ORA-01406

[Nobody/Anonymous]

Hello.

My application returns an ORA-01406 when i try to execute OCI_FetchNext(rs)
function.
Here the part of C code:

OCI_ExecuteStmt(st, "select * from zep_v_zw");
    rs = OCI_GetResultset(st);
    while(OCI_FetchNext(rs)) {
        printw("%s  %s  %s  %s\n", OCI_GetString(rs, 1),
                OCI_GetString(rs, 2), OCI_GetString(rs, 3),
                OCI_GetString(rs, 4));
    }

And view code:

SELECT
    lpad(num, 5, ' ') num,
    rpad(code, 4, ' ') code,
    TO_CHAR(idate, 'DD-MM-YY') idate,
    rpad(info, 75, ' ') info
FROM
    zw
WHERE
    status <> 'Z'
    AND code IN ( 'P', 'A' )
    AND idate > ADD_MONTHS(TRUNC(SYSDATE, 'DD'), -2)
ORDER BY num DESC, code

There's a problem with

rpad(info, 75, ' ') info

line.
Without rpad() everything is fine. Info column is 500 bytes long in
table.

Database version 9i Release 9.2.0.6.0
Client: Instant Client 11.2.0.1
Ocilib version: 3.5.1
OS: Debian 5.0.4 x86

Any suggestions?

[Vincent Rogier]

Hi,

It migth be an oracle bug (found something similar in metalink bug database).

I'll test it on my config this afternoon and let you know.

[Vincent Rogier]

hi,

i've just tried it on my local config (10gR2 client and server).
I created the same view (mapped to one of my tables where num and code are
varchar2(15), idate a date and info varchar2(200).

Is your database charset UTF8 ?

I build the test suite in ANSI charset and no error.

[Nobody/Anonymous]

No, it's EE8ISO8859P2

[Nobody/Anonymous]

When I unset NLS_LANG (I'm using POLISH_POLAND.UTF8 value) there's no error
anymore, but there's no polish letters.

[Vincent Rogier]

Ok, you're using UTF8 as client charset. So you need to use ocilib v3.6.0 that
has proper support for UTF8

[Nobody/Anonymous]

With v3.6.0 it works like a charm :)

Thank you very much!

[Vincent Rogier]

Good :)