OCI: I18N and storing wide char data

Hello,

We are developing an OCI based C++ application that requires internationalization support. One of the things we will do is to shuffle character data around using wchar_t* (instead of char*) and wstring (instead of string).

We are currently using 8.1.6 on HPUX 11.00.

I noticed the nvarchar2 and nchar data types, which appear to tailored to storing wide character (or multibyte?) character data. So, as a simple test I came up with a table that has an nvarchar2 in it and a sql statement that looks like

insert into tbl values ( :placeholder )

static sb2 indicator;
static wchar_t wc[20];
mbstowcs(wc,"hello",6);
if (OCIBindByName(

			m_stmt,


			&m_bindHandle,

			m_dbPtr->ociErr(),
			(text *)":placeholder",
			-1,
			(dvoid *)wc,
			10,
			SQLT_STR,


			&indicator,

			0,0,0,0,OCI_DEFAULT) !=
			OCI_SUCCESS)
{
	printf("bind failed\n");

The bind succeeds.

When I execute the statement I get the following error code

ORA-12704: character set mismatch

I also tried SQLCS_NCHAR instead of SQLT_STR in the bind statement.

Also, not sure what data type I should be binding, a "multibyte" char*, a wchar_t* ?

The OCI docs are lacking in this area,
or I missed something, several times :)

Clues appreciated...

Mark, Received on Sun Dec 03 2000 - 22:50:35 CST