At 03:43 +0300 on 01/08/1999, Gilles Darold wrote:
> I don't know about windows-1251. Perhaps this can't help you. But if
> you have japanese in you database, you can proceed a search on it so
> why not windows-1251 ?
Because the backend has to know that the lowercase for char NNN in this
codepage is MMM. This is different from one locale to the next. If it were
windows-1255 (Hebrew), there would be no lowercase at all for any character
above 224. So of course it's different than Japanese.
There can be two solutions to this problem:
1) Write a function using SPI, install it on the backend, and use
it for the comparison.
2) Create the new locale, or at least the LC_CTYPE part of the locale,
on the unix you are using.
Herouth
--
Herouth Maoz, Internet developer.
Open University of Israel - Telem project
http://telem.openu.ac.il/~herutma