

Thanks,

Arnold

P.S. I'm curious what current GNU grep does with such things?  Thanks