Re: [PATCH] Remove workarounds to format [u]int64's

пн, 21 мар. 2022 г. в 12:52, Aleksander Alekseev <aleksander(at)timescale(dot)com>:

> Hi hackers,
>
> I learned from Tom [1] that we can simplify the code like:
>
> ```
> char buff[32];
> snprintf(buf, sizeof(buf), INT64_FORMAT, ...)
> ereport(WARNING, (errmsg("%s ...", buf)));
> ```
>
> ... and rely on %lld/%llu now as long as we explicitly cast the
> argument to long long int / unsigned long long. This was previously
> addressed in 6a1cd8b9 and d914eb34, but I see more places where we
> still use an old approach.
>
> Suggested patch fixes this. Tested locally - no warnings; passes all the
> tests.
>
> [1]
> https://www.postgresql.org/message-id/771048.1647528068%40sss.pgh.pa.us
>
> Hi, Alexander!
Probably you can do (long long) instead of (long long int). It is shorter
and this is used elsewhere in the code.

--
Best regards,
Pavel Borisov

Postgres Professional: http://postgrespro.com <http://www.postgrespro.com>