@TableGenerator无法与EclipseLink JPA 2.0一起使用 [英] @TableGenerator not working with EclipseLink JPA 2.0
问题描述
我有一个奇怪的问题。
我在JPA应用程序中使用表生成策略。直到最近，我还是使用Toplink持久性提供程序作为JPA实现。一切正常。
刚才我添加了TopLink不支持的某些继承功能（每个类继承的表），因此我不得不移至EclipseLink。
除ID Generation之外，其他所有方法都可以正常工作。
这是我用于所有唯一ID的代码示例：
I am having a weird issue. i am using Table Generation strategy in my JPA application. until recently, I used Toplink persistence provider as the JPA implementation. All worked fine. Just now I added some inheritance features that are not supported by TopLink (Table per Class Inheritance), so I had to move to EclipseLink. All works fine, except for the ID Generation. Here is the code example which I use for all my unique IDs:
@Id
@TableGenerator(name="INCOME_SEQ", table="SEQUENCE_TABLE", pkColumnName="SEQUENCE_NAME",
valueColumnName="SEQUENCE_COUNT", pkColumnValue="INCOME_SEQ", allocationSize=1)
@GeneratedValue(strategy=GenerationType.TABLE, generator="INCOME_SEQ")
@Column(name = "INCOME_ID")
private Integer incomeId;
完全相同的代码适用于Toplink，但不适用于EclipseLink。
我得到的错误是：
The exact same code works with Toplink but does not work with EclipseLink. The Error I get is:
Internal Exception: com.mysql.jdbc.exceptions.jdbc4.MySQLIntegrityConstraintViolationException: Column 'INCOME_ID' cannot be null
Error Code: 1048
似乎表代是根本不使用...
It seems like the Table generation is not used at all...
这是MySql数据库中的表说明：
表名：SEQUENCE_TABLE
Here is the Table Description in my MySql database: Table name: SEQUENCE_TABLE
+----------------+--------------+------+-----+---------+-------+
| Field | Type | Null | Key | Default | Extra |
+----------------+--------------+------+-----+---------+-------+
| SEQUENCE_NAME | varchar(64) | NO | PRI | NULL | |
| SEQUENCE_COUNT | int(11) | NO | | 0 | |
| localized_name | varchar(128) | YES | | NULL | |
| USER_ID | int(11) | NO | MUL | 1 | |
+----------------+--------------+------+-----+---------+-------+
有人可以吗指导我解决问题？？还是至少获得未创建ID的实际原因的方法？
Can anyone please direct me to the problem?? or at least a way to get the actual reason the ID is not being created?
编辑：这是收入类和表结构：
This is the Income class and Table structure:
+-------------------------+-------------------------------------------------------------------------------+------+-----+---------+-------+
| Field | Type | Null | Key | Default | Extra |
+-------------------------+-------------------------------------------------------------------------------+------+-----+---------+-------+
| INCOME_ID | int(11) | NO | PRI | NULL | |
| REFERENCE_DOCUMENT_ID | varchar(32) | YES | | NULL | |
| CLIENT_ID | int(9) | NO | MUL | NULL | |
| income_date | datetime | NO | | NULL | |
| MOVEMENT_CATEGORY_ID | int(11) | NO | MUL | NULL | |
| SUM_AFTER_TAX | decimal(10,2) | YES | | 0.00 | |
| TOTAL_SUM | decimal(10,2) | NO | | NULL | |
| TAX_SUM | decimal(10,2) | NO | | 0.00 | |
| VAT_PERCENT | decimal(10,2) | NO | | NULL | |
| DESCRIPTION | varchar(100) | YES | | NULL | |
| STATUS | smallint(5) unsigned | YES | | 0 | |
| REFERENCE_DOCUMENT_TYPE | enum('MANUAL_INVOICE','MANUAL_RECEIPT','INVOICE','RECEIPT','INVOICE_RECEIPT') | YES | | NULL | |
| USER_ID | int(11) | NO | MUL | 1 | |
| UNIQUE_USER_SEQUENCE | int(10) unsigned | YES | | NULL | |
+-------------------------+-------------------------------------------------------------------------------+------+-----+---------+-------+
编辑：我在EclipseLink日志记录级别设置为最佳的情况下运行了该应用程序，但仍然看不到任何错误。我什至可以看到序列已正确初始化：
I ran the app with EclipseLink Logging level set to finest, but I still can't see anything wrong. I even can see the sequencing being correctly initialized:
[EL Finest]: 2012-11-07 20:37:00.359--ServerSession(1832413009)--Connection(1484802679)--Thread(Thread[pool-1-thread-2,5,main])--Connection released to connection pool [default].
[EL Finest]: 2012-11-07 20:37:00.405--ServerSession(1832413009)--Thread(Thread[pool-1-thread-2,5,main])--sequencing connected, state is NoPreallocation_State
[EL Finest]: 2012-11-07 20:37:00.406--ServerSession(1832413009)--Thread(Thread[pool-1-thread-2,5,main])--sequence SEQ_GEN_IDENTITY: preallocation size 1
[EL Finest]: 2012-11-07 20:37:00.406--ServerSession(1832413009)--Thread(Thread[pool-1-thread-2,5,main])--sequencing connected, state is Preallocation_Transaction_NoAccessor_State
[EL Finest]: 2012-11-07 20:37:00.406--ServerSession(1832413009)--Thread(Thread[pool-1-thread-2,5,main])--sequence RECEIPTS_SEQ: preallocation size 1
[EL Finest]: 2012-11-07 20:37:00.406--ServerSession(1832413009)--Thread(Thread[pool-1-thread-2,5,main])--sequence INCOME_SEQ: preallocation size 1
[EL Finest]: 2012-11-07 20:38:09.264--UnitOfWork(514772947)--Thread(Thread[AWT-EventQueue-1,6,main])--persist() operation called on: 555 וובה.
[EL Finer]: 2012-11-07 20:38:09.265--UnitOfWork(514772947)--Thread(Thread[AWT-EventQueue-1,6,main])--begin unit of work commit
[EL Finest]: 2012-11-07 20:38:09.288--UnitOfWork(514772947)--Thread(Thread[AWT-EventQueue-1,6,main])--Execute query DoesExistQuery(referenceClass=MovementCategories )
[EL Finest]: 2012-11-07 20:38:09.29--UnitOfWork(514772947)--Thread(Thread[AWT-EventQueue-1,6,main])--Execute query InsertObjectQuery(חש' ידנית 555 וובה)
[EL Finest]: 2012-11-07 20:38:09.291--ServerSession(1832413009)--Connection(1484802679)--Thread(Thread[AWT-EventQueue-1,6,main])--Connection acquired from connection pool [default].
[EL Finer]: 2012-11-07 20:38:09.291--ClientSession(2117615354)--Connection(1484802679)--Thread(Thread[AWT-EventQueue-1,6,main])--begin transaction
[EL Fine]: 2012-11-07 20:38:09.293--ClientSession(2117615354)--Connection(1484802679)--Thread(Thread[AWT-EventQueue-1,6,main])--INSERT INTO income (INCOME_ID, DESCRIPTION, INCOME_DATE, REFERENCE_DOCUMENT_ID, REFERENCE_DOCUMENT_TYPE, STATUS, SUM_AFTER_TAX, TAX_SUM, TOTAL_SUM, UNIQUE_USER_SEQUENCE, VAT_PERCENT, CLIENT_ID, MOVEMENT_CATEGORY_ID) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
bind => [null, null, 2012-11-07 20:37:55.225, 555, MANUAL_INVOICE, 1, 100.0, 16.5, 116.5, null, 16.5, 334, 3]
我看到的唯一线索是 Preallocation_Transaction_NoAccessor_State 属性...
也许有人可以帮助我理解这个。如果序列化主键ID，则日志输出应该是什么？
The only clue I see is the Preallocation_Transaction_NoAccessor_State property...
Maybe someone can help me understand this. What should be the logging output in case of sequenced primary key ID?
推荐答案
找到了解决方案，希望它对更多人有用：
由于某些原因，我有以下内容为我的持久性单元设置的持久性配置设置：
Found The Solution, hope it will be useful for some more people: For some reason I had the following persistence config setting set for my persistence unit:
persistenceMap.put(PersistenceUnitProperties.ID_VALIDATION,"NONE");
显然，在EclipseLink中关闭ID验证也会关闭ID生成。
Apparently turning off ID validation in EclipseLink also turns off ID Generation.
感谢所有尝试提供帮助的人！
Thanks to all who tried to help!
这篇关于@TableGenerator无法与EclipseLink JPA 2.0一起使用的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！
