MIME-Version: 1.0
References: <CALvfUkBxTYy5uWPFVwpk_7ii2zgT07t3d-yR_cy4sfrrLU=kcg@mail.gmail.com>
 <fc1e39f8-274f-4be6-8a70-9fb0d9fe84a2@aklaver.com> <CAFCRh-_LhP5Ln4-beVz=1pQ9VycU3hAxnJHXhPHZRmX0BSCVEQ@mail.gmail.com>
 <CAFCRh--Vhx0105S7SWAYHGcVMTSRtbcoArDUGSbYSWRUGunW_g@mail.gmail.com>
 <CAKFQuwZb3zaSqkOvGXRRLZEW61az+4xyHZ7tBPR91JfrcdCKvA@mail.gmail.com>
 <3853387.1701096982@sss.pgh.pa.us> <CAFCRh-_GdiUvjd5z5FfvTfhruOnYqBu163XU47zZE8RNATCJGQ@mail.gmail.com>
 <3a98decf-3fe3-4b49-9b68-fda01338872c@sedlakovi.org> <CALvfUkCq3pvkmniWEKZwkbA6hNNvkANF3OamMfhC170RgY+HJg@mail.gmail.com>
 <24e3ee88-ec1e-421b-89ae-8a47ee0d2df1@joeconway.com> <20231201230958.GA1786735@nathanxps13>
 <411395e0-697e-44ab-be98-dd2047f9a10e@joeconway.com> <398150.1701529880@sss.pgh.pa.us>
In-Reply-To: <398150.1701529880@sss.pgh.pa.us>
From: Hannu Krosing <hannuk@google.com>
Date: Sat, 9 Dec 2023 12:56:23 +0100
Message-ID: <CAMT0RQScQz_+nUnC4WSmkVN4KZTJMBYAFcDOn2F8gW57pP_D9w@mail.gmail.com>
Subject: Re: Emitting JSON to file using COPY TO
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Joe Conway <mail@joeconway.com>, Nathan Bossart <nathandbossart@gmail.com>, 
	Davin Shearer <davin@apache.org>, PostgreSQL-development <pgsql-hackers@postgresql.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Archived-At: <https://www.postgresql.org/message-id/CAMT0RQScQz_%2BnUnC4WSmkVN4KZTJMBYAFcDOn2F8gW57pP_D9w%40mail.gmail.com>
Precedence: bulk

>

On Sat, Dec 2, 2023 at 4:11=E2=80=AFPM Tom Lane <tgl@sss.pgh.pa.us> wrote:
>
> Joe Conway <mail@joeconway.com> writes:
> >> I noticed that, with the PoC patch, "json" is the only format that mus=
t be
> >> quoted.  Without quotes, I see a syntax error.


In longer term we should move any specific COPY flag names and values
out of grammar and their checking into the parts that actually
implement whatever the flag is influencing

Similar to what we do with OPTIONS in all levels of FDW  definitions
(WRAPPER itself, SERVER, USER MAPPING, FOREIGN TABLE)

[*] https://www.postgresql.org/docs/current/sql-createforeigndatawrapper.ht=
ml

> >> I'm assuming there's a
> >> conflict with another json-related rule somewhere in gram.y, but I hav=
en't
> >> tracked down exactly which one is causing it.
>
> While I've not looked too closely, I suspect this might be due to the
> FORMAT_LA hack in base_yylex:
>
>             /* Replace FORMAT by FORMAT_LA if it's followed by JSON */
>             switch (next_token)
>             {
>                 case JSON:
>                     cur_token =3D FORMAT_LA;
>                     break;
>             }

My hope is that turning the WITH into a fully independent part with no
grammar-defined keys or values would also solve the issue of quoting
"json".

For backwards compatibility we may even go the route of keeping the
WITH as is  but add the OPTIONS which can take any values at grammar
level.

I shared my "Pluggable Copy " talk slides from Berlin '22 in another thread

--
Hannu