I want to save the SQL create statement as following:
CREATE TABLE tbl AS
SELECT
*
FROM
tbl_info;
Is there a way to save the previous query in a table like this:
table_name query
tbl CREATE TABLE tbl AS SELECT * FROM tbl_info;
which table_name, query are the columns of this table,
i tried to look up for triggers but there is nothing about trigger before or after on create statement,
please your help.
Indeed there is a way of doing this, but it is not recommendable. Oracle offers out of the box audit capabilities much better than any custom solution you can think of.
Having said that, one option would be to use a DDL TRIGGER. In the example below you have an audit table to store the create event and a trigger to record them.
Keep in mind that I use ON SCHEMA, so it will only affect the CREATE events for the schema which the trigger belongs to.
Base code
CREATE TABLE AUDIT_DDL (
D DATE,
OSUSER VARCHAR2(255),
CURRENT_USER VARCHAR2(255),
SYSEVENT VARCHAR2(30),
STATEMENTS VARCHAR2(1000)
);
CREATE OR REPLACE TRIGGER AUDIT_DDL_TRG
AFTER DDL ON SCHEMA
DECLARE
sql_text ora_name_list_t;
v_stmt VARCHAR2(2000);
n PLS_INTEGER;
BEGIN
n := ora_sql_txt(sql_text);
FOR i IN 1 .. n LOOP
v_stmt := v_stmt || sql_text(i);
END LOOP;
v_stmt :=regexp_replace(v_stmt,
'rename[[:space:]]+.*[[:space:]]+to[[:space:]]+([a-z0-9_]+)',
'\1',
1,
1,
'i');
IF (ORA_SYSEVENT = 'CREATE')
THEN
INSERT INTO AUDIT_DDL
(D,
OSUSER,
CURRENT_USER,
SYSEVENT,
STATEMENTS)
VALUES
(SYSDATE,
SYS_CONTEXT('USERENV', 'OS_USER'),
SYS_CONTEXT('USERENV', 'CURRENT_USER'),
ORA_SYSEVENT,
v_stmt);
END IF;
END;
/
Let's check how it works
sqlplus test1/Oracle_123
SQL*Plus: Release 19.0.0.0.0 - Production on Sun Oct 3 11:13:21 2021
Version 19.6.0.0.0
Copyright (c) 1982, 2019, Oracle. All rights reserved.
Last Successful login time: Wed Sep 22 2021 08:08:57 +02:00
Connected to:
Oracle Database 19c Enterprise Edition Release 19.0.0.0.0 - Production
Version 19.6.0.0.0
SQL> create table t1 as select * from all_objects ;
create table t1 as select * from all_objects
*
ERROR at line 1:
ORA-00955: name is already used by an existing object
SQL> drop table t1 purge;
Table dropped.
SQL> create table t1 as select * from all_objects ;
Table created.
SQL> select d,statements from AUDIT_DDL ;
D
---------
STATEMENTS
--------------------------------------------------------------------------------
03-OCT-21
create table t1 as select * from all_objects
As you can see above ( on purpose ) , I did several statements ( a create failing, a drop table, and finally a CTAS statement ). However, our event trigger is only looking for the sysevent create, and because it is after ddl on schema, it will only store the data after the command is executed successfully.
You can store many other properties of the default context sys_context.
Nevertheless, this has a huge impact in performance and it is not recommendable
Related
On an Oracle 19c DB I need to sample on customers whilst keeping a uniorm distribution to model the moving of customers.
There are (on purpose) multiple rows with the same customers but changeing adderssses to model moving. Thus, for the sampling i need to group first.
Now what I got is that for the SAMPLE clause the source has to be materialized. So in the PL/SQL script I generate a temporary table that i want to use afterwards with SAMPLE.
But even a simple SELECT INTO afterwards doens't work.
set SERVEROUT on;
DECLARE
v_cust_name VARCHAR2(100);
cmd_creation VARCHAR2(500):='CREATE PRIVATE TEMPORARY TABLE ORA$PTT_temp_cust AS(
SELECT cust_id, MIN(name) as name
FROM customers
GROUP BY cust_id)';
BEGIN
EXECUTE IMMEDIATE cmd_creation;
dbms_output.put_line('temp created');
SELECT name
INTO v_cust_name
FROM (SELECT *
FROM ORA$PTT_temp_cust SAMPLE(5))
WHERE ROWNUM =1;
EXECUTE IMMEDIATE 'DROP TABLE ORA$PTT_temp_cust';
dbms_output.put_line('temp dropped');
END;
What I get is this
ORA-06550: line 15, column 18:
PL/SQL: ORA-00942: table or view does not exist
The table gets created. That far I got when I only executed the String for the creation and nothing else. Then I can access the table in a desired script from a different point.
Does this have to do with the compiling? Is there some different way to solve this?
Your PL/SQL Block simple does not compile as the table you query does not exists at the time of compilation.
You must perform event the query with execute immediate
Simplified Example
DECLARE
v_cust_name VARCHAR2(100);
cmd_creation VARCHAR2(500):='CREATE PRIVATE TEMPORARY TABLE ORA$PTT_temp_cust AS select * from dual';
BEGIN
EXECUTE IMMEDIATE cmd_creation;
dbms_output.put_line('temp created');
EXECUTE IMMEDIATE 'SELECT DUMMY FROM ORA$PTT_temp_cust' INTO v_cust_name;
dbms_output.put_line('name' || v_cust_name);
EXECUTE IMMEDIATE 'DROP TABLE ORA$PTT_temp_cust';
dbms_output.put_line('temp dropped');
END;
/
An other caveat the can lead to ORA-00942: table or view does not existin your setup is that you performs commit in your script.
The default definition of the ON COMMIT clause is DROP DEFINITION so you must use
CREATE PRIVATE TEMPORARY TABLE ORA$PTT_temp_cust
ON COMMIT PRESERVE DEFINITION
...
Dynamic SQL is evil. The fact that you created the table using dynamic SQL (your 1st execute immediate) doesn't mean that Oracle "predicts" you'll actually succeed with that and "presume" that statements that follow are correct. Nope - that table doesn't exist yet, so everything has to be moved to dynamic SQL.
Something like this (disregard changes in table and column names I used and global vs. private temporary table; this is 11gXE):
SQL> DECLARE
2 v_cust_name VARCHAR2 (100);
3 cmd_creation VARCHAR2 (500)
4 := 'CREATE global TEMPORARY TABLE PTT_temp_cust AS
5 SELECT empno, MIN(ename) as name
6 FROM emp
7 GROUP BY empno';
8 BEGIN
9 EXECUTE IMMEDIATE cmd_creation;
10
11 EXECUTE IMMEDIATE '
12 SELECT max(name)
13 FROM (SELECT *
14 FROM PTT_temp_cust SAMPLE(5))
15 WHERE ROWNUM = 1'
16 INTO v_cust_name;
17
18 EXECUTE IMMEDIATE 'DROP TABLE PTT_temp_cust';
19
20 DBMS_OUTPUT.put_line ('Result = ' || v_cust_name);
21 END;
22 /
Result =
PL/SQL procedure successfully completed.
SQL>
I got no result, though - but you should (at least, I hope so).
I have a trigger whose purpose is to fire whenever there is a DELETE on a particular table and insert the deleted data into another table in json format.
The trigger works fine if I am specifying each column explicitly. Is there any way to access the entire table row?
This is my code.
TRIGGER1
AFTER DELETE
ON QUESTION
FOR EACH ROW
DECLARE
json_doc CLOB;
BEGIN
select json_arrayagg (
json_object ('code' VALUE :old.id,
'name' VALUE :old.text,
'description' VALUE :old.text) returning clob
) into json_doc
from dual;
PROCEDURE1(json_doc);
END;
This works fine. However, what I want is something like this. Instead of explicity specifying each column, I want to convert the entire :OLD data
TRIGGER1
AFTER DELETE
ON QUESTION
FOR EACH ROW
DECLARE
json_doc CLOB;
BEGIN
select json_arrayagg (
json_object (:old) returning clob
) into json_doc
from dual;
PROCEDURE1(json_doc);
END;
Any suggestion please.
The short and correct answer is you can't. We have a few tables in our application where we do this and the developer is responsible for updating the trigger when they add a column: this is enforced with code reviews and is probably the cleanest solution for this scenario.
The long answer is you can get close, but I wouldn't do this in production for several reasons:
Triggers are terrible for performance
Triggers are terrible for code clarity
This requires reading the row again using flashback query so
You aren't getting the values of this row from inside your current transaction: if you update the row in your transaction and then delete it the JSON will show what the values were BEFORE your update
There is a performance penalty for reading from UNDO
There is potential that UNDO won't be available and your trigger will fail
Your user needs permission to execute flashback queries
Your database needs to meet all the perquisites to support flashback queries
Deleting a lot of rows will cause the ROWID collection to get large and consume PGA
There are probably more reasons, but in the interest of "can it be done" here you go...
DROP TABLE t1;
DROP TABLE t2;
DROP TRIGGER t1_ad;
CREATE TABLE t1 (
id NUMBER,
name VARCHAR2(100),
description VARCHAR2(100)
);
CREATE TABLE t2 (
dt TIMESTAMP(9),
json_data CLOB
);
INSERT INTO t1 VALUES (1, 'A','aaaa');
INSERT INTO t1 VALUES (2, 'B','bbbb');
INSERT INTO t1 VALUES (3, 'C','cccc');
INSERT INTO t1 VALUES (4, 'D','dddd');
CREATE OR REPLACE TRIGGER t1_ad
FOR DELETE ON t1
COMPOUND TRIGGER
TYPE t_rowid_tab IS TABLE OF ROWID;
v_rowid_tab t_rowid_tab := t_rowid_tab();
AFTER EACH ROW IS
BEGIN
v_rowid_tab.extend;
v_rowid_tab(v_rowid_tab.last) := :old.rowid;
END AFTER EACH ROW;
AFTER STATEMENT IS
v_scn v$database.current_scn := dbms_flashback.get_system_change_number;
v_json_data CLOB;
v_sql CLOB;
BEGIN
FOR i IN 1 .. v_rowid_tab.count
LOOP
SELECT 'SELECT json_arrayagg(json_object(' ||
listagg('''' || lower(t.column_name) || ''' VALUE ' ||
lower(t.column_name),
', ') within GROUP(ORDER BY t.column_id) || ') RETURNING CLOB) FROM t1 AS OF SCN :scn WHERE rowid = :r'
INTO v_sql
FROM user_tab_columns t
WHERE t.table_name = 'T1';
EXECUTE IMMEDIATE v_sql
INTO v_json_data
USING v_scn, v_rowid_tab(i);
INSERT INTO t2
VALUES
(current_timestamp,
v_json_data);
END LOOP;
END AFTER STATEMENT;
END t1_ad;
/
UPDATE t1
SET NAME = 'zzzz' -- not captured
WHERE id = 2;
DELETE FROM t1 WHERE id < 3;
SELECT *
FROM t2;
-- 13-NOV-20 01.08.15.955426000 PM [{"id":1,"name":"A","description":"aaaa"}]
-- 13-NOV-20 01.08.15.969755000 PM [{"id":2,"name":"B","description":"bbbb"}]
How to truncate any table using its synonym in oracle?
-- in Server_A
Create Table table_a ( col int);
-- in server_B
CREATE SYNONYM syn_table_a FOR table_a#dblink_server_a;
--insert into
INSERT INTO syn_table_a values (1);
--Truncate
How to truncate table using synonym only?.
A truncate statement cannot be used on a synonym.
Synonyms cannot be used in a drop table, drop view or truncate
table/cluster statements. If this is tried, it results in a ORA-00942:
table or view does not exist
For example,
SQL> CREATE TABLE t(col NUMBER);
Table created.
SQL>
SQL> CREATE SYNONYM t_syn FOR t;
Synonym created.
SQL>
SQL> TRUNCATE TABLE t_syn;
TRUNCATE TABLE t_syn
*
ERROR at line 1:
ORA-00942: table or view does not exist
SQL>
You could use dynamic SQL to do it, e.g.:
declare
d varchar2(1000);
begin
select 'TRUNCATE TABLE "' || table_owner || '"."' || table_name || '"'
into d
from all_synonyms
where synonym_name = 'MYSYNONYM';
execute immediate d;
end;
If the table is accessed via a database link, this will not work. In that case, you could create a procedure on the remote instance that does the truncate, then call that procedure across the database link, e.g.
begin
truncate_my_table#dblinkname;
end;
In Oracle, you can also get ORA-14410 while trying to drop/truncate a table using synonym.
The alert log:
ORA-00604: error occurred at recursive SQL level 1
ORA-14410: RPI LOCK TABLE issued to table referenced through synonym
Follow the above dynamic SQl to drop/truncate it.
I want to use an if statement inside trigger but the value if comparison will come from an other select statement.
I have done the following:
create or replace
Trigger MYTRIGGER
After Insert On Table1
Referencing Old As "OLD" New As "NEW"
For Each Row
Begin
Declare Counter Int;
Select Count(*) From Table2 Where Table2."Email" = :New.U_MAIL Into Counter;
IF Counter < 1 THEN
//INSERT Statement here...
END IF;
End;
My logic is simple, if same email user exists, insert will not work.
Above code did not work. How can we do this?
A few syntax errors. Would be closer to something like this:
create or replace
Trigger MYTRIGGER
After Insert On Table1
Referencing Old As "OLD" New As "NEW"
For Each Row
DECLARE
v_count NUMBER;
BEGIN
SELECT COUNT(*)
INTO v_count
FROM Table2
WHERE Email = :New.U_MAIL
;
IF v_count > 0
THEN
RAISE_APPLICATION_ERROR(-20000, 'Not inserted...');
END IF;
END;
Your approach is wrong. Referential integrity should not be made using triggers, it just cannot work as required. See example:
Connected to Oracle Database 12c Enterprise Edition Release 12.1.0.2.0
Connected as test#soft12c1
SQL> create table mail_1 (email varchar2(100));
Table created
SQL> create table mail_2 (email varchar2(100));
Table created
SQL> create trigger mail_1_check
2 before insert on mail_1
3 for each row
4 declare
5 cnt integer;
6 begin
7 select count(*) into cnt from mail_2 where email = :new.email;
8 if cnt > 0 then
9 raise_application_error(-20100, 'Email already exists');
10 end if;
11 end;
12 /
Trigger created
SQL> insert into mail_2 values ('president#gov.us');
1 row inserted
SQL> insert into mail_1 values ('king#kingdom.en');
1 row inserted
SQL> insert into mail_1 values ('president#gov.us');
ORA-20100: Email already exists
ORA-06512: at "TEST.MAIL_1_CHECK", line 6
ORA-04088: error during execution of trigger 'TEST.MAIL_1_CHECK'
It looks like trigger works right, but it's not true. See what happens when several users will works simultaneously.
-- First user in his session
SQL> insert into mail_2 values ('dictator#country.by');
1 row inserted
-- Second user in his session
SQL> insert into mail_1 values ('dictator#country.by');
1 row inserted
-- First user is his session
SQL> commit;
Commit complete
-- Second user is his session
SQL> commit;
Commit complete
-- Any user in any session
SQL> select * from mail_1 natural join mail_2;
EMAIL
--------------------------------------------------------------------------------
dictator#country.by
If using triggers for this task, you should serialize any attempts to use this data, say, execute LOCK TABLE IN EXCLUSIVE MODE unless commit. Generally it's a bad decision. For this concrete task you can use much better approach:
Connected to Oracle Database 12c Enterprise Edition Release 12.1.0.2.0
Connected as test#soft12c1
SQL> create table mail_1_2nd(email varchar2(100));
Table created
SQL> create table mail_2_2nd(email varchar2(100));
Table created
SQL> create materialized view mail_check
2 refresh complete on commit
3 as
4 select 1/0 data from mail_1_2nd natural join mail_2_2nd;
Materialized view created
OK. Let's see, what if we try to use same email:
-- First user in his session
SQL> insert into mail_1_2nd values ('dictator#gov.us');
1 row inserted
-- Second user in his session
SQL> insert into mail_2_2nd values ('dictator#gov.us');
1 row inserted
SQL> commit;
Commit complete
-- First user in his session
SQL> commit;
ORA-12008: error in materialized view refresh path
ORA-01476: divisor is equal to zero
SQL> select * from mail_1_2nd natural join mail_2_2nd;
EMAIL
--------------------------------------------------------------------------------
no rows selected
I want to create a simple function based index on a simple table but i get error.
So, first of all I created a function
CREATE OR REPLACE FUNCTION promo_function(p_promo_category VARCHAR2)
RETURN VARCHAR2 DETERMINISTIC IS
BEGIN
RETURN UPPER(p_promo_category);
END promo_function;
Then I would execute this, but fails
CREATE INDEX promotions_fbi
ON SH.PROMOTIONS (promo_function (promo_category));
Why? The error is ORA-00904:"PROMO_FUNCTION": Invalid identifier
But the function works well in a query:
SELECT *
FROM sh.sales s,
sh.promotions p,
sh.times t
WHERE s.promo_id = p.promo_id
AND s.time_id = t.time_id
AND t.time_id BETWEEN DATE '2000-01-01' AND DATE '2000-03-31'
AND promo_function(p.promo_category) = 'AD NEWS';
Many thanks!
There is nothing inherently wrong with your code. I can create a function-based index like this:
SQL> create table promotions (promo_category varchar2(10))
2 /
Table created.
SQL> CREATE OR REPLACE FUNCTION promo_function
2 (p_promo_category in VARCHAR2)
3 RETURN VARCHAR2 DETERMINISTIC
4 IS
5 BEGIN
6 RETURN UPPER(p_promo_category);
7 END promo_function;
8 /
Function created.
SQL> CREATE INDEX promotions_fbi
2 ON PROMOTIONS (promo_function (promo_category));
Index created.
SQL>
The only difference between my code and yours is that I don't prefix the table in the CREATE INDEX statement. Everything is in the same schema, so I don't need to.
So, can I re-create your scenario? Here's one way. I drop the index and function, then give another uses all privileges on teh table...
SQL> drop index promotions_fbi;
Index dropped.
SQL> drop function PROMO_FUNCTION;
Function dropped.
SQL> grant all on promotions to B;
Grant succeeded.
SQL>
As that user I can create a normal index ...
SQL> conn b/b
Connected.
SQL> select * from apc.promotions;
no rows selected
SQL> CREATE INDEX promotions_i
2 ON APC.PROMOTIONS (promo_category);
Index created.
SQL>
However, if I create a function I cannot create a function-based index using it....
SQL> conn b/b
Connected.
SQL> CREATE INDEX promotions_fbi
2 ON APC.PROMOTIONS (promo_function (promo_category));
ON APC.PROMOTIONS (promo_function (promo_category))
*
ERROR at line 2:
ORA-00904: : invalid identifier
SQL>
The invalid identifier fingers the function name. Why? Because although schema B would own the index schema APC owns the table, and needs to be able to execute the function too.
The solution is to grant execute rights on the function to the table owner:
SQL> conn b/b
Connected.
SQL> grant execute on promo_function to APC;
Grant succeeded.
SQL> CREATE INDEX promotions_fbi
2 ON APC.PROMOTIONS (B.promo_function (promo_category));
Index created.
SQL>
Note that we must explicitly reference the function owner as well as the table owner in this statement. It's a bit nasty, and that's why it's generally a bad idea to spread privileges across two schemas in this fashion.
Not sure how #zaratustra gets their findings, as I can definitely create function-based in indexes using the word FUNCTION in the name...
SQL> r
1 select i.table_owner, i.owner as index_owner, i.index_name
2 , i.index_type, e.column_expression
3 from all_indexes i
4 left join all_ind_expressions e
5 on i.owner = e.index_owner
6 and i.index_name = e.index_name
7* where i.table_name = 'PROMOTIONS'
TABLE_OWNER INDEX_OWNER
------------------------------ ------------------------------
INDEX_NAME INDEX_TYPE
------------------------------ ---------------------------
COLUMN_EXPRESSION
--------------------------------------------------------------------------------
APC APC
PROMOTIONS_FBI FUNCTION-BASED NORMAL
"APC"."PROMO_FUNCTION"("PROMO_CATEGORY")
APC A
PROMO_B_I FUNCTION-BASED NORMAL
"A"."B_FUNCTION"("PROMO_CATEGORY")
APC APC
PROMOTIONS_I NORMAL
SQL>
Although I am on a different point release so that might explain it
SQL> select banner from v$version;
BANNER
--------------------------------------------------------------------------------
Oracle Database 11g Enterprise Edition Release 11.2.0.2.0 - Production
SQL>
During my experiments I realized that Oracle can't let you create function-based index with the function word in the name of the procedure. Looks like a bug for me:
CREATE OR REPLACE FUNCTION func_function(p_promo_category IN VARCHAR2)
RETURN VARCHAR2 DETERMINISTIC IS
BEGIN
RETURN UPPER(p_promo_category);
END func_function;
create table t1 (
promo_category varchar2(4000)
);
Table created
CREATE INDEX promotions_fbi ON t1 (func_function (promo_category));
ORA-00911: invalid character
Let's create a function without the function word in the name:
CREATE OR REPLACE FUNCTION func_functio(p_promo_category IN VARCHAR2)
RETURN VARCHAR2 DETERMINISTIC IS
BEGIN
RETURN UPPER(p_promo_category);
END func_functio;
CREATE INDEX promotions_fbi ON t1 (func_functio (promo_category));
Index created
select index_name, index_type
from user_indexes
where lower(index_name) = 'promotions_fbi'
INDEX_NAME INDEX_TYPE
-------------------------------------
PROMOTIONS_FBI FUNCTION-BASED NORMAL
My Oracle version:
select banner from v$version
BANNER
----------------------------------------------------------------------------
Oracle Database 11g Enterprise Edition Release 11.2.0.4.0 - 64bit Production