Andrew Dunstan <andrew@dunslane.net> writes:
> I think there's a good case for some functions implementing the various
> Unicode normalization functions, though.
I have no objection to that so long as the code footprint is in line
with the utility gain (i.e. not all that much). If we have to bring in
ICU or something similar to make it happen, the cost/benefit ratio looks
pretty bad.
regards, tom lane