Return-Path: 
 <jdo-dev-return-791-apmail-db-jdo-dev-archive=www.apache.org@db.apache.org>
Delivered-To: apmail-db-jdo-dev-archive@www.apache.org
Received: (qmail 72134 invoked from network); 21 Jul 2005 10:04:12 -0000
Received: from hermes.apache.org (HELO mail.apache.org) (209.237.227.199)
  by minotaur.apache.org with SMTP; 21 Jul 2005 10:04:12 -0000
Received: (qmail 71856 invoked by uid 500); 21 Jul 2005 10:04:10 -0000
Mailing-List: contact jdo-dev-help@db.apache.org; run by ezmlm
Precedence: bulk
List-Help: <mailto:jdo-dev-help@db.apache.org>
List-Unsubscribe: <mailto:jdo-dev-unsubscribe@db.apache.org>
List-Post: <mailto:jdo-dev@db.apache.org>
List-Id: <jdo-dev.db.apache.org>
Reply-To: jdo-dev@db.apache.org
Delivered-To: mailing list jdo-dev@db.apache.org
Received: (qmail 71831 invoked by uid 99); 21 Jul 2005 10:04:10 -0000
Received: from asf.osuosl.org (HELO asf.osuosl.org) (140.211.166.49)
    by apache.org (qpsmtpd/0.29) with ESMTP; Thu, 21 Jul 2005 03:04:10 -0700
X-ASF-Spam-Status: No, hits=0.0 required=10.0
	tests=
X-Spam-Check-By: apache.org
Received-SPF: neutral (asf.osuosl.org: local policy)
Received: from [81.174.133.99] (HELO roobarb.ajsoft.net) (81.174.133.99)
    by apache.org (qpsmtpd/0.29) with ESMTP; Thu, 21 Jul 2005 03:04:03 -0700
Received: from localhost.localdomain (localhost.localdomain [127.0.0.1])
	by roobarb.ajsoft.net (Postfix) with ESMTP id 9FB021112A
	for <jdo-dev@db.apache.org>; Thu, 21 Jul 2005 11:04:00 +0100 (BST)
From: Andy Jefferson <andy@jpox.org>
To: jdo-dev@db.apache.org
Subject: Re: equivalent mappings of join condition?
Date: Thu, 21 Jul 2005 11:03:57 +0100
User-Agent: KMail/1.7.2
References: <42DECA93.1050608@sun.com>
In-Reply-To: <42DECA93.1050608@sun.com>
Organization: JPOX
MIME-Version: 1.0
Content-Type: text/plain;
  charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
Message-Id: <200507211103.58544.andy@jpox.org>
X-Virus-Checked: Checked by ClamAV on apache.org
X-Spam-Rating: minotaur.apache.org 1.6.2 0/1000/N

Hi Michelle,

> Are the following mappings equivalent?  They differ in the placement of
> the <join> element.

Seem different to me, but since we weren't involved in the design of the ORM 
we only have the spec to go by here

> Mapping 1:
>         <class name="Person" table="persons">
>             <join column="EMPID" table="employee_phoneno_type"/>
>             <field name="phoneNumbers" table="employee_phoneno_type" >
>                 <key column="TYPE"/>
>                 <value column="PHONENO"/>
>             </field>
>         </class>
>
> Mapping 2:
>         <class name="Person" table="persons">
>             <field name="phoneNumbers" table="employee_phoneno_type" >
>                 <join column="EMPID"/>
>                 <key column="TYPE"/>
>                 <value column="PHONENO"/>
>             </field>
>         </class>
>
> JPOX appears to treat them differently. 

All in the JPOX docs :-

Mapping 1 has a "Secondary Table" (spec section 15.2), which has nothing in it 
- well it has a Map, but the Map has no columns of its own, so the Secondary 
Table just has a PK. The Map in Mapping 1 is a "ForeignKey" Map (since 
there's no <join> within <field> - meaning that the keys/values are stored in 
the value table.
See here http://www.jpox.org/docs/1_1/secondary_tables.html
and here http://www.jpox.org/docs/1_1/relationships_1_N_map.html#fk_uni

In Mapping 2, the Map is a "JoinTable" Map (since there's a <join> within the 
<field>) - meaning that the Map has its key and value (or FK equivalents if 
the key/value are PC) stored in the join table.
See here http://www.jpox.org/docs/1_1/relationships_1_N_map.html#join_pc_pc


Again, this comes down to the interpretation of the <join> within <field>. 
Since JPOX (and other JDO impls for that matter) supports having Maps formed 
with a FK in the value, we interpret the lack of a <join> within the <field> 
as saying that we should store the linkage between map and value as a FK in 
the value (This is consistent with Collections where there is no <join> 
within <field> being interpreted as having a FK in the element).

Criticise away ;-)


It's good to see that the TCK is flushing out all possible differences of 
interpretation in ORM. There are so many ways of representing relationships, 
and JPOX supports what our users feel are the most common.

-- 
Andy