| From: | Michael Paquier <michael(at)paquier(dot)xyz> |
|---|---|
| To: | Rikard Falkeborn <rikard(dot)falkeborn(at)gmail(dot)com> |
| Cc: | pgsql-hackers(at)postgresql(dot)org |
| Subject: | Re: Wrong dead return value in jsonb_utils.c |
| Date: | 2019-05-12 09:07:38 |
| Message-ID: | 20190512090738.GB2120@paquier.xyz |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Sun, May 12, 2019 at 03:18:08AM +0200, Rikard Falkeborn wrote:
> Returning -1 from a function with bool as return value is the same as
> returning true. Now, the code is dead (since elog(ERROR, ...) does not
> return) so it doesn't matter to the compiler, but changing to false is less
> confusing for the programmer. Appologies if this is seen as unnecessary
> churn.
>
> The same code is present since 9.4, but perhaps it's not really worth
> backporting since it is more of an aesthetic change?
This is an aesthetic change in the fact that elog(ERROR) would not
cause -1 to be returned, still I agree that it is cleaner to do things
the way your patch does. And the origin of the issue is I think that
the code of equalsJsonbScalarValue() has been copy-pasted from
compareJsonbScalarValue().
As that's really cosmetic, I would just change that on HEAD, or
perhaps others feel differently?
--
Michael
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Julien Rouhaud | 2019-05-12 09:16:28 | cleanup & refactoring on reindexdb.c |
| Previous Message | Fabien COELHO | 2019-05-12 09:02:16 | make libpq documentation navigable between functions |