From: | Peter Michaux <petermichaux(at)gmail(dot)com> |
---|---|
To: | pgsql-general(at)postgresql(dot)org |
Subject: | join table design question |
Date: | 2006-01-14 23:51:44 |
Message-ID: | 3cbaf1c80601141551r659e6857mefaa44e83499862c@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Hi,
I have tried two db designs and both can work but I wonder which one
the db gurus would go with. This is a general db design question
that has cropped up in a few situation. Here is the example I've been
working on...
In an online store, categories and images have a many-to-many relationship and
products and images have a many-to-many relationship. (It may be that
many other tables can have many-to-many relationship with images.) The
db has the basic three
tables
categories
- id
- name
products
- id
- name
images
- id
- filename
What I can't decide on is how to do the join tables.
Option 1 - would have many join tables and each would require a class in my ORM.
categories_images
- category_id
- image_id
images_products
- product_id
- image_id
Option 2 - have a one generic join table and only one join class
image_joins
- owner_class
- owner_id
- image_id
In option 2 a record could be (Category, 2, 3) or (Product, 4, 1)
--------
Option 1 seems simpler but could lead to some repetitive code (at
least in the data base definition).
Option 2 has only one join table and class. It is a little more
complicated and I don't know how it fits
into the normalization stuff.
How do I choose?
Thanks,
Peter
From | Date | Subject | |
---|---|---|---|
Next Message | Robert Paulsen | 2006-01-15 00:12:31 | Fwd: Re: sort character data in arbitrary order? |
Previous Message | Zlatko Matić | 2006-01-14 23:42:44 | visual query builder for Postgres? |