Multi-Language upper(),lower()

Ara Anjargolian <ara818@yahoo.com> writes:

Am I mistaken, or was multi-language support for
upper(), lower() and friends recently added to 7.5?

I fixed upper/lower so that they would work on multi-byte character set
encodings, given that the platform supports the <wctype.h> functions
towupper, towlower, etc. This means they can work on Unicode (UTF-8)
data. However, this doesn't do anything to improve the fundamental
issue that we are dependent on the C library's API for locale-specific
behavior, and that that API does *not* easily support dealing with
multiple locales at once. See the nearby "unicode and sorting" thread.

regards, tom lane

Tom Lane wrote:

Ara Anjargolian <ara818@yahoo.com> writes:

Am I mistaken, or was multi-language support for
upper(), lower() and friends recently added to 7.5?

I fixed upper/lower so that they would work on multi-byte character set
encodings, given that the platform supports the <wctype.h> functions
towupper, towlower, etc. This means they can work on Unicode (UTF-8)
data. However, this doesn't do anything to improve the fundamental
issue that we are dependent on the C library's API for locale-specific
behavior, and that that API does *not* easily support dealing with
multiple locales at once. See the nearby "unicode and sorting" thread.

OK, I marked the item as done.

-- 
  Bruce Momjian                        |  http://candle.pha.pa.us
  pgman@candle.pha.pa.us               |  (610) 359-1001
  +  If your life is a hard drive,     |  13 Roberts Road
  +  Christ can be your backup.        |  Newtown Square, Pennsylvania 19073