Re: "truncate all"?
| От | Tom Lane |
|---|---|
| Тема | Re: "truncate all"? |
| Дата | |
| Msg-id | 22159.1060092981@sss.pgh.pa.us обсуждение |
| Ответ на | Re: "truncate all"? (Andreas <e9625203@student.tuwien.ac.at>) |
| Список | pgsql-hackers |
Andreas <e9625203@student.tuwien.ac.at> writes:
> Agreed, workarounds are easy. The original suggestion of "TRUNCATE ALL"
> (or "TRUNCATE table CASCADE"), however, was motivated by the search for a
> simple and efficient truncation of all tables to accelerate unit-testing.
I still think the best suggestion for you is the one someone made
already: create a schema dump (pg_dump -s), then do DROP DATABASE,
CREATE DATABASE, load the schema dump. This will be somewhat slower
than a hypothetical TRUNCATE ALL, but it has the great advantage that
your starting point is well-documented and guaranteed to be the same
on every iteration. TRUNCATE ALL would provide no such guarantee.
And, of course, that solution exists today ...
regards, tom lane
В списке pgsql-hackers по дате отправления: