Date: Tue, 4 Feb 2025 06:46:41 -0800 (PST)
From: Rich Shepard <rshepard@appl-ecosys.com>
To: pgsql-general@postgresql.org
Subject: Re: Lookup tables
In-Reply-To: <0def87f9-6fa6-49a7-b0bc-f313c066d387@gmail.com>
Message-ID: <c4fcf2f3-dda-e52e-4140-e16f7e95a580@appl-ecosys.com>
References: <4e7a338-c7d3-e944-20bd-a6e346e175c7@appl-ecosys.com>
 <0def87f9-6fa6-49a7-b0bc-f313c066d387@gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; format=flowed; charset=ISO-8859-15
Content-Transfer-Encoding: 8BIT
Archived-At: 
 <https://www.postgresql.org/message-id/c4fcf2f3-dda-e52e-4140-e16f7e95a580%40appl-ecosys.com>
Precedence: bulk

On Tue, 4 Feb 2025, Rob Sargent wrote:

> I would definitely add an Id for each of the looked up values. The code
> can use the Id (for the join or lookup) and the string value can safely be
> changed (think typo) without ripple effect.  It also prevents other tables
> from referencing the lookup witrh bogus values (those not in the lookup
> table)
>
> Unless your lookup tables are huge I would create a new table matching
> your current table but with an identity column and load from you original
> table.
> https://www.postgresql.org/docs/current/ddl-identity-columns.html

Thanks, Rob. Each lookup table has only a very few rows,

I'll make new tables today.

Regards,

Rich