[Sorry if the post is duplicated. But I don't receive and ACK from majordamo.]
On 11/27/05, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Doesn't this just move the failure cases around?
I don't think so. I've tried to fix every tolower/toupper related
problem (depending on the MB characters) I found in the code. If
there's still a problem with them, it should be related with the
pg_wchar_tb concept in utils/mb/wchar.c (AFAIC, the last remaning part
about this case conversion mess is the one related with PQfnumber -
which needs a very different approach.)
> The really fundamental problem is that tolower/toupper don't work
> on wchar_t.
Yes, indeed. I agree with you to find a proper solution for character
set handling. But, IMHO, it's better to have a-patchy working system
instead of a not working one. Put yourself in our position, we're
trying to give enterprise solutions to our customers and ILIKE,
lower/upper doesn't work in the product that we support. We can work
with a small patch instead of waiting next releases for an exact
solution to the problem.
