From: | Stephen Frost <sfrost(at)snowman(dot)net> |
---|---|
To: | Magnus Hagander <magnus(at)hagander(dot)net> |
Cc: | PostgreSQL Developers <pgsql-hackers(at)lists(dot)postgresql(dot)org> |
Subject: | Re: Function to execute a program |
Date: | 2020-09-14 14:33:01 |
Message-ID: | 20200914143300.GG3063@tamriel.snowman.net |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
Greetings,
* Magnus Hagander (magnus(at)hagander(dot)net) wrote:
> Would it make sense to have a pg_execute_program() that corresponds to COPY
> FROM PROGRAM? This would obviously have the same permissions restrictions
> as COPY FROM PROGRAM.
Eh, perhaps.
> The usecase would be to for example execute a command that returns json
> format output, which could then be parsed and processed as part of a query.
> Today, COPY FROM PROGRAM cannot do this, as we can't read a multiline json
> value with COPY.
I'd rather come up with a way to import this kind of object into PG by
using COPY rather than adding a different way to pull them in.
In the "this is a crazy idea" category- have a way to specify an input
function to COPY, or a target data type (and then use it's input
function) to which COPY just keeps feeding bytes until the function
comes back with "ok, I got a value to return" or something.
Would be really neat to have a way to COPY in a JSON, XML, or whatever
even if it's multi-line and even if it has multiple objects (ie: 10
complete JSON documents in a single file would become 10 rows of jsonb).
Thanks,
Stephen
From | Date | Subject | |
---|---|---|---|
Next Message | Pavel Stehule | 2020-09-14 14:34:22 | Re: On login trigger: take three |
Previous Message | Justin Pryzby | 2020-09-14 14:31:03 | Re: 回复:how to create index concurrently on partitioned table |