Thread: Raise Notice

Raise Notice

From
"Prasad dev"
Date:
Hi,
 
 I have been trying to figure out why this simple query doesn't execute if no record exist !
 
CREATE OR REPLACE FUNCTION del_rest() RETURNS TRIGGER AS 'DECLARE    t record; 
BEGIN     SELECT * INTO t FROM del2 WHERE (d2=OLD.d2 or old.d2 is null) and (d3=OLD.d3
or old.d2 is null);         IF NOT FOUND
THEN          RAISE NOTICE ''No such record exits in table
del2'';        END IF;         
returnnull;      END;'LANGUAGE 'plpgsql'; 
 
Looking forward for suggestions. Thanks
 
Prasad.

Re: Raise Notice

From
Michael Fuhr
Date:
[Please don't post in HTML.]

On Mon, Jun 20, 2005 at 12:27:41AM +0000, Prasad dev wrote:

> I have been trying to figure out why this simple query doesn't
> execute if no record exist !
>
> CREATE OR REPLACE FUNCTION del_rest() RETURNS TRIGGER AS '
> DECLARE
>     t record;
>   BEGIN
>       SELECT * INTO t FROM del2 WHERE (d2=OLD.d2 or old.d2 is null)
> and (d3=OLD.d3 or  old.d2 is null);

In the last line above, should the last part of the expression be
"old.d3 is null" instead of "old.d2 is null"?  Perhaps that's why
the subsequent IF NOT FOUND isn't doing what you expect, if that's
indeed the problem you mean.

If the expression is correct as shown, then please post a minimal
but complete example that somebody could load into an empty database
to reproduce the problem.  That is, all SQL statements to create
the necessary tables, functions, and triggers, and to populate the
tables with a sample data set; also the query that's not behaving
the way you expect and an explanation of what you'd like it to do.

--
Michael Fuhr
http://www.fuhr.org/~mfuhr/

Re: Raise Notice

From
Michael Fuhr
Date:
[Please copy the mailing list on replies, and please don't post in HTML.
Ordinarily I'd trim the message I'm replying to, but since it wasn't
sent to the mailing list I'll post it here in its entirety.  See my
response below.]

On Tue, Jun 21, 2005 at 10:49:35PM +0000, Prasad dev wrote:
>
> After posting i realised what i asked for was very unclear so here it
> goes,
> create table del2 (d2 int,d3 int,d4 int,primary key(d2,d3));
> create table del1 (d1 int, d2 int,d3 int,primary key(d1),foreign key
> (d2,d3) references del2(d2,d3));
> insert into del2 values (1,1,1);
> insert into del2 values (2,2,2);
> insert into del2 values (3,3,3);
> insert into del1 values (1,1,1);
> -------------------------------------------------------------------------
> CREATE OR REPLACE FUNCTION del_rest() RETURNS TRIGGER AS '
> DECLARE
>     t record;
>   BEGIN
>       SELECT * INTO t FROM del2 WHERE d2=OLD.d2  and d3=OLD.d3;
>          IF NOT FOUND THEN
>              RAISE NOTICE ''No such record exits in table del2'';
>       return null;
>            END IF;
>      END;
> 'LANGUAGE 'plpgsql';
> CREATE TRIGGER delrest BEFORE  DELETE ON del2 FOR EACH ROW EXECUTE
> PROCEDURE del_rest();
>
> delete from del2 where d2=44 and d3=44;
> Here record 44 doesn't exist in the table del2 i just want it to show
> me the message in raise notice, i am using postgre - 7.3. When i
> run the query this is the output without the message.
>
> data=> delete from del2 where d2=44 and d3=44;
> DELETE 0
> data=>

A row-level trigger calls the trigger function once for each row
that the operation would affect.  Since there are no matching rows,
the trigger function is never called.

What problem are you trying to solve?  The "DELETE 0" response
already shows that no rows were deleted -- what purpose would the
notice serve?

--
Michael Fuhr
http://www.fuhr.org/~mfuhr/

Re: Raise Notice

From
"Prasad dev"
Date:
Hi,

Basically i am dealing with triggers which addresses referential integrity,
lets take a delete restrict case,  I have a parent table - "del2", child -
"del1"

when i try to delete a record from del2 it should perform following steps.

begin
perform lookup in del2
    if found then
      perform lookup in del1
        if found then
            restrict delete
       else
            delete from del2
        end if
     else
          message " no record found"
     end if
end

While posting i removed a few lines concerning delete restrict and just
placed raise notice query. SO in the end what i look for is a proper message
instead of "DELETE 0". Hope you get what i was looking for, by the way what
do you mean by dont post in HTML ?

Cheers
Prasad.


>From: Michael Fuhr <mike@fuhr.org>
>Reply-To: pgsql-novice@postgresql.org
>To: Prasad dev <esteem3300@hotmail.com>
>CC: pgsql-novice@postgresql.org
>Subject: Re: [NOVICE] Raise Notice
>Date: Tue, 21 Jun 2005 18:18:46 -0600
>
>[Please copy the mailing list on replies, and please don't post in HTML.
>Ordinarily I'd trim the message I'm replying to, but since it wasn't
>sent to the mailing list I'll post it here in its entirety.  See my
>response below.]
>
>On Tue, Jun 21, 2005 at 10:49:35PM +0000, Prasad dev wrote:
> >
> > After posting i realised what i asked for was very unclear so here it
> > goes,
> > create table del2 (d2 int,d3 int,d4 int,primary key(d2,d3));
> > create table del1 (d1 int, d2 int,d3 int,primary key(d1),foreign key
> > (d2,d3) references del2(d2,d3));
> > insert into del2 values (1,1,1);
> > insert into del2 values (2,2,2);
> > insert into del2 values (3,3,3);
> > insert into del1 values (1,1,1);
> >
>-------------------------------------------------------------------------
> > CREATE OR REPLACE FUNCTION del_rest() RETURNS TRIGGER AS '
> > DECLARE
> >     t record;
> >   BEGIN
> >       SELECT * INTO t FROM del2 WHERE d2=OLD.d2  and d3=OLD.d3;
> >          IF NOT FOUND THEN
> >              RAISE NOTICE ''No such record exits in table del2'';
> >       return null;
> >            END IF;
> >      END;
> > 'LANGUAGE 'plpgsql';
> > CREATE TRIGGER delrest BEFORE  DELETE ON del2 FOR EACH ROW EXECUTE
> > PROCEDURE del_rest();
> >
> > delete from del2 where d2=44 and d3=44;
> > Here record 44 doesn't exist in the table del2 i just want it to show
> > me the message in raise notice, i am using postgre - 7.3. When i
> > run the query this is the output without the message.
> >
> > data=> delete from del2 where d2=44 and d3=44;
> > DELETE 0
> > data=>
>
>A row-level trigger calls the trigger function once for each row
>that the operation would affect.  Since there are no matching rows,
>the trigger function is never called.
>
>What problem are you trying to solve?  The "DELETE 0" response
>already shows that no rows were deleted -- what purpose would the
>notice serve?
>
>--
>Michael Fuhr
>http://www.fuhr.org/~mfuhr/
>
>---------------------------(end of broadcast)---------------------------
>TIP 9: In versions below 8.0, the planner will ignore your desire to
>        choose an index scan if your joining column's datatypes do not
>        match



Re: Raise Notice

From
Michael Fuhr
Date:
On Wed, Jun 22, 2005 at 11:49:59PM +0000, Prasad dev wrote:
>
> Basically i am dealing with triggers which addresses referential integrity,

Are you familiar with foreign key constraints?  PostgreSQL can
automatically do referential integrity checks for you.

http://www.postgresql.org/docs/7.3/static/tutorial-fk.html
http://www.postgresql.org/docs/7.3/static/ddl-constraints.html#DDL-CONSTRAINTS-FK
http://www.postgresql.org/docs/7.3/static/sql-createtable.html

(Links are to the 7.3 documentation because you said you were running
that version.)

> SO in the end what i look for is a proper message instead of "DELETE 0".

You could do this in application code, or you could wrap the delete
in a function that checks how many rows were deleted and raises a
notice if the count was zero.

http://www.postgresql.org/docs/7.3/static/plpgsql-statements.html#PLPGSQL-STATEMENTS-DIAGNOSTICS

If you were using 7.4 or later, you could use a statement-level
AFTER trigger to raise a notice if no rows were deleted.  But it's
still not clear what value such a notice would add.

> by the way what do you mean by dont post in HTML ?

Your previous messages were HTML-formatted.  Some people use
text-based mail readers, so they have to take extra steps to convert
the HTML to something legible; also, some people's spam filters
might automatically delete HTML messages.  Plain text is more likely
to be acceptable to some people.

--
Michael Fuhr
http://www.fuhr.org/~mfuhr/