Markus Schiltknecht <markus(at)bluegap(dot)ch> writes:
> CREATE TABLE category (
> id INT PRIMARY KEY,
> name TEXT);
> CREATE TABLE category_todo (
> cat_id INT REFERENCES category(id)
> DEFERRABLE INITIALLY DEFERRED
> INSERT INTO category (id, name) VALUES (0, 'test');
> INSERT INTO category_todo (cat_id) VALUES (0);
> TRUNCATE category_todo;
> -- COMMIT fails with: 'failed to fetch new tuple for AFTER trigger'
Hm. At least for this case, it seems the nicest behavior would be for
TRUNCATE to scan the deferred-triggers list and just throw away any
pending trigger firings for the target table(s). I wonder however
whether there are cases where that would be a bad idea. It might be
safer for the TRUNCATE to error out if there are any pending triggers.
Stephan, any thoughts about it?
regards, tom lane
In response to
pgsql-bugs by date
|Next:||From: Stephan Szabo||Date: 2006-08-21 20:43:58|
|Subject: Re: truncate in combination with deferred triggers |
|Previous:||From: Markus Schiltknecht||Date: 2006-08-21 18:24:12|
|Subject: truncate in combination with deferred triggers|