In a DB2 database, I have the following table:
CREATE TABLE MyTestTable
(
MYPATH VARCHAR(512) NOT NULL,
MYDATA BLOB,
CONSTRAINT MYTESTTABLE_PK PRIMARY KEY (MYPATH)
);
Using Java, I wish to update an existing row in this table with new blob data. My preferred way is to obtain an OutputStream to the BLOB column & write my data to the OutputStream.
Here is the test code I am using:
Connection connection = null;
PreparedStatement pStmnt = null;
ResultSet rSet = null;
try {
connection = ... // get db connection
String id = ... // set the MYPATH value
String sql = "SELECT MYDATA FROM MyTestTable WHERE MYPATH='"+id+"' FOR UPDATE";
pStmnt = connection.prepareStatement(sql);
rSet = pStmnt.executeQuery();
while (rSet.next()) {
Blob blobData = rSet.getBlob("MYDATA"); // this is a java.sql.Blob
OutputStream blobOutputStream = blobData.setBinaryStream(1);
blobOutputStream.write(data);
blobOutputStream.close();
connection.commit();
}
}
// close ResultSet/PreparedStatement/etc in the finally block
The above code works for the Oracle DB.
However, in DB2, calling setBinaryStream to get the OutputStream does not seem to work. The data does not get updated, and I do not get any error messages.
Qs: How can I get an OutputStream to the BLOB column of a DB2 table? What might need to be changed in the above code?
You are probably getting the data written to the Blob object successfully, but you need to do more with the PreparedStatement and ResultSet in order to actually update the value in the database.
First, your PreparedStatement must be instantiated using a version of
Connection.prepareStatement()
that takes a resultSetConcurrency parameter, which you must set to the valueResultSet.CONCUR_UPDATABLE
. (I don't know that the SQL SELECT actually needs to specify theFOR UPDATE
clause - see the tutorial at the link at the end of this answer.)Second, after you close blobOutputStream, you need to update the value in the ResultSet using
updateBlob(int columnIndex, Blob x)
orupdateBlob(String columnLabel, Blob x)
, then invokeResultSet.updateRow()
before doing aConnection.commit()
.I haven't updated Blob values this way myself, but it should work. If you run into any issues trying to reuse the Blob originally read from the ResultSet (which you probably don't need to do if you're not actually using the original data), you can use
Connect.createBlob()
to make an empty one to start with. You can learn more about updating ResultSets from this tutorial.