In oracle procedure how to fetch the PKId just added to be added as FK ID for next table - sql

Question

In an oracle procedure, I need to insert data in EmployeeHeader table and then later insert the PK id of this table as the FK id of EmployeeDetails table. How can we achieve this?
INSERT INTO EmployeeHeader(
HEADER_PK_ID
empNo
)
VALUES(
HEADER_PK_ID_SEQ.NEXTVAL,
'SOMETHING'
);
INSERT INTO EmployeeDetails (
DTLHEADER_PK_ID,
HEADER_fK_ID
empname,
age
)
VALUES(
DTLHEADER_PK_ID_SEQ.NEXTVAL,
HEADER_PK_IDn, -- (THIS NEEDS TO BE FETCHED FROM EmployeeHeader)
21
);

Answer 1

You can use currval in most cases:
select HEADER_PK_ID_SEQ.CURRVAL
from dual;
You might need to wrap the two inserts in a single transaction, if you want the values to be safe for concurrent inserts.

Answer 2

Use the RETURNING clause of the INSERT statement:
DECLARE
nHeader_pk_id NUMBER;
BEGIN
INSERT INTO EmployeeHeader
(HEADER_PK_ID, EMPNO)
VALUES
(HEADER_PK_ID_SEQ.NEXTVAL, 'SOMETHING')
RETURNING HEADER_PK_ID INTO nHeader_pk_id;
INSERT INTO EmployeeDetails
(DTLHEADER_PK_ID, HEADER_FK_ID, EMPNAME, AGE)
VALUES
(DTLHEADER_PK_ID_SEQ.NEXTVAL, nHeader_pk_id, 'Somebody', 21);
END;
My personal preference is to use ON INSERT triggers to handle the population of primary key fields, in the following manner:
CREATE OR REPLACE TRIGGER EMPLOYEEHEADER_BI
BEFORE INSERT ON EMPLOYEEHEADER
FOR EACH ROW
BEGIN
IF :NEW.HEADER_PK_ID IS NULL THEN
:NEW.HEADER_PK_ID := HEADER_PK_ID_SEQ.NEXTVAL;
END IF;
END EMPLOYEEHEADER_BI;
CREATE OR REPLACE TRIGGER EMPLOYEEDETAILS_BI
BEFORE INSERT ON EMPLOYEEDETAILS
FOR EACH ROW
BEGIN
IF :NEW.DTLHEADER_PK_ID IS NULL THEN
:NEW.DTLHEADER_PK_ID := DTLHEADER_PK_ID_SEQ.NEXTVAL;
END IF;
END EMPLOYEEDETAILS_BI;
and the INSERT statements become:
DECLARE
nHeader_pk_id NUMBER;
nDtlheader_pk_id NUMBER;
BEGIN
INSERT INTO EmployeeHeader
(EMPNO) -- Note: PK field not mentioned - will be populated by trigger
VALUES
('SOMETHING')
RETURNING HEADER_PK_ID INTO nHeader_pk_id;
INSERT INTO EmployeeDetails
(HEADER_FK_ID, EMPNAME, AGE) -- Note: PK field not mentioned - will be populated by trigger
VALUES
(nHeader_pk_id, 'Somebody', 21)
RETURNING DTLHEADER_PK_ID INTO nDtlheader_pk_id;
END;
(I use the IF pk_field IS NULL THEN construct because I often need to copy data from production to development databases and wish to preserve any key values pulled from production to simplify debugging. If you don't have this requirement you can eliminated the IS NULL check and just assign the sequence's NEXTVAL directly to the column).
Done in this manner the application code doesn't need to know or care about which sequence is used to generate the primary key value for a particular table, and the primary key field is always going to end up populated.
Best of luck.

Answer 3

You can use the RETURNING clause.
Here's an example:
create sequence test_seq1
start with 1
maxvalue 999
minvalue 1
nocycle
cache 20
noorder;
create sequence test_seq2
start with 100
maxvalue 999
minvalue 1
nocycle
cache 20
noorder;
create table test_tab_p (col1 number, col2 number);
create table test_tab_c (col1 number, col2 number, col3 number);
declare
v_p_col2 number := 1;
v_c_col3 number := 10;
v_p_col1 number;
begin
insert into test_tab_p (col1, col2)
values (test_seq1.nextval, v_p_col2)
returning col1 into v_p_col1;
insert into test_tab_c (col1, col2, col3)
values (test_seq2.nextval, v_p_col1, v_c_col3);
commit;
end;
/
select * from test_tab_p;
COL1 COL2
---------- ----------
1 1
select * from test_tab_c;
COL1 COL2 COL3
---------- ---------- ----------
100 1 10

Answer 4

Use currval
INSERT INTO EmployeeHeader
(header_pk_id, empNo)
VALUES
(header_pk_id_seq.nextval, 'SOMETHING');
INSERT INTO EmployeeDetails
(dtlheader_pk_id, header_fk_id, empname, age)
VALUES
(dtlheader_pk_id_seq.nextval, header_pk_id_seq.currval, 21);
currval is safe to use here. It always returns the last value obtained by nextval for the current connection. So even if other transactions (which means a different connection) calls nextval between those two statements, currval will still reflect the value of "this" nextval call.
Quote from the manual:
Each user that references a sequence has access to his or her current sequence number, which is the last sequence generated in the session. A user can issue a statement to generate a new sequence number or use the current number last generated by the session. After a statement in a session generates a sequence number, it is available only to this session.

Answer 5

The very efficient way will be just save the seq id value into a variable and then use it for the whole transaction. Below is the mentioned example. Let me know for any issues.
DECLARE
lv_pkid PLS_INTEGER:=SEQ.NEXTVAL;
BEGIN
INSERT
INTO EmployeeHeader
(
HEADER_PK_ID,
empNo
)
VALUES
(
lv_pkid,
'SOMETHING'
);
INSERT
INTO EmployeeDetails
(
DTLHEADER_PK_ID,
HEADER_fK_ID,
empname,
age
)
VALUES
(
DTLHEADER_PK_ID_SEQ.NEXTVAL,
lv_pkid, -- (THIS NEEDS TO BE FETCHED FROM EmployeeHeader)
21
);
COMMIT;
END;