Yes, thank you, I corrected my function from statement level to row level.
This did get rid of the error message. However, I still get no output from
an OLD variable that should contain data: see the test variable in the
simple case below.

How else can I test OLD variables? This is the simplest test case I can
think of. Any suggestions would be appreciated!

Thanks,
Rick

> I think you have created a statement level trigger (If they existed in
> 7.4.7...)  by not including FOR EACH ROW in your create statement.  In
> statement level triggers, there is no OLD or NEW.
>
>>>> Rick Casey <[EMAIL PROTECTED]> 02/24/05 1:22 PM >>>
> Hello all,
>
> I am trying to a simple thing: create a log history of deletes, and
> updates; but which I am having trouble getting to work in PG 7.4.7
> (under Debian Linux 2.6.8).
>
> I have reduced my code to the following trivial case:
>
> Here is the code that creates the delete trigger:
> create trigger PEDIGREES_hist_del_trig
> AFTER DELETE
> on PEDIGREES
> EXECUTE PROCEDURE logPedigreesDel();
>
>
> Here is the trigger code: (famindid is an integer field in the Pedigrees
>
> table):
>
> CREATE OR REPLACE FUNCTION logPedigreesDel() RETURNS TRIGGER AS '
> DECLARE
>     test integer;
> begin
>     test := OLD.famindid;
>     RAISE EXCEPTION ''OLD.famindid = '', test;
>     return OLD;
> end;
> ' LANGUAGE plpgsql;
>
>
> Here is the error message returned:
> psql:testphdtrig.sql:1: ERROR:  record "old" is not assigned yet
> DETAIL:  The tuple structure of a not-yet-assigned record is
> indeterminate.
> CONTEXT:  PL/pgSQL function "logpedigreesdel" line 4 at assignment
>
> Would *really appreciate* any suggestions! This could help us decide
> whether to PostGres for a major project...
>
> thanks --rick
>
> ---------------------------(end of broadcast)---------------------------
> TIP 8: explain analyze is your friend
>
>
> ---------------------------(end of broadcast)---------------------------
> TIP 2: you can get off all lists at once with the unregister command
>     (send "unregister YourEmailAddressHere" to [EMAIL PROTECTED])
>


---------------------------(end of broadcast)---------------------------
TIP 9: the planner will ignore your desire to choose an index scan if your
      joining column's datatypes do not match

Reply via email to