Reputation: 1656
I have a table with many columns, and what I would like to do is duplicate all of the rows in the table, but also update one of the columns to a new value.
For example lets say I have the table below. I want to add to my table a duplicate of each row, except instead of BASIC access, it will have 'ADVANCED':
Before:
NAME, GENDER, ACCESS
----------------------
STEVE, MALE, BASIC
MOLLY, FEMALE, BASIC
After
NAME, GENDER, ACCESS
----------------------
STEVE, MALE, BASIC
MOLLY, FEMALE, BASIC
STEVE, MALE, ADVANCED
MOLLY, FEMALE, ADVANCED
Is there a way to do this without specifying all columns? I have 60 columns in the table, and the structure can change (meaning columns may be added, removed, renamed, etc).
Is it possible in Oracle SQL to automate this?
Upvotes: 0
Views: 130
Reputation: 442
Try below method with anonymous block to avoid listing columns in insert statements
CREATE TABLE ACCESS_CHN
(NAAME VARCHAR2(100),
GENDER VARCHAR2(20),
ACCCESS VARCHAR2(30))
INSERT into ACCESS_CHN values('STEVE','MALE','BASIC');
INSERT into ACCESS_CHN values('MOLLY','FEMALE','BASIC');
COMMIT;
DECLARE
column_list varchar2(2000):=NULL;
plsql_block VARCHAR2(1000);
BEGIN
select LISTAGG(column_name,',') within group (order by column_id)
into column_list
from user_tab_columns
where table_name='ACCESS_CHN';
plsql_block := 'CREATE TABLE ACCESS_CHN_BKP as select '|| column_list || ' from ACCESS_CHN';
EXECUTE IMMEDIATE plsql_block;
plsql_block := 'UPDATE ACCESS_CHN_BKP set ACCCESS=''ADVANCED'' ';
EXECUTE IMMEDIATE plsql_block;
COMMIT;
plsql_block := 'CREATE TABLE ACCESS_CHN_FINAL as select * from ACCESS_CHN
union all
select * from ACCESS_CHN_BKP';
EXECUTE IMMEDIATE plsql_block;
END;
--To rerun drop tables ACCESS_CHN_BKP and ACCESS_CHN_FINAL
Upvotes: 0
Reputation: 142705
Without specifying all the columns? With some help of a "temporary" table, here's how:
Your current table:
SQL> create table test
2 (name varchar2(10),
3 gender varchar2(20),
4 caccess varchar2(20));
Table created.
SQL> insert into test
2 select 'steve', 'male', 'basic' from dual union all
3 select 'molly', 'female', 'basic' from dual;
2 rows created.
SQL> create table test_temp as select * From test;
Table created.
SQL> update test_temp set caccess = 'advanced';
2 rows updated.
SQL> insert into test select * From test_temp;
2 rows created.
SQL> drop table test_Temp;
Table dropped.
SQL> select * From test;
NAME GENDER CACCESS
---------- -------------------- --------------------
steve male basic
molly female basic
steve male advanced
molly female advanced
SQL>
Apparently, that works, but - what if the original table is huge? It takes a lot of space, and its copy takes approximately twice as much. Why are you doing that, anyway?
Upvotes: 1
Reputation: 1269703
Just use insert . . . select
:
insert into t (name, gender, access)
select name, gender, 'ADVANCED'
from t;
You need to list all the columns. You can shorten the manual process by using a query to generate the list. If you had to do this a lot and always knew you were leaving out access
and access
is the last column, you could use a view:
create view v_t as
select . . . -- all but access
from t;
insert into t ( . . . )
select v.*, 'ADVANCED'
from v_t;
Or you could use dynamic SQL to generate the statement.
However, I don't recommend any of those. Instead I would be concerned about a data model where you are regularly adding and modifying the columns in a table. That sounds dangerous.
Upvotes: 2