| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Adam Dziedzic <adam(dot)dziedzi(at)gmail(dot)com> |
| Cc: | pgsql-novice(at)postgresql(dot)org |
| Subject: | Re: PostgreSQL source code: copy command |
| Date: | 2015-09-11 18:12:22 |
| Message-ID: | 7409.1441995142@sss.pgh.pa.us |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-novice |
Adam Dziedzic <adam(dot)dziedzi(at)gmail(dot)com> writes:
> in file: /src/backend/command/copy.c
> I found that it can happen that there is nothing in the buffer
> (fe_msgbuf) to be written to a file and there is only a check of count
> (which in this case would be 0):
> 498 if (fwrite(fe_msgbuf->data, fe_msgbuf->len, 1,
> 499 cstate->copy_file) != 1 ||
> 500 ferror(cstate->copy_file))
On what grounds do you claim that control can reach there without any
data in the buffer? AFAICS, there must be at least either a row header
(if binary mode) or trailer (if not).
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Bastiaan Olij | 2015-09-17 03:21:46 | Combining insert rules on views with functions |
| Previous Message | Adam Dziedzic | 2015-09-11 17:46:24 | PostgreSQL source code: copy command |