Hibernate 3.1

[Shubh Bhat]

Hi all,
I am a Hibernate newbie. Had a simple task.


I am using the following mapping query and class:

<hibernate-mapping>
<class name="com.TableA" table="TableA">
<id name="id" column="id" type="string"></id>
<property name="fun"/>
<property name="foo"/>
</class>
</hibernate-mapping>

---
Query
----
SQLQuery q = this.sessionFactory.getCurrentSession().createSQLQuery(
"select {r.*}, b.rate from TableA r, TableB b " +
"where b.id=r.testId"
);


q.addEntity("r", TableA.class);
q.addJoin("b", "r.rate");
--
TableA class
---
Has all the setter and getter for, id, foo, fun AND ALSO rate



My task is:

1. when I fetch data, I do a join on two tables to get all the fields, lets say TableA & TableB.

2. when I update and save I just want to TableA to be updated.

From the above I get and error that rate is not mapped.
How do I put rate in the mapping and when I do an update I don't really affect TableB.

Shubhra

[Travis Hein]

The hibernate mapping file:

<hibernate-mapping>
 <class name="com.TableA" table="TableA">
 <id name="id" column="id" type="string"></id>
 <property name="fun"/>
 <property name="foo"/>
 
 <property name="testId"/>
 
     <!-- read only bean mapped for TableB
        Since we have the column id above as a column, we must specify the insert=false & update=false here.
        Alternatively, we could move them to the testId property, which makes the testId property read only and then the TableB is always saved when TableA is saved, but you didn't want that.
     -->
    <many-to-one name="test"
                 column="testId"
                 class="com.TableB"
                 cascade="all"
                 fetch="join"
                 unique="true"
                 not-null="true"
                 insert="false"
                 update="false"
                 />
                 
 </class>
 
 <query name="selectAll"><![CDATA[from TableA]]></query>
 <query name="findById"><![CDATA[from TableA where id = ?]]></query>
</hibernate-mapping>
 
 
<hibernate-mapping>
 <class name="com.TableB" table="TableB">
 <id name="id" column="id" type="string"></id>
 <property name="faz"/>
 </class>
</hibernate-mapping>

your beans:

package com;
public classs TableA {
    String id;
    String testId;
    String fun;
    String foo;
 
    TableB test; // instance of our joined table
 
    // AND getters/setters
}
 
// other one
package com;
public class TableB {
    String id;
    // other columns
 
    // AND getters/setters
    
}

then your query:

// get our named query. Note that the wasy we have the query inside the mapping for TableA,
// requires us to enter table a's package in the query name.
Query q = this.sessionFactory.getCurrentSession.getNamedQuery("com.TableA.selectAll");

then when you do a q.list() you will get instances of TableA beans, that contain a nested (joined) TableB, that you can access like

for (Iterator it = q.list(); it.hasNext() ;)  {
    TableA tableAitem = (TableA) it.next();
    tableAitem.getTest(); // since i just mapped it to a property called test.
    // stuff
}

and later when you save

this.sessionFactory.getCurrentSession.saveOrUpdate(tableAitem)

It will save just the tableA, but not the nested TableB

Note also that i used named queries as opposed to building the query string in the code. Unless there is a good reason, like a complete free form dynamic query, I have found it better to use named queries as they are validated when hibernate boots up, as opposed to when they are run, which saves a lot of sanity when you are investigating a mature app and someone changed the table name but missed the one screen that was doing the query in the code using the old column name.

The other thing i did differntly was use the hibernate query as opposed to the sql query. Again unless there is a good reason to use sql queries, like adhoc join of different tables, it is better to use hibernate mappings and hibernate queries.
[ July 03, 2006: Message edited by: Travis Hein ]

[Vijay Kashyap]

Travis,

Sorry for jumping in.

I could not understand your remarks.

<!-- read only bean mapped for TableB
Since we have the column id above as a column, we must specify the insert=false & update=false here.
Alternatively, we could move them to the testId property, which makes the testId property read only and then the TableB is always saved when TableA is saved, but you didn't want that.
-->

Can you explain bit further.

Regards,
Vijay