Tom Lane wrote:
> Alvaro Herrera <alvherre@commandprompt.com> writes:
>> My thinking is that a TRUNCATE trigger is a per-statement trigger which
>> doesn't have access to the set of deleted rows (Replicator uses it that
>> way -- we replicate the truncate action, and replay it on the replica).
>> In that way it would be different from a per-statement trigger for
>> DELETE.
>
> Ah, right. I was thinking in terms of having TRUNCATE actually fire the
> existing ON DELETE-type triggers, but that's not really helpful --- you'd
> need a separate trigger-event type. So we could just say by fiat that
> an ON TRUNCATE trigger doesn't get any rowset information, even after we
> add that for the other types of statement-level triggers.
I've always considered TRUNCATE to be DDL rather than DML. I mentally
group it with DROP TABLE rather than DELETE>
-- Richard Huxton Archonet Ltd