| From: | Peter Eisentraut <peter(at)eisentraut(dot)org> |
|---|---|
| To: | Florents Tselai <florents(dot)tselai(at)gmail(dot)com>, Pavel Stehule <pavel(dot)stehule(at)gmail(dot)com> |
| Cc: | Chao Li <li(dot)evan(dot)chao(at)gmail(dot)com>, pgsql-hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org> |
| Subject: | Re: Add jsonb_translate(jsonb, from, to) |
| Date: | 2025-10-01 11:55:55 |
| Message-ID: | 8d3c7094-4b22-4c6c-a9e7-3f0b55f5ec04@eisentraut.org |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On 29.09.25 14:34, Florents Tselai wrote:
>> Cannot be better to use JsonPath for specification what should be
>> replaced?
>
> Fair point.
> The main purpose of this patch is to provide a recursive, global
> replacement across all values and arrays,
> which is not as straightforward to express in JSONPath today.
> I understand that some may find this too case-specific, so I’m just
> leaving it out there for consideration.
> That said, I believe it can be quite useful in domains where documents
> carry many tags or labels that need to be translated or normalized
> consistently.
Oracle has a json_transform function, which has also been added to the
SQL standard draft recently. That appears to be more in the direction
that Pavel is suggesting. Maybe someone wants to take a stab at it.
(In that case, having both a json_transform and a json_translate might
be confusing.)
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Greg Burd | 2025-10-01 11:56:07 | Re: [PATCH] Add tests for Bitmapset |
| Previous Message | Aleksander Alekseev | 2025-10-01 11:45:59 | Re: The ability of postgres to determine loss of files of the main fork |