BUG #2402: case insensitive match for unicode doesn't work

The following bug has been logged online:

Bug reference: 2402
Logged by: Balzs Klein
Email address: balazs(dot)klein(at)t-online(dot)hu
PostgreSQL version: 8.1
Operating system: Gentoo Linux
Description: case insensitive match for unicode doesn't work
Details:

case conversion and case insensitive match (with or without regular
expression) does not work with non standard (Greek, Turkish, Hungarian,
etc.) characters

Examples:

Select 'a' ILIKE 'A'
--True

Select '' ILIKE ''
--False

Select 'a' ~* 'A'
--True

Select '' ~* ''
--False

Select upper('a')='A'
--True

Select upper('')=''
--False

You would get the same result with ,,,, or with many Greek
characters that I tried.

I raised this issue on pgsql.general -
http://groups.google.com/group/pgsql.general/browse_frm/thread/20aed89ab0e19
e3d/4771fb1be397afea#4771fb1be397afea
but there didn't seem to be an easy workaround for this

On a sidenote SQLServer2000 and Access also make mistakes with the
equivalent of upper/lower function but they give correct result with more
characters than PG. Which at least consistently fails:)

Regards.
Balzs