Re: Extend file_fdw wrapper

Attached patch.

------------
pasman

2011/10/10 pasman pasmański <pasman(dot)p(at)gmail(dot)com>:
> Attached patch.

... and what are these new options intended to do?

--
Thom Brown
Twitter: @darkixion
IRC (freenode): dark_ixion
Registered Linux user: #516935

EnterpriseDB UK: http://www.enterprisedb.com
The Enterprise PostgreSQL Company

At a quick glance, this patch seems to have an issue about priority.
Which value is used if an option has been set both on a foreign table
and a foreign server?

Also I think documents and regression tests would be required for
this kind of change.

Regards,
--
Shigeru Hanada

On Mon, Oct 10, 2011 at 10:51:03PM +0900, Shigeru Hanada wrote:
> At a quick glance, this patch seems to have an issue about priority.
> Which value is used if an option has been set both on a foreign table
> and a foreign server?

I believe that the finer-grained setting should always override the
coarser, so in this case, the setting for the table should take
precedence over the setting for the server.

> Also I think documents and regression tests would be required for
> this kind of change.

+1 :)

Cheers,
David.
--
David Fetter <david(at)fetter(dot)org> http://fetter.org/
Phone: +1 415 235 3778 AIM: dfetter666 Yahoo!: dfetter
Skype: davidfetter XMPP: david(dot)fetter(at)gmail(dot)com
iCal: webcal://www.tripit.com/feed/ical/people/david74/tripit.ics

Remember to vote!
Consider donating to Postgres: http://www.postgresql.org/about/donate

On 10/10/2011 09:51 AM, Shigeru Hanada wrote:
> At a quick glance, this patch seems to have an issue about priority.
> Which value is used if an option has been set both on a foreign table
> and a foreign server?
>
> Also I think documents and regression tests would be required for
> this kind of change.
>
>

I'm not even sure I understand why we should want this anyway. The
closest analog I can think of to a more conventional server is that the
whole file system is the foreign server, and there just don't seem to be
any relevant options at that level. All the options being supplied seem
much saner left as just foreign table options.

cheers

andrew

On Mon, Oct 10, 2011 at 10:23:51AM -0400, Andrew Dunstan wrote:
> On 10/10/2011 09:51 AM, Shigeru Hanada wrote:
> >At a quick glance, this patch seems to have an issue about priority.
> >Which value is used if an option has been set both on a foreign table
> >and a foreign server?
> >
> >Also I think documents and regression tests would be required for
> >this kind of change.
>
> I'm not even sure I understand why we should want this anyway. The
> closest analog I can think of to a more conventional server is that
> the whole file system is the foreign server, and there just don't
> seem to be any relevant options at that level. All the options being
> supplied seem much saner left as just foreign table options.

You raise an excellent point, which is that there probably should be
options at that level which override the (settable) generic file_fdw
options.

Cheers,
David.
--
David Fetter <david(at)fetter(dot)org> http://fetter.org/
Phone: +1 415 235 3778 AIM: dfetter666 Yahoo!: dfetter
Skype: davidfetter XMPP: david(dot)fetter(at)gmail(dot)com
iCal: webcal://www.tripit.com/feed/ical/people/david74/tripit.ics

Remember to vote!
Consider donating to Postgres: http://www.postgresql.org/about/donate

On 10/10/2011 11:59 AM, David Fetter wrote:
> On Mon, Oct 10, 2011 at 10:23:51AM -0400, Andrew Dunstan wrote:
>> On 10/10/2011 09:51 AM, Shigeru Hanada wrote:
>>> At a quick glance, this patch seems to have an issue about priority.
>>> Which value is used if an option has been set both on a foreign table
>>> and a foreign server?
>>>
>>> Also I think documents and regression tests would be required for
>>> this kind of change.
>> I'm not even sure I understand why we should want this anyway. The
>> closest analog I can think of to a more conventional server is that
>> the whole file system is the foreign server, and there just don't
>> seem to be any relevant options at that level. All the options being
>> supplied seem much saner left as just foreign table options.
> You raise an excellent point, which is that there probably should be
> options at that level which override the (settable) generic file_fdw
> options.
>
>

That's not my point at all.

cheers

andrew