Postgres Pro
Downloads
Home   >   mailing lists

Thread: initdb: Refactor PG_CMD_PUTS loops

initdb: Refactor PG_CMD_PUTS loops

From
Peter Eisentraut
Date:
Keeping the SQL commands that initdb runs in string arrays before
feeding them to PG_CMD_PUTS() seems unnecessarily verbose and
inflexible.  In some cases, the array only has one member.  In other
cases, one might want to use PG_CMD_PRINTF() instead, to parametrize a
command, but that would require breaking up the loop or using
workarounds like replace_token().  This patch unwinds all that; it's 
much simpler that way.
Attachment

Re: initdb: Refactor PG_CMD_PUTS loops

From
John Naylor
Date:

On Thu, Dec 1, 2022 at 5:02 PM Peter Eisentraut <peter.eisentraut@enterprisedb.com> wrote:
>
> Keeping the SQL commands that initdb runs in string arrays before
> feeding them to PG_CMD_PUTS() seems unnecessarily verbose and
> inflexible.  In some cases, the array only has one member.  In other
> cases, one might want to use PG_CMD_PRINTF() instead, to parametrize a
> command, but that would require breaking up the loop or using
> workarounds like replace_token().  This patch unwinds all that; it's
> much simpler that way.

+1,  I can't think of a reason to keep the current coding

--
John Naylor
EDB: http://www.enterprisedb.com

Re: initdb: Refactor PG_CMD_PUTS loops

From
Andrew Dunstan
Date:
On 2022-12-01 Th 05:02, Peter Eisentraut wrote:
> Keeping the SQL commands that initdb runs in string arrays before
> feeding them to PG_CMD_PUTS() seems unnecessarily verbose and
> inflexible.  In some cases, the array only has one member.  In other
> cases, one might want to use PG_CMD_PRINTF() instead, to parametrize a
> command, but that would require breaking up the loop or using
> workarounds like replace_token().  This patch unwinds all that; it's
> much simpler that way.


Looks reasonable. (Most of this dates back to 2003/2004, the very early
days of initdb.c.)


cheers


andrew

--
Andrew Dunstan
EDB: https://www.enterprisedb.com

Re: initdb: Refactor PG_CMD_PUTS loops

From
Peter Eisentraut
Date:
On 02.12.22 15:07, Andrew Dunstan wrote:
> On 2022-12-01 Th 05:02, Peter Eisentraut wrote:
>> Keeping the SQL commands that initdb runs in string arrays before
>> feeding them to PG_CMD_PUTS() seems unnecessarily verbose and
>> inflexible.  In some cases, the array only has one member.  In other
>> cases, one might want to use PG_CMD_PRINTF() instead, to parametrize a
>> command, but that would require breaking up the loop or using
>> workarounds like replace_token().  This patch unwinds all that; it's
>> much simpler that way.
> 
> Looks reasonable. (Most of this dates back to 2003/2004, the very early
> days of initdb.c.)

committed