Database Software

Hi,

We are facing problem in using unicode on HP-UX 11.11 with Oracle 10g combination.
Following sample proc pgm tries to insert a sample record in NVARCHAR2 column using uvarchar proc datatype.

But before reaching to SQL level ir crashes giving a segmentation fault at wcscpy line.
We observed that
sizeof(wchar_t) = 4 // c++ datatype
sizeof(utext) = 2 on HP-UX. Will it create any problem?
We tried setting NLS_LANG on server to AMERICAN_AMERICA.WE8ISO8859P1 and AMERICAN_AMERICA.UTF8.
and DB characteristics are as follows.

SQL*Plus: Release 10.2.0.1.0 - Production on Wed Oct 4 18:09:48 2006

Copyright (c) 1982, 2005, Oracle. All rights reserved.


Connected to:
Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - 64bit Production
With the Partitioning, OLAP and Data Mining options

SQL> select * from nls_database_parameters;

PARAMETER VALUE
-------------------------------------------------------------------
NLS_LANGUAGE AMERICAN
NLS_TERRITORY AMERICA
NLS_CURRENCY $
NLS_ISO_CURRENCY AMERICA
NLS_NUMERIC_CHARACTERS .,
NLS_CHARACTERSET WE8ISO8859P1
NLS_CALENDAR GREGORIAN
NLS_DATE_FORMAT DD-MON-RR
NLS_DATE_LANGUAGE AMERICAN
NLS_SORT BINARY
NLS_TIME_FORMAT HH.MI.SSXFF AM

PARAMETER VALUE
------------------------------ -----------------------------------
NLS_TIMESTAMP_FORMAT DD-MON-RR HH.MI.SSXFF AM
NLS_TIME_TZ_FORMAT HH.MI.SSXFF AM TZR
NLS_TIMESTAMP_TZ_FORMAT DD-MON-RR HH.MI.SSXFF AM TZR
NLS_DUAL_CURRENCY $
NLS_COMP BINARY
NLS_LENGTH_SEMANTICS BYTE
NLS_NCHAR_CONV_EXCP FALSE
NLS_NCHAR_CHARACTERSET AL16UTF16
NLS_RDBMS_VERSION 10.2.0.1.0

20 rows selected.


#include
#include
#include
#include
#include

#include "/opt/db/oracle/product/10g/xdk/include/oratypes.h"
#include "/opt/db/oracle/product/10g/precomp/public/sqlucs2.h"
#include "/opt/db/oracle/product/10g/precomp/public/sqlca.h"

EXEC SQL INCLUDE SQLCA;
int main()
{

EXEC SQL BEGIN DECLARE SECTION;
uvarchar pref_name_u[35] ;
uvarchar pref_name_g[35] ;
EXEC SQL END DECLARE SECTION;

printf("Before wcscpy \n");
wcscpy((wchar_t *)pref_name_u.arr,L"SAMPLE");---------> This line gives a segmentation fault.
printf("After wcscpy \n");
pref_name_u.len = 6;
pref_name_u.arr[6]= L'\0';

EXEC SQL DELETE mc_dummy;

EXEC SQL COMMIT ;

printf("WIDECHAR Data to be inserted is : %S\n", (wchar_t* )pref_name_u.arr );
EXEC SQL INSERT INTO mc_dummy VALUES ( ref_name_u );

printf("sql retcode for insertion is : %d \n", sqlca.sqlcode );
EXEC SQL COMMIT ;

EXEC SQL SELECT name INTO ref_name_g FROM mc_dummy;

printf("sql retcode for fetch is : %d \n", sqlca.sqlcode );
printf("fetched data length is :%d\n",pref_name_g.len);
pref_name_g.arr[pref_name_g.len]= L'\0';

printf("fetched data is : %S \n", (wchar_t *)pref_name_g.arr );

return 0;
}
Table details :
SQL> desc mc_dummy;
Name Null? Type
---------------------------- -------- ----------------------------
NAME NVARCHAR2(100)

However when we made the following changes to the proc program we were able to insert and fetch the data.
#define utext unsigned int
This results in making sizeof(utext) same as sizeof(wchar_t) = 4.

But the data inserted into DB consumed morethan 1 byte per character. So though the column size is specified as 100, we were not able to insert 100. ( not even 50 ).

This is our urgent project requirement.
Could any one guide us in case we are missing anything.
Are we missing any DB settings?
Thanks
--------
Makarand Kulkarni