Re: Asssociative Arrays: Best practices / snippets? - Mailing list pgsql-general

From Steve Atkins
Subject Re: Asssociative Arrays: Best practices / snippets?
Date
Msg-id EE7F465B-04BA-4589-84E8-2810B7F36871@blighty.com
Whole thread Raw
In response to Asssociative Arrays: Best practices / snippets?  (Stefan Keller <sfkeller@gmail.com>)
Responses Re: Asssociative Arrays: Best practices / snippets?
List pgsql-general
On May 23, 2009, at 2:25 AM, Stefan Keller wrote:

> I have a use case where the I want to put an unforeseable number of
> key/value pairs in a column.
> Now, PostgreSQL has arrays as first class types.
> Are there any best practices and snippets (preferrably in plpgsql)
> for handling key/value pairs?

Take a look at "hstore". It's a data type in contrib that's intended
to store a set of key, value
pairs in a single column.

It's well suited for the case where you're not searching on the keys
and values. If you do
need to search on them then you can either index the hstore field with
a gin or gist index,
or use a separate EAV table for the keys and values.

Cheers,
   Steve


pgsql-general by date:

Previous
From: Ludwig Kniprath
Date:
Subject: Re: 8.3: timestamp subtraction
Next
From: Ivan Sergio Borgonovo
Date:
Subject: psql Windows and variables