Details
-
Bug
-
Status: Open (View Workflow)
-
Major
-
Resolution: Unresolved
-
10.0.17, 10.3, 10.4
-
CentOS release 6.5 (X86_64), PureData Release 7.0 (P-3) , NetezzaSQL ODBCDriver
Description
I am trying to make a CONNECT table to a table on an IBM database.
It seems to me no test has been done yet for the IBM server.
The original table includes columns in NVARCHAR(N) for UTF-8 characters.
I have two problems.
1.
When I try creating a CONNECT table without specifying the column definition,
I get the below error.
Error (Code 1105): Unsupported SQL type -9
It doesn't happen when the table does not have any NVARCHAR(N) columns,
so I guess the CONNECT engine is not familiar with this data type
2.
When I specify the column definition such as VARCHAR(32),
the CONNECT table is created with no error.
But I get ? marks for all UTF-8 characters.
MariaDB [test]> show variables like '%char%';
|
+--------------------------+----------------------------+
|
| Variable_name | Value |
|
+--------------------------+----------------------------+
|
| character_set_client | utf8 |
|
| character_set_connection | utf8 |
|
| character_set_database | utf8 |
|
| character_set_filesystem | binary |
|
| character_set_results | utf8 |
|
| character_set_server | utf8 |
|
| character_set_system | utf8 |
|
| character_sets_dir | /usr/share/mysql/charsets/ |
|
+--------------------------+----------------------------+
|
8 rows in set (0.06 sec)
|
 |
MariaDB [test]> SELECT c1 FROM CONNECT_TABLE;
|
+-------+
|
| c1 |
|
+-------+
|
| ??? |
|
| JCB |
|
| ??? |
|
| ??? |
|
| ??? |
|
| ??? |
|
| ??? |
|
| ??? |
|
| UFJ |
|
| ??? |
|
+-------+
|
10 rows in set (0.00 sec)
|
I found it is not a problem of the ODBC driver, queries on isql returns the UTF-8 values correctly.
Changing the table's DEFAULT CHARSET or DATA_CHARSET did not work.