May 2007 - hibernate-issues - Jboss List Archives

[Hibernate-JIRA] Created: (HHH-2318) Sometimes wrong classloader used for proxy interfaces

by Jan Wiemer (JIRA)

Sometimes wrong classloader used for proxy interfaces ----------------------------------------------------- Key: HHH-2318 URL: http://opensource.atlassian.com/projects/hibernate/browse/HHH-2318 Project: Hibernate3 Type: Bug Components: core Versions: 3.2.1 Reporter: Jan Wiemer For some of our business classes we used a Mapping declaring the class to be lazy initialized and providing a proxy interface like e.g.: <class name="TestClassImpl" proxy="TestClass" table="testTable" lazy="true"> ... </class> Using classes mapped this way - e.g. as endpoint of a one to one relation - sporadically leads to exceptions like the following: org.hibernate.PropertyAccessException: IllegalArgumentException occurred while calling setter of test.AbstractTestClass2Impl.testField at org.hibernate.property.BasicPropertyAccessor$BasicSetter.set(BasicPropertyAccessor.java:104) at org.hibernate.tuple.entity.AbstractEntityTuplizer.setPropertyValues(AbstractEntityTuplizer.java:337) at org.hibernate.tuple.entity.PojoEntityTuplizer.setPropertyValues(PojoEntityTuplizer.java:200) at org.hibernate.persister.entity.AbstractEntityPersister.setPropertyValues(AbstractEntityPersister.java:3514) at org.hibernate.engine.TwoPhaseLoad.initializeEntity(TwoPhaseLoad.java:129) at org.hibernate.loader.Loader.initializeEntitiesAndCollections(Loader.java:842) at org.hibernate.loader.Loader.doQuery(Loader.java:717) at org.hibernate.loader.Loader.doQueryAndInitializeNonLazyCollections(Loader.java:224) at org.hibernate.loader.Loader.doList(Loader.java:2211) at org.hibernate.loader.Loader.listIgnoreQueryCache(Loader.java:2095) at org.hibernate.loader.Loader.list(Loader.java:2090) at org.hibernate.loader.hql.QueryLoader.list(QueryLoader.java:388) at org.hibernate.hql.ast.QueryTranslatorImpl.list(QueryTranslatorImpl.java:338) at org.hibernate.engine.query.HQLQueryPlan.performList(HQLQueryPlan.java:172) at org.hibernate.impl.SessionImpl.list(SessionImpl.java:1121) at org.hibernate.impl.QueryImpl.list(QueryImpl.java:79) at org.hibernate.impl.AbstractQueryImpl.uniqueResult(AbstractQueryImpl.java:804) Note that the application is residing in a different classloader than hibernate (and CGLIB...). Examining the situation we see that the value passed to the setter was a CGLIB proxy. Internally the proxy stores an array of interfaces the proxy should implement. In our situation this interface contains the HibernateProxy interface and our business interface provided as proxy interface in the mapping. In this interface array we checked the classloader of the interfaces. As expected the HibernateProxy interface is loaded by the system classloader and our interface was loaded by our custom classloader. However examining the actual interfaces of the proxy (with proxy.getClass().getInterfaces()[i].getClassLoader() for all i) shows that all interfaces are loaded with the system classloader. This causes the exception above. Doing some more experiments we experience that the problem does not occur all the time. Sometimes the actual proxy interfaces are as expected (the HibernateProxy interface is loaded by the system classloader and our interface was loaded by our custom classloader). We notice that each time the test failed the HibernateProxy was the first interface in the interface array stored in the proxy. Some experiments with the CGLIB (the Enhancer class) shows us that (if there is no superclass given) they use the classloader of the first passed interface as default classloader (compare method net.sf.cglib.proxy.Enhancer.getDefaultClassLoader()). Finally we find out that hibernate passes the proxy interfaces in an arbitrary order since they are using a HashSet for the proxy interfaces in the method org.hibernate.tuple.entity.PojoEntityTuplizer.buildProxyFactory(PersistentClass persistentClass, Getter idGetter, Setter idSetter). This Hash set is passed to the method org.hibernate.proxy.pojo.cglib.CGLIB_ProxyFactory.postInstantiate(...). There it is simply converted to an Array (leading to a randomized order). As a workaround it is possible to patch the class org.hibernate.proxy.pojo.cglib.CGLIB_ProxyFactory and add reorganize the array if the HibernateProxy interface is the first one: this.interfaces = (Class[]) interfaces.toArray(NO_CLASSES); //----->PATCH<-------- if(this.interfaces.length > 1) { Class firstIfc = this.interfaces[0]; if(firstIfc.getName().startsWith("org.hibernate")) { this.interfaces[0] = this.interfaces[1]; this.interfaces[1] = firstIfc; System.err.println("Replace: " + firstIfc.getName() + " by " + this.interfaces[0].getName()); } } //-------------------- After applying this patch everything woks as expected. Compare with the discussion in: http://forum.hibernate.org/viewtopic.php?p=2334617#2334617 -- 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.... - For more information on JIRA, see: http://www.atlassian.com/software/jira

12 years, 4 months

4
11
0 / 0

[Hibernate-JIRA] Created: (HHH-2394) Support filter tag in subclass

by Welkin Hu (JIRA)

Support filter tag in subclass ------------------------------ Key: HHH-2394 URL: http://opensource.atlassian.com/projects/hibernate/browse/HHH-2394 Project: Hibernate3 Type: Improvement Components: core Versions: 3.0 beta 4 Reporter: Welkin Hu Assigned to: Steve Ebersole Fix For: 3.0 rc 1 According to a user: http://forum.hibernate.org/viewtopic.php?p=2229539#2229539 It's not possible to get a filter to apply to a subclass. This is Bad ;-) Steve, is that correct? -- 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.... - For more information on JIRA, see: http://www.atlassian.com/software/jira

12 years, 5 months

7
9
0 / 0

[Hibernate-JIRA] Created: (HHH-2158) incorrect hql query on one-to-one with property-ref

by Sebastien Cesbron (JIRA)

incorrect hql query on one-to-one with property-ref ---------------------------------------------------- Key: HHH-2158 URL: http://opensource.atlassian.com/projects/hibernate/browse/HHH-2158 Project: Hibernate3 Type: Bug Components: core Versions: 3.2.0.ga Environment: hibernate 3.2.0.ga with firebird 1.5.3 and jaybird 1.5.5 driver on windows XP Reporter: Sebastien Cesbron Attachments: testhib.zip I have a one-to-one relationship with property-ref between Master and Slave2. I want to find all Master instances that have a null Slave2 instance associated. To do so my query is select master from Master master where master.slave2 is null The sql generated is select master0_.oid as oid0_, master0_.libelle as libelle0_ from Master master0_ where master0_.oid is null which seems incorrect. It checks here Master instances with null id (config files are listed below). If I do my query like this select master from Master master where master.slave2.oid is null the generated sql is ok : select master0_.oid as oid0_, master0_.libelle as libelle0_ from Master master0_, Slave2 slave2x1_ where master0_.oid=slave2x1_.myMaster and (slave2x1_.oid is null) I have attached a small eclipse project that reproduces the problem This problem may-be related to the one I have submitted as issue HHH-1849 -- 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.... - For more information on JIRA, see: http://www.atlassian.com/software/jira

12 years, 5 months

6
7
0 / 0

[Hibernate-JIRA] Created: (ANN-609) Need a way to specify "unsaved-value" strategy using annotations

by Suji Suresh (JIRA)

Need a way to specify "unsaved-value" strategy using annotations ---------------------------------------------------------------- Key: ANN-609 URL: http://opensource.atlassian.com/projects/hibernate/browse/ANN-609 Project: Hibernate Annotations Issue Type: Bug Environment: Hibernate 3.2.2, Hibernate-annotations 3.2.1 Reporter: Suji Suresh Hibernate has two ways of specifying a value for primary key: 1. Assign a value before handing it over to Hibernate 2. Have Hibernate generate the value before persisting In my project I have objects of both of the above specified types. In other words I assign value for the primary key for some the objects (lets call these objects "assigned") and for others I have Hibernate generate the value before persisting (lets call these objects "generated"). Since Hibernate annotations does not support "unsaved-value", I do not have a way of specifying "unsaved-value" strategy. While my "generated" objects work perfectly with Hibernate's default "unsaved-value" strategy, I see wrong (but expected) behaviour when I try to persist an "assigned" object that is already present in the database in that, I get StaleStateException instead of DataIntegrityViolationException -- 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.... - For more information on JIRA, see: http://www.atlassian.com/software/jira

12 years, 6 months

6
9
0 / 0

[Hibernate-JIRA] Created: (HSEARCH-5) Only index an entity when it's index properties have physically changed

by Emmanuel Bernard (JIRA)

Only index an entity when it's index properties have physically changed ----------------------------------------------------------------------- Key: HSEARCH-5 URL: http://opensource.atlassian.com/projects/hibernate/browse/HSEARCH-5 Project: Hibernate Search Type: New Feature Components: engine Reporter: Emmanuel Bernard Priority: Minor -- 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.... - For more information on JIRA, see: http://www.atlassian.com/software/jira

12 years, 7 months

3
11
0 / 0

[Hibernate-JIRA] Created: (HHH-2614) Blob Length Set to 255 By Default With Derby DB

by Sean Sylvis (JIRA)

Blob Length Set to 255 By Default With Derby DB ----------------------------------------------- Key: HHH-2614 URL: http://opensource.atlassian.com/projects/hibernate/browse/HHH-2614 Project: Hibernate3 Issue Type: Bug Components: core Affects Versions: 3.2.3 Environment: Apache Derby 10.2.2, Java 1.6.0 Reporter: Sean Sylvis When the (Hibernate) type "blob" is used in an object mapping file and a length is not specified, Hibernate converts this type as "blob (255)" to the Apache Derby database table definition. (This seems to be the same as the conversion behavior of a Java String to "varchar (255).") The same behavior results when using a CustomType that returns a Java SQL Types.BLOB. This is not the behavior I expected. Hibernate converts the same type as simply "blob" (which has a maximum length of 65535 bytes when no length is specified) when using MySQL version 5.0 (http://dev.mysql.com/doc/refman/5.0/en/storage-requirements.html). Furthermore, the Derby specification (http://db.apache.org/derby/docs/10.2/ref/rrefblob.html) states that a blob type without a length value will default to a length of 2M-1 bytes. I could not find a length specification for the Java SQL Blob type (http://java.sun.com/javase/6/docs/api/), but the current default length does seem to conform to other Hibernate type standards. That is, if a length is not specified, the length defaults to the maximum length defined for that type by the database. -- 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.... - For more information on JIRA, see: http://www.atlassian.com/software/jira

12 years, 8 months

9
13
0 / 0

[Hibernate-JIRA] Created: (HHH-2085) org.hibernate.event.def.DefaultLoadEventListener logs exception at info level

by Don Smith (JIRA)

org.hibernate.event.def.DefaultLoadEventListener logs exception at info level ----------------------------------------------------------------------------- Key: HHH-2085 URL: http://opensource.atlassian.com/projects/hibernate/browse/HHH-2085 Project: Hibernate3 Type: Bug Components: core Versions: 3.2.0.cr2 Environment: Hibernate 3.2.0.cr2, MySQL 4.0.18 Reporter: Don Smith Priority: Minor Prior to schema generation, accessing a persistent object throws an exception, as expected. The org.hibernate.event.def.DefaultLoadEventListener logs the exception at the info level, which just adds noise to the application console, and prevents application exception handling (in my case it's valid if the table doesn't exist, so I just want to catch the exception and continue). This log statement should be removed, or moved to a debug or trace level, so it won't show up in normal execution. Code, from line 94 of org.hibernate.event.def.DefaultLoadEventListener: catch(HibernateException e) { log.info("Error performing load command", e); throw e; } -- 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.... - For more information on JIRA, see: http://www.atlassian.com/software/jira

12 years, 8 months

2
1
0 / 0

[Hibernate-JIRA] Created: (HHH-2548) when "order by" a composit object, which may be null, the result list returns only when the composit objcts not null

by Yu Deng (JIRA)

when "order by" a composit object, which may be null, the result list returns only when the composit objcts not null -------------------------------------------------------------------------------------------------------------------- Key: HHH-2548 URL: http://opensource.atlassian.com/projects/hibernate/browse/HHH-2548 Project: Hibernate3 Issue Type: Bug Affects Versions: 3.2.3 Environment: tomcat 5.5, eclipse3.2, jre1.5, postgreSQL8.1, hibernate3 and hibernate3.2 Reporter: Yu Deng Priority: Critical <hibernate-mapping package="com.grapevinecs.ams.domain.perbranch.inv"> <class name="InvItem"> <id name="id" column="id"> <generator class="hilo"/> </id> ... <many-to-one name="transferredDetails"/>  .... </class> (another mapping file) </hibernate-mapping>> <hibernate-mapping package="com.grapevinecs.ams.domain.perbranch.inv"> <class name="InvTransferDetails" > <id name="id" column="id"> <generator class="hilo"/> </id> <property name="returnedToICBC" type="com.grapevinecs.ams.dao.BooleanInteger" not-null="true"/> ... </class> </hibernate-mapping> Probelm: 1) The result returns 3000 records ---- select count(*) from InvItem c where upper(c.invType) LIKE '%D%' and c.invSeries.closed = '0' and c.branch =60551 order by c.transferredDetails.returnedToICBC asc, c.stamp asc,c.invNmbr asc 2) result List return size 0 by this query --- from InvItem c where upper(c.invType) LIKE '%D%' and c.invSeries.closed = '0' and c.branch =60551 order by c.transferredDetails.returnedToICBC asc, c.stamp asc,c.invNmbr asc 1) is correct but 2) is not correct. only when transferredDetails in InvItem not null be returned. in case 2), I have no transferredDetails not null. after I excuted the query 2), even I remove "c.transferredDetails.returnedToICBC asc", the result is still return nothing. -- 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.... - For more information on JIRA, see: http://www.atlassian.com/software/jira

12 years, 9 months

3
2
0 / 0

[Hibernate-JIRA] Created: (EJB-265) Incorrect SQL generated for JPA QL Update statement involving multiple entities

by Alex Savitsky (JIRA)

Incorrect SQL generated for JPA QL Update statement involving multiple entities ------------------------------------------------------------------------------- Key: EJB-265 URL: http://opensource.atlassian.com/projects/hibernate/browse/EJB-265 Project: Hibernate Entity Manager Type: Bug Versions: 3.2.1 Environment: Hibernate 3.2.2 GA, Oracle 9i Reporter: Alex Savitsky When generating SQL for statements like this: UPDATE Issue i SET i.dueDate = i.dueDate + 10 WHERE i.visit.branch.id = :branchId, an extra comma is generated at the end of table list: update Issue, set DUE_DATE=DUE_DATE+10 where BRANCH_ID=? -- 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.... - For more information on JIRA, see: http://www.atlassian.com/software/jira

12 years, 10 months

7
18
0 / 0

[Hibernate-JIRA] Created: (HHH-2125) Hibernate XML Mapping should not depend on dom4j

by Don Smith (JIRA)

Hibernate XML Mapping should not depend on dom4j ------------------------------------------------ Key: HHH-2125 URL: http://opensource.atlassian.com/projects/hibernate/browse/HHH-2125 Project: Hibernate3 Type: Improvement Components: core Versions: 3.2.0 cr1, 3.2.0.cr2, 3.2.0.cr3, 3.2.0.cr4 Environment: all Reporter: Don Smith Hibernate XML mapping requires dom4j, which is one of the packages that causes the Java version of "dll hell" - applications requiring conflicting versions of jars with no way to resolve the problem in a given release of an application server or package. HIbernate should be converted to use the standard XML APIs included in the JRE 5.0 specification, when Hibernate no longer needs to support JRE 1.4.x. -- 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.... - For more information on JIRA, see: http://www.atlassian.com/software/jira

12 years, 11 months

6
10
0 / 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

hibernate-issues May 2007