Re: Stored Procedure / Trigger Strangeness

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: laotse(at)lumberjack(dot)snurgle(dot)org
Cc: pgsql-general(at)postgresql(dot)org
Subject: Re: Stored Procedure / Trigger Strangeness
Date: 2001-12-21 15:41:40
Message-ID: 18388.1008949300@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

laotse(at)lumberjack(dot)snurgle(dot)org writes:
> ------------------------------------------------------------------ CREATE
> TRIGGER fti_employee_lastname AFTER UPDATE OR INSERT OR DELETE ON person
> FOR EACH ROW EXECUTE PROCEDURE fti(fti, lastname);

> CREATE TRIGGER fti_employee_firstname AFTER UPDATE OR INSERT OR DELETE ON
> person FOR EACH ROW EXECUTE PROCEDURE fti(fti, firstname);

> CREATE TRIGGER fti_employee_screenname AFTER UPDATE OR INSERT OR DELETE ON
> person FOR EACH ROW EXECUTE PROCEDURE fti(fti, screenname);

This will not work because there's no guarantee about the order of the
execution of the triggers. I haven't worked with fti much, but it's
obvious that it expects you to have only *one* trigger relating a given
indextable to the master --- on update, the trigger deletes all existing
indextable rows for that master row.

It looks like the intended way to index multiple columns using a single
indextable is

CREATE TRIGGER fti_person AFTER UPDATE OR INSERT OR DELETE ON person
FOR EACH ROW EXECUTE PROCEDURE fti(fti, firstname, lastname, screenname);

Or you could use a separate indextable for each column, but that might
not be what you want.

regards, tom lane

In response to

Responses

Browse pgsql-general by date

  From Date Subject
Next Message laotse 2001-12-21 15:47:00 Re: Stored Procedure / Trigger Strangeness
Previous Message laotse 2001-12-21 14:39:39 Stored Procedure / Trigger Strangeness