PL/SQL or Trigger

Hi guys,
Somebody help me to write trigger or pl/sql which automatically adds record in a table on the 1st day of every year.
Thanks

You can do it by
DBMS_JOB package or you can simply do it by korne shell script in unix as follows.
create newyrsql.ksh on unix prompt by following.
vi newyrsql.ksh
#!/bin/ksh
sqlplus -s scott/tiger@test << EOF
create table emp2006 as select * from emp;
-- OR next one
insert into emp (select * from empold);
exit
save the file.
create "newyrjob" file as follows on unix prompt
vi newyrjob
0 1 1 1 * newyrsql.ksh
save the file
Now add the job in crontab by following
crontab newyrjob
That's it, this way this job will run at 1 am on 1st january on every year
JP

Similar Messages

Can I call a Java program from a SQL Server Trigger?

Hello,
I want to encrypt some data in a database column in SQL Server. Today I am using java code to encrypt the value and store it in the database using JDBC.
Now I want to use a VB client to store the encrypted value in the SQL Server DB. Since the encryption is handled by a java class, can I write a trigger in SQL Server that while inserting the raw data, calls the java class for encrypting the value and then inserts the encrypted value into the column?
In general, is it possible to call a java class from a SQL Server trigger?
Thanks
Bipin

Here are 3 examples of code for insert, update and delete:
CREATE TRIGGER [PLI_INSERT_TRIGGER] ON [dbo].[PLI]
FOR INSERT
AS
Declare @cmd sysname, @code sysname, @list sysname
     Select @code = PLI_K_COD, @list = PLI_K_LISTINO from inserted
     Set @cmd = 'java mirrorDb.Copy PLI INSERT ' + @code + ' ' + @list
     EXEC master..xp_cmdshell @cmd
CREATE TRIGGER [PLI_UPDATE_TRIGGER] ON [dbo].[PLI]
FOR UPDATE
AS
Declare @cmd sysname, @code sysname, @list sysname
     Select @code = PLI_K_COD, @list = PLI_K_LISTINO from inserted
     Set @cmd = 'java mirrorDb.Copy PLI UPDATE ' + @code + ' ' + @list
     EXEC master..xp_cmdshell @cmd
CREATE TRIGGER [PLI_DELETE_TRIGGER] ON [dbo].[PLI]
FOR DELETE
AS
Declare @cmd sysname, @code sysname, @list sysname
     Select @code = PLI_K_COD, @list = PLI_K_LISTINO from deleted
     Set @cmd = 'java mirrorDb.Copy PLI DELETE ' + @code + ' ' + @list
     EXEC master..xp_cmdshell @cmd
you must go "sql server entreprise manager" right click on the table you want to add triggers and select: all activities, manage triggers.
You have 3 examples: for an insert, for an update and for a delete
ON [dbo].[PLI] specify the table on which you want to setup trigger.
FOR DELETE, INSERT, UPDATE specify the event.
The Declare statement create the variables in which I want to put some values to pass to the java program, for example which table, which event, which key fields.
the "Select @code = PLI_K_COD, @list = PLI_K_LISTINO from inserted" set the variables with the value of the columns of the table I am interested to read from my java program, for example the variable @code receive the value of the column pli_k_kod (is the key) of the table PLI.
The "Set @cmd = 'java mirrorDb.Copy PLI DELETE ' + @code + ' ' + @list " prepared the variable @cmd with the java command followed by the package.classname and parameters.
The EXEC launch the command to the operating system.
Daniele

Executing dynamic SQL in trigger

Hi !
I just wanted to ask the community if it is ok using dynamic SQL
with execute immediate in triggers, like:
if (deleting or updating)
and nvl(:old.col_val, 0) != 0
then
/* Update */
sSql :=
'update TMP_STAT_TEST ' ||
'set col_val_' || to_char(:old.col_id) || ' = col_val_' ||
to_char(:old.col_id) || ' - :1 ';
execute immediate sSql using :old.col_val;
end if;
All dyn SQL inside the trigger does not reference the triggered
table, so I will not get a mutating trigger.
Thanks for your feedback.
Stefan.

Hi Andrew !
I have a "master" table with date and amount and some more
columns. I have a statistic table which must get updated
accordingly each time the master table is changed (ins, upd,
del). In this stats table there a cols like "sum_2000, sum_2001"
etc. The dynamic statement depends on date to figure out the
column name for the sum_yyyy cols, and amount will be
added/subtracted. Of course there is a check that the trigger
cannot handle when the year of the date is smaller than 2000 or
greater than 2009.
That's why I want to use dyn SQL.
Stefan.

MS SQL-server trigger

Hi,
I wonder if there is anyone who know if there is any method as listen on a table in a ms sql-server. I want my program to run if a new row appear in the database.
/Rikard

Why would you need to do that? Put all theprocessing
into the trigger, that's where it belongs.Possibly a bad assumption on my part, but since this
is op posted in a java forum and said:
"I want my program to run if a new row appear in the
database"
I assumed he has a program and that it is probably
made with Java.
I think nasch_ meant basically that perhaps the business logic that this 'app' does, really belongs in the database trigger.

PowerShell and SQL Server Trigger

I had the following trigger written but apparently our IT department has not set up server/account or whatever needs to be done to use SQL Server's DB Mail.
ALTER TRIGGER [dbo].[Faxiles_Trigger]
ON [dbo].[DummyFQF]
AFTER INSERT
AS
BEGIN
DECLARE @tableHTML NVARCHAR(MAX) ;
SET @tableHTML =
N'<H1>Fax Error Report</H1>' +
N'<table border="1">' +
N'<tr><th>[FileID]</th>
<th>[FaxID]</th>
<th>[FilePath]</th></tr>' +
CAST ( (
SELECT
td=T1.[FileID],' ',
td=T1.[FaxID],' ',
td=T1.[FilePath],' '
FROM
FaxQueueFiles as T1
INNER JOIN
FaxQueue as T2
ON
T1.[FaxID] = T2.[FaxID]
WHERE
T2.[DateAdded] between Dateadd(hour,-1,GETDATE()) and Getdate()
AND
T1.[FilePath] NOT LIKE 'C:\%'
FOR XML PATH('tr'), TYPE
) AS NVARCHAR(MAX) ) +
N'</table>' ;
EXEC msdb.dbo.sp_send_dbmail
@recipients='[email protected]',
@subject = 'Fax Error Report',
@body = @tableHTML,
@body_format = 'HTML';
END
How can I instead make the trigger envoke/call a powershell script that can perform the same query and send an e-mail?
Apparently we do have e-mail functionality for our scripts, so that part I can just copy/paste from a pre-existing script. What I don't know how to do (since I am brand new to powershell scripting) is save the results of a query into a HTML-f0rmatted table
as the body of the e-mail like I did with SQL.
Thanks in advance for any assistance you can provide.
# Connect to the database.
$dbConnSites = New-Object System.Data.SqlClient.SqlConnection("Data Source=localhost;Initial Catalog=Main;Integrated Security=SSPI");
# Open database connection.
$dbConnSites.Open();
# Query String
$dbQry_BadTIFs = "
SELECT
T1.[FileID], T1.[FaxID], T1.[FilePath]
FROM
Main.FaxQueueFiles as T1
INNER JOIN
FaxQueue as T2
ON
T1.FaxID = T2.FaxID
WHERE
T2.[DateAdded] between DATEADD(hour,-1,GETDATE()) and GETDATE()
AND
T1.[FilePath] NOT LIKE 'C:\%'
# Create an SQL Command to execute the above query using the database connection object
$dbGetBadTIFs = New-Object System.Data.SqlClient.SqlCommand(dbQry_BadTIFs, $dbConnSites);
# Reader object to read the results of the executed SQL Command
$dbGetBadTIFsReader = $dbGetBadTIFs.ExecuteReader();
# Loop through the results of GetBadTIFs
while ($dbGetBadTFIsReader.Read())
# Close all the database connections
$dbConnSites.Close();
The above code is what I've come up with so far.... So within my while loop I presume the E-mail will some how be generated?
Should I declare a new string object and concatenate within the loop?
# Start the HTML Table for the e-mail
$str_EmailBody = "
<H1>Fax Error Report<H1>
<TABLE border=`"1`">
<TR>
<TH>[FileID]</TH>
<TH>[FaxID]</TH>
<TH>[FilePath]</TH>
</TR>
</TABLE>
# Loop through the results of GetBadTIFs
# to fill out the HTML Table
while ($dbRdr_GetBadTIFs.Read())
$str_EmailBody += "???";
# Finish the HTML Table for the e-mail
$str_EmailBody += "
</TR>
</TABLE>";
-Nothing to see. Move along.

Hi Blacksaibot,
If you want to query the sql cmd and send eamil with the result, please refer this script, to save the result of the query, plrease try to create a sql datatable:
PowerShell Display Table In HTML Email
If I have any misunderstanding, please let me know.
Best Regards,
Anna

PL/SQL for Trigger can't see a table?

I am trying to write a simple log-in script that limits the number of log-in attempts that are allowed in a period of time. To support this I created a tiny 3 column table to record the failed log-in attempts -- but my WHEN-BUTTON-PRESSED PL/SQL trigger code can't seem to see the table? It always throws a SQL%NOTFOUND exception when I try to use this table. I've tried cnt, logons.cnt, and nw.logons.cnt to identify the columns but nothing has worked.
What am I doing wrong?
DECLARE
dummy1 NUMBER;
BEGIN
SELECT nw.logons.cnt
INTO dummy1
FROM nw.logons
WHERE nw.logons.userid = '2';
LOGONS table in the NW schema
USERID VARCHAR2(6 BYTE)
LASTTIME DATE
CNT NUMBER

Oh GOD, I forgot to COMMIT. That is all it was. Sorry.

SQL 2012 Trigger bulk insert lock

Hi
I have built a proc which basically captures some invalid codes and what I would like to do is use a trigger to fire an email everytime this the table receives information .
;WITH MatchesCTE ( SK_Partial, Matchcode ) AS
(SELECT 1, '1234')
SELECT * INTO match FROM MatchesCTE
ALTER TRIGGER EmailInavildPartials1
   ON [dbo].[match]
   AFTER INSERT
AS
BEGIN
   -- SET NOCOUNT ON added to prevent extra result sets from
   -- interfering with SELECT statements.
   IF (@@rowcount > 1)
   SET NOCOUNT ON;
   --SELECT GETDATE() ;
   EXEC Send_email_Invalid_Partials1
   END
ALTER PROCEDURE [dbo].[Send_email_Invalid_Partials1]
AS
DECLARE @p_body as nvarchar(max), @p_subject as nvarchar(max)
DECLARE @p_recipients as nvarchar(max), @p_profile_name as nvarchar(max), @p_query as nvarchar(max),
@p_subject_date as nvarchar(max), @fileName as nvarchar(max)
SET @p_profile_name = N'DEV'
SET @p_recipients = N'ROBERT@BLAH;'
Select @p_subject_date = (select DATENAME (weekday, getdate ()))+ ' ' + substring (convert(varchar,getdate(),12), 5,2) +'/' + substring (convert(varchar,getdate(),12), 3,2)+ '/' +
   substring (convert(varchar,getdate(),12), 1,2)
SET @filename = 'INVALID MATCH CODES ' + CONVERT(CHAR(8),GETDATE(),112)+ '.csv'
SET @p_subject = @p_subject_date + N' INVALID MATCH CODES'
SET @p_body = 'Please see the invaild Partials List.'
SET @p_query = 'SET NOCOUNT ON;Select * FROM DATABASE..match '
EXEC msdb.dbo.sp_send_dbmail
   @profile_name = @p_profile_name,
   @recipients = @p_recipients,
   @body = @p_body,
   @body_format = 'HTML',
   @subject = @p_subject,
   @query = @p_query,
   @attach_query_result_as_file = 1,
@importance = 'High',
@query_attachment_filename =@filename,
@query_result_separator = ',',
@query_result_no_padding = 1
you''ll need to change
@p_profile_name
@p_recipients
@p_query
when you then run the following
INSERT INTO match
SELECT 1,'1234'
UNION
SELECT 1, '5678'
It locks.    Is there any way around this ? Any help would be great
Thanks

While I can debate the wisdom of this approach in general, as well as the selectivity of your logic with respect to the trigger, the rows inserted, the contents of the entire table - I won't.
The short story is that this approach will not work. Your trigger fires within the context of an insert (presumably - perhaps a merge) statement that inserts some unknown number of rows into the target table. Your trigger logic will then execute
a procedure if more than one (but not one or zero rows) is inserted. Your procedure will then attempt to execute a query against the same table and attempt to dump ALL rows in the table into a file for the email. This "dumping" executes in a different
process - hence the blocking.
Is it correct that your procedure will send the email with an attachment of ALL rows, not just the inserted ones (but ignore the insertion of just one row)? I can't say. It would be unusual but stranger things have been done.   To get
around that you can use a locking hint as suggested - but that may not be what you desire depending on other factors (is anything else manipulating the contents of the table while this process is running? Are you certain of that answer after evaluating the
range of possibilities?).
Perhaps the best approach is to use a very different one. It appears that you insert rows into the table for a single purpose. So why not just turn that process into the one that does the notification?

Another sql query--trigger

I want to write a trigger, usig DUAL table if possible, which shows me the name of a user who make changes in a table.
User names are stored in "users" table (Smith, Jhon,etc,etc) and the trigger that I created looks like this:
CREATE OR REPLACE TRIGGER ITEMS_after_update
AFTER UPDATE
ON ITEMS
FOR EACH ROW
DECLARE
v_username varchar2(50);
BEGIN
-- Find username of person performing UPDATE into table
-- SELECT useFNAME INTO v_username
-- FROM USERS;
-- Insert record into audit table
INSERT INTO new_ITEMS
( ITEMKEY_new,
ITEMKEY,
ITEMNAME_new,
ITEMNAME,SORTGROUP,
SORTGROUP_new )
VALUES
( :new.ITEMKEY,
:OLD.ITEMKEY,
:NEW.ITEMNAME,
:OLD.ITEMNAME,
:new.SORTGROUP,
:OLD.SORTGROUP );
END;
If I'm using SELECT user INTO v_username FROM dual;
instead of SELECT useFNAME INTO v_username FROM USERS I get the "company1" name and doesn't help.I would like to receive Jhon or Smith....
Can anyone help me with this trigger ?
Regards
Robert B

Where exactly are you trying to derive the user from? Based on your query
"SELECT useFNAME INTO v_username FROM USERS;" that would bring back all rows from that table, which is not legitimate when using an INTO clause.
Maybe you are wanting to do something like this?
SELECT useFNAME INTO v_username
from USERS where USERNAME = USER
where USER is an Oracle function returning the user.

Dynamic SQL in trigger

In a before insert/update on each row trigger I do the following:
I loop over the user_tab_columns to get the column names for a table.
In the loop I want to call :new.<columnname> with <columnname> a variable in which I stocked the current columnname.
It has to be dynamic, a direct call to the column does not solve my problem.
How can this be done?
Thanks in advance!
Marianne

Have a look at
re:Dynamic references to trigger vars :new and :old
Cheers,
Alex

SQL 2008 Trigger to handle multi rows scenario

I have created below trigger to start logging the company changes from the table1 into another audit table. It works fine with single row but crashing with identical change with multiple rows. Can you please help me to update the trigger to handle multi-row
scenario. Thanks.
GO
IF
NOT EXISTS
(SELECT
* FROM sys.objects
WHERE object_id
= OBJECT_ID(N'[dbo].[Company_AuditPeriod]')
AND type
in (N'U'))
CREATE
TABLE [dbo].[Company_AuditPeriod](
      [Client] [varchar](25)
NOT NULL,
      [Period] [varchar](25),
      [Table_Name] [varchar](25),
      [Field_Name] [varchar](25),
      [Old_Value] [varchar](25),
      [New_Value] [varchar](25),
      [User_ID] [varchar](25)
      [Last_Update] [datetime],
      [agrtid] [bigint]
IDENTITY(1,1)
NOT NULL,
ON [PRIMARY]
GO
--create trigger
SET
QUOTED_IDENTIFIER ON
GO
CREATE
TRIGGER [dbo].[Table1_Update]
ON [dbo].[Table1]
FOR
UPDATE
NOT
FOR REPLICATION
AS
BEGIN
DECLARE
      @status
varchar(3),
      @user_id
varchar(25),
      @period
varchar(25),
      @client
varchar(25),
      @last_update
datetime
DECLARE
      @Old_status
varchar(3),
      @Old_user_id
varchar(25),
      @Old_period
varchar(25),
      @Old_client
varchar(25)
SELECT
      @status
= status,
      @user_id
= user_id,
      @period
= period,
      @client
= client,
      @last_update
= last_update
FROM Inserted
SELECT
      @Old_status
= status,
      @Old_user_id
= user_id,
      @Old_period
= period,
      @Old_client
= client
FROM Deleted
If @Old_status <> @status
INSERT INTO Company_AuditPeriod
VALUES ( @client, @period,
'Table1',
'period',@old_status, @status, @user_id, @last_update)
END
GO

Sorry for the confusion.
I just made sure the table name is same in sys.objects statement and create table statement (there was a typo)
IF
NOT EXISTS
(SELECT
* FROM sys.objects
WHERE object_id
= OBJECT_ID(N'[dbo].[Company_AuditPeriod]')
AND type
in (N'U'))
CREATE
TABLE [dbo].[ Company_AuditPeriod](
Earlier you created Trigger on Company_AuditPeriod but
We have to create trigger on Table1 please with multi row scenario. Thanks.
--Company_AuditPeriod DDL
GO
IF
NOT EXISTS
(SELECT
* FROM sys.objects
WHERE object_id
= OBJECT_ID(N'[dbo].[Company_AuditPeriod]')
AND type
in (N'U'))
CREATE
TABLE [dbo].[ Company_AuditPeriod](
      [Client] [varchar](25)
NOT NULL,
      [Period] [varchar](25),
      [Table_Name] [varchar](25),
      [Field_Name] [varchar](25),
      [Old_Value] [varchar](25),
      [New_Value] [varchar](25),
      [User_ID] [varchar](25)
      [Last_Update] [datetime],
      [agrtid] [bigint]
IDENTITY(1,1)
NOT NULL,
ON [PRIMARY]
GO
--Table1 DDL
CREATE TABLE [dbo].[Table1](
[bflag] [int] NOT NULL,
[client] [varchar](25) NOT NULL,
[copies] [int] NOT NULL,
[cost_bio] [decimal](28, 8) NOT NULL,
[cost_cpu] [decimal](28, 8) NOT NULL,
[cost_dio] [decimal](28, 8) NOT NULL,
[date_ended] [datetime] NOT NULL,
[date_started] [datetime] NOT NULL,
[description] [varchar](255) NOT NULL,
[expire_days] [int] NOT NULL,
[func_arg] [varchar](255) NOT NULL,
[func_id] [int] NOT NULL,
[ing_status] [int] NOT NULL,
[invoke_time] [datetime] NOT NULL,
[last_update] [datetime] NOT NULL,
[mail_flag] [tinyint] NOT NULL,
[me_mail_flag] [tinyint] NOT NULL,
[module] [char](3) NOT NULL,
[order_date] [datetime] NOT NULL,
[orderno] [int] NOT NULL,
[output_id] [int] NOT NULL,
[poll_status] [char](1) NOT NULL,
[printer] [char](16) NOT NULL,
[priority] [char](1) NOT NULL,
[priority_no] [int] NOT NULL,
[process_id] [int] NOT NULL,
[report_cols] [int] NOT NULL,
[report_id] [varchar](255) NOT NULL,
[report_name] [varchar](25) NOT NULL,
[report_type] [char](1) NOT NULL,
[server_queue] [char](12) NOT NULL,
[status] [char](1) NOT NULL,
[step_id] [char](8) NOT NULL,
[system_name] [char](8) NOT NULL,
[used_bio] [int] NOT NULL,
[used_cpu] [int] NOT NULL,
[used_dio] [int] NOT NULL,
[user_id] [varchar](25) NOT NULL,
[variant] [int] NOT NULL,
[agrtid] [bigint] IDENTITY(1,1) NOT FOR REPLICATION NOT NULL,
CONSTRAINT [PK_acrrepord001] PRIMARY KEY NONCLUSTERED
[agrtid] ASC
)WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, IGNORE_DUP_KEY = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON, FILLFACTOR = 90) ON [INDEX]
) ON [PRIMARY]
G0<o:p></o:p>
--create trigger
SET
QUOTED_IDENTIFIER ON
GO
CREATE
TRIGGER [dbo].[Table1_Update]
ON [dbo].[Table1]
FOR
UPDATE
NOT
FOR REPLICATION
AS
BEGIN
DECLARE
      @status
varchar(3),
      @user_id
varchar(25),
      @period
varchar(25),
      @client
varchar(25),
      @last_update
datetime
DECLARE
      @Old_status
varchar(3),
      @Old_user_id
varchar(25),
      @Old_period
varchar(25),
      @Old_client
varchar(25)
SELECT
      @status
= status,
      @user_id
= user_id,
      @period
= period,
      @client
= client,
      @last_update
= last_update
FROM Inserted
SELECT
      @Old_status
= status,
      @Old_user_id
= user_id,
      @Old_period
= period,
      @Old_client
= client
FROM Deleted
If @Old_status <> @status
INSERT INTO Company_AuditPeriod
VALUES ( @client, @period,
'Table1',
'period',@old_status, @status, @user_id, @last_update)
END
go

Migrate a SQL-Server trigger

Parse error in line 10, column 9. encountered INSERT.
here is the code :
CREATE TRIGGER delCategoria ON dbo.CATEGORIA
FOR DELETE
AS
declare @usuarioTMS varchar(50)
     IF(UPPER(substring(APP_NAME(),1,4))='TMS\')
          set @usuarioTMS=UPPER(substring(APP_NAME(),1,50))
     Else
          set @usuarioTMS=UPPER(substring(SYSTEM_USER,1,50))
     INSERT INTO LOG
     (responsable,
     modulo,
     tipo,
     descripcion,
     datos_modificados)
     SELECT @usuarioTMS,
     'Administracion',
     'E',
     'EliminaciÃ³n en Categoria',
     'Id CategorÃa: ' + convert(varchar,deleted.id_categoria) + ' - Nombre: ' + deleted.nombre
     FROM delete
Colud you help me, i dont find any error in this code.

I think the last line should be 'FROM deleted'. Also try the parser option Allow 'Reserved Words' in table names, in this case the parser is confused by the table name LOG (you could change it to myLOG and edit the result back to LOG in the oracle model if you prefer.)
I hope that gets you over that issue.
Regards,
Turloch
Oracle Migration Workbench Team

Error in execuitng a trigger

I am getting the following error in executing while executing the "dropping the trigger " , "creating a trigger " . or executing any stored proc .
Please advise .
I am on 10g Release 2 on Redhat linux .
SQL> drop trigger logon_trigger ;
drop trigger logon_trigger
ORA-00604: error occurred at recursive SQL level 1
ORA-01008: not all variables bound
ORA-06512: at line 21

I was able to resolve this by setting a undocumented parameter "_system_trig_enabled"=FALSE .
But , I have a logon trigger . If I set this to FALSE , logon trigger does not kick in .
How to have a logon trigger working and to avoid this error ?
QL> drop table t;
drop table t
ORA-00604: error occurred at recursive SQL level 1
ORA-01008: not all variables bound
ORA-06512: at line 21
SQL> alter system set "_system_trig_enabled"=FALSE;
System altered
SQL> drop table t;
Table dropped

Default value? Trigger?

Hi all,
I got one Oracle 10g problem here, not sure you can help me or not.
If a table has a column which has "default value" setup.
Example
CREATE TABLE KENTEMP
refno number(10),
user_create varchar2(30) Default USER not null
And it also has a trigger with following contents:
If :new.user_create is Null then
:new.user_create := 'TEST';
end if;
If I going insert a record with null value on user_create column.
In some Oracle database, the create_user will become USER, some will become "TEST".
Do you understand why?
Thanks

In some Oracle database, the create_user will become USER, some will become "TEST". Are you running the same INSERT statement on all these databases? Because the default value gets applied before the trigger fires, and I would expect that to be consistent across all databases.
SQL> create table defval (col1 number, col2 varchar2(30) default user)
2 /
Table created.
SQL> insert into defval (col1) values (1)
2 /
1 row created.
SQL> select * from defval
2 /
      COL1 COL2
         1 APC
SQL> create trigger defval_trg before insert on defval for each row
2 begin
3    if :new.col2 is null then
4       :new.col2 := 'TEST';
5    end if;
6 end;
7 /
Trigger created.
SQL> insert into defval (col1) values (2)
2 /
1 row created.
SQL> select * from defval
2 /
      COL1 COL2
         1 APC
         2 APC
SQL> insert into defval (col1,col2) values (3, null)
2 /
1 row created.
SQL> select * from defval
2 /
      COL1 COL2
         1 APC
         2 APC
         3 TEST
SQL> Cheers, APC
Blog : http://radiofreetooting.blogspot.com/

Help me write a trigger : I have no idea how to :(

Hi,
I have to design a trigger based on the following requirement:
I have an ASSET table ASSET (asset_id number(10) (primary key) , site_id (foreign key) references SITE(site_id) , cost_ctr_id (foreign key) references COST_CENTER(cost_ctr_id) )
I have a COST_CENTER table COST_CENTER (cost_center_id number(10) primary key) I have a SITE table SITE (site_id number(10) primary key)
I have a SITE_COST_CTR table SITE_COST_CTR (site_id (foreign key) references SITE(site_id ) , cost_ctr_id (foreign key) references COST_CENTER(cost_ctr_id) ) here (site_id, cost_ctr_id) is the composite primary key .
Now I want a trigger which should satisfy the following : Whenever there is an insert or update in the ASSET table, it should first check in the SITE table and if the site_id is not there it should insert that site_id in the site table, similarly for cost_center_id it should see the cost center table and insert cost_center_id there, followed by an insert into SITE_COST_CTR table (It should insert site_id, cost_center_id in SITE_COST_CTR table)
Can anyone please help me with this trigger ? I have no idea, never ever done it till date :'(. Please help me.
Here's my effort so far, and I am not even a novice in this area!
CREATE OR REPLACE TRIGGER TR_SITE_COST_CTR
BEFORE INSERT OR UPDATE ON ASSET
REFERENCING NEW AS NEW_ASSET
FOR EACH ROW
WHEN ( (SELECT COUNT(*) FROM SITE WHERE SITE_ID = NEW_ASSET.SITE_ID) > 1 AND (SELECT COUNT(*) FROM COST_CENTER WHERE COST_CENTER_ID = NEW_ASSET.COST_CENTER_ID) > 1 )
BEGIN
INSERT INTO SITE (SITE_ID) VALUES (:NEW_ASSET.SITE_ID);
INSERT INTO COST_CENTER VALUES (:NEW_ASSET.COST_CENTER_ID , :NEW_ASSET.COST_CENTER_ID); INSERT INTO SITE_COST_CENTER VALUES (:NEW_ASSET.SITE_ID, :NEW_ASSET.COST_CENTER_ID);
END
Thanks in advance :)

SQL> create table site (site_id number(10) primary key)
2 /
Tabel is aangemaakt.
SQL> create table cost_center (cost_ctr_id number(10) primary key)
2 /
Tabel is aangemaakt.
SQL> create table site_cost_ctr
2 ( site_id     number(10) references site(site_id)
3 , cost_ctr_id number(10) references cost_center(cost_ctr_id)
4 , primary key (site_id,cost_ctr_id)
5 )
6 /
Tabel is aangemaakt.
SQL> create table asset
2 ( asset_id    number(10) primary key
3 , site_id     number(10) references site(site_id)
4 , cost_ctr_id number(10) references cost_center(cost_ctr_id)
5 )
6 /
Tabel is aangemaakt.
SQL> create trigger tr_site_cost_ctr
2 before insert or update on asset for each row
3 declare
4    l_dummy number;
5 begin
6    begin
7      select site_id
8        into l_dummy
9        from site
10       where site_id = :new.site_id
11      ;
12    exception
13    when no_data_found then
14      insert into site (site_id) values (:new.site_id);
15    end
16    ;
17    begin
18      select cost_ctr_id
19        into l_dummy
20        from cost_center
21       where cost_ctr_id = :new.cost_ctr_id
22      ;
23    exception
24    when no_data_found then
25      insert into cost_center (cost_ctr_id) values (:new.cost_ctr_id);
26    end
27    ;
28    begin
29      select site_id
30        into l_dummy
31        from site_cost_ctr
32       where site_id = :new.site_id
33         and cost_ctr_id = :new.cost_ctr_id
34      ;
35    exception
36    when no_data_found then
37      insert into site_cost_ctr (site_id,cost_ctr_id) values (:new.site_id,:new.cost_ctr_id);
38    end
39    ;
40 end;
41 /
Trigger is aangemaakt.
SQL> insert into asset values (1, 11, 21)
2 /
1 rij is aangemaakt.
SQL> insert into asset values (2, 11, 22)
2 /
1 rij is aangemaakt.
SQL> insert into asset values (3, 12, 21)
2 /
1 rij is aangemaakt.
SQL> insert into asset values (4, 11, 21)
2 /
1 rij is aangemaakt.
SQL> select * from site
2 /
                               SITE_ID
                                    11
                                    12
2 rijen zijn geselecteerd.
SQL> select * from cost_center
2 /
                           COST_CTR_ID
                                    21
                                    22
2 rijen zijn geselecteerd.
SQL> select * from site_cost_ctr
2 /
                               SITE_ID                            COST_CTR_ID
                                    11                                     21
                                    11                                     22
                                    12                                     21
3 rijen zijn geselecteerd.
SQL> select * from asset
2 /
                              ASSET_ID                                SITE_ID                            COST_CTR_ID
                                     1                                     11                                 21
                                     2                                     11                                 22
                                     3                                     12                                 21
                                     4                                     11                                 21
4 rijen zijn geselecteerd.Regards,
Rob.

What is wrong in this trigger

Hello All,
please tell me whats wrong in the following trigger,
what effectively i am trying here is I wanted to insert the old records
before update to a temporary table, if this is not the way, then what is the correct way.
SQL> desc ttt1;
Name Null? Type
C1 NUMBER
C2 NUMBER
C3 NUMBER
BMU_ID NUMBER
SQL> desc temp_changes;
Name Null? Type
BMU_ID NUMBER
C2 NUMBER
C3 NUMBER
SQL> ed
Wrote file afiedt.buf
1 create or replace trigger tri_ttt1
2 before update of c3, c2 on ttt1 REFERENCING OLD AS old_ttt1 NEW AS new_ttt1
3 begin
4 insert into temp_changes
5 SELECT BMU_ID,C2,C3
6 FROM ttt1
7 WHERE old_ttt1.c2<>new_ttt1.c2
8 AND old_ttt1.C3<>new_ttt1.C3;
9* end;
SQL> /
Warning: Trigger created with compilation errors.
SQL> sho err
Errors for TRIGGER TRI_TTT1:
LINE/COL ERROR
2/1 PL/SQL: SQL Statement ignored
6/18 PL/SQL: ORA-00904: "NEW_TTT1"."C3": invalid identifier

you forgot the colon (:) before new and old...

PL/SQL or Trigger

Similar Messages

Maybe you are looking for