Message-ID: <65fab458-fd9d-1f0c-6d6a-d9726cb43044@dunslane.net>
Date: Sun, 3 Dec 2023 08:46:28 -0500
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.15.1
Content-Language: en-US
To: Joe Conway <mail@joeconway.com>, Maciek Sakrejda <m.sakrejda@gmail.com>
Cc: Davin Shearer <davin@apache.org>,
 PostgreSQL-development <pgsql-hackers@postgresql.org>
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>
 <CAOtHd0C+k8M1HYPubnH6J_7B5Eh5_nrun-n3fc9ZsSb_Tq9ZHA@mail.gmail.com>
 <70055855-8ea1-43dc-85c2-630b8cf3bbbd@joeconway.com>
From: Andrew Dunstan <andrew@dunslane.net>
Subject: Re: Emitting JSON to file using COPY TO
In-Reply-To: <70055855-8ea1-43dc-85c2-630b8cf3bbbd@joeconway.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Archived-At: <https://www.postgresql.org/message-id/65fab458-fd9d-1f0c-6d6a-d9726cb43044%40dunslane.net>
Precedence: bulk


On 2023-12-02 Sa 17:43, Joe Conway wrote:
> On 12/2/23 13:50, Maciek Sakrejda wrote:
>> On Fri, Dec 1, 2023 at 11:32 AM Joe Conway <mail@joeconway.com> wrote:
>>> 1. Is supporting JSON array format sufficient, or does it need to
>>> support some other options? How flexible does the support scheme 
>>> need to be?
>>
>> "JSON Lines" is a semi-standard format [1] that's basically just
>> newline-separated JSON values. (In fact, this is what
>> log_destination=jsonlog gives you for Postgres logs, no?) It might be
>> worthwhile to support that, too.
>>
>> [1]: https://jsonlines.org/
>
>
> Yes, I have seen examples of that associated with other databases 
> (MSSQL and Duckdb at least) as well. It probably makes sense to 
> support that format too.


You can do that today, e.g.


copy (select to_json(q) from table_or_query q) to stdout


You can also do it as a single document as proposed here, like this:


copy (select json_agg(q) from table_or_query q) to stdout


The only downside to that is that it has to construct the aggregate, 
which could be ugly for large datasets, and that's why I'm not opposed 
to this patch.


cheers


andrew

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