How to support German, French and other characters.

[Andrei Bintintan]

Hi to all,

We are using pgsql as a "multilanguage" database. But I noticed yesterday a strange problem with the german umlaut characters. I cannot convert the in upper case in lowercase. Probably there are also other bad functionalities.

Now. I searched the internet for answers but they were not quite exact, so I could't find a solution.

We use ASCII encoding for the database, but I tried the Latin1 -> Latin 10 and the behavior is the same. In some forums there is written something about some "locale" setting... etc etc.

What settings do I have to make so that we won't have these problems in the future. We intend to work with French characters also, and in the future with Hungarian.

As system the pgsql is v.7.4 running on a Suse 9.1 machine. The locale command gives me:

linz:/var/lib/pgsql/data # locale
LANG=
LC_CTYPE=en_US.UTF-8
LC_NUMERIC="POSIX"
LC_TIME="POSIX"
LC_COLLATE="POSIX"
LC_MONETARY="POSIX"
LC_MESSAGES="POSIX"
LC_PAPER="POSIX"
LC_NAME="POSIX"
LC_ADDRESS="POSIX"
LC_TELEPHONE="POSIX"
LC_MEASUREMENT="POSIX"
LC_IDENTIFICATION="POSIX"
LC_ALL=


Thank you in advance.
Andy.

[Andrei Bintintan]

Hi Ivo,

If I use UNICODE encoding for the DB I get some errors:
For example the following query: select lower('MöBÜEL')
returns ERROR: Unicode characters greater than or equal to 0x10000 are not
supported.

I found this on a forum:
----------------------------------------------------------------
----------------------------------------------------------------

postgres 7.4 on linux, glibc 2.2.4-6
I've a table containing unicode-data and the lower()-function does not
work proper. While it lowers standard letters like A->a,B->b ... it
fails on special letters like german umlauts (Ä , Ö ...) that are

simply

keeped untouched.

upper() and lower() didn't support multibyte character sets before 8.0.

regards, tom lane
----------------------------------------------------------------
----------------------------------------------------------------

We are using these king of comparations and character translations in our
DB.

I really cannot figure out what solution to use.

Best regards,
Andy.



----- Original Message -----
From: "Ivo Rossacher" <***@bluewin.ch>
To: "Andrei Bintintan" <***@ar-sd.net>
Sent: Wednesday, November 10, 2004 9:57 PM
Subject: Re: [ADMIN] How to support German, French and other characters.

Dear Andy,
ASCII encoding means that the database does not care (and know) about the
encoding. So the client is in full charge to deal with the encoding issue.
This is very uncomfortable within a multilanguage enviroment. Without the
encoding the caption can not be determined correctly by the database it

self.

Suse 9.1 does use unicode as the default encoding for all the desctop. For

my

multilanguage projects I do use therefore UNICODE as encoding for the
database. (createdb -EUNICODE dbname will generate a unicode database more
precisly a UTF8 database)
Most of the Microsoft clients are internal UNICODE anyway and can deal

with

this setting. Older Unix or Linux installations need some tweaking

probably.

Best regards
Ivo Rossacher

Post by Andrei Bintintan
Hi to all,
We are using pgsql as a "multilanguage" database. But I noticed

yesterday a

Post by Andrei Bintintan
strange problem with the german umlaut characters. I cannot convert the

in

Post by Andrei Bintintan
upper case in lowercase. Probably there are also other bad

functionalities.

Post by Andrei Bintintan
Now. I searched the internet for answers but they were not quite exact,

so

Post by Andrei Bintintan
I could't find a solution.
We use ASCII encoding for the database, but I tried the Latin1 -> Latin

10

Post by Andrei Bintintan
and the behavior is the same. In some forums there is written something
about some "locale" setting... etc etc.
What settings do I have to make so that we won't have these problems in

the

Post by Andrei Bintintan
future. We intend to work with French characters also, and in the future
with Hungarian.
As system the pgsql is v.7.4 running on a Suse 9.1 machine. The locale
linz:/var/lib/pgsql/data # locale
LANG=
LC_CTYPE=en_US.UTF-8
LC_NUMERIC="POSIX"
LC_TIME="POSIX"
LC_COLLATE="POSIX"
LC_MONETARY="POSIX"
LC_MESSAGES="POSIX"
LC_PAPER="POSIX"
LC_NAME="POSIX"
LC_ADDRESS="POSIX"
LC_TELEPHONE="POSIX"
LC_MEASUREMENT="POSIX"
LC_IDENTIFICATION="POSIX"
LC_ALL=
Thank you in advance.
Andy.

--
Ivo Rossacher

---------------------------(end of broadcast)---------------------------
TIP 2: you can get off all lists at once with the unregister command
(send "unregister YourEmailAddressHere" to ***@postgresql.org)