[hibernate-issues] [Hibernate-JIRA] Commented: (ANN-632) @IndexColumn doesn't set value of index column

[Al Le (JIRA)]

    [ http://opensource.atlassian.com/projects/hibernate/browse/ANN-632?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_30964 ] 

Al Le commented on ANN-632:
---------------------------

I found a solution (or, rather, a workaround). I made the column 'child_index' in the DB nullable. Now the index is correctly set.

The reason why it works now is that when the association is managed by the 'one' side, the insert is done in two stages. First, an INSERT is done for each child element, and then an UPDATE is executed that sets the child_index. This two-stage procedure doesn't work if the index column does not allow NULL values.

>From the DB model perspective, it would be desirable that the column does _not_ allow NULL value.

One way to achieve this (IIUC) would be to make the 'many' side manage the association. But then I'd have to manually set the value of the index column.

Ideally, the two things (automatic setting of the index column and having it as NOT NULL in the DB) would be possible at the same time.

> @IndexColumn doesn't set value of index column
> ----------------------------------------------
>
>                 Key: ANN-632
>                 URL: http://opensource.atlassian.com/projects/hibernate/browse/ANN-632
>             Project: Hibernate Annotations
>          Issue Type: Improvement
>          Components: documentation
>    Affects Versions: 3.3.0.ga
>            Reporter: Dan Allen
>            Assignee: Diego Plentz
>            Priority: Minor
>
> I'm sure I will get screamed at for this, but the @IndexColumn just doesn't work with @OneToMany.  When I say it doesn't work, it means that I am a reasonable person and I have studied the documentation for at least 4 hours and I just cannot figure out how to make it work.  So either the documentation needs to be improved, or there is something wrong with Hibernate. I refuse to believe that I am this stupid.
> Here is my problem in a nutshell.  I have a Person and a collection of Jobs. The Jobs should be an indexed list based on the history that the person holds them.
> @Entity
> public class Person {
>     @Id @GeneratedValue
>     private long id;
>     @Column
>     private String name;
>     @OneToMany(cascade=ALL, fetch=LAZY, mappedBy = "job")
>     @IndexColumn(base = 1, name = "order")
>     private List<Job> jobs = new ArrayList<Job>();
>    // getters and setters
> }
> @Entity
> public class Job {
>     @Id @GeneratedValue
>     private long id;
>     @Column
>     private String name;
>     @ManyToOne
>     @JoinColumn(name="person_id")
>     private Person person;
>     @Column
>     private Integer order;
>    // getters and setters
> }
> If I do the following, I get NULL for order.
> Person person = new Person();
> person.setName("Chuck")
> Job job1 = new Job();
> job1.setName("sysadmin")
> job1.setPerson(person);
> person.getJobs().add(job1);
> Job job2 = new Job();
> jobs2.setName("network admin")
> job2.setPerson(person);
> person.getJobs().add(job2);
> entityManager.persist(person);
> Assume that the reason I am not assigning an order is more complex than this example. The point is that we want to see the order column populated with the index of the list.
> Now, if you give me the business about removing mappedBy, to that I will respond that by removing mappedBy, Hibernate tries to work with a person_job table, which I don't want. I want two tables, one for person and one for job.

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://opensource.atlassian.com/projects/hibernate/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira