================================================== ==========================
POSTGRESQL BUG REPORT TEMPLATE
================================================== ==========================

Your name : Matthew Cooper
Your email address : matty (at) cloverworxs (dot) com

System Configuration
---------------------
Architecture (example: Intel Pentium) : Intel Pentium
Operating System (example: Linux 2.0.26 ELF) : Redhat 8.0 / 9.0
PostgreSQL version (example: PostgreSQL-7.2.2): PostgreSQL-7.2.2 / 7.3.2
Compiler used (example: gcc 2.95.2) : none

Please enter a FULL description of your problem:
------------------------------------------------
I have a database with UNICODE encoding set. In it is a table with a
varchar(10) column. If I insert 10 western characters into it, it is OK. If
I insert 10 chinese characters it says:
postgresql value too long for type character varying 10
when using 7.3.2. If I use 7.2.2 it works fine.
Please describe a way to repeat the problem. Please try to provide a
concise reproducible example, if at all possible:
----------------------------------------------------------------------
createdb -E UNICODE mydb
Then in psql...
create table mgc (c1 varchar(10));
insert into mgc values('0123456789');
This all works fine.
Now I put the following command into a file (say my.sql) which is UTF-8
encoded and the literal is 10 chinese characters.
(I don't know if once emailed this command will be readable so you may have
to re-create the command by pasting 10 chinese characters into your
favourite UTF-8 compatible editor.)
insert into mgc values ('åˆâ€*钟练ä¹ åˆâ€*钟练ä¹ 练ä¹ ');
I then run psql -f my.sql and get the error for 7.3.2 but it works for
7.2.2.

If you know how this problem might be fixed, list the solution below:
---------------------------------------------------------------------
I am guessing it is incorrectly counting the bytes and not the characters.
Presumably a workaround is to double the length of the field.


---------------------------(end of broadcast)---------------------------
TIP 1: subscribe and unsubscribe commands go to majordomo (AT) postgresql (DOT) org

"Matthew Cooper" <matty (AT) cloverworxs (DOT) com> writes:
Ah, no doubt it did.

This works fine for me, using either 7.3.4 or CVS tip. Are you sure
that the system knows your client-side encoding is supposed to be UTF8?

uc=# show client_encoding ;
client_encoding
-----------------
UNICODE
(1 row)

uc=# create table mgc(f1 varchar(10));
CREATE TABLE
uc=# \i mgc.sql
INSERT 328444 1
uc=# select * from mgc;
f1
----------------------
͈†Õ’ŸÓ£ƒð¿*͈†Õ’ŸÓ£ƒð¿*Ó£ƒð¿*
(1 row)

uc=# select length(f1) from mgc;
length
--------
10
(1 row)


regards, tom lane

---------------------------(end of broadcast)---------------------------
TIP 6: Have you searched our list archives?

http://archives.postgresql.org