Message-ID: <8b9985e6e88e9f9519d6e415c9d6408116668c6b.camel@yahoo.co.uk>
Subject: Re: database encoding migration from a pg_dumpall set
From: Steve T <stevetucknott@yahoo.co.uk>
Reply-To: stevetucknott@yahoo.co.uk
To: pgsql-novice@lists.postgresql.org
In-Reply-To: <287927.1660404827@sss.pgh.pa.us>
References: <ecf6066e7e1372c849f9d36477f999b0ad627785.camel.ref@yahoo.co.uk>
	 <ecf6066e7e1372c849f9d36477f999b0ad627785.camel@yahoo.co.uk>
	 <287927.1660404827@sss.pgh.pa.us>
Organization: Personal
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Date: Sun, 14 Aug 2022 07:17:25 +0100
MIME-Version: 1.0
User-Agent: Evolution 3.44.4 (3.44.4-1.fc36) 
Content-Length: 657
Archived-At: 
 <https://www.postgresql.org/message-id/8b9985e6e88e9f9519d6e415c9d6408116668c6b.camel%40yahoo.co.uk>
Precedence: bulk

On Sat, 2022-08-13 at 11:33 -0400, Tom Lane wrote:
> Not via pg_dumpall.=C2=A0 If you dump the databases individually,
> then you can create them manually on the target installation
> with whatever encoding (or other properties) you want,
> finishing by restoring the per-DB dumps into the respective
> new databases.
>=20
> Another way is to manually edit the pg_dumpall output
> script and change the encoding options in the CREATE
> DATABASE commands (but NOT the "SET client_encoding"
> commands).

Thanks Tom.
I did try doing another dumpall with the encoding set on that - but no joy.
I'll edit the dumpall if that's a legit way.

Thanks again.
Steve