Re: [HACKERS] Patch: Add --no-comments to skip COMMENTs with pg_dump

On 7 August 2017 at 16:14, Fabrízio de Royes Mello
<fabriziomello@gmail.com> wrote:
>
> On Mon, Aug 7, 2017 at 10:43 AM, Robins Tharakan <tharakan@gmail.com> wrote:
>>
>> On 20 July 2017 at 05:14, Robins Tharakan <tharakan@gmail.com> wrote:
>>>
>>> On 20 July 2017 at 05:08, Michael Paquier <michael.paquier@gmail.com>
>>> wrote:
>>>>
>>>> On Wed, Jul 19, 2017 at 8:59 PM,
>>>> Fabrízio de Royes Mello
>>>> > You should add the properly sgml docs for this pg_dumpall change also.
>>>>
>>>> Tests of pg_dump go to src/bin/pg_dump/t/ and tests for objects in
>>>> extensions are in src/test/modules/test_pg_dump, but you just care
>>>> about the former with this patch. And if you implement some new tests,
>>>> look at the other tests and base your work on that.
>>>
>>>
>>> Thanks Michael /
>>> Fabrízio.
>>>
>>> Updated patch (attached) additionally adds SGML changes for pg_dumpall.
>>> (I'll try to work on the tests, but sending this
>>> nonetheless
>>> ).
>>>
>>
>> Attached is an updated patch (v4) with basic tests for pg_dump /
>> pg_dumpall.
>> (Have zipped it since patch size jumped to ~40kb).
>>
>
> The patch applies cleanly to current master and all tests run without
> failures.
>
> I also test against all current supported versions (9.2 ... 9.6) and didn't
> find any issue.
>
> Changed status to "ready for commiter".

I get the problem, but not this solution. It's too specific and of
zero other value, yet not even exactly specific to the issue. We
definitely don't want --no-extension-comments, but --no-comments
removes ALL comments just to solve a weird problem. (Meta)Data loss,
surely?

Thinking ahead, are we going to add a new --no-objecttype switch every
time someone wants it?

It would make more sense to add something more general and extensible
such as --exclude-objects=comment
or similar name

--
Simon Riggs                http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services