Re: [dbunit-user] Using DbUnit with Hibernate &javax.persistence.TableGenerator

[Davis F. <dav...@gm...>]

Hi Mohamed -- I solved the issue.  I had changed
DatabaseOperation.CLEAN_INSERT.execute() to
DatabaseOperation.INSERT.execute() earlier when I was trying to debug
a different issue.  Moving it back to CLEAN_INSERT removes the
duplicates in the ID_GEN table.

Regards,
Davis

On Thu, Feb 21, 2008 at 12:42 PM, Ragab, Mohamed H
<moh...@ed...> wrote:
> Hi Davis,
>
>  Have you tried including the table ID_GEN in xml dataset for the test,
>  so that DBUNIT cleans it as well
>
>  Regards,
>
>  Mohamed H. Ragab
>
>
>
>
>  -----Original Message-----
>  From: dbu...@li...
>  [mailto:dbu...@li...] On Behalf Of Davis
>  Ford
>  Sent: Thursday, February 21, 2008 6:12 PM
>  To: dbunit-user
>  Subject: [dbunit-user] Using DbUnit with Hibernate
>  &javax.persistence.TableGenerator
>
>  Hi, I'm using Hibernate with JPA annotations.  For my PK_ID generation,
>  I'm using TableGenerator ->
>
>  http://java.sun.com/javaee/5/docs/api/javax/persistence/TableGenerator.h
>  tml
>
>  Here's an example for an entity class name QRule (maps to table QRULE)
>  ->
>
>     /** The database primary key for this object. */
>     @Id
>     @GeneratedValue(strategy = GenerationType.TABLE, generator =
>  "qRuleGen")
>     @TableGenerator(name = "qRuleGen", table = "ID_GEN", pkColumnName =
>  "GEN_KEY", valueColumnName = "GEN_VALUE", pkColumnValue = "QRULE_ID",
>  allocationSize = 1)
>     private long id;
>
>  So, this means there is a table in the database ID_GEN with GEN_KEY =
>  "QRULE" and GEN_VALUE = (next valid pk_id).
>
>  This makes things tricky for dbunit datasets.  I have found that if I
>  create an XML dataset like this:
>
>  For a particular unit test, I use hibernate's create-drop feature to
>  export the DDL.  This happens only for test setup (i.e. @BeforeClass).
>   Then for each test method (i.e. @Before), I do CLEAN_INSERT.execute(),
>  but what ends up happening is that the ID_GEN table is not cleaned --
>  and I end up getting duplicates in it.
>
>  Has anyone else run into this, and is there some way around this issue?
>
>  Thanks in advance,
>  Davis
>
>  ------------------------------------------------------------------------