From: | Geoff Winkless <pgsqladmin(at)geoff(dot)dj> |
---|---|
To: | pgsql-general(at)lists(dot)postgresql(dot)org |
Subject: | Re: ADD OR ALTER column |
Date: | 2018-01-12 14:39:17 |
Message-ID: | CAEzk6fcdirrfQ40aj=OKnAgbzOURy-daOYbcdwTs6uK26pG9CQ@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
On 12 January 2018 at 13:56, Rakesh Kumar <rakeshkumar464(at)mail(dot)com> wrote:
> >I'm probably missing something, but can anyone see a reason why adding an
> ADD OR >ALTER COLUMN syntax to the ALTER TABLE command isn't either a)
> achievable or b) >desirable?
> >It seems to me to be eminently useful and not overly difficult, while
> potentially >saving a significant amount of effort on the part of users.
>
> what is your use case ? Not sure what benefit you are talking about.
>
Having one script to ensure the user has the latest version of a db,
without a significant amount of to-and-fro effort.
Actually I would probably rather have CREATE OR ALTER TABLE instead, with
the syntax of CREATE TABLE but with an optional USING clause for each
column for use if the column already existed and was of a
non-implicitly-castable type, along with an optional WITH DROP COLUMNS [ *
| name [,...] ] clause (to allow user to drop any (or specific) columns not
in the new definition).
That's likely to be rather more effort though, at first glance.
Geoff
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Kazimiers | 2018-01-12 14:53:09 | Get IEEE754 bytes of real and double types for msgpack encoding |
Previous Message | Andreas Joseph Krogh | 2018-01-12 14:14:22 | Sv: ADD OR ALTER column |