Hello,

there seems to be a bug with Oracle-JDBC-Driver. We are using an UTF8-Database with NLS_LENGTH_SEMANTICS set to CHAR.

Since DATABASEMETADATA.GETCOLUMNS() returns for COLUMNS_SIZE ..

DECODE (t.data_precision, null, t.data_length, t.data_precision)
AS column_size

We alwasy get the size of the column in byte. And that is 3times bigger than the size in character. So for a table TEST with one column TESTCHAR declared as CHAR (1 CHAR), the size returned by getColumns is 3 instead of expected 1.

Does anyone know of that problem and has a workaround for that?

We have opened a bug at the oracleside. But since it is a big company, we expect the problem not to be solved very quickly.

Thanks in advance
Christian