Why do I get the error message that my cursor is invalid? - sql

Question

SET SERVEROUTPUT ON
SET VERIFY OFF
DECLARE
v_deptno empcopy.deptno%TYPE;
v_empno empcopy.empno%TYPE;
v_sal empcopy.sal%TYPE;
v_bonus NUMBER(7,2);
CURSOR emp_cursor IS
SELECT deptno, empno, sal
FROM empcopy
WHERE v_deptno < 25;
BEGIN
FETCH emp_cursor INTO v_deptno, v_empno, v_sal;
FOR r_emp in emp_cursor LOOP
IF v_sal < 3000 THEN
v_bonus := v_sal * 1.1;
ELSE
v_bonus := v_sal * 1.12;
v_deptno := v_deptno;
v_empno := v_empno;
v_sal := v_sal;
END IF;
INSERT INTO emp3
VALUES(v_empno, v_deptno, v_sal);
FETCH emp_cursor INTO v_deptno, v_empno, v_sal;
END LOOP;
CLOSE emp_cursor;
DBMS_OUTPUT.PUT_LINE(v_deptno || v_empno || v_bonus);
END;
/
SET SERVEROUTPUT OFF
SET VERIFY ON
SQL> # emp3
DECLARE
*
ERROR at line 1:
ORA-01001: invalid cursor
ORA-06512: at line 11
I was wondering why it says I have an invalid cursor. I have tried to change the name of the value that has to be < 25 but it didn't work.

Answer 1

Code you wrote is full of errors (you already know that), I don't know whether I'll catch them all, but I'll try:
cursor selects data whose v_deptno < 25. But, v_deptno looks like a locally declared variable (the first one you declared), so - unless there's a column whose name is v_deptno, it'll fail.
besides, hardcoding deptno value to "25" doesn't scale at all. I'd suggest you to switch to a stored procedure which accepts department number as a parameter
the first executable line is FETCH, but it should have been OPEN (if you choose to do everything
manually:
declare a cursor
declare cursor variable (or separate variables, as you did)
open the cursor
loop
fetch from the cursor
exit when cursor%notfound
do something
end loop
close cursor
or, better, using cursor FOR loop where Oracle does all the dirty job for you:
FOR cursor in (SELECT statement) LOOP
do something
end loop
when you're checking salary (in IF), ELSE is ... more than strange. What's the purpose of v_deptno := v_deptno; (as well as other 3 variables you set to what they were)?
INSERT statement: it lacks column list you're inserting into. This is not an error (might fail if you did it wrong), but - it is confusing, especially when there are many columns involved).
but, you're inserting V_SAL. OK. Why did you compute V_BONUS, then? You never, ever used it (except in DBMS_OUTPUT, but that just displays the value, never does anything with it as far as data in the database is concerned)
OK, now my attempt, if I may.
First, code you wrote, fixed (I'm inserting V_BONUS value):
SQL> declare
2 v_bonus number;
3 begin
4 for cur_r in (select deptno, empno, sal
5 from empcopy
6 where deptno < 25
7 )
8 loop
9 if cur_r.sal < 3000 then
10 v_bonus := cur_r.sal * 1.1;
11 else
12 v_bonus := cur_r.sal * 1.12;
13 end if;
14
15 insert into emp3 (empno, deptno, sal)
16 values (cur_r.empno, cur_r.deptno, v_bonus);
17 end loop;
18 end;
19 /
PL/SQL procedure successfully completed.
SQL> select * From emp3;
EMPNO DEPTNO SAL
---------- ---------- ----------
7369 20 1100
7566 20 3272,5
7782 10 2695
7788 20 3360
7839 10 5600
7876 20 1210
7902 20 3360
7934 10 1430
8 rows selected.
SQL>
However, that can be done with a single INSERT statement at the SQL layer, you don't need PL/SQL at all. So, unless you're just practicing your PL/SQL skills, I'd suggest you to use this option:
SQL> rollback;
Rollback complete.
SQL> insert into emp3 (empno, deptno, sal)
2 select empno,
3 deptno,
4 case when sal < 3000 then sal * 1.1
5 else sal * 1.12
6 end
7 from empcopy
8 where deptno < 25;
8 rows created.
SQL> select * From emp3;
EMPNO DEPTNO SAL
---------- ---------- ----------
7369 20 1100
7566 20 3272,5
7782 10 2695
7788 20 3360
7839 10 5600
7876 20 1210
7902 20 3360
7934 10 1430
8 rows selected.
SQL>
Finally, a stored procedure option I suggested at the beginning; instead of < in the where clause, use =.
SQL> create or replace procedure p_bonus (par_deptno in empcopy.deptno%type)
2 is
3 v_bonus number;
4 begin
5 for cur_r in (select deptno, empno, sal
6 from empcopy
7 where deptno = par_deptno --> "=" instead of "<"
8 )
9 loop
10 if cur_r.sal < 3000 then
11 v_bonus := cur_r.sal * 1.1;
12 else
13 v_bonus := cur_r.sal * 1.12;
14 end if;
15
16 insert into emp3 (empno, deptno, sal)
17 values (cur_r.empno, cur_r.deptno, v_bonus);
18 end loop;
19 end;
20 /
Procedure created.
SQL> begin
2 p_bonus (par_deptno => 25);
3 end;
4 /
PL/SQL procedure successfully completed.

Answer 2

The table EMP3 apparently has fewer columns than the INSERT command provides data for (4).
Check the definition of that EMP3 table and make sure the table and the insert command can be matched.
After the comments/discussion, this is the code is a more readable form (I added indentation):
SET SERVEROUTPUT ON
SET VERIFY OFF
DECLARE
v_deptno empcopy.deptno%TYPE;
v_empno empcopy.empno%TYPE;
v_sal empcopy.sal%TYPE;
v_bonus NUMBER(7,2);
CURSOR emp_cursor IS
SELECT deptno, empno, sal
FROM empcopy
WHERE v_deptno < 25;
BEGIN
FETCH emp_cursor INTO v_deptno, v_empno, v_sal;
FOR r_emp in emp_cursor LOOP
IF v_sal < 3000 THEN
v_bonus := v_sal * 1.1;
ELSE
v_bonus := v_sal * 1.12;
v_deptno := v_deptno;
v_empno := v_empno;
v_sal := v_sal;
END IF;
INSERT INTO emp3
VALUES(v_empno, v_deptno, v_sal);
FETCH emp_cursor INTO v_deptno, v_empno, v_sal;
END LOOP;
CLOSE emp_cursor;
DBMS_OUTPUT.PUT_LINE(v_deptno || v_empno || v_bonus);
END;
/
SET SERVEROUTPUT OFF
SET VERIFY ON
What this code appears to be doing is:
go over every "employee" record in table EMPCOPY
for each of those records, check if the salary is lower than 3000 and calculate the bonus by multiplying salary by 1.1 if it is
if the salary is not lower than 3000, v_bonus is calculated by 1.12
after this, the new values should be INSERTed into table EMP3
finally, the values of the very last inserted record are printed out.
First off: the poor folks that make less than 3000 also only get 110% bonus while the already better off employees get 2% more on top?
I'm aware that this is common practice, but I'd advocate aiming for a better, fairer world - at least in simple coding examples like this.
Now to the code.
What this code should do, does not require the use of cursors at all. In fact, it is not a good practice in this case.
Instead, just use plain SQL to achieve the desired effect:
INSERT INTO EMP3
(empno, deptno, sal, bonus)
(SELECT
empno, deptno, sal
, case
when sal < 3000 then sal * 1.1
else sal * 1.12
end as bonus
FROM
empcopy
WHERE
deptno < 25);
The caveat here is that the target table EMPCOPY needs to have a column for BONUS. That this column is missing was likely the cause for the original issue.
Also worth noting that this approach does not print out anything.
If that is a requirement for the solution, then this should be done after the INSERT (a simple FOR LOOP over the target table should do).
I hope that helps.