Re: Backpatching of "Teach the regular expression functions to do case-insensitive matching"

Andres Freund <andres(at)anarazel(dot)de> writes:
> In my opinion this is actually a bug in < 9.0. As its a (imo) low impact fix
> thats constrained to two files it seems sensible to backpatch it now that the
> solution has proven itself in the field?

FWIW, I still don't trust that patch a lot (and I was the one who wrote it).
The question is whether you believe that every platform uses Unicode
code points directly as the wchar_t representation in UTF8-based locales.
Although we've not heard any complaints suggesting that that assumption
doesn't hold, I don't feel that 9.0 has been out long enough to prove
much. The complaints about the old code were rare enough to make me
think people just don't exercise the case too much, or don't notice if
the behavior is wrong. So it likely hasn't had that much exercise in
9.0 either.

In short, I'm pretty wary of back-patching it.

regards, tom lane