| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | "Karl O(dot) Pinc" <kop(at)meme(dot)com> |
| Cc: | Gilles Darold <gilles(dot)darold(at)dalibo(dot)com>, Christoph Berg <myon(at)debian(dot)org>, Michael Paquier <michael(dot)paquier(at)gmail(dot)com>, "Shulgin, Oleksandr" <oleksandr(dot)shulgin(at)zalando(dot)de>, Robert Haas <robertmhaas(at)gmail(dot)com>, PostgreSQL Hackers <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Patch to implement pg_current_logfile() function |
| Date: | 2016-11-23 15:26:29 |
| Message-ID: | 27584.1479914789@sss.pgh.pa.us |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
"Karl O. Pinc" <kop(at)meme(dot)com> writes:
> Maybe on the 2nd call to strtok() you could pass ""
> as the 2nd argument? That'd be a little wonky but
> the man page does not say you can't have an empty
> set of delimiters. On the other hand strtok() is
> not a perfect choice, you don't want to "collapse"
> adjacent delimiters in the parsed string or ignore
> leading spaces. I'd prefer a strstr() solution.
I'd stay away from strtok() no matter what. The process-wide static
state it implies is dangerous: if you use it, you're betting that
you aren't interrupting some caller's use, nor will any callee decide
to use it. In a system as large as Postgres, that's a bad bet, or
would be if we didn't discourage use of strtok() pretty hard.
As far as I can find, there are exactly two users of strtok() in
the backend, and they're already playing with fire because one
calls the other (look in utils/misc/tzparser.c). I don't want the
hazard to get any larger.
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Robert Haas | 2016-11-23 15:36:11 | Re: Patch to implement pg_current_logfile() function |
| Previous Message | Tomas Vondra | 2016-11-23 15:12:02 | Re: [RFC] Should we fix postmaster to avoid slow shutdown? |