Re: Refcursor bug

[Dave Cramer <pg(at)fastcrypt(dot)com>]

Did this work in previous versions ?

I made a change recently which closes the refcursor immediately after reading it, because it was leaking on large batches

Dave

On 27-Apr-07, at 8:20 AM, Altaf Malik wrote:

Hi,
 I encountered a bug with refcursors with the latest JDBC driver. I cannot get two refcursors as OUT parameter if both of them point to the same underlying refcursor.

Steps to produce:
1- Create table:
CREATE TABLE testref ( a  int4, b  varchar);

2-Create the function:
CREATE OR REPLACE FUNCTION reftwo(
 A OUT REFCURSOR, B OUT REFCURSOR)
RETURNS record AS $$  DECLARE
  V_REF REFCURSOR;
BEGIN
     OPEN V_REF FOR SELECT * FROM testref;
     A:=V_REF;
     B:=V_REF;
END;  $$ LANGUAGE 'plpgsql'

3- Run the following code:

        con.setAutoCommit(false);
        CallableStatement stm = con.prepareCall("{call reftwo(?,?)}");       
        stm.registerOutParameter(1,Types.OTHER);
        stm.registerOutParameter(2,Types.OTHER);      
        stm.execute();      
        ResultSet rs = (ResultSet) stm.getObject(1);
        ResultSet rs2 = (ResultSet) stm.getObject(2);      
        con.setAutoCommit(true);

When i execute this code, it reports the following problem:
org.postgresql.util.PSQLException: ERROR: cursor "<unnamed portal 1>" does not exist

Am i doing something wrong?

--Altaf Malik

---

Ahhh...imagining that irresistible "new car" smell?
Check out new cars at Yahoo! Autos.