Horton

Re: Emitting JSON to file using COPY TO

Andrew Dunstan <andrew@dunslane.net>

View thread
From: Andrew Dunstan <andrew@dunslane.net>
To: Joe Conway <mail@joeconway.com>, Davin Shearer <davin@apache.org>, PostgreSQL-development <pgsql-hackers@postgresql.org>
Date: 2023-12-06T15:32:41Z
Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. Add option force_array for COPY JSON FORMAT

  2. json format for COPY TO

  3. introduce CopyFormat, refactor CopyFormatOptions

  4. Doc: add IDs to copy.sgml's <varlistentry> and <refsect1>

On 2023-12-06 We 08:49, Joe Conway wrote:
> On 12/6/23 07:36, Andrew Dunstan wrote:
>>
>> On 2023-12-05 Tu 16:46, Joe Conway wrote:
>>> On 12/5/23 16:20, Andrew Dunstan wrote:
>>>> On 2023-12-05 Tu 16:09, Joe Conway wrote:
>>>>> On 12/5/23 16:02, Joe Conway wrote:
>>>>>> On 12/5/23 15:55, Andrew Dunstan wrote:
>>>>>>> and in any other case (e.g. LINES) I can't see why you
>>>>>>> would have them.
>>>>>
>>>>> Oh I didn't address this -- I saw examples in the interwebs of 
>>>>> MSSQL server I think [1] which had the non-array with commas 
>>>>> import and export style. It was not that tough to support and the 
>>>>> code as written already does it, so why not?
>>>>
>>>> That seems quite absurd, TBH. I know we've catered for some 
>>>> absurdity in
>>>> the CSV code (much of it down to me), so maybe we need to be 
>>>> liberal in
>>>> what we accept here too. IMNSHO, we should produce either a single 
>>>> JSON
>>>> document (the ARRAY case) or a series of JSON documents, one per row
>>>> (the LINES case).
>>>
>>> So your preference would be to not allow the non-array-with-commas 
>>> case but if/when we implement COPY FROM we would accept that format? 
>>> As in Postel'a law ("be conservative in what you do, be liberal in 
>>> what you accept from others")?
>>
>>
>> Yes, I think so.
>
> Awesome. The attached does it that way. I also ran pgindent.
>
> I believe this is ready to commit unless there are further comments or 
> objections.


Sorry to bikeshed a little more, I'm a bit late looking at this.

I suspect that most users will actually want the table as a single JSON 
document, so it should probably be the default. In any case FORCE_ARRAY 
as an option has a slightly wrong feel to it. I'm having trouble coming 
up with a good name for the reverse of that, off the top of my head.


cheers


andrew

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