On 12-Oct-08, at 1:19 PM, Aleksey Cheusov wrote:
libc's regexp doesn't support UTF-8 and therefore usr.bin/grep will
not too.
I'd strongly suggest that's actually more of an incentive to move to a
BSD grep. (primarily because it would mean grep/egrep would then
conform to the standard libc regexp implementation since obviously it
doesn't do so now)
It's also an incentive to improve (eg. add UTF-8 support to) the base
regexp code too, if that's deemed desirable and possible to do without
causing major confusion and bloat.