Re: timing for \copy
| От | Neil Conway |
|---|---|
| Тема | Re: timing for \copy |
| Дата | |
| Msg-id | 1166031616.5901.63.camel@localhost.localdomain обсуждение исходный текст |
| Ответ на | timing for \copy (Andrew Dunstan <andrew@dunslane.net>) |
| Ответы |
Re: timing for \copy
|
| Список | pgsql-patches |
On Wed, 2006-12-13 at 11:36 -0500, Andrew Dunstan wrote:
> + if (pset.timing)
> + {
> + GETTIMEOFDAY(&after);
> + elapsed_msec = DIFF_MSEC(&after, &before);
> + if (success)
> + printf(_("Time: %.3f ms\n"), elapsed_msec);
> +
> + }
Not exactly a big deal, but
if (pset.timing && success)
{
/* ... */
}
seems a bit clearer, and avoids unnecessary work on failure. You can
also move "elapsed_msec" and "after" inside the "if" block.
Also, I believe headers should #include the headers that they themselves
depend on, so the inclusion of the getttimeofday() or _ftime() headers
could be moved to common.h
-Neil
В списке pgsql-patches по дате отправления: