Alvaro Herrera escribió:
> However, I think previous discussions have conflated many different
> things, and we've been slowly fixing them one by one; so perhaps at this
> point it does make sense to have a new object-drop event. Let's discuss
> it -- we would define it as taking place just before ddl_command_end,
> and firing any time a command (with matching tag?) has called
> performDeletion or performMultipleDeletions. Does that sound okay?
Here's another version of this patch. I hope this is really final now
... but then, that's what I thought of the previous two versions, too.
I have re-instated event sql_drop, which takes place just before
ddl_command_end, and is called a single time per command (not once per
object dropped, as the old definition would have had it). Within a
function running in that event, you can call
pg_event_trigger_dropped_object(); that will give you a list of all
objects that have been dropped. Since the deletion command has already
been run, the objects are not in catalogs anymore. There are no magic
TG_* variables about objects deleted. I'm a bit unhappy about having to
add calls to EventTriggerSQLDrop() just before each call to
EventTriggerDDLCommandEnd(), but I didn't see any way to make this less
duplicative.
Docs and regression tests have been minimally fixed.
The new event is called sql_drop, note.
--
Álvaro Herrera http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
