Thread: AppendStringInfoChar instead of appendStringInfoString

Hi

In (/src/backend/replication/backup_manifest.c)

I found the following code:

    appendStringInfoString(&buf, "\n");
    appendStringInfoString(&buf, "\"");

Since only one bit string is appended here,
I think it will be better to call appendStringInfoChar.

Best reagrds,
houzj

On Fri, Sep 25, 2020 at 08:49:57AM +0000, Hou, Zhijie wrote:
> In (/src/backend/replication/backup_manifest.c)
> 
> I found the following code:
> 
>     appendStringInfoString(&buf, "\n");
>     appendStringInfoString(&buf, "\"");

Good point.  There's another one:

$ git grep -E 'appendStringInfoString.*".{,1}");'
src/backend/utils/adt/ruleutils.c:      appendStringInfoString(buf, "(");

I noticed you added a similar thread here.
https://commitfest.postgresql.org/30/

I think this one could be combined as a single patchset with the existing CF
entry for the other thread.

https://www.postgresql.org/message-id/flat/cb172cf4361e4c7ba7167429070979d4%40G08CNEXMBPEKD05.g08.fujitsu.local
https://www.postgresql.org/message-id/ce9a8564ccf1435eacf14bb7364ae9de%40G08CNEXMBPEKD05.g08.fujitsu.local

-- 
Justin

On 2020-Sep-27, Justin Pryzby wrote:

> On Fri, Sep 25, 2020 at 08:49:57AM +0000, Hou, Zhijie wrote:
> > In (/src/backend/replication/backup_manifest.c)
> > 
> > I found the following code:
> > 
> >     appendStringInfoString(&buf, "\n");
> >     appendStringInfoString(&buf, "\"");
> 
> Good point.  There's another one:

These have been handled by David Rowley in commit 110d81728a0a.