I have a trigger that performs an UPDATE query on the same table that
the Update trigger is defined. Of course, it loops infinitely, is there
a way to do this?
CREATE OR REPLACE FUNCTION "public"."clear_common_groups" () RETURNS
trigger AS'
DECLARE
norows integer;
BEGIN
RAISE NOTICE ''Looking for common areas in Building
%'',NEW.hud_building_id;
IF NEW.common_area = ''t'' THEN
UPDATE tblhudunits SET common_area = NULL WHERE unit_id <>
NEW.unit_id AND hud_building_id = NEW.hud_building_id;
END IF;
IF NEW.exterior_area = ''t'' THEN
UPDATE tblhudunits SET exterior_area = NULL WHERE unit_id <>
NEW.unit_id AND hud_building_id = NEW.hud_building_id;
END IF;
GET DIAGNOSTICS norows = ROW_COUNT;
IF NOT FOUND THEN
RAISE NOTICE ''Nothing updated'';
ELSE
RAISE NOTICE ''% rows updated'',norows;
END IF;
RETURN NULL;
END;
'LANGUAGE 'plpgsql' IMMUTABLE CALLED ON NULL INPUT SECURITY INVOKER;
CREATE TRIGGER "insert_new_common_area" AFTER INSERT OR UPDATE
ON "public"."tblhudunits" FOR EACH ROW
EXECUTE PROCEDURE "public"."clear_common_groups"();
--
Robert