Hi!
I need have records on different languages simultaneously on some DB
for i18n of my application. Application is multilingual.
I decide use for this UNICODE. (UNICODE is encoding for my cluster)
But PostgreSQL do not work properly with UNICODE :(
I have the following problems:
1) ORDER BY is not alphabetical, sorting is by symbol code :(
2) ILIKE do not work case insensitive, it is work case sensitive
3) ~* work case sensitive
4) do not work UPPER() and LOWER() functions
I suppose all of this problems has some basis.
Possible Postgres can not do Unicode Collation and Case Mappings
correctly :(
For example:
If I have ru_RU.UNOCODE for DB, then ORDER BY and ILIKE work properly
for KOI8 charset. But UPPER() and LOWER() do not work still.
If I have KOI8 for cluster all work good (ILIKE, ~*, upper()/lower())
But solution with KOI8 for cluster or ru_RU.UNOCODE is not appropriate
for my application, because I need have many encodings on some DB simultaneously.
I think it is bug.
How can I fix this currently?
Will be it fixed on future versions and which version of PostgreSQL will work with
UNICODE correctly?
The same question is present in mailing list
http://archives.postgresql.org/pgsql-general/2003-03/msg00167.php
but it is unreciprocated.
Best regards,
Alexander Litvin.
