I just currently learning about Oracle PL/SQL. I wanna create store procedure with variable and then call it with another script. Is it possible?
I tried use simple script without variable and it works:
CREATE OR REPLACE PROCEDURE testmyproc AS
BEGIN
INSERT INTO tes_table(dt)
VALUES (sysdate);
commit;
END testmyproc;
Then I call it with another script abc.sql
begin
testmyproc;
end;
It works successfully.
But, unfortunately if I use DECLARE (variable) at my PROCEDURE, it show error when I execute (but it success in create procedure).
Here's my PROCEDURE (no error):
CREATE OR REPLACE PROCEDURE sp_testmyproc AS
DECLARE
job_name varchar(100);
status_key number;
status_desc varchar(100);
notes varchar(250);
BEGIN
status_key := 1;
status_desc := 'SUCCESS';
notes := 'Process Completed';
INSERT INTO automation_log(job_name, dt, status_key, status_desc, notes)
VALUES (job_name, sysdate, status_key, status_desc, notes);
commit;
END sp_testmyproc;
Here's my execure script abc.sql (show error when i execute it)
-without DECLARE
begin
sp_testmyproc;
end;
-I tried to execute it with DECLARE
DECLARE
job_name varchar(100);
status_key number;
status_desc varchar(100);
notes varchar(250);
begin
status_key := 1;
status_desc := 'SUCCESS';
notes := 'Process Completed';
sp_testmyproc;
end;
It show error like this:
> ORA-06550: line 8, column 11:
> PLS-00905: SP_TESTMYPROC is invalid
> ORA-06550: line 8, column 3:
> PL/SQL: Statement ignored
Can I call Procedure for another script? Is It best practice?
I just think PROCEDURE can be used for many cases (something like function in programming).
Thank you!
You need to learn the syntax of the procedure.
In Procedure, You should not use the keyword DECLARE. Any variables you want to declare must be between AS and BEGIN in the procedure.
Your procedure should look like follows:
CREATE OR REPLACE PROCEDURE sp_testmyproc AS
--DECLARE
job_name varchar(100);
status_key number;
status_desc varchar(100);
notes varchar(250);
BEGIN
status_key := 1;
.....
.....
.....
Refer to this document for the syntax of the oracle procedure as it is very easy to follow.
Please note the difference when creating a stored procedure in SQL*Plus:
SQL> create or replace procedure test_ok as
2 v number;
3 begin
4 v:=0;
5 end;
6 /
Procedure created.
SQL> show errors
No errors.
SQL> create or replace procedure test_ko as
2 declare
3 v number;
4 begin
5 v:=0;
6 end;
7 /
Warning: Procedure created with compilation errors.
SQL> show errors
Errors for PROCEDURE TEST_KO:
LINE/COL ERROR
-------- -----------------------------------------------------------------
2/1 PLS-00103: Encountered the symbol "DECLARE" when expecting one of
the following:
begin function pragma procedure subtype type <an identifier>
<a double-quoted delimited-identifier> current cursor delete
exists prior external language
SQL>
If you have compilation errors you get at least Warning: Procedure created with compilation errors. If you have compilation errors and use show errors you get all error messages.
Here's my stored procedure:
CREATE OR REPLACE PROCEDURE STATS_SD
(
P_ID IN NUMBER,
PRC OUT SYS_REFCURSOR
)
IS
BEGIN
OPEN PRC FOR
SELECT
ID,
SESID
FROM RESPONSES
WHERE ID IN (P_ID)
END;
When I try to execute it using
EXEC EXAM_STATS_STUDENTS_SD('6901');
I get the following error:
PLS-00306: wrong number or types of arguments in call to 'STATS_SD'
Do you have any ideas why?
Here is an example using an OUT parameter that is a sys_refcursor. Note that I close the cursor in the pl/sql block that uses it (which is important!):
create or replace procedure get_data(o_cur OUT SYS_REFCURSOR) as
begin
OPEN o_cur FOR
select * from emp;
end;
And using the get_data procedure:
declare
l_cur sys_refcursor;
l_row emp%rowtype;
begin
get_data(l_cur);
LOOP
fetch l_cur
into l_row;
exit when l_cur%notfound;
-- do something with l_row here
END LOOP;
close l_cur;
end;
You are passing a wrong datatype to your procedure.
According to your declaration a NUMBER is expected:
P_ID IN NUMBER
However, you pass a VARCHAR2 in your exec command:
EXEC EXAM_STATS_STUDENTS_SD('6901');
Note the '' around the value.
Try calling this instead:
EXEC EXAM_STATS_STUDENTS_SD(6901);
Apart from that you are missing the second parameter completely.
I previously asked a question in stackoverflow regarding a certain function that I was having a hard time with and I found the answer when I saw the RETURN statement missing.
Now, I am dealing with a procedure that calls for the input text letters to be converted to UPPER if in the LOWER case and vice versa.
e.g IF I put in 'AbC' it should return 'aBc'
So far this is my code
CREATE OR REPLACE PROCEDURE Opposite_Case (p_string IN VARCHAR2)
IS
var_string VARCHAR2(20);
var_contain VARCHAR2(20);
i NUMBER;
BEGIN
var_string:=substr(Opposite_Case.p_string,i,1);
var_contain:= var_string || var_contain;
FOR i in 1.. length(var_string)
LOOP
BEGIN
IF var_string IN ('ABCDEFGHIJKLMNOPQRSTUVWXYZ') THEN
SELECT LOWER(var_string) INTO var_contain FROM dual;
ELSE
SELECT UPPER(var_string) INTO var_contain FROM dual;
END IF;
END;
END LOOP;
END;
/
BUT the following error is returned:
EXECUTE Opposite_Case('AbC')
begin Opposite_Case('AbC'); end;
ORA-06502: PL/SQL: numeric or value error
ORA-06512: at "SYS.OPPOSITE_CASE", line 10
ORA-06512: at line 1
Thanks in advance.
PS. This is just my 3rd day with a PL/SQL language so please bear with me.
EDIT: I got it to work finally thanks to #Satya's help. Now I get to convert them like it should but how do I output my selects in one line though?
I'm almost there. Appreciate the help a lot from this community!!
What I have so far:
SQL> EXECUTE Opposite_Case('AbC')
A
b
C
PL/SQL procedure successfully completed
You should investigate the TRANSLATE function. To use it to switch the case of the characters in your string you'd do something like the following:
SELECT TRANSLATE('AbC',
'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz',
'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ')
FROM DUAL;
which should return 'aBc'.
SQLFiddle here
Share and enjoy.
Finally found the answer! Thanks guys! (without TRANSLATE though)!
CREATE OR REPLACE PROCEDURE Opposite_Case (p_string IN VARCHAR2)
IS
var_string VARCHAR2(20);
var_contain VARCHAR2(20);
i NUMBER;
BEGIN
FOR i in 1.. length(p_string)
LOOP
BEGIN
var_string:=substr(p_string,i,1);
var_contain:= var_string;
IF var_string IN ('A','B','C','D','E','F','G','H','I','J','K','L','M','N','O','P','Q','R','S','T','U','V','W','X','Y','Z') THEN
SELECT LOWER(var_string) INTO var_contain FROM dual;
DBMS_OUTPUT.PUT(var_contain);
ELSE
SELECT UPPER(var_string) INTO var_contain FROM dual;
DBMS_OUTPUT.PUT(var_contain);
END IF;
END;
END LOOP;
dbms_output.new_line;
END;
/
Here I have a stored procedure in Oracle:
CREATE OR REPLACE PROCEDURE StP_COMPS
IS
CV_1 SYS_REFCURSOR;
BEGIN
OPEN CV_1 FOR SELECT * FROM COMPUTERS;
END;
When I execute the procedure like EXEC SP_COMPS I get no error, the SQL Developer just shows "ananymous block completed". Then I change the procedure to a
CREATE OR REPLACE PROCEDURE SP_COMPS
(cv_1 OUT SYS_REFCURSOR)
IS
BEGIN
OPEN CV_1 FOR SELECT * FROM COMPUTERS;
END;
and when I execute I get error stating that the number of type of the arguments are wrong. I'm very curious what I could send as an argument to the procedure if it's just an output parameter. I want to get the result set of the query run inside the procedure. What am I doing wrong here?
P.S. When I try to run the procedure by right clicking the procedure and selecting Run I get:
DECLARE
CV_2 sys_refcursor;
BEGIN
SP_COMPS(
CV_2 => CV_2
);
:CV_2 := CV_2; -- <--Can't understand this part
END;
You have a variable, you should execute the procedure like:
DECLARE
CV_1 SYS_REFCURSOR;
BEGIN
SP_COMPS(CV_1);
--use cv_1
END;
UPDATE(after OP update): That's a simple template for testing. As explained here: Easiest method to test an Oracle Stored Procedure, just run that code, and select ref_cursor as type of cv2 variable.
I am trying to run a stored procedure that has multiple in and out parameters. The procedure can only be viewed in my Connections panel by navigating
Other Users | <user> | Packages | <package> | <procedure>
If I right click , the menu items are "Order Members By..." and "Create Unit Test" (greyed out). The ability to "Run" the procedure does not seem possible when it's accessed by user.
I have been trying to find an example of how to create an anonymous block so that I can run the procedure as a SQL file, but haven't found anything that works.
Does anyone know how I can execute this procedure from SQL Developer? I am using Version 2.1.1.64.
EDIT 1:
The procedure I want to call has this signature:
user.package.procedure(
p_1 IN NUMBER,
p_2 IN NUMBER,
p_3 OUT VARCHAR2,
p_4 OUT VARCHAR2,
p_5 OUT VARCHAR2,
p_6 OUT NUMBER)
If I write my anonymous block like this:
DECLARE
out1 VARCHAR2(100);
out2 VARCHAR2(100);
out3 VARCHAR2(100);
out4 NUMBER(100);
BEGIN
EXECUTE user.package.procedure (33,89, :out1, :out2, :out3, :out4);
END;
I get the error:
Bind Varialbe "out1" is NOT DECLCARED
anonymous block completed
I've tried initializing the out* variables:
out1 VARCHAR2(100) := '';
but get the same error:
EDIT 2:
Based on Alex's answer, I tried removing the colons from in front of the params and get this:
Error starting at line 1 in command:
DECLARE
out1 VARCHAR2(100);
out2 VARCHAR2(100);
out3 VARCHAR2(100);
out4 NUMBER(100);
BEGIN
EXECUTE user.package.procedure (33,89, out1, out2, out3, out4);
END;
Error report:
ORA-06550: line 13, column 17:
PLS-00103: Encountered the symbol "USER" when expecting one of the following:
:= . ( # % ; immediate
The symbol ":=" was substituted for "USER" to continue.
06550. 00000 - "line %s, column %s:\n%s"
*Cause: Usually a PL/SQL compilation error.
*Action:
With simple parameter types (i.e. not refcursors etc.) you can do something like this:
SET serveroutput on;
DECLARE
InParam1 number;
InParam2 number;
OutParam1 varchar2(100);
OutParam2 varchar2(100);
OutParam3 varchar2(100);
OutParam4 number;
BEGIN
/* Assign values to IN parameters */
InParam1 := 33;
InParam2 := 89;
/* Call procedure within package, identifying schema if necessary */
schema.package.procedure(InParam1, InParam2,
OutParam1, OutParam2, OutParam3, OutParam4);
/* Display OUT parameters */
dbms_output.put_line('OutParam1: ' || OutParam1);
dbms_output.put_line('OutParam2: ' || OutParam2);
dbms_output.put_line('OutParam3: ' || OutParam3);
dbms_output.put_line('OutParam4: ' || OutParam4);
END;
/
Edited to use the OP's spec, and with an alternative approach to utilise :var bind variables:
var InParam1 number;
var InParam2 number;
var OutParam1 varchar2(100);
var OutParam2 varchar2(100);
var OutParam3 varchar2(100);
var OutParam4 number;
BEGIN
/* Assign values to IN parameters */
:InParam1 := 33;
:InParam2 := 89;
/* Call procedure within package, identifying schema if necessary */
schema.package.procedure(:InParam1, :InParam2,
:OutParam1, :OutParam2, :OutParam3, :OutParam4);
END;
/
-- Display OUT parameters
print :OutParam1;
print :OutParam2;
print :OutParam3;
print :OutParam4;
Executing easy. Getting the results can be hard.
Take a look at this question I asked Best way/tool to get the results from an oracle package procedure
The summary of it goes like this.
Assuming you had a Package named mypackage and procedure called getQuestions. It returns a refcursor and takes in string user name.
All you have to do is create new SQL File (file new). Set the connection and paste in the following and execute.
var r refcursor;
exec mypackage.getquestions(:r, 'OMG Ponies');
print r;
For those using SqlDeveloper 3+, in case you missed that:
SqlDeveloper has feature to execute stored proc/function directly, and output are displayed in a easy-to-read manner.
Just right click on the package/stored proc/ stored function, Click on Run and choose target to be the proc/func you want to execute, SqlDeveloper will generate the code snippet to execute (so that you can put your input parameters). Once executed, output parameters are displayed in lower half of the dialog box, and it even have built-in support for ref cursor: result of cursor will be displayed as a separate output tab.
Open the procedure in SQL Developer and run it from there. SQL Developer displays the SQL that it runs.
BEGIN
PROCEEDURE_NAME_HERE();
END;
Use:
BEGIN
PACKAGE_NAME.PROCEDURE_NAME(parameter_value, ...);
END;
Replace "PACKAGE_NAME", "PROCEDURE_NAME", and "parameter_value" with what you need. OUT parameters will need to be declared prior to.
Though this question is quite old, I keep stumbling into same result without finding an easy way to run from sql developer.
After couple of tries, I found an easy way to execute the stored procedure from sql developer itself.
Under packages, select your desired package and right click on the package name (not on the stored procedure name).
You will find option to run. Select that and supply the required arguments. Click OK and you can see the output in output variables section below
I'm using SQL developer version 4.1.3.20
None of these other answers worked for me. Here's what I had to do to run a procedure in SQL Developer 3.2.20.10:
SET serveroutput on;
DECLARE
testvar varchar(100);
BEGIN
testvar := 'dude';
schema.MY_PROC(testvar);
dbms_output.enable;
dbms_output.put_line(testvar);
END;
And then you'd have to go check the table for whatever your proc was supposed to do with that passed-in variable -- the output will just confirm that the variable received the value (and theoretically, passed it to the proc).
NOTE (differences with mine vs. others):
No : prior to the variable name
No putting .package. or .packages. between the schema name and the procedure name
No having to put an & in the variable's value.
No using print anywhere
No using var to declare the variable
All of these problems left me scratching my head for the longest and these answers that have these egregious errors out to be taken out and tarred and feathered.
Can't believe, this won't execute in SQL Developer:
var r refcursor;
exec PCK.SOME_SP(:r,
'02619857');
print r;
BUT this will:
var r refcursor;
exec TAPI_OVLASCENJA.ARH_SELECT_NAKON_PRESTANKA_REG(:r, '02619857');
print r;
Obviously everything has to be in one line..
Using SQL Developer Version 4.0.2.15 Build 15.21 the following works:
SET SERVEROUTPUT ON
var InParam1 varchar2(100)
var InParam2 varchar2(100)
var InParam3 varchar2(100)
var OutParam1 varchar2(100)
BEGIN
/* Assign values to IN parameters */
:InParam1 := 'one';
:InParam2 := 'two';
:InParam3 := 'three';
/* Call procedure within package, identifying schema if necessary */
schema.package.procedure(:InParam1, :InParam2, :InParam3, :OutParam1);
dbms_output.enable;
dbms_output.put_line('OutParam1: ' || :OutParam1);
END;
/
To run procedure from SQL developer-only execute following command
EXECUTE PROCEDURE_NAME;
I had a stored procedure that returned a cursor, in my case it was actually of a custom package type (T_CURSOR, looks like a convention to me) that is defined as REF CURSOR.
There may be a better way to do this, but I defined variables for all the columns of the table that the cursor was iterating, looped the cursor fetching each row into those variables, then printed them out.
SET serveroutput on;
DECLARE
testvar number;
v_cur SYS_REFCURSOR;
ORIGINAL_EMP_NUM NUMBER;
TEMPORARY_EMP_NUM NUMBER;
ORG_UNIT_CODE VARCHAR2(2 BYTE);
MRU_CODE VARCHAR2(10 BYTE);
CTRL_COMPANY_CODE VARCHAR2(10 BYTE);
IS_TEMP_FLAG VARCHAR2(1 BYTE);
BEGIN
testvar := 420;
foo.updates.get_temporary_authorisations(testvar, v_cur);
dbms_output.enable;
dbms_output.put_line(testvar);
LOOP
FETCH v_cur INTO ORIGINAL_EMP_NUM, TEMPORARY_EMP_NUM, ORG_UNIT_CODE, MRU_CODE, CTRL_COMPANY_CODE, IS_TEMP_FLAG;
EXIT WHEN v_cur%NOTFOUND;
dbms_output.put_line(ORIGINAL_EMP_NUM || ',' || TEMPORARY_EMP_NUM || ',' || ORG_UNIT_CODE || ',' || MRU_CODE|| ',' || CTRL_COMPANY_CODE|| ',' || IS_TEMP_FLAG);
END LOOP;
CLOSE v_cur;
END;
I wasn't able to get #Alex Poole answers working. However, by trial and error, I found the following works (using SQL Developer version 3.0.04). Posting it here in case it helps others:
SET serveroutput on;
DECLARE
var InParam1 number;
var InParam2 number;
var OutParam1 varchar2(100);
var OutParam2 varchar2(100);
var OutParam3 varchar2(100);
var OutParam4 number;
BEGIN
/* Assign values to IN parameters */
InParam1 := 33;
InParam2 := 89;
/* Call procedure within package, identifying schema if necessary */
schema.package.procedure(InParam1, InParam2,
OutParam1, OutParam2, OutParam3, OutParam4);
/* Display OUT parameters */
dbms_output.put_line('OutParam1: ' || OutParam1);
dbms_output.put_line('OutParam2: ' || OutParam2);
dbms_output.put_line('OutParam3: ' || OutParam3);
dbms_output.put_line('OutParam4: ' || OutParam4);
END;
--for setting buffer size needed most of time to avoid `anonymous block completed` message
set serveroutput on size 30000;
-- declaration block in case output need to catch
DECLARE
--declaration for in and out parameter
V_OUT_1 NUMBER;
V_OUT_2 VARCHAR2(200);
BEGIN
--your stored procedure name
schema.package.procedure(
--declaration for in and out parameter
V_OUT_1 => V_OUT_1,
V_OUT_2 => V_OUT_2
);
V_OUT_1 := V_OUT_1;
V_OUT_2 := V_OUT_2;
-- console output, no need to open DBMS OUTPUT seperatly
-- also no need to print each output on seperat line
DBMS_OUTPUT.PUT_LINE('Ouput => ' || V_OUT_1 || ': ' || V_OUT_2);
END;
Creating Pl/SQL block can be painful if you have a lot of procedures which have a lot of parameters. There is an application written on python that do it for you.
It parses the file with procedure declarations and creates the web app for convenient procedure invocations.
var out_para_name refcursor;
execute package_name.procedure_name(inpu_para_val1,input_para_val2,... ,:out_para_name);
print :out_para_name;