Links: Table of Contents | Single HTML | Single PDF

JAXB Release Documentation


Table of Contents

Overview
1. Documentation
2. Software Licenses
3. Sample Apps
3.1. Using the Runtime Binding Framework
Release Notes
1. Java™ 2 Platform, Standard Edition (J2SE™) Requirements
2. Identifying the JAR Files
3. Locating the Normative Binding Schema
4. Changelog
4.1. Changes between 2.2.6 to 2.2.7
4.2. Changes between 2.2.5-2 to 2.2.6
4.3. Changes between 2.2.5-1 to 2.2.5-2
4.4. Changes between 2.2.5 to 2.2.5-1
4.5. Changes between 2.2.4-1 to 2.2.5
4.6. Changes between 2.2.4 to 2.2.4-1
4.7. Changes b±etween 2.2.3u2 to 2.2.4
4.8. Notable Changes between 2.2.3u1 to 2.2.3u2
4.9. Notable Changes between 2.2.3 to 2.2.3u1
4.10. Notable Changes between 2.2.2 to 2.2.3
4.11. Notable Changes between 2.2.1.1 to 2.2.2
4.12. Notable Changes between 2.2.1 to 2.2.1.1
4.13. Notable Changes between 2.2 to 2.2.1
4.14. Notable Changes between 2.1.12 to 2.2
4.15. Notable Changes between 2.1.12 to 2.1.13
4.16. Notable Changes between 2.1.11 to 2.1.12
4.17. Notable Changes between 2.1.10 to 2.1.11
4.18. Notable Changes between 2.1.9 to 2.1.10
4.19. Notable Changes between 2.1.8 to 2.1.9
4.20. Notable Changes between 2.1.7 to 2.1.8
4.21. Notable Changes between 2.1.6 to 2.1.7
4.22. Notable Changes between 2.1.5 to 2.1.6
4.23. Notable Changes between 2.1.4 to 2.1.5
4.24. Notable Changes between 2.1.3 to 2.1.4
4.25. Notable Changes between 2.1.2 to 2.1.3
4.26. Notable Changes between 2.1.1 to 2.1.2
4.27. Notable Changes between 2.1 First Customer Ship (FCS) to 2.1.1
4.28. Notable Changes between 2.1 Eary Access 2 to 2.1 First Customer Ship (FCS)
4.29. Notable Changes between 2.1 Eary Access to 2.1 Early Access 2
4.30. Notable Changes between 2.0.2 to 2.1 Early Access
4.31. Notable Changes between 2.0.2 to 2.0.3
4.32. Notable Changes between 2.0.1 to 2.0.2
4.33. Notable Changes between 2.0 to 2.0.1
4.34. Notable Changes between 2.0 Early Access 3 and 2.0 First Customer Ship (FCS)
4.35. Notable Changes between 2.0 Early Access 2 and 2.0 Early Access 3
4.36. Notable Changes between 2.0 Early Access and 2.0 Early Access 2
4.37. Notable Changes between 1.0.x FCS and 2.0 Early Access
JAXB Users Guide
1. Compiling XML Schema
1.1. Dealing with errors
1.2. Fixing broken references in schema
1.3. Mapping of <xs:any />
1.4. Mapping of <xs:element /> to JAXBElement
1.5. How modularization of schema interacts with XJC
1.6. Adding behaviors
1.7. Avoid strong databinding
1.8. Working with generated code in memory
2. Customization of Schema Compilation
2.1. Customizing Java packages
2.2. Using SCD for customizations
2.3. Using different datatypes
3. Annotating Your Classes
3.1. Mapping your favorite class
3.2. Mapping interfaces
3.3. Evolving annotated classes
3.4. XML layout and in-memory data layout
3.5. Mapping cyclic references to XML
4. Unmarshalling
4.1. @XmlRootElement and unmarshalling
4.2. Unmarshalling is not working! Help!
4.3. Element default values and unmarshalling
4.4. Dealing with large documents
5. Marshalling
5.1. Changing prefixes
5.2. Element default values and marshalling
5.3. Different ways of marshalling
5.4. Interaction between marshalling and DOM
6. Schema Generation
6.1. Invoking schemagen programatically
6.2. Generating Schema that you want
7. Deployment
7.1. Migrating JAXB 2.0 applications to JavaSE 6
7.2. Which JAXB RI is included in which JDK?
7.3. Running JAXB 1.0 and 2.x side by side
7.4. Migrating JAXB 1.0 applications to JAXB 2.x
7.5. Runtime Errors
8. Other Miscellaneous Topics
8.1. Performance and thread-safety
8.2. Compiling DTD
8.3. Using JAXB from Maven
8.4. Designing a client/server protocol in XML
Tools
1. XJC
1.1. xjc Overview
1.2. Launching xjc
1.3. xjc Syntax
1.4. Compiler Restrictions
1.5. Generated Resource Files
2. XJC Ant Task
2.1. xjc Task Overview
2.2. xjc Task Attributes
2.3. Generated Resource Files
2.4. Up-to-date Check
2.5. Schema Language Support
2.6. xjc Examples
3. SchemaGen
3.1. schemagen Overview
3.2. Launching schemagen
3.3. schemagen Syntax
3.4. Generated Resource Files
4. SchemaGen Ant Task
4.1. schemagen Task Overview
4.2. schemagen Task Attributes
4.3. schemagen Examples
5. 3rd Party Tools
5.1. Maven JAXB Plugin
5.2. JAXB Workshop
5.3. XJC Plugins
5.4. RDBMS Persistence
JAXB RI Extensions
1. Overview
2. Runtime Properties
2.1. Marshaller Properties
3. XJC Customizations
3.1. Customizations
4. DTD
4.1. DTD
5. Develop Plugins
5.1. What Can A Plugin Do?
Frequently Asked Questions
Related Articles
1. Introductory
2. Blogs
3. Interesting articles

Overview

The Java™ Architecture for XML Binding (JAXB) provides an API and tools that automate the mapping between XML documents and Java objects.

The JAXB framework enables developers to perform the following operations:

  • Unmarshal XML content into a Java representation

  • Access and update the Java representation

  • Marshal the Java representation of the XML content into XML content

JAXB gives Java developers an efficient and standard way of mapping between XML and Java code. Java developers using JAXB are more productive because they can write less code themselves and do not have to be experts in XML. JAXB makes it easier for developers to extend their applications with XML and Web Services technologies.

1. Documentation

Documentation for this release consists of the following:

2. Software Licenses

3. Sample Apps

This page summarizes basic use-cases for Java-2-Schema, Schema-2-Java, and lists all of the sample applications that ship with JAXB.

3.1. Using the Runtime Binding Framework

3.1.1. Schema-2-Java

Schema-2-Java is the process of compiling one or more schema files into generated Java classes. Here are some of the basic steps for developing an app:

  1. Develop/locate your schema

  2. Annotate the schema with binding customizations if necessary (or place them in an external bindings file)

  3. Compile the schema with the XJC binding compiler

  4. Develop your JAXB client application using the Java content classes generated by the XJC binding compiler along with the javax.xml.bind runtime framework

  5. Set your CLASSPATH to include all of the Identifying the JAR Files

  6. Compile all of your Java sources with javac

  7. Run it!

3.1.2. Java-2-Schema

Java-2-Schema is the process of augmenting existing Java classes with the annotations defined in the javax.xml.bind.annotation package so that the JAXB runtime binding framework is capable of performing the (un)marshal operations. Here are the basic steps for developing an app:

  1. Develop your data model in Java

  2. Apply the javax.xml.bind.annotation annotations to control the binding process

  3. Set your CLASSPATH to include all of the Identifying the JAR Files

  4. Compile your data model with javac

    Important

    Make sure that you CLASSPATH includes jaxb-xjc.jar before running javac.

  5. The resulting class files will contain your annotations as well other default annotations needed by the JAXB runtime binding framework

  6. Develop your client application that uses the data model and develop the code that uses the JAXB runtime binding framework to persist your data model using the (un)marshal operations.

  7. Compile and run your client application!

For more information about this process, see the the Java WSDP Tutorial and the extensive Sample Apps documentation.

3.1.3. Building and Running the Sample Apps with Ant

To run the sample applications, just go into each sample directory, and run ant without any option.

A few sample applications do not use Ant. For those samples, refer to the included readme.txt files for instructions.

3.1.4. List of Sample Apps

samples/catalog-resolver

This example demonstrates how to use the -catalog compiler switch to handle references to schemas in external web sites.

samples/character-escape

This example shows how you can use the new JAXB RI Marshaller property com.sun.xml.bind.characterEscapeHandler to change the default character escaping behavior.

samples/class-resolver

This little DI-container-by-JAXB example demonstrates how one can avoid passing in a list of classes upfront, and instead load classes lazily.

samples/create-marshal

This sample application demonstrates how to use the ObjectFactory class to create a Java content tree from scratch and marshal it to XML data. It also demonstrates how to add content to a JAXB List property.

samples/cycle-recovery

JAXB RI's vendor extension CycleRecoverable provides application a hook to handle cycles in the object graph. Advanced.

samples/datatypeconverter

This sample application is very similar to the inline-customize sample application (formerly SampleApp6), but illustrates an easier, but not as robust, <jaxb:javaType> customization.

samples/dtd

This sample application illustrate some of the DTD support available in the JAXB RI's extension mode. Please refer to the JAXB RI Extensions page for more detail.

samples/element-substitution

This sample application illustrates how W3C XML Schema substitution groups are supported in JAXB RI's extension mode. Please refer to the JAXB RI Extensions page for more detail.

samples/external-customize

This sample application is identical to the datatypeconverter sample application (formerly SampleApp7) except that the binding customizations are contained in an external binding file.

samples/fix-collides

Another binding customization example that illustrates how to resolve name conflicts. Running this sample without the binding file will result in name collisions (see readme.txt) . Running ant will use the binding customizations to resolve the name conflicts while compiling the schema.

samples/inline-customize

This sample application demonstrates how to customize the default binding produced by the XJC binding compiler.

samples/j2s-create-marshal

This sample application demonstrates marshalling, unmarshalling and unmarshal validation with existing Java classes annotated with JAXB annotations.

samples/j2s-xmlAccessorOrder

This sample application demonstrates the use of mapping annotations @XmlAccessorOrder and @XmlType.propOrder in Java classes for ordering properties and fields in Java to schema bindings.

samples/j2s-xmlAdapter

This sample application demonstrates the use of interface XmlAdapter and annotation XmlJavaTypeAdapter for custom marshaling/unmarshaling XML content into/out of a Java type.

samples/j2s-xmlAttribute

This sample application demonstrates the use of annotation @XmlAttribute for defining Java properties and fields as XML attributes.

samples/j2s-xmlRootElement

This sample application demonstrates the use of annotation @XmlRootElement to define a class to be an XML element.

samples/j2s-xmlSchemaType

This sample application demonstrates the use of annotation @XmlSchemaType to customize the mapping of a property or field to an XML built-in type.

samples/j2s-xmlType

This sample application demonstrates the use of mapping annotations @XmlAccessorOrder and @XmlType.propOrder in Java classes for ordering properties and fields in Java to schema bindings.

samples/locator-support

This sample shows how to use the new non-standard locator support. By following the instructions in the readme.txt file, you can cause all of the generated impl classes to implement a new interface that provides more information about error locations. When a ValidationEvent happens on your content tree, simply retrieve the object and cast it down to com.sun.xml.bind.extra.Locatable.

samples/modify-marshal

This sample application demonstrates how to modify a java content tree and marshal it back to XML data.

samples/namespace-prefix

This sample application demonstrates how to use the new JAXB RI Marshaller property com.sun.xml.bind.namespacePrefixMapper to customize the namespace prefixes generated during marshalling.

samples/partial-unmarshalling

In this example, the input document will be unmarshalled a small chunk at a time, instead of unmarshalling the whole document at once.

samples/pull-parser

This sample app demonstrates how a pull-parser can be used with JAXB to increase the flexibility of processing.

samples/streaming-unmarshalling

This example illustrates a different approach to the streaming unmarshalling, which is suitable for processing a large document.

samples/synchronized-methods

This sample shows how to use the new non-standard synchronized method support. By following the instructions in the readme.txt, you can cause all of the generated impl class methods signatures to contain the synchronized keyword.

samples/type-substitution

This sample app demonstrates type substitution using the W3C XML Schema Part 0: Primer international purchase order schema.

samples/ubl

This project processes a UBL (Universal Business Language) order instance and prints a report to the screen.

samples/unmarshal-read

This sample application demonstrates how to unmarshal an instance document into a Java content tree and access data contained within it.

samples/unmarshal-validate

This sample application demonstrates how to enable validation during the unmarshal operations.

samples/updateablePartialBind

This sample application demonstrates how to partially map a DOM tree to JAXB (using JAXP 1.3 XPath), modify JAXB mapped instance and then update modifications back to the DOM tree.

samples/vendor-extensions

This example demonstrates how to use <xjc:superClass> vendor extensions provided by Sun's JAXB RI, as well as <jaxb:serializable> customization.

samples/xml-channel

This example demonstrates how one can use one communication channel (such as a socket) to send multiple XML messages, and how it can be combined with JAXB.

samples/xml-stylesheet

A common customization need for the marshalling output is about introducing extra processing instruction and/or DOCTYPE declaration. This example demonstrates how such modification can be done easily.

Release Notes

Table of Contents

1. Java™ 2 Platform, Standard Edition (J2SE™) Requirements
2. Identifying the JAR Files
3. Locating the Normative Binding Schema
4. Changelog
4.1. Changes between 2.2.6 to 2.2.7
4.2. Changes between 2.2.5-2 to 2.2.6
4.3. Changes between 2.2.5-1 to 2.2.5-2
4.4. Changes between 2.2.5 to 2.2.5-1
4.5. Changes between 2.2.4-1 to 2.2.5
4.6. Changes between 2.2.4 to 2.2.4-1
4.7. Changes between 2.2.3u2 to 2.2.4
4.8. Notable Changes between 2.2.3u1 to 2.2.3u2
4.9. Notable Changes between 2.2.3 to 2.2.3u1
4.10. Notable Changes between 2.2.2 to 2.2.3
4.11. Notable Changes between 2.2.1.1 to 2.2.2
4.12. Notable Changes between 2.2.1 to 2.2.1.1
4.13. Notable Changes between 2.2 to 2.2.1
4.14. Notable Changes between 2.1.12 to 2.2
4.15. Notable Changes between 2.1.12 to 2.1.13
4.16. Notable Changes between 2.1.11 to 2.1.12
4.17. Notable Changes between 2.1.10 to 2.1.11
4.18. Notable Changes between 2.1.9 to 2.1.10
4.19. Notable Changes between 2.1.8 to 2.1.9
4.20. Notable Changes between 2.1.7 to 2.1.8
4.21. Notable Changes between 2.1.6 to 2.1.7
4.22. Notable Changes between 2.1.5 to 2.1.6
4.23. Notable Changes between 2.1.4 to 2.1.5
4.24. Notable Changes between 2.1.3 to 2.1.4
4.25. Notable Changes between 2.1.2 to 2.1.3
4.26. Notable Changes between 2.1.1 to 2.1.2
4.27. Notable Changes between 2.1 First Customer Ship (FCS) to 2.1.1
4.28. Notable Changes between 2.1 Eary Access 2 to 2.1 First Customer Ship (FCS)
4.29. Notable Changes between 2.1 Eary Access to 2.1 Early Access 2
4.30. Notable Changes between 2.0.2 to 2.1 Early Access
4.31. Notable Changes between 2.0.2 to 2.0.3
4.32. Notable Changes between 2.0.1 to 2.0.2
4.33. Notable Changes between 2.0 to 2.0.1
4.34. Notable Changes between 2.0 Early Access 3 and 2.0 First Customer Ship (FCS)
4.35. Notable Changes between 2.0 Early Access 2 and 2.0 Early Access 3
4.36. Notable Changes between 2.0 Early Access and 2.0 Early Access 2
4.37. Notable Changes between 1.0.x FCS and 2.0 Early Access

This document contains information that should help you use this software library more effectively. See the Frequently Asked Questions for additional information.

The most up-to-date version of this document can be found on-line.

1. Java™ 2 Platform, Standard Edition (J2SE™) Requirements

This release of the JAXB Reference Implementation requires Java SE 6 or higher.

2. Identifying the JAR Files

Use

Description

Jars

Runtime

Jars required to deploy a JAXB 2.x client

jaxb-api.jar(in endorsed dir), jaxb-core.jar, jaxb-impl.jar,

Compiler

Jars required at your development environment (but not runtime)

jaxb-core.jar jaxb-xjc.jar jaxb-jxc.jar jaxb-impl.jar

3. Locating the Normative Binding Schema

You may find information about the normative binding schema defined in the JAXB Specification at http://java.sun.com/xml/ns/jaxb.

4. Changelog

The JAXB 2.x RI is a major reimplementation to meet the requirements of the 2.x specifications.

Please refer to the ??? for older releases.

4.1. Changes between 2.2.6 to 2.2.7

  • JAXB 1.0 runtime is restored. It's bundled into the jaxb-extra-osgi.jar now. Required libraries are located in $JAXB_HOME/tools/jaxb1_libs folder.
  • The processing of -source XJC option is removed. We use 2.0 version of the schema compiler. For generating 1.0.x code, please use an installation of the 1.0.x codebase.
  • resolver.jar is not bundled by JAXB anymore. It's already part of JDK. If you are using JDK which doesn't contain it - you have to download it manually (e.g. from maven: com.sun.org.apache.xml.internal:resolver).
  • JAXB project was split into several modules: core, impl, xjc and jxc. So now it is possible to use XJC with other JAXB runtime implementation (e.g. EclipseLink MOXy). While using XJC be aware of having any runtime JAXB implementation (respectively jaxb-impl.jar) on classpath.
  • Bug fixes:
                * [JAXB-452] - Embedded DOM Elements may lose locally defined namespace prefixes
                * [JAXB-915] - Cannot generate java classes from xjc
                * [JAXB-919] - xjc - proxy authentication fails
                * [JAXB-922] - JAXB samples fail
                * [JAXB-939] - '#' at the end of a namespace : different of behavior, sometimes a underscore is added,
                    sometimes not.
                * [JAXB-940] - wrong default value for "target" parameter for xjc ant task
                * [JAXB-948] - JAXB differences in JDK 7 in
                    com/sun/xml/internal/bind/v2/runtime/reflect/Lister.java#Lister.CollectionLister.endPacking
            
  • Improvement:
                * [JAXB-912] - XJC split
            

4.2. Changes between 2.2.5-2 to 2.2.6

  • Project cleaning done:
    • JAXB 1.0 is not bundled anymore.
    • Libraries jaxb1-xjc.jar, msv.jar was removed.
    • Package com/sun/msv/datatype was extracted to jaxb-extra-osgi.jar
  • Bug fixes:
                * [JAXB-890] - JAXB Unmarshaller tries to instantiate abstract class ignoring xsi:type if nillable="true".
                * [JAXB-871] - Disabled fields and multiple-inherence (override once => override for sub-classes).
                * [JAXB-900] - MarshalException for XmlIDREF used on interfaces.
            

4.3. Changes between 2.2.5-1 to 2.2.5-2

  • Fixed version number in OSGi bundles

4.4. Changes between 2.2.5 to 2.2.5-1

  • JAX_WS-1059 - wsimport Ant tasks causes NoClassDefFoundError

4.5. Changes between 2.2.4-1 to 2.2.5

    * [JAXB-415] - Marshaller.marshall throws NPE if an adapter adapts a non-null bound value to null.
    * [JAXB-488] - compatibility issue with JAXB 2.0
    * [JAXB-608] - javax.xml.bind.DatatypeConverter.parseBoolean incorrect
    * [JAXB-616] - XMLCatalog not used from xjc/xjctask when strict validation is enabled
    * [JAXB-617] - setting Marshaller CharacterEncodingHandler with utf-8 does not work
    * [JAXB-790] - Whitespace processing for xs:anyURI type
    * [JAXB-794] - Classes annotated with @XmlTransient are still mapped to XML
    * [JAXB-795] - Catalog passed to XJC is not used during the schema correctness check
    * [JAXB-814] - Failing with 'no-arg default constructor' on @XmlTransient 
    * [JAXB-825] - Attribute with default generates different types for get and set methods
    * [JAXB-831] - Memory leak in com.sun.xml.bind.v2.ClassFactory
    * [JAXB-834] - Cannot use binding file on Windows
    * [JAXB-836] - CLONE -Marshaller.marshall throws NPE if an adapter adapts a non-null bound value to null.
    * [JAXB-837] - XJC generated code misses out some fields from XML
    * [JAXB-843] - Marshaller holds reference after marshalling for the writer that was used to marshal into
    * [JAXB-844] - Memory Leak in com.sun.xml.bind.v2.runtime.Coordinator
    * [JAXB-847] - DataTypeConverterImpl patch causes corrupted stream generation in some cases
    * [JAXB-849] - JAXB:Invalid boolean values added to lists as 'false'
    * [JAXB-856] - xs:import namespace="http://www.w3.org/2005/05/xmlmime" is not generated
    * [JAXB-858] - xjc generates no JAXBElement for nillable element with required attribute 
    * [JAXB-859] - Corrupt license file in the distribution
    * [JAXB-860] - NullPointerException
        com.sun.xml.bind.v2.runtime.ClassBeanInfoImpl.checkOverrideProperties(ClassBeanInfoImpl.java:205)
    * [JAXB-867] - jax-ws / jax-b / glassfish 3.1.1 web services fail to send data from beans based on variable name.
    * [JAXB-868] - Escape policy for quote (") is different when the serialization is performed to OutputStream or Writer
    * [JAXB-869] - Multiple <any></any> elements on a schema : second element is not loaded
    * [JAXB-882] - Marshalling Objects extending JAXBElement linked by @XmlElementRef
    * [JAXB-445] - Generated episode bindings should contain target package name
    * [JAXB-499] - Umbrella issue for all XJC related encoding issues wrt xjc itself and maven plugins/ant tasks
    * [JAXB-839] - More Schema Annotations/Documentation to Javadoc

4.6. Changes between 2.2.4 to 2.2.4-1

  • 834 - Cannot use binding file on Windows

  • 835 - XJC fails with project path that contains spaces

4.7. Changes between 2.2.3u2 to 2.2.4

  • 413 - Redundant cast to byte[] in code generated by XJCTask

  • 416 - Ant XJC task problems with spaces in paths

  • 450 - Reusing unmarshaller results in an unexpected result

  • 549 - Unescaped javadoc in @XmlElements

  • 602 - Different unmarshalling behavior between primitive type and simpletype with enumeration under restriction

  • 618 - XJC generates certain code lines in a random order

  • 620 - Problems with abstract classes and @XMLValue

  • 652 - JAXB 2.2 API should be changed to work with old 2.1 jaxb implementation in JDK

  • 696 - Incorrect implementation/documentation of javax.xml.bind.annotation.adapters.NormalizedStringAdapter

  • 726 - Missing keyword 'throw'

  • 761 - DatatypeConverterInterface.printDate inconsistencies

  • 774 - Extra slash in XSD location prevents customization

  • 803 - 2.2.2 strips schemaLocation in binding files

  • 804 - JAXB 2.x : How to override an XmlElement annotation from parent class - Mission Impossible?

  • 813 - Preserving white-space in XML (xs:string enumeration value) does not work

  • 815 - Binding file cannot refer to schema file with space in file name.

  • 816 - Incorrect System property to define the provider factory class

  • 821 - Global customization are not applied to xjc when input document is WSDL

  • 824 - API files in javax.xml.bind need to be updated for references to JLS with editions/hyperlinks

  • 826 - JAXB fails to unmarshal attributes as properties on aix

  • 268 - Map handling broken

  • 470 - Potential changes to make JAXB work better with OSGi

  • 478 - jaxb2-sources : Allow compilation under Java 6

  • 633 - JDefinedClass getMods()

  • 768 - Mailing list consolidation suggestions

  • 784 - JAnnotationUse should provide getters for clazz and memberValues properties

  • 406 - Allow setting of access modifiers in JMods

  • 769 - Update to command-line help text

  • 772 - Updates to XJC.html page and -Xpropertyaccessors plugin

  • 783 - I18N: xjc generates localized strings of AM/PM so compilation fails

4.8. Notable Changes between 2.2.3u1 to 2.2.3u2

  • 826 - JAXB fails to unmarshal attributes as properties on AIX

4.9. Notable Changes between 2.2.3 to 2.2.3u1

  • 805 - SpecJ performance regression

4.10. Notable Changes between 2.2.2 to 2.2.3

  • 6975714 - JAXB 2.2 throws IAE for invalid Boolean values

  • 620 - Problems with abstract classes (xsi type processing)

  • Regression in Jersey JSON mapping

4.11. Notable Changes between 2.2.1.1 to 2.2.2

4.12. Notable Changes between 2.2.1 to 2.2.1.1

  • This minor-minor release contains only changes relevant to GlassFish OSGi environment

4.13. Notable Changes between 2.2 to 2.2.1

4.14. Notable Changes between 2.1.12 to 2.2

4.15. Notable Changes between 2.1.12 to 2.1.13

4.17. Notable Changes between 2.1.10 to 2.1.11

  • Minor licencing and legal fixes. No changes to source code.

4.18. Notable Changes between 2.1.9 to 2.1.10

4.19. Notable Changes between 2.1.8 to 2.1.9

4.20. Notable Changes between 2.1.7 to 2.1.8

4.21. Notable Changes between 2.1.6 to 2.1.7

4.22. Notable Changes between 2.1.5 to 2.1.6

4.23. Notable Changes between 2.1.4 to 2.1.5

4.24. Notable Changes between 2.1.3 to 2.1.4

4.25. Notable Changes between 2.1.2 to 2.1.3

4.26. Notable Changes between 2.1.1 to 2.1.2

4.27. Notable Changes between 2.1 First Customer Ship (FCS) to 2.1.1

4.28. Notable Changes between 2.1 Eary Access 2 to 2.1 First Customer Ship (FCS)

4.34. Notable Changes between 2.0 Early Access 3 and 2.0 First Customer Ship (FCS)

  • Java to schema samples are added

  • Added <xjc:javaType> vendor customization

  • Added experimental <xjc:simple> vendor customization, which brings in a new simpler and better binding mode

  • The spec has renamed AccessType to XmlAccessType, and @AccessorType to @XmlAccessorType.

  • Various error handling improvements

  • Experimental canonicaliztion support is added.

  • The '-b' option can now take a directory and recursively search for all "*.xjb" files.

  • Fixed various issues regarding using JAXB from codef inside a restricted security sandbox.

  • Added more pluggability points for plugins to customize the code generation behavior.

  • Some of the code is split into a separate istack-commons project to promote more reuse among projects.

  • Made a few changes so that RetroTranslator can translate the JAXB RI (and its generated code) to run it on JDK 1.4 and earlier

  • Improved the quality of the generated code by removing unnecessary imports.

  • Other countless bug fixes

4.35. Notable Changes between 2.0 Early Access 2 and 2.0 Early Access 3

  • Map property can be now correctly bound to XML Schema

  • Default marshaller error handling behavior became draconian (previously errors were ignored.)

  • @link to a parameterized type is now correctly generated

  • started producing architecture document for those who want to build plugins or play with the RI internal.

  • XJC now uses the platform default proxy setting by default.

  • @XmlAccessorOrder, @XmlSchemaType and @XmlInlineBinaryData are implemented

  • @XmlJavaTypeAdapter on a class/package is implemented

  • Marshaller life-cycle events are implemented

  • Integration to FastInfoset is improved in terms of performance

  • XJC can generate @Generated

  • The unmarshaller is significantly rewritten for better performance

  • Added schemagen tool and its Ant task

  • Various improvements in error reporting during unmarshalling/marshalling

  • JAXB RI is now under CDDL

4.36. Notable Changes between 2.0 Early Access and 2.0 Early Access 2

  • The default for @XmlAccessorType was changed to PUBLIC_MEMBER

  • Optimized binary data handling enabled by callbacks in package javax.xml.bind.attachment. Standards supported include MTOM/XOP and WS-I AP 1.0 ref:swaRef.

  • Unmarshal/marshal support of element defaulting

  • Improved the quality of the generated Java code

  • Fixed bugs in default value handling

  • Runtime performance improvements, memory usage improvements

  • Added support for <xjc:superInterface> vendor extension

  • Migrated source code to http://jaxb2-sources.dev.java.net

  • Published NetBeans project file for JAXB RI

  • Added more support to the schema generator: anonymous complex types, attribute refs, ID/IDREF, etc

  • Implemented javax.xml.bind.Binder support (not 100% done yet)

  • Implemented marshal-time validation

  • Improved xjc command line interface - better support for proxy options, more options for specifying schema files

  • Added schema-2-Java support for simple type substitution

  • Added support for the new <jaxb:globalBindings localScoping="nested" | "toplevel"> customization which helps control deeply nested classes

  • Made the default ValidationEventHandler more forgiving in 2.0 than it was in 1.0 (The class still behaves the same as it did when used by a 1.0 app)

  • Added wildcard support for DTD

  • Numerous other small changes and bugfixes....

4.37. Notable Changes between 1.0.x FCS and 2.0 Early Access

  • Support for 100% W3C XML Schema (not all in EA, but planned for FCS)

  • Support for binding Java to XML

  • Addition of javax.xml.bind.annotation package for controling the binding from Java to XML

  • Significant reduction in the number of generated schema-derived classes

    • Per complex type definition, generate one value class instead of an interface and implementation class.

    • Per global element declaration, generate a factory method instead of generating a schema-derived interface and implementation class.

  • Replaced the validation capabilities in 1.0 with JAXP 1.3 validation API's

  • Smaller runtime libraries

  • Numerous other changes...

JAXB Users Guide

Abstract

This document explains various interesting/complex/tricky aspects of JAXB, based on questions posted on the JAXB users forum and answers I provided. This is an ongoing work-in-progress. Any feedback appreciated.

1. Compiling XML Schema

1.1. Dealing with errors

1.1.1. Schema errors

Because XML Schema is so complicated, and because there are a lot of tools out there do not implement the spec correctly, it is often the case that a schema you are trying to compile has some real errors in it. When this is the case, you'll see XJC reporting somewhat cryptic errors such as rcase-RecurseLax.2: There is not a complete functional mapping between the particles.

The JAXB RI uses the schema correctness checker from the underlying JAXP implementation, which is the JAXP RI in a typical setup. The JAXP RI is one of the most conformant schema validators, and therefore most likely correct. So the first course of action usually is to fix problems in the schema.

However, in some situations, you might not have an authority to make changes to the schema. If that is the case and you really need to compile the schema, you can bypass the correctness check by using the -nv option in XJC. When you do this, keep in mind that you are possibly feeding "garbage" in, so you may see XJC choke with some random exception.

1.1.2. Property 'fooBarZot' is already defined

One of the typical errors you'll see when compiling a complex schema is:

Example 1. Multiple property definitions error

parsing a schema...
[ERROR] Property "MiOrMoOrMn" is already defined.
  line 132 of
file:/C:/kohsuke/Sun/JAXB/jaxb-unit/schemas/individual/MathML2/presentation/scripts.xsd

[ERROR] The following location is relevant to the above error
  line 138 of
file:/C:/kohsuke/Sun/JAXB/jaxb-unit/schemas/individual/MathML2/presentation/scripts.xsd

This is an actual example of the offending part of a schema, taken from MathML. If you go to line 132 of scripts.xsd, you'll see that it has a somewhat complicated content model definition:

Example 2. Multiple property definitions in MathML

<xs:group name="mmultiscripts.content">
    <xs:sequence>
        <xs:group ref="Presentation-expr.class"/>
        <xs:sequence minOccurs="0" maxOccurs="unbounded">      <!-- line 132 -->
            <xs:group ref="Presentation-expr-or-none.class"/>
            <xs:group ref="Presentation-expr-or-none.class"/>
        </xs:sequence>
        <xs:sequence minOccurs="0">
            <xs:element ref="mprescripts"/>
            <xs:sequence maxOccurs="unbounded">                 <!-- line 138 -->
                <xs:group ref="Presentation-expr-or-none.class"/>
                <xs:group ref="Presentation-expr-or-none.class"/>
            </xs:sequence>
        </xs:sequence>
    </xs:sequence>
</xs:group>

This is a standard technique in designing a schema. When you want to say "in this element, B can occur arbitrary times, but C can occur only up to once", you write this as B*,(C,B*)?. This, however, confuses JAXB, because it tries to bind the first B to its own property, then C to its own property, then the second B to its own property, and so we end up having a collision again.

In this particular case, B isn't a single element but it's a choice of large number of elements abstracted away in <xs:group>s, so they are hard to see. But if you see the same content model referring to the same element/group twice in a different place, you can suspect this.

In this case, you'd probably want the whole thing to map to a single list so that you can retain the order those elements show up in the document. You can do this by putting the same <jaxb:property> customization on the whole "mmultiscripts.content" model group, like this (or you can do it externally with XPath):

Example 3. How to fix the problem?

<xs:groupname="mmultiscripts.content">
<xs:annotation>
    <xs:appinfo>
        <jaxb:propertyname="content"/>
    </xs:appinfo>
</xs:annotation>
<xs:sequence>
<xs:groupref="Presentation-expr.class"/>

Another way to fix this problem is to use the simpler and better binding mode in XJC, which is a JAXB RI vendor extension.

1.1.3. Two declarations cause a collision in the ObjectFactory class

When schemas contain similar looking element/type names, they can result in "Two declarations cause a collision in the ObjectFactory class" errors. To be more precise, for each of all types and many elements (exactly what elements get a factory and what doesn't is bit tricky to explain), XJC produces one method on the ObjectFactory class in the same package. The ObjectFactory class is created for each package that XJC generates some files into. The name of the method is derived from XML element/type names, and the error is reported if two elements/types try to generate the same method name.

There are two approaches to fix this problem. If the collision is coming from two different schemas with different target namespaces, then you can easily avoid the collision by compiling them into different Java packages. To do this, use <schemabindings> customization on two schemas and specify the package name.

Another way to fix this problem is to use <factoryMethod> customization on two conflicting elements/types to specify different factory method names. This can be used in all cases, but if you have a large number of conflicts, you'll have to specify this customization one by one.

Notice that <class> customization doesn't affect the ObjectFactory method name by itself.

1.1.4. Customization errors

1.1.4.1. XPath evaluation of ... results in empty target node

External JAXB customizations are specified by using XPath (or using SCD.) This works by writing an XPath expression that matches a particular element in the schema document. For example, given the following schema and binding file:

Example 4. Schema and external binding file

test.xsd

<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema">
    <xs:complexTypename="foo"/>
</xs:schema>

test.xjb

<bindings version="2.0" xmlns="http://java.sun.com/xml/ns/jaxb" xmlns:xs="http://www.w3.org/2001/XMLSchema">
    <bindings schemaLocation="test.xsd">
        <bindings node="//xs:complexType[@name='foo']">
            <classname="Bar"/>
        </bindings>
    </bindings>
</bindings>

will be interpreted as if the class customization is attached to the complex type 'foo'.

For this to work, the XPath expression needs to match one and only one element in the schema document. When the XPath expression is incorrect and it didn't match anything, you get this "XPath evaluation of ... results in empty target node" problem.

Common causes of this problem include typos, incorrect namespace URI declarations, and misunderstanding of XPath.

1.2. Fixing broken references in schema

Sometimes a schema may refer to another schema document without indicating where the schema file can be found, like this:

Example 5. Schema reference without location

<xs:import namespace="http://www.w3.org/1999/xlink" />

In other cases, a schema may refer to another schema on the network, which often slows down your compilation process and makes it unreliable. Yet in some other cases, a schema may reference another schema in relative path, and that may not match your directory structure.

XJC bundles a catalog resolver so that you can work around these situations without changing the schema documents. The main idea behind the catalog is "redirection" --- when XJC is about to fetch resources, it will consult the catalog resolver to see if it can find the resource elsewhere (which is usually your local resources.)

1.2.1. Catalog format

The catalog resolver supports many different formats, but the easiest one is a line based *.cat format. Other than comments and empty lines, the file mainly consists of two kinds of declarations, SYSTEM, and PUBLIC.

Example 6. sample-catalog.cat

--
  sample catalog file.

  double hyphens are used to begin and end a comment section.
--

SYSTEM "http://www.w3.org/2001/xml.xsd" "xml.xsd"

PUBLIC "-//W3C//DTD XMLSCHEMA 200102//EN" "s4s/XMLSchema.dtd"

1.2.2. Resolve by system ID

The SYSTEM entry has the format of "SYSTEM REFERENCE ACTUAL-LOCATION", which defines a simple redirection. Every time XJC loads any resource (be it schemas, DTDs, any entities referenced within), it will first resolve relative paths to absolute paths, then looks for a matching REFERENCE line. If it is found, the specified actual location is read instead. Otherwise XJC will attempt to resolve the absolutepath.

ACTUAL-LOCATION above accepts relative paths, and those are resolved against the catalog file itself (so in the above example, xml.xsd is assumed to be in the same directory with sample-catalog.cat.

What you need to be careful is the fact that the REFERENCE portion must be absolute, and when XJC finds a reference in schema, it will first convert that to the absolute path before checking the catalog. So what this means is that if your schema is written like this:

Example 7. Schema reference by relative path

<xs:import namespace="http://www.w3.org/1999/xlink" schemaLocation="xlink.xsd" />

Then your catalog entry would have to look like this:

Example 8. xlink.cat

-- this doesn't work because xlink.xsd will be turned into absolute path --
SYSTEM "xlink.xsd" "http://www.w3.org/2001/xlink.xsd"

-- this will work, assuming that the above schema is in /path/to/my/test.xsd --
SYSTEM "/path/to/my/xlink.xsd" "http://www.w3.org/2001/xlink.xsd"

1.2.3. Resolve by public ID / namespace URI

Another kind of entry has the format of "PUBLIC PUBLICID ACTUAL-LOCATION" or "PUBLIC NAMESPACEURI ACTUAL-LOCATION".

The "PUBLICID" version is used to resolve DTDs and entities in DTDs. But this type of entry is also used to resolve <xs:import> statements. XJC will match the value of the namespace attribute and see if there's any matching entry. So given a schema like this:

Example 9. Schema import

<xs:import namespace="http://www.w3.org/1999/xlink" schemaLocation="xlink.xsd" />
<xs:import namespace="http://www.w3.org/1998/Math/MathML" />

The following catalog entries will match them.

Example 10. by-publicid.cat

PUBLIC "http://www.w3.org/1999/xlink" "http://www.w3.org/2001/xlink.xsd"
PUBLIC "http://www.w3.org/1998/Math/MathML" "/path/to/my/mathml.xsd"

As you can see, XJC will check the PUBLIC entries regardless of whether <xs:import> has the schemaLocation attribute or not. As with the case with the SYSTEM entry, the ACTUAL-LOCATION part can be relative to the location of the catalog file.

1.2.4. Specifying the catalog file

Once you write a catalog file, you'd need to specify that when you invoke XJC.

CLI

To do this from the CLI, use the -catalog option. See xjc -help for more details.

Ant

Use the catalog attribute on the <xjc> task. See XJC ant task documentation for more details.

Maven

For the Maven plugin, use the <catalog> element in the configuration:

<plugin>
    <groupId>org.jvnet.jaxb2.maven2</groupId>
    <artifactId>maven-jaxb2-plugin</artifactId>
    <configuration>
        <!-- relative to the POM file -->
        <catalog>mycatalog.cat</catalog>
    </copnfiguration>
</plugin>

1.2.5. Debugging catalog file

If you are trying to write a catalog file and banging your head against a wall because it's not working, you should enable the verbose option of the catalog resolver. How you do this depends on what interface you use:

CLI

Specify export XJC_OPTS="-Dxml.catalog.verbosity=999" then run XJC.

Ant/Maven

Add -Dxml.catalog.verbosity=999 as a command line option to Ant/Maven.

If you are otherwise invoking XJC programmatically, you can set the above system property before invoking XJC.

1.3. Mapping of <xs:any />

XJC binds <xs:any /> in the following ways:

1.3.1. processContents="skip"

<xs:any /> with processContents=skip means any well-formed XML elements can be placed. Therefore, XJC binds this to DOM Element interface.

Example 11. Any/Skip schema

<xs:element name="person">
  <xs:complexType>
    <xs:sequence>
      <xs:element name="name" type="xs:string" />
      <xs:any processContents="skip" maxOccurs="unbounded" minOccurs="0" />
    </xs:sequence>
  </xs:complexType>
</xs:element>

Example 12. Any/Skip binding

import org.w3c.dom.Element;

@XmlRootElement
class Person {
  public String getName();
  public void setName(String);

  @XmlAnyElement
  public List<Element> getAny();
}

1.3.2. processContents="strict"

<xs:any /> with processContents=strict (or <xs:any /> without any processContents attribute, since it defaults to "strict") means any XML elements placed here must have corresponding schema definitions. This mode is not what people typically expect as "wildcard", but this is the default. The following shows this binding. (lax=true is unintuitive, but it's not an error in this document):

Example 13. Any/Strict schema

<xs:element name="person">
  <xs:complexType>
    <xs:sequence>
      <xs:element name="name" type="xs:string" />
      <xs:any maxOccurs="unbounded" minOccurs="0" />
    </xs:sequence>
  </xs:complexType>
</xs:element>

Example 14. Any/Strict binding

@XmlRootElement
class Person {
  public String getName();
  public void setName(String);

  @XmlAnyElement(lax=true)
  public List<Object> getAny();
}

JAXB binds any such element to an Object, and during unmarshalling, all elements encountered are unmarshalled into corresponding JAXB objects (including JAXBElements if necessary) and placed in this field. If it encounters elements that cannot be unmarshalled, DOM elements are produced instead.

At runtime, you can place either DOM elements or some JAXB objects that map to elements. A typical mistake is to put a String that contains XML fragment, but this won't work; you'd have to first read that into a DOM.

1.3.3. processContents="lax"

<xs:any /> with processContents=lax means any XML elements can be placed here, but if their element names match those defined in the schema, they have to be valid. XJC actually handles this exactly like processContents='strict', since the strict binding allows unknown elements anyway.

1.4. Mapping of <xs:element /> to JAXBElement

Sometimes XJC binds an element declaration to JAXBElement. Sometimes XJC binds an element declaration to a Java class. What makes this difference?

TODO: Copy from http://weblogs.java.net/blog/kohsuke/archive/2006/03/why_does_jaxb_p.html

1.5. How modularization of schema interacts with XJC

Over time schema authors have developed several techniques to modularize large schemas. Some of those techniques have some noteworthy interactions with XJC.

1.5.1. Chameleon schema

Chameleon schema" (read more, in particular this) is a technique used to define multiple almost-identical sets of definitions into multiple namespaces from a single schema document.

For example, with this technique, you can write just one "foo" complex type and define it into namespace X and Y. In this case, one tends to hope that XJC will only give you one Foo class for this, but unfortunately because it's actually defined in two namespaces, JAXB needs two Java classes to distinguish X:foo and Y:foo, so you'll get multiple copies.

If you find this to be problematic, there are a few ways to work around the problem.

  1. If you are in control of the schema, see if you can rewrite the schema to avoid using this technique. In some cases, the schema doesn't actually exploit the additional power of this technique, so this translation can be done without affecting XML instance documents. In some other cases, the chameleon schema can be argued as a bad schema design, as it duplicates definitions in many places.

  2. If you are not in control of the schema, see if you can rewrite the schema nevertheless. This will only work if your transformation doesn't affect XML instance documents.

  3. Perhaps there can be a plugin that eases the pain of this, such as by defining common interfaces among copies.

1.6. Adding behaviors

Adding behaviors to the generated code is one area that still needs improvement. Your feedback is appreciated.

Suppose if JAXB generated the following classes.

Example 15. Simple JAXB Generated Code

package org.acme.foo;

@XmlRootElement
class Person {
  private String name;

  public String getName() { return name; }
  public void setName(String) { this.name=name; }
}

@XmlRegistry
class ObjectFactory {
  Person createPerson() { ... }
}

To add a behavior, first write a class that extends from Person. You also need to extend ObjectFactory to return this new class. Notice that neither classes have any JAXB annotation, and I put them in a separate package. This is because we'd like PersonEx class to be used in place of Person, and we don't want PersonEx to be bound to its own XML type.

Example 16. Extended Person class

package org.acme.foo.impl;

class PersonEx extends Person {
  @Override
  public void setName(String name) {
    if(name.length()<3) throw new IllegalArgumentException();
    super.setName(name);
  }
}

@XmlRegistry
class ObjectFactoryEx extends ObjectFactory {
  @Override
  Person createPerson() {
    return new PersonEx();
  }
}

At runtime, you can create JAXBContext normally, like this.

Example 17. Creating JAXBContext

JAXBContext context = JAXBContext.newInstance(ObjectFactory.class);
// or JAXBContext.newInstance("org.acme.foo");

PersonEx can be marshalled out just like Person:

Example 18. Marshalling

Person p = new PersonEx();
context.createMarshaller().marshal(p,System.out);
// this will produce <person />

To unmarshal XML documents into PersonEx, you'll need to configure the unmarshaller to use your ObjectFactoryEx as the factory, like this:

Example 19. Unmarshalling

Unmarshaller u = context.createUnmarshaller();
u.setProperty("com.sun.xml.bind.ObjectFactory",new ObjectFactoryEx());
PersonEx p = (PersonEx)u.unmarshal(new StringReader("<person />"));

If you have multiple packages and thus multiple ObjectFactorys, you can pass in an array of them (new Object[]{new OFEx1(),new OFEx2(),...}.)

1.6.1. Inserting your class in the middle

If you have a type hierarchy and would like to insert your class in the middle, you can use the combination of XmlTransient and @implClass of <class> customization. See the following example:

Example 20. Hierarchy of types and <jaxb:class implClass>

<xs:schema ...>
  <xs:complexType name="vehicle">
    <xs:annotation><xs:appinfo>
      <jaxb:class implClass="MyVehicle" />
    </xs:appinfo></xs:annotation>
  </xs:complexType>

  <xs:complexType name="car">
    <xs:complexContent>
      <xs:extension base="vehicle" />
    </xs:complexContent>
  </xs:complexType>

  <xs:complexType name="bicycle">
    <xs:complexContent>
      <xs:extension base="vehicle" />
    </xs:complexContent>
  </xs:complexType>
</xs:schema>

Example 21. This creates a class hierarchy like the following (among the generated Java code):

            Vehicle
               ^
               |
            MyVehicle
               ^
          _____|______
         |            |
        Car          Bicycle

You'll then manually write MyVehicle class that extends from Vehicle. Annotate this class with XmlTransient to achieve the desired effect.

1.7. Avoid strong databinding

Under some limited circumstances, a weaker databinding is preferable for various reasons. JAXB does offer a few ways for you to achieve this.

1.7.1. Avoid mapping to enum

The following customization will stop binding a simple type to a type-safe enum. This can be convenient when number of constants is too large to be an useful enum (by default, the JAXB spec won't generate enum with more than 256 constants, but even 100 might be too large for you.)

Example 22. Avoid mapping one simple type

<xs:simpleType name="foo">
  <xs:annotation><xs:appinfo>
    <jaxb:typesafeEnumClass map="false" />
  </xs:appinfo></xs:annotation>
  <xs:restriction base="xs:string">
    <xs:enumeration value="x" />
    <xs:enumeration value="y" />
    <xs:enumeration value="z" />
  </xs:restriction>
</xs:simpleType>

To disable such type-safe enum binding altogether for the entire schema, use a global binding setting like this (this is actually telling XJC not to generate enums if a simple type has more than 0 constants --- the net effect is no enum generation):

Example 23. Avoid generating enums at all

<xs:schema ...>
  <xs:annotation><xs:appinfo>
    <jaxb:globalBindings typesafeEnumMaxMembers="0" />
  </xs:appinfo></xs:annotation>
  ...
</xs:schema>

1.7.2. Mapping to DOM

The <jaxb:dom>customization allows you to map a certain part of the schema into a DOM tree. This customization can be attached to the following schema components:

  • Wildcards (<xs:any>)

  • Type definitions (<xs:complexType> and <xs:simpleType>)

  • Model groups (<xs:choice>,<xs:all>,<xs:sequence>)

  • Model group declarations (<xs:group>)

  • Particles

  • Element declarations (<xs:element>)

In the following example, a wildcard is mapped to a DOM node. Each element that matches to the wildcard will be turned into a DOM tree.

Example 24. Dom Customization example

<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema"
               xmlns:jaxb="http://java.sun.com/xml/ns/jaxb"
               jaxb:version="2.0">

        <xs:element>
           <xs:complexType>
              <xs:sequence>
                 <xs:any maxOccurs="unbounded" processContents="skip">
                    <xs:annotation><xs:appinfo>
                      <jaxb:dom/>
                    </xs:appinfo></xs:annotation>
                 </xs:any>
              </xs:sequence>
           </xs:complexType>
        </xs:element>
    .
    .
    .
    </xs:schema>

This extension can be used to access wildcard content or can be used to process a part of a document by using other technologies that require "raw" XML. By default, JAXB generates a getContent() method for accessing wildcard content, but it only supports "lax" handling which means that unknown content is discarded. You may find more information in 7.12 chapter of JAXB 2 specification.

1.8. Working with generated code in memory

1.8.1. Cloning

The generated beans (and in particular the JAXBElement class) do not support the clone operation. There was a suggestion by another user that beanlib has been used successfully to clone JAXB objects.

2. Customization of Schema Compilation

2.1. Customizing Java packages

The JAXB specification provides a <jaxb:schemaBindings> customization so that you can control which namespace goes to which package. See the example below:

Example 25. package customization

    <jaxb:schemaBindings>
      <jaxb:package name="org.acme.foo"/>
    </jaxb:schemaBindings>

You can do this as an internal customization (in which case you put this in <xs:annotation><xs:appinfo> under place it right under the <xs:schema> element), or do this as an external customization, like this:

Example 26. External package customization

<bindings xmlns="http://java.sun.com/xml/ns/jaxb" version="2.1">
  <bindings schemaLocation="../path/to/my.xsd">
    <schemaBindings>
      <package name="org.acme.foo"/>
    </schemaBindings>
  </bindings>
</bindings>

Note that this customization is per namespace. That is, even if your schema is split into multiple schema documents, you cannot put them into different packages if they are all in the same namespace.

2.1.1. Tip: get rid of the org.w3._2001.xmlschema package

Under some rare circumstances, XJC will generate some Java classes into a package called org.w3._2001.xmlschema. This happens when XJC decides that it needs some Java artifacts for the XML Schema built-in namespace of http://www.w3.org/2001/XMLSchema.

Since this package name is most often problematic, you can rename this by simply saving the following text in an .xsd file and submitting it to XJC along with the other schemas you have:

Example 27. Schemalet to get rid of org.w3._2001.xmlschema

<schema xmlns="http://www.w3.org/2001/XMLSchema"
  targetNamespace="http://www.w3.org/2001/XMLSchema"
  xmlns:jaxb="http://java.sun.com/xml/ns/jaxb"
  jaxb:version="2.0">
  <annotation><appinfo>
    <jaxb:schemaBindings>
      <jaxb:package name="org.acme.foo"/>
    </jaxb:schemaBindings>
  </appinfo></annotation>
</schema>

This is bit tricky, but the idea is that since you can define a schema for one namespace in multiple schema documents, this makes XJC think that this schema is a part of the built-in "XML Schema for XML Schema".

2.2. Using SCD for customizations

When using an external customization file, the JAXB spec requires that you use XPath as a means to specify what your customization is attached to. For example, if you want to change the class name generated from a complex type, you'd write something like:

Example 28. External customization example

<bindings xmlns="http://java.sun.com/xml/ns/jaxb" version="2.0" xmlns:xs="http://www.w3.org/2001/XMLSchema">
  <bindings schemaLocation="../path/to/my.xsd" node="/xs:schema/xs:complexType[@name='foo']">
    <class name="FooType"/>
  </bindings>
</bindings>

While the above process works, the problem with this is that the XPath+ schemaLocation combo tends to be verbose and error prone. It's verbose, because often a trivial target schema component like this "global complex type foo" takes up a lot of characters. The xs namespace declaration also takes up some space, although in this case we managed to avoid declaring the "tns" namespace (that represents the namespace that the schema defines.)

It's also error prone, because it relies on the way schema documents are laid out, because the schemaLocation attribute needs to point to the right schema document file. When a schema is split into multiple files for modularity (happens especially often with large schemas), then you'd have to find which schema file it is. Even though you can use relative paths, this hard-coding of path information makes it hard to pass around the binding file to other people.

JAXB RI 2.1 and onward offers a better way to do this as a vendor extension.

The key technology to solve this problem is a "schema component designator" (SCD.) This is a path language just like XPath, but whereas XPath is designed to refer to XML infoset items like elements and attributes, SCD is designed to refer to schema components like element declarations or complex types.

With SCD, the above binding can be written more concisely as follows:

Example 29. External customization by SCD

<bindings xmlns="http://java.sun.com/xml/ns/jaxb" version="2.1" xmlns:tns="http://my.namespace/">
  <bindings scd="/type::tns:foo">
    <class name="FooType"/>
  </bindings>
</bindings>

/type::tns:foo can be written more concisely as /~tns:foo, too. If you are interested in more about the syntax of SCDs, read the example part of the spec, and maybe EBNF. If you know XPath, I think you'll find this fairly easy to learn.

Another benefit of an SCD is that tools will have easier time generating SCDs than XPath, as XPaths are often vulnerable to small changes in the schema document, while SCDs are much more robust. The downside of using SCD is as of JAXB 2.1, this feature is a vendor extension and not defined in the spec.

2.3. Using different datatypes

JAXB has a built-in table that determines what Java classes are used to represent what XML Schema built-in types, but this can be customized.

One of the common use cases for customization is to replace the XMLGregorianCalendar with the friendlier Calendar or Date. XMLGregorianCalendar is designed to be 100% compatible with XML Schema's date/time system, such as providing infinite precision in sub-seconds and years, but often the ease of use of those familiar Java classes win over the precise compatibility.

One very easy way to do this is to simply use your IDE (or even "sed") to replace all the references to XMLGregorianCalendar by Calendar. This is of course not a very attractive option if your build process runs XJC as a part of it.

Alternatively, the following customization file can be used to do this. When using external customization file, the JAXB spec requires you to use XPath as a means to specify what your customization is attached to. For example, if you want to change the class name generated from a complex type, you'd use the following customization:

Example 30. Customization to use Calendar for xs:date

<bindings xmlns="http://java.sun.com/xml/ns/jaxb" version="2.0" xmlns:xs="http://www.w3.org/2001/XMLSchema">
  <globalBindings>
    <javaType name="java.util.Calendar" xmlType="xs:date"
      parseMethod="javax.xml.bind.DatatypeConverter.parseDate"
      printMethod="javax.xml.bind.DatatypeConverter.printDate"
    />
  </globalBindings>
</bindings>

Save this in a file and specify this to JAXB with the "-b" option.

To use the Date class, you'll need to do a bit more work. First, put the following class into your source tree:

Example 31. Adapter for Date

package org.acme.foo;
public class DateAdapter {
  public static Date parseDate(String s) {
    return DatatypeConverter.parseDate(s).getTime();
  }
  public static String printDate(Date dt) {
    Calendar cal = new GregorianCalendar();
    cal.setTime(dt);
    return DatatypeConverter.printDate(cal);
  }
}

... then your binding file will be the following:

Example 32. Customization to use Date for xs:date

<bindings xmlns="http://java.sun.com/xml/ns/jaxb" version="2.0" xmlns:xs="http://www.w3.org/2001/XMLSchema">
  <globalBindings>
    <javaType name="java.util.Date" xmlType="xs:date"
      parseMethod="org.acme.foo.DateAadpter.parseDate"
      printMethod="org.acme.foo.DateAdapter.printDate"
    />
  </globalBindings>
</bindings>

If you are using JAXB 2.0, and not 2.1, see this blog entry for how to do this for 2.0.

3. Annotating Your Classes

3.1. Mapping your favorite class

3.1.1. ResultSet

JAXB (or any other databinding engine, for that matter) is for binding strongly-typed POJO-like objects to XML, such as AddressBook class, PurchaseOrder class, and so on, where you have fields and methods that shape a class.

There are other kinds of classes that are more close to reflection. Those classes don't have methods like getAddress, and instead you'd do get("Address"). JDBC ResultSet is one of those classes. It's one class that represents million different data structures, be it a customer table or a product table. Generally speaking, these classes does not allow JAXB to statically determine what the XML representation should look like. Instead, you almost always need to look at an instance to determine the shape of XML.

These classes are not really suitable for binding in JAXB. If this is the only object that you'd want to write out, then you'd be better off using XMLStreamWriter or some such XML infoset writing API. There are a few online articles that cover this topic. Also, many modern database offers a native ability to export a query into XML, which is supposed to work a lot faster than you'd do in Java (and saves your time of writing code.)

If you are using ResultSet as a part of your object tree that you want to marshal to JAXB, then you can use XmlJavaTypeAdapter.

3.1.2. HashMap

JAXB spec defines a special handling for Map when it's used as a propety of a bean. For example, the following bean would produce XMLs like the following:

Example 33. Bean with Map

@XmlRootElement
class Foo {
  public HashMap<String,Integer> map;
}

Example 34. XML representation

<foo>
  <map>
    <entry>
      <key>a</key>
      <value>1</value>
    </entry>
    <entry>
      <key>b</key>
      <value>2</value>
    </entry>
  </map>
</foo>

Unfortunately, as of 2.1, this processing is only defined for bean properties and not when you marshal HashMap as a top-level object (such as a value in JAXBElement.) In such case, HashMap will be treated as a Java bean, and when you look at HashMap as a bean it defines no getter/setter property pair, so the following code would produce the following XML:

Example 35. Bean with Map

m = new HashMap();
m.put("abc",1);
marshaller.marshal(new JAXBElement(new QName("root"),HashMap.class,m),System.out);

Example 36. XML representation

<root />

This issue has been recorded as #223 and the fix needs to happen in later versions of the JAXB spec.

In the mean time, such top-level objects have to be first adapted to a bean that JAXB can process. This has added benefit of being able to control XML representation better. The following code illustrates how to do this:

Example 37. Adapting HashMap

public class MyHashMapType {
    public List<MyHashMapEntryType> entry = new ArrayList<MyHashMapEntryType>();
    public MyHashMapType(Map<String,Integer> map) {
        for( Map.Entry<String,Integer> e : map.entrySet() )
            entry.add(new MyHashMapEntryType(e));
    }
    public MyHashMapType() {}
}

public class MyHashMapEntryType {
    @XmlAttribute // @XmlElement and @XmlValue are also fine
    public String key;

    @XmlAttribute // @XmlElement and @XmlValue are also fine
    public int value;

    public MyHashMapEntryType() {}
    public MyHashMapEntryType(Map.Entry<String,Integer> e) {
       key = e.getKey();
       value = e.getValue();
    }
}

marshaller.marshal(new JAXBElement(new QName("root"),MyHashMapType.class,new MyHashMapType(m)),System.out);

If you have a lot of difference kinds of Map, you can instead use Object as the key and the value type. In that way, you'll be able to use maps with different type parameters, at the expense of seeing xsi:type attribute on the instance document.

3.2. Mapping interfaces

Because of the difference between the XML type system induced by W3C XML Schema and the Java type system, JAXB cannot bind interfaces out of the box, but there are a few things you can do.

3.2.1. Use @XmlRootElement

When your interface is implemented by a large number of sub-classes, consider using XmlRootElement annotation like this:

Example 38. XmlRootElement for open-ended interfaces

@XmlRootElement
class Zoo {
  @XmlAnyElement
  public List<Animal> animals;
}

interface Animal {
  void sleep();
  void eat();
  ...
}

@XmlRootElement
class Dog implements Animal { ... }

@XmlRootElement
class Lion implements Animal { ... }

This will produce XML documents like this:

Example 39. XML for XmlRootElement

<zoo>
    <lion> ... </lion>
    <dog> ... </dog>
</zoo>

The key characteristics of this approach is:

  1. Implementations are open-ended; anyone can implement those interfaces, even by different people from different modules, provided they are all given to the JAXBContext.newInstance method. There's no need to list all the implementation classes in anywhere.

  2. Each implementation of the interface needs to have an unique element name.

  3. Every reference to interface needs to have the XmlElementRef annotation. The type=Object.class portion tells JAXB that the greatest common base type of all implementations would be java.lang.Object.

@XmlElementWrapper is often useful with this, as it allows you need to group them. Such as:

Example 40. XmlRootElement for open-ended interfaces

@XmlRootElement
class Zoo {
  @XmlElementWrapper
  @XmlAnyElement
  public List<Animal> onExhibit;
  @XmlElementWrapper
  @XmlAnyElement
  public List<Animal> resting;
}

Example 41. Effect of XmlElementWrapper

<zoo>
    <onExhibit>
        <lion> ... </lion>
        <dog> ... </dog>
    </onExhibit>
    <resting>
        <lion> ... </lion>
        <dog> ... </dog>
    </resting>
</zoo>

3.2.2. Use @XmlJavaTypeAdapter

When you use interfaces just to hide your implementation classes from exposure, and when there's 1-to-1 (or close to 1-on-1) relationship between a class and an interface, XmlJavaTypeAdapter can be used like below.

Example 42. XmlJavaTypeAdapter for interfaces

@XmlJavaTypeAdapter(FooImpl.Adapter.class)
interface IFoo {
  ...
}
class FooImpl implements IFoo {
  @XmlAttribute
  private String name;
  @XmlElement
  private int x;

  ...

  static class Adapter extends XmlAdapter<FooImpl,IFoo> {
    IFoo unmarshal(FooImpl v) { return v; }
    FooImpl marshal(IFoo v) { return (FooImpl)v; }
  }
}

class Somewhere {
  public IFoo lhs;
  public IFoo rhs;
}

Example 43. XML of XmlJavaTypeAdapter

<somewhere>
  <lhs name="...">
    <x>5</x>
  </lhs>
  <rhs name="...">
    <x>5</x>
  </rhs>
</somewhere>

The key characteristics of this approach is:

  1. Interface and implementation will be tightly coupled through an adapter, although changing an adapter code will allow you to support multiple implementations.

  2. There's no need of any annotation in where interfaces are used.

A variation of this technique is when you have a few implementations for interface, not just one.

Example 44. XmlJavaTypeAdapter for interfaces with multiple implementations

@XmlJavaTypeAdapter(AbstractFooImpl.Adapter.class)
interface IFoo {
  ...
}
abstract class AbstractFooImpl implements IFoo {
  ...

  static class Adapter extends XmlAdapter<AbstractFooImpl,IFoo> {
    IFoo unmarshal(AbstractFooImpl v) { return v; }
    AbstractFooImpl marshal(IFoo v) { return (AbstractFooImpl)v; }
  }
}

class SomeFooImpl extends AbstractFooImpl {
  @XmlAttribute String name;
  ...
}

class AnotherFooImpl extends AbstractFooImpl {
  @XmlAttribute int id;
  ...
}

class Somewhere {
  public IFoo lhs;
  public IFoo rhs;
}

Example 45. XML of XmlJavaTypeAdapter with multiple implementations

<somewhere>
  <lhs xsi:type="someFooImpl" name="...">
  </lhs>
  <rhs xsi:type="anotherFooImpl" id="3" />
</somewhere>

Note that SomeFooImpl and AnotherFooImpl must be submitted to JAXBContext.newInstance one way or the other.

To take this example a bit further, you can use Object instead of AbstractFooImpl. The following example illustarates this:

Example 46. XmlJavaTypeAdapter for interfaces with multiple implementations

@XmlJavaTypeAdapter(AnyTypeAdapter.class)
interface IFoo {
  ...
}
public class AnyTypeAdapter extends XmlAdapter<Object,Object> {
  Object unmarshal(Object v) { return v; }
  Object marshal(Object v) { return v; }
}

class SomeFooImpl implements IFoo {
  @XmlAttribute String name;
  ...
}

class Somewhere {
  public IFoo lhs;
  public IFoo rhs;
}

Example 47. Corresponding schema

<xs:complexType name="somewhere">
  <xs:sequence>
    <xs:element name="lhs" type="xs:anyType" minOccurs="0"/>
    <xs:element name="rhs" type="xs:anyType" minOccurs="0"/>
  </xs:sequence>
</xs:complexType>

As you can see, the schema will generated to accept xs:anyType which is more relaxed than what the Java code actually demands. The instance will be the same as the above example. Starting from JAXB RI 2.1, we bundle the com.sun.xml.bind.AnyTypeAdapter class in the runtime that defines this adapter. So you won't have to write this adapter in your code.

3.2.3. Use @XmlElement

If the use of interface is very little and there's 1-to-1 (or close to) relationship between interfaces and implementations, then you might find XmlElement to be the least amount of work.

Example 48. XmlElement for interfaces

interface IFoo {
  ...
}
class FooImpl implements IFoo {
  ...
}

class Somewhere {
  @XmlElement(type=FooImpl.class)
  public IFoo lhs;
}

Example 49. XML of XmlElement

<somewhere>
  <lhs> ... </lhs>
</somewhere>

This effectively tells JAXB runtime that "even though the field is IFoo, it's really just FooImpl.

In this approach, a reference to an interface has to have knowledge of the actual implementation class. So while this requires the least amount of typing, it probably wouldn't work very well if this crosses module boundaries.

Like the XmlJavaTypeAdapter approach, this can be used even when there are multiple implementations, provided that they share the common ancestor.

The extreme of this case is to specify @XmlElement(type=Object.class).

3.2.4. Hand-write schema

Occasionally the above approaches cause the generated schema to become somewhat ugly, even though it does make the JAXB runtime work correctly. In such case you can choose not to use the generated schema and instead manually modify/author schemas tht better match your needs.

3.2.5. Do schema-to-java

With sufficient knowlege, one can also use <jaxb:class ref="..."/> annotation so that you can cause XJC to use the classes you already wrote. TODO: more details and perhaps an example.

3.2.6. DOESN'T WORK: Have JAXB generate interaces and swap different implementations

Some users attempted to use the "generateValueClass" customization and see if they can completely replace the generated implementations with other implementations. Unfortunately, this does not work.

Even with the interface/implementation mode, JAXB runtime still requires that the implementation classes have all the JAXB annotations. So just implementing interfaces is not sufficient. (This mode is mainly added to simplify the migration from JAXB 1.0 to JAXB 2.0, and that's a part of the reason why things are done this way.)

3.3. Evolving annotated classes

Here is the basic problem of evolution. You got your CoolApp v1, which contains class Foo that has some JAXB annotations. Now you are working towawrd CoolApp v2, and you want to make some changes to Foo. But you want to do so in such a way that v1 and v2 can still talk to each other.

The evolution compatibility has two different aspects. One is the schema compatibility, which is about the relationship between the v1 schema and the v2 schema. The other is about runtime compatibility, which is about reading/writing documents between two versions.

3.3.1. Runtime compatibility

There are two directions in the runtime compatibility. One is whether v1 can still read what v2 write (forward compatible), and the other is whether v2 can read what v1 wrote (backward compatible).

3.3.2. "Semi-compatible"

JAXB can read XML documents that don't exactly match what's expected. This is the default behavior of the JAXB unmarshaller, yet you can change it to a more draconian behavior (TODO: pointer to the unmarshalling section.)

When we are talking about evolving classes, it's often convenient to leave it in the default behavior, as that would allow JAXB to nicely ignore elements/attributes newly added in v2. So we call it backward semi-compatible if v2 can read what v1 wrote in this default unmarshalling mode, and similarly forward semi-compatible if v1 can read what v2 wrote in this default unmarshalling mode.

Technically, these are weaker than true backward/forward compatibility (since you can't do a draconian error detection), yet in practice it works just fine.

3.3.3. Adding/removing/changing non-annotated things

You can add, remove, or change any non-annotated fields, methods, inner/nested types, constructors, interfaces. Those changes are both backward and forward compatible, as they don't cause any change to the XML representation.

Adding super class is backward compatible and forward semi-compatible. Similarly, removing super class is forward compatible and backward semi-compatible.

3.3.4. Adding/removing/changing properties

Adding new annotated fields or methods is backward compatible and forward semi-compatible. Similarly, removing them is forward compatible and backward semi-compatible.

Changing a property is bit more tricky.

  1. If you change the property name from X to Y, that would be the equivalent of deleting X and adding Y, so it would be backward and forward semi-compatible. What JAXB really cares is properties' XML names and not Java names, so by using the name parameter of XmlElement , XmlAttribute et al, you can change Java property names without affecting XML, or change XML without affecting Java properties. These are backward and forward semi-compatible. See below:

  2. Example 50. Changing Java without affecting XML

    // BEFORE
    public class Foo {
        public String abc;
    }
    // AFTER: Java name changed, but XML remains the same
    public class Foo {
        @XmlElement(name="abc")
        public String def;
    }

    Example 51. Changing XML without affecting Java

    // BEFORE
    public class Foo {
        public String abc;
    }
    // AFTER: no Java change, but XML will look different
    public class Foo {
        @XmlElement(name="def")
        public String abc;
    }

  3. If you change a property type, generally speaking it will be not compatible at all. For example, you can't change from java.util.Calendar to int and expect it to work. To make it a somewhat compatible change, the old type and the new type has to be related. For example, String can represent all int values, so changing int to String would be a backward compatible and forward semi-compatible change. XmlJavaTypeAdapter allows you to make changes to Java without affecting XML (or vice versa.)

3.3.5. Changing class names

XmlType and XmlRootElement allows you to change a class name without affecting XML.

Example 52. Changing class name without affecting XML (1)

// BEFORE
@XmlRootElement
public class Foo { ... }

// AFTER: no XML change
@XmlRootElement(name="foo")
@XmlType(name="foo")
public class Bar { ... }

Example 53. Changing class name without affecting XML (2)

// BEFORE
public class Foo { ... }

// AFTER: no XML change
@XmlType(name="foo")
public class Bar { ... }

3.3.6. Schema Compatibility

TODO.

3.4. XML layout and in-memory data layout

Your program sometimes needs to have a different in-memory data structure from its XML representation. JAXB has a few different ways to achieve this.

3.4.1. XmlJavaTypeAdapter

XmlJavaTypeAdapter allows you to de-couple the in-memory representation and the XML representation by introducing an intermediate representation. The basic model is as follows:

In-memory objects  <===>  Intermediate objects   <===>
XML
                  adapter                         JAXB

Your adapter code will be responsible for converting in-memory objects to/from intermediate objects. Intermediate objects are then bound to XML by following the standard JAXB rules. See XmlAdapter for a general description of how adapters works.

Adapters extend from the XmlAdapter class and provide two methods "unmarshal" and "marshal" that converts values in both directions, and then the XmlJavaTypeAdapter annotation is used to tell JAXB where and what adapters kick in.

(TODO: more info about XmlJavaTypeAdapter needed)

  1. adapting a class

  2. adapting a property

  3. adapting an external class

  4. adapting a collection and its effect

  5. adapting and using interfaces

3.4.2. Using XmlJavaTypeAdapter for element/attribute values

One of the common use cases of XmlJavaTypeAdapter is to map a "value object" to a string in XML. The following example illustrates how to do this, by using java.awt.Color as an example.

Example 54. Mapping Color to #RRGGBB

@XmlRootElement
class Box {
  @XmlJavaTypeAdapter(ColorAdapter.class)
  @XmlElement
  Color fill;
}

class ColorAdapter extends XmlAdapter<String,Color> {
  public Color unmarshal(String s) {
    return Color.decode(s);
  }
  public String marshal(Color c) {
    return "#"+Integer.toHexString(c.getRGB());
  }
}

This maps to the following XML representation:

Example 55. Box instance

<box>
  <fill>#112233</fill>
</box>

Since XmlJavaTypeAdapter is on a field, this adapter only kicks in for this particular field. If you have many Color fields and would like them all to use the same adapter, you can move the annotation to a package:

Example 56. package-info.java

@XmlJavaTypeAdapter(type=Color.class,value=ColorAdapter.class)
package foo;

Example 57. Box.java

@XmlRootElement
class Box {
  @XmlElement Color fill;
  @XmlElement Color border;
}

This causes all the fields in the classes in the foo package to use the same specified adapter.

Also see the DatatypeConverter class that defines a series of basic conversion routines that you may find useful.

3.4.3. Pair property

Another useful technique is to define two properties, one for JAXB and the other for your application. See the following example:

Example 58. Pair property sample

@XmlRootElement
class Person {
  private int age;

  // This public property is for users
  @XmlTransient
  public int getAge() {
    return age;
  }
  public void setAge(int age) {
    this.age = age;
  }

  // This property is for JAXB
  @XmlAttribute(name="age")
  private String getAge_() {
    if(age==-1)  return "dead";
    else         return String.valueOf(age);
  }
  private void setAge_(String v) throws NumberFormatException {
    if(v.equals("dead"))   this.age=-1;
    else                   this.age=Integer.parseInt(age);
}

The main "age" property is public, but marked as XmlTransient , so it's exposed in your program, but JAXB will not map this to XML. There's another private "age_" property. Since this is marked with XmlAttribute , this is what JAXB is going to use to map to the attribute. The getter and setter methods on this property will handle the conversion between the in-memory representation and the XML representation.

3.5. Mapping cyclic references to XML

Object models designed in Java often have cycles, which prevent straight-forward conversion to XML by JAXB. In fact, when you try to marshal an object tree that contains a cycle, the JAXB marshaller reports an error, pointing out the objects that formed the cycle. This is because JAXB by itself cannot figure out how to cut cycles into a tree.

Thus it is your responsibility to annotate classes and use other means to "tell" JAXB how to handle a cycle. This chapter talks about various techniques to do this.

3.5.1. Parent pointers

One of the very common forms of cycle is a parent pointer. The following example illustrates a typical parent pointer, and how this can be turned into "natural" XML:

Example 59. Classes with parent pointer

@XmlRootElement
class Department {
  @XmlAttribute
  String name;
  @XmlElement(name="employee")
  List<Employee> employees;
}

class Employee {
  @XmlTransient
  Department department;  // parent pointer
  @XmlAttribute
  String name;

  public void afterUnmarshal(Unmarshaller u, Object parent) {
    this.department = (Department)parent;
  }
}

This will produce the following XML:

Example 60. XML view of department

<department name="accounting">
  <employee name="Joe Chin" />
  <employee name="Adam Smith" />
  ...
</department>

And reading this document back into Java objects will produce the expected tree with all the proper parent pointers set up correctly.

The first technique here is the use of XmlTransient on the parent pointer. This tells JAXB that you don't need this parent pointer to be represented explicitly in XML, because the fact that employee is always contained inside department implies the parent/child relationship. This causes the marshaller to produce the expected XML. However, when you unmarshal it, since this field is not bound, the Employee.department field will be left null.

That's where the second technique comes in, which is the use of the afterUnmarshal callback. This method is invoked by the JAXB implementation on each instance when the unmarshalling of a Employee object completes. Furthermore, the second parameter to the method is the parent object, which in this case is a Department object. So in this example, this sets up the parent pointer correctly.

This callback can be also used to perform other post-unmarshalling set up work.

3.5.2. Many-to-many relationship

TBD

3.5.3. @XmlID and @XmlIDREF

When a reference to another object is annotated with XmlIDREF , its corresponding XML it will be referenced by xs:IDREF, instead of containment. See below for an example:

Example of @XmlID and @XmlIDREF

@XmlRootElement
class Root {
  List<Foo> foos;
  List<Bar> bars;
}
class Foo {
  // you don't have to make it an attribute, but that's more common
  @XmlAttribute @XmlIDREF Bar bar;
}
class Bar {
  // you don't have to make it an attribute, but that's more common
  @XmlAttribute @XmlID String id;
}

Example 61. Schema for above

<xs:complexType name="foo">
  <xs:sequence/>
  <xs:attribute name="bar" type="xs:IDREF"/>
  </xs:sequence>
</xs:complexType>
<xs:complexType name="bar">
  <xs:sequence/>
  <xs:attribute name="id" type="xs:ID"/>
</xs:complexType>

Example 62. A sample instance

<root>
  <foo bar="x"/>
  <foo bar="y"/>
  <bar id="x"/>
  <bar id="y"/>
</root>

There are a few things to consider when you do this. First, the object to be referenced must have an ID that is unique within the whole document. You'd also need to ensure that the referenced objects are contained somewhere else (like in the Root class in this case), or else Bar objects will never be marshalled. This technique can be used to remove the cyclic references, but it's only possible when your object model has an easy cut point.

3.5.4. Use the CycleRecoverable interface

Starting 2.1 EA2, the JAXB RI exposes CycleRecoverable interface. Applications can choose to implement this interface in some of its objects. When a cyclic reference is detected during marshalling, and if the object that formed a cycle implements this interface, then the method on this interface is called to allow an application to nominate its replacement to be written to XML. In this way, the application can recover from a cycle gracefully.

This technique allows you to cope with a situation where you cannot easily determine upfront as to where a cycle might happen. On the other hand, this feature is a JAXB RI feature. Another downside of this is that unless you nominate your replacement carefully, you can make the marshalling output invalid with respect to the schema, and thus you might hit another problem when you try to read it back later.

4. Unmarshalling

4.1. @XmlRootElement and unmarshalling

Classes with XmlRootElement can be unmarshalled from XML elements simply by invoking the unmarshal method that takes one parameter. This is the simplest mode of unmarshalling.

Unmarshalling with @XmlRootElement

@XmlRootElement
class Foo {
  @XmlAttribute
  String name;
  @XmlElement
  String content;
}

Unmarshaller u = ...;
Foo foo = (Foo)u.unmarshal(new File("foo.xml"));

Example 63. foo.xml

<foo name="something">
  <content>abc</content>
</foo>

However, sometimes you may need to unmarshal an instance of a type that does not have an XmlRootElement . For example, you might dynamically find out at the runtime that a certain element has a certain type. For example, the following document illustrates an XML instance where the content of <someOtherTagName> element is represented by the Foo class.

Example 64. foo2.xml

<someOtherTagName name="something">
  <content>abc</content>
</someOtherTagName>

To unmarshal this into a Foo class, use the version of the unmarshal method that takes the 'expectedType' argument, as follows:

Example 65. Unmarshalling into a known type

Unmarshaller u = ...;
JAXBElement<Foo> root = u.unmarshal(new StreamSource(new File("foo.xml")),Foo.class);
Foo foo = root.getValue();

To reduce the number of the unmarshal methods, this two-argument version is not defined for every single-argument version. So as in this example, you might need to perform additional wrapping of the input parameter.

This instructs JAXB that the caller is expecting to unmarshal Foo instance. JAXB returns a JAXBElement of Foo, and this JAXBElement captures the tag name of the root element.

4.2. Unmarshalling is not working! Help!

There are a few common causes for this problem. These causes often exhibit similar symptoms:

  1. Instance documents are invalid

  2. JAXBContext is not created correctly.

4.2.1. Make sure your instance document is valid

First, use an independent schema validator to check if your document is really valid with respect to the schema you compiled. When the root element of a document is invalid, then the unmarshaller will issue "unexpected element" errors. When a portion of a document is invalid, JAXB skips that portion, so the end result is that the unmarshalling returns normally, yet you notice that a part of the content tree is missing. This is often the desirable behavior, but it sometimes ends up masking a problem.

Also, try to install ValidationEventHandler on the unmarshaller. When a portion of a document is skipped, the unmarshaller notifies a ValidationEventHandler, so it allows you to see what's going on.

Example 66. Installing ValidationEventHandler

Unmarshaller u = ...;
// this implementation is a part of the API and convenient for trouble-shooting,
// as it prints out errors to System.out
u.setEventHandler(new javax.xml.bind.helpers.DefaultValidationEventHandler());

u.unmarshal(new File("foo.xml"));

Also consider installing a Schema object to the unmarshaller, so that the unmarshaller performs a schema validation while unmarshalling. Earlier I suggested that you try an independent schema validator, but for various reasons (not all tools are reliable, you might have made an error and used a different schema/instance), using validating unmarshalling is a better way to guarantee the validity of your instance document being unmarshalled. Please follow the JAXP tutorial for more about how to construct a Schema object from your schema.

If you are unmarshalling from XML parser APIs (such as DOM, SAX, StAX), then also make sure that the parser/DOM is configured with the namespace enabled.

4.2.2. Check if your JAXBContext is correct

(TODO: This also applies to the marshaller. Think about moving it.)

The other possibility is that JAXBContext is not set up correctly. JAXBContext "knows" a set of classes, and if it doesn't know a class that it's supposed to know, then the unmarshaller may fail to perform as you expected.

To verify that you created JAXBContext correctly, call JAXBContext.toString(). It will output the list of classes it knows. If a class is not in this list, the unmarshaller will never return an instance of that class. Make you see all the classes you expect to be returned from the unmarshaller in the list. When dealing with a large schema that spans across a large number of classes and packages, this is one possible cause of a problem.

If you noticed that a class is missing, explicitly specify that to JAXBContext.newInstance. If you are binding classes that are generated from XJC, then the easiest way to include all the classes is to specify the generated ObjectFactory class(es).

4.3. Element default values and unmarshalling

Because of the "strange" way that element default values in XML Schema work, people often get confused about their behavior. This section describes how this works.

When a class has an element property with the default value, and if the document you are reading is missing the element, then the unmarshaller does not fill the field with the default value. Instead, the unmarshaller fills in the field when the element is present but the content is missing. See below:

Example 67. XML instance 1

<foo />

Example 68. XML instance 2

<foo>
  <a/>  <!-- or <a></a> -->
</foo>

Example 69. XML instance 3

<foo>
  <a>abc</a>
</foo>

Example 70. Element defaults and XML

@XmlRootElement
class Foo {
  @XmlElement(defaultValue="value") public String a=null;
}

Foo foo = unmarshaller.unmarshal("instance1.xml");
System.out.println(foo.a);   // null

Foo foo = unmarshaller.unmarshal("instance2.xml");
System.out.println(foo.a);   // "value". The default kicked in.

Foo foo = unmarshaller.unmarshal("instance3.xml");
System.out.println(foo.a);   // "abc". Read from the instance.

This is consistent with the XML Schema spec, where it essentially states that the element defaults do not kick in when the element is absent, so unfortunately we can't change this behavior.

Depending on your expectation, using a field initializer may achieve what you are looking for. See below:

Example 71. Possible changes by using field initializer

@XmlRootElement
class Foo {
  @XmlElement public String a="value";
}

Foo foo = unmarshaller.unmarshal("instance1.xml");
System.out.println(foo.a);   // "value", because JAXB didn't overwrite the value

Foo foo = unmarshaller.unmarshal("instance2.xml");
System.out.println(foo.a);   // "", because <a> element had 0-length string in it

Foo foo = unmarshaller.unmarshal("instance3.xml");
System.out.println(foo.a);   // "abc". Read from the instance.

Alternatively, attribute default values work in a way that agrees with the typical expectation, so consider using that. Also, see Element default values and marshalling.

4.4. Dealing with large documents

JAXB API is designed to make it easy to read the whole XML document into a single tree of JAXB objects. This is the typical use case, but in some situations this is not desirable. Perhaps:

  1. A document is huge and therefore the whole may not fit the memory.

  2. A document is a live stream of XML (such as XMPP) and therefore you can't wait for the EOF.

  3. You only need to databind the portion of a document and would like to process the rest in other XML APIs.

This section discusses several advanced techniques to deal with these situations.

4.4.1. Processing a document by chunk

When a document is large, it's usually because there's repetitive parts in it. Perhaps it's a purchase order with a large list of line items, or perhaps it's an XML log file with large number of log entries.

This kind of XML is suitable for chunk-processing; the main idea is to use the StAX API, run a loop, and unmarshal individual chunks separately. Your program acts on a single chunk, and then throws it away. In this way, you'll be only keeping at most one chunk in memory, which allows you to process large documents.

See the streaming-unmarshalling example and the partial-unmarshalling example in the JAXB RI distribution for more about how to do this. The streaming-unmarshalling example has an advantage that it can handle chunks at arbitrary nest level, yet it requires you to deal with the push model --- JAXB unmarshaller will "push" new chunk to you and you'll need to process them right there.

In contrast, the partial-unmarshalling example works in a pull model (which usually makes the processing easier), but this approach has some limitations in databinding portions other than the repeated part.

4.4.2. Processing a live stream of XML

The techniques discussed above can be used to handle this case as well, since they let you unmarshal chunks one by one. See the xml-channel example in the JAXB RI distribution for more about how to do this.

4.4.3. Creating virtual infosets

For further advanced cases, one could always run a streaming infoset conversion outside JAXB API and basically curve just the portion of the infoset you want to data-bind, and feed it as a complete infoset into JAXB API. JAXB API accepts XML infoset in many different forms (DOM, SAX, StAX), so there's a fair amount of flexibility in choosing the right trade off between the development effort in doing this and the runtime performance.

For more about this, refer to the respective XML infoset API.

5. Marshalling

5.1. Changing prefixes

By default, a JAXB marshaller uses random namespace prefixes (such as ns1, ns2, ...) when it needs to declare new namespace URIs. While this is perfectly valid XML wrt the schema, for human readability, you might want to change them to something that makes more sense.

The JAXB RI defines NamespacePrefixMapper to allow you to do this. See the namespace-prefix sample in the distribution for more details.

5.2. Element default values and marshalling

Because of a "strange" way element default values in XML Schema work, people often get confused about its behavior. This section describes how this works.

When a class has an element property with the default value, and if a value is null, then the marshaller will not produce the corresponding element in XML:

Example 72. Element defaults and XML

@XmlRootElement
class Foo {
  @XmlElement(defaultValue="value") public String a=null;
}

marshaller.marshal(new Foo(),System.out);

Example 73. Marshalling output from above

<foo />

This is consistent with the XML Schema spec, where it essentially states that the element defaults do not kick in when the element is absent. Attribute default values do not have this problem, so if you can change the schema, changing it to an attribute is usually a better idea. Alternatively, depending on your expectation, setting the field to a default value in Java may achieve what you are looking for.

Example 74. Possible changes

@XmlRootElement
class Foo {
  @XmlElement public String a="value";
}
@XmlRootElement
class Bar {
  @XmlAttribute public String a;
}

marshaller.marshal(new Foo(),System.out);
marshaller.marshal(new Bar(),System.out);

Example 75. Marshalling output from above

<foo>
    <a>value</a>
</foo>

<bar/>

Also, see Element default values and unmarshalling.

5.3. Different ways of marshalling

5.3.1. Different output media

The most basic notion of the marshalling is to take a JAXB-bound object that has @XmlRootElement, and write it out as a whole XML document. So perhaps you have a class like this:

Example 76. JAXB POJO

class Point {
  @XmlElement
  public int x;
  @XmlElement
  public int y;
  Point(...) { ... }
}

Then you can do:

Example 77. Plain marshalling

marshaller.marshal( new Point(1,3), System.out );
marshaller.marshal( new Point(1,3), new File("out.xml") );

.. and so on. There're seven Marshaller.marshal methods that takes different output media as the second parameter. If you are writing to a file, a socket, or memory, then you should use the version that takes OutputStream. Unless you change the target encoding to something else (default is UTF-8), there's a special marshaller codepath for OutputStream, which makes it run really fast. You also don't have to use BufferedOutputStream, since the JAXB RI does the adequate buffering.

You can also write to Writer, but in this case you'll be responsible for encoding characters, so in general you need to be careful. If you want to marshal XML into an encoding other than UTF-8, it's best to use the JAXB_ENCODING property and then write to OutputStream, as it escapes characters to things like &#x1824; correctly.

The next medium we support is W3C DOM. This is bit unintuitive, but you'll do it like this:

Example 78. Marshal to DOM

DocumentBuilderFactory dbf = DocumentBuilderFactory.newInstance();
dbf.setNamespaceAware(true);
Document doc = dbf.newDocumentBuilder().newDocument();

marshaller.marshal( new Point(1,3), doc );

And after the method invocation you get a complete DOM tree that represents the marshalled document.

The other versions of the marshal methods are there to write XML documents in terms of other XML APIs, such as SAX and StAX. The version that takes ContentHandler is useful when you need a custom formatting needs (like you want each attribute to be in new line, etc), but otherwise they are not very interesting if you are writing a whole document.

5.3.2. Marshalling into a subtree

Another common use of JAXB is where you are writing a bigger document, and you use JAXB to generate part(s) of it. The JAX-WS RI is the prime example. It produces a SOAP message, and JAXB is only used to produce the body. When you are doing this, you first set JAXB_FRAGMENT property on the marshaller. This changes the behaviors of the marshaller so that it works better in this situation.

If you are writing to an OutputStream or Writer and generally sending it to someone else, you can do something like this:

Example 79. Marshalling into a subtree

System.out.println("<envelope>");
marshaller.marshal( object, System.out );
System.out.println("</envelope>");

Like I mentioned, this is probably the fastest, even though println isn't very pretty. JAXB_FRAGMENT prevents the marshaller from producing an XML declaration, so the above works just fine. The downside of this approach is that if the ancestor elements declare the namespaces, JAXB won't be able to take advantage of them.

You can also marshal an object as a subtree of an existing DOM tree. To do this, you pass the Element object as the second parameter, and the marshaller will marshal an object as a child of this node.

StAX is also very convenient for doing this sort of things. You can create XMLStreamWriter, write some stuff, and then pass that to the marshaller. JAXB_FRAGMENT prevents the marshaller from producing startDocument and endDocument token. When doing this sub-tree marshaling to DOM and StAX, JAXB can take advantage of available in-scope namespace bindings.

Finally, you can marshal an object as a subtree into ContentHandler, but it requires a fair amount of SAX programming experience, and it goes beyond the scope of this entry.

5.3.3. Marshalling a non-element

Another common use case is where you have an object that doesn't have @XmlRootElement on it. JAXB allows you to marshal it like this:

Example 80. Marshalling a non-element

marshaller.marshal( new JAXBElement(
  new QName("","rootTag"),Point.class,new Point(...)));

This puts the <rootTag> element as the root element, followed by the contents of the object, then </rootTag>. You can actually use it with a class that has @XmlRootElement, and that simply renames the root element name.

At the first glance the second Point.class parameter may look redundant, but it's actually necessary to determine if the marshaller will produce (infamous) @xsi:type. In this example, both the class and the instance are Point, so you won't see @xsi:type. But if they are different, you'll see it.

This can be also used to marshal a simple object, like String or an integer.

Marshalling a non-element with @xsi:type

marshaller.marshal( new JAXBElement(
  new QName("","rootTag"),String.class,"foo bar"));

But unfortunately it cannot be used to marshal objects like List or Map, as they aren't handled as the first-class citizen in the JAXB world.

5.3.4. Connecting to other XML APIs

Because of the Source and Result support, JAXB objects can be easily marshalled into other XML APIs that are not mentioned here. For example, dom4j has DocumentResult that extends Result, so you can do:

Example 81. Marshalling to dom4j

DocumentResult dr = new DocumentResult();
marshaller.marshal( object, dr );
o = dr.getDocument();

Similar mechanism is available for JDOM and XOM. This conversion is much more efficient than first marshalling to ByteArrayOutputStream and then read it back into these DOMs. The same mechanism can be used to marshal to FastInfoset or send the marshaled document to an XSLT engine (TransformerHandler.)

The other interesting connector is JAXBSource, which wraps a marshaller and allows a JAXB object to be used as a "source" of XML. Many XML APIs take Source as an input, and now JAXB object can be passed to them directly.

For example, you can marshal a JAXB object and unmarshal it into another JAXBContext like this:

Example 82. Loading into a different JAXBContext

JAXBContext context1 = ... ;
JAXBContext context2 = ... ;

context1.createUnmarshaller().unmarshal( new JAXBSource(context2,object) );

This amounts to looking at the same XML by using different schema, and again this is much more efficient than going through ByteArrayOutputStream.

5.4. Interaction between marshalling and DOM

Sometimes you may notice that JAXB is producing XML with seemingly unnecessary namespace declarations. In this section, we'll discuss the possible causes and how to resolve this.

5.4.1. Caused by DOM mapping

The #1 cause of extra namespace declarations is due to the DOM mapping. This mainly happens because of a schema construct that forces XJC to generate a property with DOM. This includes the use of wildcard <xs:any/> (see more about this Mapping of <xs:any />), as well as xs:anyType (which can also happen by omission, such as <xs:element name="foo"/>, which is interpreted as <xs:element name="foo" type="xs:anyType" />.

During unmarshalling, when a subtree of the input XML is converted into XML, JAXB copies all the in-scope namespace bindings active at that time to the root of the DOM element. So for example, given the following Java class and XML, the DOM tree that the child field will get will look like the following:

Example 83. Bean with wildcard

@XmlRootElement
class Foo {
  @XmlAnyElement
  public Element child;
}

Example 84. Instance with subtree matching wildcard

<foo xmlns:a="a" xmlns:b="b" xmlns:c="c">
  <subtree xmlns:c="cc">
    <data>a:xyz</data>
  </subtree>
</foo>

Example 85. DOM tree to be stored in Foo.child

<subtree xmlns:a="a" xmlns:b="b" xmlns:c="cc">
    <data>a:xyz</data>
  </subtree>

Note that the two namespace declarations are copied over, but c is not because it's overridden. Also not that JAXB is not touching the whitespace in document. This copying of namespace declarations is necessary to preserve the infoset in the input document. For example, if the <data> is a QName, its meaning would change if JAXB unmarshaller doesn't copy it.

Now, imagine what happens when you marshal this back to XML. Despite the fact that in this example neither b nor c prefixes are in use, JAXB cannot delete them, because it doesn't know if those attributes are significant to the application or not. Therefore, this could end up producing XML with "extra namespace declarations" like:

Example 86. DOM tree to be stored in Foo.child

<foo>
  <subtree xmlns:a="a" xmlns:b="b" xmlns:c="cc">
    <data>a:xyz</data>
  </subtree>
</foo>

Resolving this problem is not possible in the general case, but sometimes one of the following strategy works:

  1. Sometimes schema author incorrectly assumes that <xs:element name="foo"/> means <xs:element name="foo" type="xs:string"/>, because attribute declarations work somewhat like this. In such a case, adding explicit type attribute avoids the use of DOM, so things will work as expected.

  2. The wildcard processing mode " strict" would force a typed binding, and thereby eliminate any DOM mapping.

  3. You might be able to manulally go into the DOM tree and remove unnecessary namespace declarations, if your application knows what are necessary and what are not.

6. Schema Generation

6.1. Invoking schemagen programatically

Schemagen tools by default come in as CLI, ant task, and Maven plugin. These interfaces allow you to invoke schemagen functionality from your program.

6.1.1. At runtime

If the classes you'd like to generate schema from are already available as java.lang.Class objects (meaning they are already loaded and resolved in the current JVM), then the easiest way to generate a schema is to use the JAXB API:

Example 87. Generate schema at runtime

File baseDir = new File(".");

class MySchemaOutputResolver extends SchemaOutputResolver {
    public Result createOutput( String namespaceUri, String suggestedFileName ) throws IOException {
        return new StreamResult(new File(baseDir,suggestedFileName));
    }
}

JAXBContext context = JAXBContext.newInstance(Foo.class, Bar.class, ...);
context.generateSchema(new MySchemaOutputResolver());

6.1.2. CLI interface

The CLI interface (public static int com.sun.tools.jxc.SchemaGenerator.run(String[])) is the easiest API to access. You can pass in all the schemagen command-line arguments as a string array, and get the exit code as an int value. Messages are sent to System.err and System.out.

6.1.3. Ant interface

Ant task can be invoked very easily from a non-Ant program. The schemagen ant task is defined in the SchemaGenTask class,

6.1.4. Native Java API

The above two interfaces are built on top of externally committed contracts, so they'll evolve only in a compatibile way. The downside is that the amount of control you can exercise over them would be limited.

So yet another approach to invoke schemagen is to use JAXB RI's internal interfaces. But be warned that those interfaces are subject to change in the future versions, despite our best effort to preserve them. This is the API that the JAX-WS RI uses to generate schema inside WSDL when they generate WSDL, so does some other web services toolkits that work with the JAXB RI.

Most of those interfaces are defined and well-documented in the com.sun.tools.xjc.api package. You can see how the schemagen tools are eventually calling into this API at the implementaion of SchemaGenerator class.

6.2. Generating Schema that you want

This section discusses how you can change the generated XML schema. For changes that also affect the infoset (such as changing elements to attributes, namespaces, etc.), refer to a different section "XML layout and in-memory data layout".

6.2.1. Adding facets to datatypes

As of JAXB 2.1.4, currently no support for this, although there has been several discussions in the users alias.

The JAXB project is currently lacking resources to attack this problem, and therefore looking for volunteers to work on this project. The basic idea would be to define enough annotations to cover the basic constraint facets (such as length, enumerations, pattern, etc.) The schema generator would have to be then extended to honor those annotations and generate schemas accordingly.

Some users pointed out relevance of this to JSR 303: Bean Validation. If you are interested in picking up this task, let us know!

7. Deployment

7.1. Migrating JAXB 2.0 applications to JavaSE 6

JavaSE 6 ships with its own JAXB 2.0 implementation. This implementation is based on the JAXB RI, where the only differences are:

  • No RI-specific vendor extensions are supported: This is so that portability across different JavaSE 6 implementations will be guaranteed.

  • Code in JavaSE 6 is hosted in its own packages to avoid conflicts: This allows applications to continue to use a specific version of the JAXB RI that they choose to use.

Therefore, if you develop an application that uses JAXB 2.0 for JavaSE 5 today, the easiest way to upgrade to JavaSE 6 is to do nothing. You should keep the JAXB RI in your development environment, keep bundling the JAXB RI runtime jars to your app, just like you do that today.

7.1.1. Reducing footprint

If you'd like to reduce the footprint of your application by taking advantage of a JAXB implementation in JavaSE, you can take the following steps:

  1. You will no longer have to ship jaxb-api.jar in your application. This doesn't require any change to your code.

  2. If your application does not use any of the vendor extension features of the JAXB RI runtime (such as unmarshaller/marshaller properties whose names start with "com.sun."), then you will no longer have to ship jaxb-impl.jar (nor jaxb1-impl.jar, jaxb-libs.jar.) When you do this, be sure to test your application since it's not very easy to find all such dependencies.

7.1.2. Using JAXB with Java SE

JavaSE comes with JAXB 2.x API/implementation in rt.jar. Each version of JavaSE (6, 7, 8, ...) contains different version of JAXB 2.x API. Therefore, if you want to use different version of JAXB API/implementation than the one present in your version of JDK, you are required to override a portion of rt.jar with the new API. There are several ways to achieve this:

  1. Place the jaxb-api.jar into $JRE_HOME/lib/endorsed and put jaxb-impl to classpath of your application. Do not put jaxb-impl or jaxb-xjc into the endorsed directory. This essentially makes your JRE to "JRE X + JAXB 2.y". This won't affect any other applications that use this JRE, and it's easy. On the other hand, in various scenarios you may not be able to alter the JRE.

  2. Use the system property java.endorsed.dirs when you launch your application, and have it point to the directory which contains the jaxb-api.jar only. The directory must not contain any other jaxb artifacts (like jaxb-impl.jar or jaxb-xjc.jar). This allows you use to use different version of JAXB without modifying the JRE.

No matter which approach you take, make sure not to include jar files other than jaxb-api.jar. Doing so, for example including jaxb-xjc.jar, may result in classloading related errors such as "taskdef A class needed by class com.sun.tools.xjc.XJCTask cannot be found: org/apache/tools/ant/...."

See the endorsed directory mechanism for more details.

7.1.3. Where's the XJC ant task?

JavaSE has never shipped an Ant task implementation, so we are just following that tradition. There's an (process-wise) overhead of adding additional dependencies during the JavaSE build, and there would likely be some runtime dependency issues in having a class in tools.jar that would require the ant classes, due to class loader delegation.

We are thinking about perhaps releasing a small jar that only contains the ant task for JDK6.

Please also note that the syntax of <xjc> task is neither defined in the JAXB spec nor in the JavaSE spec. Therefore other JavaSE vendors may not implement that at all, or do so in a different class name, etc. Therefore, from a portability perspective, if you choose to depend on the <xjc> task you should bundle the JAXB RI.

7.2. Which JAXB RI is included in which JDK?

This table captures the history of the JAXB RI integration into JDK. This is for information purposes only. In particular, the JAXB RI in JDK is package renamed to avoid any conflict with the stand-alone JAXB RI distribution or with the jaxb2-reflection library. You can use any version of these with any version of JDK without worrying about implementation conflicts.

JDK6 first release - JDK6 u3

JAXB RI 2.0.3

JDK6 u4

JAXB RI 2.1.3

JDK6 u14

JAXB RI 2.1.10

JDK7 first release

JAXB RI 2.2.4-1

7.3. Running JAXB 1.0 and 2.x side by side

You should still be able to use the legacy compatible jaxb1-impl.jar: http://search.maven.org/remotecontent?filepath=com/sun/xml/bind/jaxb1-impl/2.2.5-1/jaxb1-impl-2.2.5-1.jar with your application. Note however that this jar is no longer distributed with JAXB 2.2.x releases and this functionality is no longer supported.

7.4. Migrating JAXB 1.0 applications to JAXB 2.x

While it's always possible to keep your JAXB 1.0 applications as-is and run them on the JAXB 2.x runtime, or have some parts of your application use JAXB 1.0 while others use 2.x, there are situations where you might prefer to migrate your existing applications from JAXB 1.0 to JAXB 2.x.

Common reasons why people would like to migrate includes:

  • Taking advantage of the flexible, robust unmarshalling in JAXB 2.0

  • Exposing the JAXB-bound objects through JAX-WS as web services

7.4.1. Typical migration work

This section outlines how the typical migration work proceeds:

  1. Replace the invocation of XJC 1.0.x with XJC 2.x. Both the CLI and ant task syntax are backward-compatible to make this easier.

  2. Re-compile your schema with XJC 2.x. This will generate a new and different set of classes and methods.

  3. Update your application code that deals with the generated code to use the newer classes and methods. This step is mostly straight-forward but may be time consuming, especially if your project is big.

  4. Remove the JAXB 1.0.x runtime from your application, and place in JAXB 2.x runtime instead. See the list for what jars were used in 1.0 and what are used in 2.0.

The classes generated from JAXB 2.x tends to be a lot more compact.

JAXB RI 2.x ships many of the same samples that JAXB RI 1.0.x shipped, except that they are updated to work with JAXB 2.0. So comparing those samples may help you better understand what this type of migration involves.

7.4.2. Other Miscellaneous Notes

  • JAXB 2.x requires JavaSE 5 or above, whereas JAXB 1.0 runs on JavaSE 1.3 or above.

  • The javax.xml.bind interfaces remain the same. So the part of the code that deals with those interfaces may remain the same.

  • JAXB 1.0 unmarshaller was a lot more draconian to the errors in the input document than 2.0 is. In a sense, you can think of the 1.0 unmarshaller as equivalent to the 2.0 unmarshaller + validator. If your application was relying on this behaivior, you may want to enable validation in 2.0 by using Unmarshaller.setSchema.

  • JAXB 2.x by default creates plain Java beans, whereas in 1.0 it created separated interfaces and implementations. Many people find beans to be easier to work with, but you can use the following customization to tell JAXB 2.x to generate interfaces and implementations in a similar fashion.

    Example 88. Interface/Implementation customization

    <jxb:bindings version="2.1"
                  xmlns:jxb="http://java.sun.com/xml/ns/jaxb">
      <jxb:globalBindings generateValueClass="false" />
    </jxb:bindings>


  • When creating instances of the generated classes, JAXB 1.0 required that you do so through ObjectFactory, like new ObjectFactory().createAbc(). With 2.0, you can simply do new Abc(). This no longer throws JAXBExceptions, either.

  • Binding customization syntax of 2.0 is backward compatible with 1.0. However, there are several customizations that were previously allowed in 1.0 but not in 2.0, such as using class customizations on model groups.

7.5. Runtime Errors

7.5.1. Illegal class modifiers for package-info: 0x1600

When you compile your sources files with JDK from Java6 and try to run it on Java5 JVMs, you may see a ClassFormatError at runtime like this:

main" java.lang.ClassFormatError: Illegal class modifiers in class
com/alu/parentalcontrol/jaxb/package-info: 0x1600
at java.lang.ClassLoader.defineClass1(Native Method)
at java.lang.ClassLoader.defineClass(ClassLoader.java:620)
at java.security.SecureClassLoader.defineClass(SecureClassLoader.java:124)
at java.net.URLClassLoader.defineClass(URLClassLoader.java:260)
at java.net.URLClassLoader.access$100(URLClassLoader.java:56)
at java.net.URLClassLoader$1.run(URLClassLoader.java:195)
at java.security.AccessController.doPrivileged(Native Method)
at java.net.URLClassLoader.findClass(URLClassLoader.java:188)
at java.lang.ClassLoader.loadClass(ClassLoader.java:306)
at sun.misc.Launcher$AppClassLoader.loadClass(Launcher.java:268)
at java.lang.ClassLoader.loadClass(ClassLoader.java:251)
at java.lang.ClassLoader.loadClassInternal(ClassLoader.java:319)
at java.lang.Class.forName0(Native Method)
at java.lang.Class.forName(Class.java:242)
at java.lang.Package.getPackageInfo(Package.java:350)
at java.lang.Package.getAnnotation(Package.java:361)
at com.sun.xml.bind.v2.model.annotation.RuntimeInlineAnnotationReader.getPackageAnnotation(RuntimeInlineAnnotationReader.java:125)
at com.sun.xml.bind.v2.model.annotation.RuntimeInlineAnnotationReader.getPackageAnnotation(RuntimeInlineAnnotationReader.java:53)
at com.sun.xml.bind.v2.model.impl.TypeInfoImpl.parseElementName(TypeInfoImpl.java:122)
at com.sun.xml.bind.v2.model.impl.ClassInfoImpl.<init>(ClassInfoImpl.java:166)

This is due to a bug, and so far the only way to fix this is to compile your project with JavaSE 5.

Here is what's happening. Java5 added a new class file modifier attribute called ACC_SYNTHETIC, whose bit value is 0x1000. This attribute marks types and methods that are not present in the source file but generated by the compiler. When package-info.java is compiled into package-info.class, javac in Java5 apparently only put 0x0600 (= ACC_ABSTRACT|ACC_INTERFACE.) Some time during Java6 development, someone filed a bug, saying it should also include ACC_SYNTHETIC bit, since it was a synthesized class.

Later it is discovered that the corresponding VM change needs to be made to allow this 0x1600 combination, but apparently no one realized the real implication of this --- namely, 0x1600 will break all the past JVMs. Of course, this problem is eventually discovered, but as of this writing there's still no fix for this.

So as you see, this is why the only workaround is to use javac from Java5.

8. Other Miscellaneous Topics

8.1. Performance and thread-safety

The JAXBContext class is thread safe, but the Marshaller, Unmarshaller, and Validator classes are not thread safe.

For example, suppose you have a multi-thread server application that processes incoming XML documents by JAXB. In this case, for the best performance you should have just one instance of JAXBContext in your whole application like this:

Example 89. Singleton JAXBContext

class MyServlet extends HttpServlet {
    static final JAXBContext context = initContext();

    private static JAXBContext initContext() {
        return JAXBContext.newInstance(Foo.class,Bar.class);
    }
}

And each time you need to unmarshal/marshal/validate a document. Just create a new Unmarshaller/Marshaller/Validator from this context, like this:

Example 90. Thread local Unmarshaller

    public void doGet( HttpServletRequest req, HttpServletResponse ) {
        Unmarshaller u = context.createUnmarshaller();
        u.unmarshal(...);
    }

This is the simplest safe way to use the JAXB RI from multi-threaded applications.

If you really care about the performance, and/or your application is going to read a lot of small documents, then creating Unmarshaller could be relatively an expensive operation. In that case, consider pooling Unmarshaller objects. Different threads may reuse one Unmarshaller instance, as long as you don't use one instance from two threads at the same time.

8.2. Compiling DTD

The JAXB RI is shipped with an "experimental" DTD support, which let's you compile XML DTDs. It is marked "experimental" not because the feature is unstable nor unreliable, but rather because it's not a part of the JAXB specification and therefore the level of commitment to compatibility is lower.

Example 91. To compile a DTD, run the XJC binding compiler as follows:

$ xjc.sh -dtd test.dtd

All the other command-line options of the XJC binding compiler can be applied. Similarly, the XJC ant task supports DTD. The generated code will be no different from what is generated from W3C XML Schema. You'll use the same JAXB API to access the generated code, and it is portable in the sense that it will run on any JAXB 2.0 implementation.

DTD long predates XML namespace, although people since then developed various techniques to use XML namespaces in conjunction with DTD. Because of this, XJC is currently unable to reverse-engineer the use of XML namespace from DTD. If you compile DTDs that use those techniques, you'd either manuallly modify the generated code, or you can try a tool like Trang that can convert DTD into XML Schema in ways that better preserves XML namespaces.

8.2.1. Customizations

The customization syntax for DTD is roughly based on the ver.0.21 working draft of the JAXB specification, which is available at xml.coverpages.org. The deviations from this document are:

  • The whitespace attribute of the conversion element takes " preserve", " replace", and " collapse" instead of " preserve", " normalize", and " collapse" as specified in the document.

  • The interface customization just generates marker interfaces with no method.

8.2.2. Compiling DTD from Maven2

Example 92. The following POM snippest describes how to invoke XJC to compile DTD from a Maven 2 project:

<plugin>
  <groupId>org.jvnet.jaxb2.maven2</groupId>
  <artifactId>maven-jaxb2-plugin</artifactId>
  <executions>
    <execution>
      <goals>
        <goal>generate</goal>
      </goals>
      <configuration>
        <!--  if you want to put DTD somewhere else
        <schemaDirectory>src/main/jaxb</schemaDirectory>
        -->
        <extension>true</extension>
        <schemaLanguage>DTD</schemaLanguage>
        <schemaIncludes>
          <schemaInclude>*.dtd</schemaInclude>
        </schemaIncludes>
        <bindingIncludes>
          <bindingInclude>*.jaxb</bindingInclude>
        </bindingIncludes>
        <args>
          <arg>-Xinject-listener-code</arg>
        </args>
      </configuration>
    </execution>
  </executions>
  <dependencies>
    <dependency>
      <groupId>org.jvnet.jaxb2-commons</groupId>
      <artifactId>property-listener-injector</artifactId>
      <version>1.0</version>
    </dependency>
  </dependencies>
</plugin>

Example 93. The dependencies section inside the plugin element can be used to specify additional XJC plugins. If you'd like to use more recent version of the JAXB RI, you can specify a dependency to XJC here to do so, like this:

<dependency>
  <groupId>com.sun.xml.bind</groupId>
  <artifactId>jaxb-xjc</artifactId>
  <version>2.1.2</version>
</dependency>

The complete sample project is available from here.

8.3. Using JAXB from Maven

Example 94. If you are using Maven 2 as your build system, you can declare the following dependencies in your POM for using the JAXB RI. Replace the version with the version of your choice.

<dependency>
  <groupId>com.sun.xml.bind</groupId>
  <artifactId>jaxb-impl</artifactId>
  <version>&jaxb.release.impl.version;</version>
</dependency>

This artifact is available in the java.net Maven repository, so you also need the following <repository> declaration in your POM:

<repository>
  <id>java.net</id>
  <url>https://maven.java.net/content/repositories/releases/</url>
  <layout>default</layout>
</repository>

For a list of available artifacts and versions, see this. In addition to the runtime jar, XJC is available under the artifact Id "jaxb-xjc".

You can also invoke XJC through a Maven plugin by using the Maven2 JAXB 2.x plugin.

8.4. Designing a client/server protocol in XML

Occasionally, people try to define a custom protocol that allows multiple XML requests/responses to be sent over a single transport channel. This section discusses the non-trivial interaction between XML and sockets, and how you can design a protocol correctly.

XML1.0 requires a conforming parser to read the entire data till end of the stream (because a parser needs to handle documents like <root/><!-- post root comment -->). As a result, a naive attempt to keep one OutputStream open and marshal objects multiple times fails.

Example 95. One easy way to work around this limitation is to design your protocol so that the data on the wire will look like the following:

<conversation>
  <!-- message 1 -->
  <message>
    ...
  </message>

  <!-- message 2 -->
  <message>
    ...
  </message>

  ...
</conversation>

The <conversation> start tag is sent immediately after the socket is opened. This works as a container to send multiple "messages", and this is also an excellent opportunity to do the hand-shaking (e.g., protocol-version='1.0' attribute.) Once the <conversation> tag is written, multiple messages can be marshalled as a tree into the channel, possibly with a large time lag in between. You can use the JAXB marshaller to produce such message. When the sender wants to disconnect the channel, it can do so by sending the </conversation> end tag, followed by the socket disconnection.

Of course, you can choose any tag names freely, and each message can have different tag names.

The receiver would use the StAX API and use XMLStreamReader to read this stream. You'd have to use this to process the first <conversation> start tag. After that, every time you call a JAXB unmarshaller, you'll get the next message.

For the concrete code, see the xml-channel example in the JAXB RI distribution.

Tools

1. XJC

1.1. xjc Overview

JAXB RI also provides an Ant task to run the binding complier - see the instructions for XJC Ant Task.

1.2. Launching xjc

The binding compiler can be launched using the appropriate xjc shell script in the bin directory for your platform.

  • Solaris/Linux

    % /path/to/jaxb/bin/xjc.sh -help
  • Windows

    > c:\path\to\jaxb\bin\xjc.bat -help

1.2.1. Execute the jaxb-xjc.jar JAR File

If all else fails, you should be able to execute the jaxb-xjc.jar file:

  • Solaris/Linux

    % java -jar $JAXB_HOME/lib/jaxb-xjc.jar -help
  • Windows

    > java -jar %JAXB_HOME%\lib\jaxb-xjc.jar -help

This is equivalent of running xjc.sh orxjc.bat, and it allows you to set the JVM parameters.

Be aware of having jaxb-impl.jar (or any other JAXB runtime implementation, e.g. EclipseLink MOXy) on the classpath.

1.3. xjc Syntax

xjc [OPTION]... <schema file/URL/dir/jar> [-b <binding>...]

Usage: xjc [-options ...] <schema file/URL/dir/jar> ... [-b <bindinfo>] ...
If dir is specified, all schema files in it will be compiled.
If jar is specified, /META-INF/sun-jaxb.episode binding file will be compiled.
Options:
  -nv                :  do not perform strict validation of the input schema(s)
  -extension         :  allow vendor extensions - do not strictly follow the
                        Compatibility Rules and App E.2 from the JAXB Spec
  -b <file/dir>      :  specify external bindings files (each <file> must have its own -b)
                        If a directory is given, **/*.xjb is searched
  -d <dir>           :  generated files will go into this directory
  -p <pkg>           :  specifies the target package
  -httpproxy <proxy> :  set HTTP/HTTPS proxy. Format is [user[:password]@]proxyHost:proxyPort
  -httpproxyfile <f> :  Works like -httpproxy but takes the argument in a file to protect password 
  -classpath <arg>   :  specify where to find user class files
  -catalog <file>    :  specify catalog files to resolve external entity references
                        support TR9401, XCatalog, and OASIS XML Catalog format.
  -readOnly          :  generated files will be in read-only mode
  -npa               :  suppress generation of package level annotations (**/package-info.java)
  -no-header         :  suppress generation of a file header with timestamp
  -target (2.0|2.1)  :  behave like XJC 2.0 or 2.1 and generate code that doesn't use any 2.2 features.
  -encoding <encoding> :  specify character encoding for generated source files
  -enableIntrospection :  enable correct generation of Boolean getters/setters to enable Bean Introspection apis 
  -disableXmlSecurity  :  disables XML security features when parsing XML documents 
  -contentForWildcard  :  generates content property for types with multiple xs:any derived elements 
  -xmlschema         :  treat input as W3C XML Schema (default)
  -relaxng           :  treat input as RELAX NG (experimental,unsupported)
  -relaxng-compact   :  treat input as RELAX NG compact syntax (experimental,unsupported)
  -dtd               :  treat input as XML DTD (experimental,unsupported)
  -wsdl              :  treat input as WSDL and compile schemas inside it (experimental,unsupported)
  -verbose           :  be extra verbose
  -quiet             :  suppress compiler output
  -help              :  display this help message
  -version           :  display version information
  -fullversion       :  display full version information

Extensions:
  -Xinject-code      :  inject specified Java code fragments into the generated code
  -Xlocator          :  enable source location support for generated code
  -Xsync-methods     :  generate accessor methods with the 'synchronized' keyword
  -mark-generated    :  mark the generated code as @javax.annotation.Generated
  -episode           :  generate the episode file for separate compilation
  -Xpropertyaccessors :  Use XmlAccessType PROPERTY instead of FIELD for generated classes

1.3.1. Summary of Command Line Options

-nv

By default, the XJC binding compiler performs strict validation of the source schema before processing it. Use this option to disable strict schema validation. This does not mean that the binding compiler will not perform any validation, it simply means that it will perform less-strict validation.

-extension

By default, the XJC binding compiler strictly enforces the rules outlined in the Compatibility chapter of the JAXB Specification. Appendix E.2 defines a set of W3C XML Schema features that are not completely supported by JAXB v1.0. In some cases, you may be allowed to use them in the "-extension" mode enabled by this switch. In the default (strict) mode, you are also limited to using only the binding customizations defined in the specification. By using the "-extension" switch, you will be allowed to use the Overview.

-b <file>

Specify one or more external binding files to process. (Each binding file must have it's own -b switch.) The syntax of the external binding files is extremely flexible. You may have a single binding file that contains customizations for multiple schemas or you can break the customizations into multiple bindings files:

xjc schema1.xsd schema2.xsd schema3.xsd -b bindings123.xjb
xjc schema1.xsd schema2.xsd schema3.xsd -b bindings1.xjb -b bindings2.xjb -b bindings3.xjb

In addition, the ordering of the schema files and binding files on the command line does not matter.

-d <dir>

By default, the XJC binding compiler will generate the Java content classes in the current directory. Use this option to specify an alternate output directory. The output directory must already exist, the XJC binding compiler will not create it for you.

-encoding <encoding>

Set the encoding name for generated sources, such as EUC-JP or UTF-8. If -encoding is not specified, the platform default encoding is used.

-p <pkg>

Specifying a target package via this command-line option overrides any binding customization for package name and the default package name algorithm defined in the specification.

-httpproxy <proxy>

Specify the HTTP/HTTPS proxy. The format is [user[:password]@]proxyHost[:proxyPort]. The old -host and -port are still supported by the RI for backwards compatibility, but they have been deprecated.

-httpproxyfile <f>

Same as the -httpproxy <proxy> option, but it takes the <proxy> parameter in a file, so that you can protect the password (passing a password in the argument list is not safe.)

-classpath <arg>

Specify where to find client application class files used by the <jxb:javaType> and <xjc:superClass> customizations.

-catalog <file>

Specify catalog files to resolve external entity references. Supports TR9401, XCatalog, and OASIS XML Catalog format. Please read the catalog-resolver sample application.

-readOnly

By default, the XJC binding compiler does not write-protect the Java source files it generates. Use this option to force the XJC binding compiler to mark the generated Java sources read-only.

-npa

Supress the generation of package level annotations into **/package-info.java. Using this switch causes the generated code to internalize those annotations into the other generated classes.

-no-header

Supress the generation of a file header comment that includes some note and timestamp. Using this makes the generated code more diff-friendly.

-target (2.0|2.1)

Avoid generating code that relies on any JAXB 2.1|2.2 features. This will allow the generated code to run with JAXB 2.0 runtime (such as JavaSE 6.)

-xmlschema

treat input schemas as W3C XML Schema (default). If you do not specify this switch, your input schemas will be treated as W3C XML Schema.

-relaxng

Treat input schemas as RELAX NG (experimental, unsupported). Support for RELAX NG schemas is provided as a Overview.

-relaxng-compact

Treat input schemas as RELAX NG compact syntax(experimental, unsupported). Support for RELAX NG schemas is provided as a Overview.

-dtd

Treat input schemas as XML DTD (experimental, unsupported). Support for RELAX NG schemas is provided as a Overview.

-wsdl

Treat input as WSDL and compile schemas inside it (experimental,unsupported).

-quiet

Suppress compiler output, such as progress information and warnings..

-verbose

Be extra verbose, such as printing informational messages or displaying stack traces upon some errors..

-help

Display a brief summary of the compiler switches.

-version

Display the compiler version information.

<schema file/URL/dir>

Specify one or more schema files to compile. If you specify a directory, then xjc will scan it for all schema files and compile them.

-Xlocator

This feature causes the generated code to expose SAX Locator information about the source XML in the Java bean instances after unmarshalling.

-Xsync-methods

This feature causes all of the generated method signatures to include the synchronized keyword.

-mark-generated

This feature causes all of the generated code to have @Generated annotation.

-episode <FILE>

Generate an episode file from this compilation, so that other schemas that rely on this schema can be compiled later and rely on classes that are generated from this compilation. The generated episode file is really just a JAXB customization file (but with vendor extensions.)

-Xinject-code

Inject specified Java code fragments into the generated code; see here for more details.

-Xpropertyaccessors>

Annotate the @XmlAccessorType of generated classes with XmlAccessType PROPERTY instead of FIELD

1.3.2. Summary of Deprecated and Removed Command Line Options

-host & -port

These options have been deprecated and replaced with the -httpproxy option. For backwards compatibility, we will continue to support these options, but they will no longer be documented and may be removed from future releases.

-use-runtime

Since the JAXB 2.0 specification has defined a portable runtime, it is no longer necessary for the JAXB RI to generate **/impl/runtime packages. Therefore, this switch is obsolete and has been removed.

1.4. Compiler Restrictions

In general, it is safest to compile all related schemas as a single unit with the same binding compiler switches.

Please keep the following list of restrictions in mind when running xjc. Most of these issues only apply when compiling multiple schemas with multiple invocations of xjc.

  • To compile multiple schemas at the same time, keep the following precedence rules for the target Java package name in mind:

    1. The -p command line option takes the highest precedence.

    2. <jaxb:package> customization

    3. If targetNamespace is declared, apply targetNamespace -> Java package name algorithm defined in the specification.

    4. If no targetNamespace is declared, use a hardcoded package named "generated".

  • It is not legal to have more than one < jaxb:schemaBindings> per namespace, so it is impossible to have two schemas in the same target namespace compiled into different Java packages.

  • All schemas being compiled into the same Java package must be submitted to the XJC binding compiler at the same time - they cannot be compiled independently and work as expected.

  • Element substitution groups spread across multiple schema files must be compiled at the same time.

1.5. Generated Resource Files

XJC produces a set of packages containing Java source files and also jaxb.properties files, depending on the binding options you used for compilation. When generated, jaxb.properties files must be kept with the compiled source code and made available on the runtime classpath of your client applications:

2. XJC Ant Task

2.1. xjc Task Overview

The jaxb-xjc.jar file contains the XJCTask.class file, which allows the XJC binding compiler to be invoked from the Ant build tool. To use XJCTask, include the following statement in your build.xml file:

<taskdef name="xjc" classname="com.sun.tools.xjc.XJCTask">
    <classpath>
        <fileset dir="path/to/jaxb/lib" includes="*.jar"/>
    </classpath>
</taskdef>

This maps XJCTask to an Ant task named xjc. For detailed examples of using this task, refer to any of the build.xml files used by the Sample Apps.

2.2. xjc Task Attributes

2.2.1. Environment Variables

  • ANT_OPTS - command-line arguments that should be passed to the JVM. For example, you can define system properties or set the maximum Java heap size here.

2.2.2. Parameter Attributes

xjc supports the following parameter attributes.

AttributeDescriptionRequired

schema

A schema file to be compiled. A file name (can be relative to the build script base directory), or an URL.

This or nested < schema> elements are required.

binding

An external binding file that will be applied to the schema file.

No

package

If specified, generated code will be placed under this Java package. This option is equivalent to the "-p" command-line switch.

No

destdir

Generated code will be written under this directory. If you specify destdir="abc/def" and package="org.acme", then files are generated to abc/def/org/acme.

Yes

encoding

Set the encoding name for generated sources, such as EUC-JP or UTF-8. If it is not specified, the platform default encoding is used.

No

readonly

Generate Java source files in the read-only mode if true is specified. false by default.

No

header

Generate a header in each generated file indicating that this file is generated by such and such version of JAXB RI when. true by default.

No

extension

If set to true, the XJC binding compiler will run in the extension mode. Otherwise, it will run in the strict conformance mode. Equivalent of the " -extension" command line switch. The default is false.

No

catalog

Specify the catalog file to resolve external entity references. Support TR9401, XCatalog, and OASIS XML Catalog format. See the catalog-resolver sample for details.

No

removeOldOutput

Used in pair with nested <produces> elements. When this attribute is specified as " yes", the files pointed to by the <produces> elements will be all deleted before the XJC binding compiler recompiles the source files. See the up-to-date check section for details.

No

target

Specifies the runtime environment in which the generated code is supposed to run. Expects 2.0 or 2.1 values. This allows more up-to-date versions of XJC to be used for developing applications that run on earlier JAXB versions.

No, defaults to "2.2"

language

Specifies the schema language to compile. Supported values are "WSDL", "XMLSCHEMA", and "WSDL." Case insensitive.

No, defaults to "XMLSCHEMA"

2.2.3. Nested Elements

xjc supports the following nested element parameters.

2.2.3.1. schema

To compile more than one schema at the same time, use a nested <schema> element, which has the same syntax as <fileset> .

2.2.3.2. binding

To specify more than one external binding file at the same time, use a nested <binding> element, which has the same syntax as <fileset> .

2.2.3.3. classpath

To specify locations of the user-defined classes necessary during the compilation (such as an user-defined type that is used through a <javaType> customization), use nested <classpath> elements. For the syntax, see "path-like structure" .

2.2.3.4. arg

Additional command line arguments passed to the XJC. For details about the syntax, see the relevant section in the Ant manual. This nested element can be used to specify various options not natively supported in the xjc Ant task. For example, currently there is no native support for the following xjc command-line options:

  • -nv

  • -use-runtime

  • -schema

  • -dtd

  • -relaxng

  • -Xlocator

  • -Xsync-methods

To use any of these features from the <xjc> Ant task, you must specify the appropriate nested < arg> elements.

2.2.3.5. depends

Files specified with this nested element will be taken into account when the XJC task does the up-to-date check. See the up-to-date check section for details. For the syntax, see <fileset> .

2.2.3.6. produces

Files specified with this nested element will be taken into account when the XJC task does the up-to-date check. See the up-to-date check section for details. For the syntax, see <fileset> .

2.2.3.7. xmlcatalog

The xmlcatalog element is used to resolve entities when parsing schema documents.

2.3. Generated Resource Files

Please see the Generated Resource Files for more detail.

2.4. Up-to-date Check

By default, the XJC binding compiler always compiles the inputs. However, with a little additional setting, it can compare timestamps of the input files and output files and skip compilation if the files are up-to-date.

Ideally, the program should be able to find out all the inputs and outputs and compare their timestamps, but this is difficult and time-consuming. So you have to tell the task input files and output files manually by using nested <depends> and <produces> elements. Basically, the XJC binding compiler compares the timestamps specified by the <depends> elements against those of the <produces> set. If any one of the "depends" file has a more recent timestamp than some of the files in the "produces" set, it will compile the inputs. Otherwise it will skip the compilation.

This will allow you to say, for example "if any of the .xsd files in this directory are newer than the .java files in that directory, recompile the schema".

Files specified as the schema files and binding files are automatically added to the "depends" set as well, but if those schemas are including/importing other schemas, you have to use a nested <depends> elements. No files are added to the <produces> set, so you have to add all of them manually.

A change in a schema or an external binding file often results in a Java file that stops being generated. To avoid such an "orphan" file, it is often desirable to isolate all the generated code into a particular package and delete it before compiling a schema. This can be done by using the removeOldOutput attribute. This option allows you to remove all the files that match the "produces" filesets before a compilation. Be careful when you use this option so that you don't delete important files.

2.5. Schema Language Support

This release of the JAXB RI includes experimental support for RELAX NG, DTD, and WSDL. To compile anything other than W3C XML Schema from the xjc Ant task, you must use the nested < arg> element to specify the appropriate command line switch, such as -dtd, -relaxng, or -wsdl. Otherwise, your input schemas will be treated as W3C XML Schema and the binding compiler will fail.

2.6. xjc Examples

Compile myschema.xsd and place the generated files under src/org/acme/foo:

<xjc schema="src/myschema.xsd" destdir="src" package="org.acme.foo"/>

Compile all XML Schema files in the src directory and place the generated files under the appropriate packages in the src directory:

<xjc destdir="src">
    <schema dir="src" includes="*.xsd"/>
</xjc>

Compile all XML Schema files in the src directory together with binding files in the same directory and places the generated files under the appropriate packages in the src directory. This example assumes that binding files contain package customizations. This example doesn't search subdirectories of the src directory to look for schema files.

<xjc destdir="src">
    <schema dir="src" includes="*.xsd"/>
    <binding dir="src" includes="*.xjb"/>
</xjc>

Compile abc.xsd with an up-to-date check. Compilation only happens when abc.xsd is newer than any of the files in the src/org/acme/foo directory (and its impl subdirectory). Files in these two directories will be wiped away before a compilation, so don't add your own code in those directories. Note that the additional mkdir task is necessary because Ant's fileset requires the directory specified by the dir attribute to exist.

<mkdir dir="src/org/acme/foo"/>
<xjc destdir="src" schema="abc.xsd" removeOldOutput="yes"
     package="org.acme.foo">
    <produces dir="src/org/acme/foo" includes="* impl/*"/>
</xjc>

More complicated example of up-to-date check. In this example, we assume that you have a large set of schema documents that reference each other, with DTDs that describe the schema documents. An explicit <depends> is necessary so that when you update one of the DTDs, XJC will recompile your schema. But <depends> don't have to re-specify all the schema files, because you've already done that via <schema>.

<mkdir dir="src/org/acme/foo"/>
<xjc destdir="src" removeOldOutput="yes"
     package="org.acme.foo">
    <schema dir="schema" includes="*.xsd"/>
    <depends dir="schema" includes="*.dtd"/>
    <produces dir="build/generated-src/org/acme/foo"
              includes="**/*"/>
</xjc>

Compile all XML Schema files in the src directory and subdirectories, excluding files named debug.xsd, and place the generated files under the appropriate packages in the src directory. This example also specifies the -nv option, which disables the strict schema correctness checking:

<xjc destdir="src">
    <schema dir="src" includes="**/*.xsd"
            excludes="**/debug.xsd"/>
    <arg value="-nv"/>
</xjc>

If you depend on a proxy server to resolve the location of imported or included schemas (as you might if you're behind a firewall), you need to make the hostname and port number accessible to the JVM hosting ant. Do this by setting the environment variable ANT_OPTS to a string containing the appropriate java options. For example, from DOS:

> set ANT_OPTS=-Dhttp.proxyHost=webcache.east
> set ANT_OPTS=%ANT_OPTS% -Dhttp.proxyPort=8080
> ant

3. SchemaGen

3.1. schemagen Overview

The current schema generator can process either Java source files or class files.

We also provide an Ant task to run the schema generator - see the instructions for SchemaGen Ant Task.

3.2. Launching schemagen

The schema generator can be launched using the appropriate schemagen shell script in the bin directory for your platform.

If your java sources/classes reference other classes, they must be accessable on your system CLASSPATH environment variable, or they need to be given to the tool by using the -classpath/ -cp options. Otherwise you will see errors when generating your schema.

  • Solaris/Linux

    % path/to/jaxb/bin/schemagen.sh Foo.java Bar.java ...
    Note: Writing schema1.xsd
  • Windows

    > path\to\jaxb\bin\schemagen.bat Foo.java Bar.java ...
    Note: Writing schema1.xsd

3.3. schemagen Syntax

schemagen [OPTION]... <java files>

Usage: schemagen [-options ...] <java files> 

Options: 
    -d <path>         :  Specify where to place processor and javac generated class files 
    -cp <path>        :  Specify where to find user specified files  
    -classpath <path> :  Specify where to find user specified files  
    -episode <file>   :  generate episode file for separate compilation
    -version          :  display version information
    -help             :  Display this usage message

3.3.1. Summary of Command Line Options

-episode

Generates the "episode file", which is really just a JAXB customization file (but with vendor extensions specific to the JAXB RI, as of 2.1.) When people develop additional schemas that depend on what this schemagen invocation produces, they can use this episode file to have their generated code refer to your classes.

3.4. Generated Resource Files

The current schema generator simply creates a schema file for each namespace referenced in your Java classes. There is no way to control the name of the generated schema files at this time. Use SchemaGen Ant Task for that purpose.

4. SchemaGen Ant Task

4.1. schemagen Task Overview

The jaxb-xjc.jar file contains the SchemaGenTask.class file, which allows the schema generator to be invoked from the Ant build tool. To use SchemaGenTask, include the following statement in your build.xml file:

<taskdef name="schemagen"
         classname="com.sun.tools.jxc.SchemaGenTask">
    <classpath>
        <fileset dir="path/to/jaxb/lib" includes="*.jar"/>
    </classpath>
</taskdef>

This maps SchemaGenTask to an Ant task named schemagen. For detailed examples of using this task, refer to the build.xml files used by the java to schema Sample Apps.

4.2. schemagen Task Attributes

4.2.1. Environment Variables

  • ANT_OPTS - command-line arguments that should be passed to the JVM. For example, you can define system properties or set the maximum Java heap size here.

4.2.2. Parameter Attributes

schemagen supports most of the attributes defined by the javac task, plus the following parameter attributes.

AttributeDescriptionRequired

destdir

Base directory to place the generated schema files

No

classpath

Works just like the nested <classpath> element

No

episode

If specified, generate an episode file in the specified name. For more about the episode file, see -episode.

No

4.2.3. Nested Elements

xjc supports all the nested elements defined by the javac task, the following nested element parameters.

4.2.3.1. schema

Control the file name of the generated schema. This element takes a mandatory namespace attribute and a mandaotry file attribute. When this element is present, the schema document generated for the specified namespace will be placed in the specified file name.

The file name is interpreted as relative to the destdir attribute. In the absence of the destdir attribute, file names are relative to the project base directory. This element can be specified multiple times.

4.2.3.2. classpath

A path-like structure that represents the classpath. If your Java sources/classes depend on other libraries, they need to be available in the classpath.

4.3. schemagen Examples

Generate schema files from source files in the src dir and place them in the build/schemas directory.

<schemagen srcdir="src" destdir="build/schemas">

Compile a portion of the source tree.

<schemagen destdir="build/schemas">
    <src path="src"/>
    <exclude name="Main.java"/>
</schemagen>

Set schema file names.

<schemagen srcdir="src" destdir="build/schemas">
    <schema namespace="http://myschema.acme.org/common"
            file="myschema-common.xsd"/>
    <schema namespace="http://myschema.acme.org/onion"
            file="myschema-onion.xsd"/>
</schemagen>

5. 3rd Party Tools

5.1. Maven JAXB Plugin

The Maven2 plugin for JAXB2 is hosted at java.net. If you are using Maven, JAXB jars are available in the java.net maven repository. This repository is synced with maven central, so usually you find all JAXB releases in maven central after ~ 1 day after java.net release.

5.2. JAXB Workshop

JAXB Workshop hosts a collection of visual tools and IDE plugins useful for working with JAXB.

5.3. XJC Plugins

Various people in the community have developed plugins for XJC that you can use today. These plugins allow you to enhance/alter the code generation of XJC in many different ways. See the list of plugins (if you are interested in hosting your plugin here, let us know.)

5.4. RDBMS Persistence

Lexi has developed the HyperJAXB3 project for RDBMS persistence support for the JAXB RI.

JAXB RI Extensions

1. Overview

This page contains information about vendor-specific features provided by the JAXB RI.

Runtime Properties

This document describes JAXB RI specific properties that affect the way that the JAXB runtime library behaves.

XJC Customizations

This document describes additional binding customizations that can be used to control the generated source code.

DTD

This document describes the JAXB RI's experimental support for W3C XML Schema features not currently described in the JAXB Specification as well as support for other schema languages (RELAX NG and DTD).

2. Runtime Properties

2.1. Marshaller Properties

The JAXB RI provides additional Marshaller properties that are not defined by the JAXB specification. These properties allow you to better control the marshalling process, but they only work with the JAXB RI; they may not work with other JAXB providers.

2.1.2. Namespace Prefix Mapping

Property name:com.sun.xml.bind.namespacePrefixMapper
Type:com.sun.xml.bind.marshaller.NamespacePrefixMapper
Default value:

null

The JAXB RI provides a mechanism for users to control declarations of namespace URIs and what prefixes they will be bound to. This is the general procedure:

  1. The application developer provides an implementation of com.sun.xml.bind.marshaller.NamespacePrefixMapper.

  2. This class is then set on the marshaller via the RI specific property com.sun.xml.bind.namespacePrefixMapper.

  3. Each time the marshaller sees a URI, it performs a callback on the mapper: "What prefix do you want for this namespace URI?"

  4. If the mapper returns something, the marshaller will try to use it.

The com.sun.xml.bind.marshaller.NamespacePrefixMapper class has the following method that you need to implement:

/**
 * Implemented by the user application to determine URI -> prefix
 * mapping.
 * 
 * This is considered as an interface, though it's implemented
 * as an abstract class to make it easy to add new methods in
 * a future. 
 * 
 * @author
 *     Kohsuke Kawaguchi (kohsuke.kawaguchi@sun.com)
 */
public abstract class NamespacePrefixMapper {

    private static final String[] EMPTY_STRING = new String[0];

    /**
     * Returns a preferred prefix for the given namespace URI.
     * 
     * This method is intended to be overrided by a derived class.
     *
     * <p>
     * As noted in the return value portion of the javadoc, there
     * are several cases where the preference cannot be honored.
     * Specifically, as of JAXB RI 2.0 and onward:
     *
     * <ol>
     * <li>
     * If the prefix returned is already in use as one of the in-scope
     * namespace bindings. This is partly necessary for correctness
     * (so that we don't unexpectedly change the meaning of QNames
     * bound to {@link String}), partly to simplify the marshaller.
     * <li>
     * If the prefix returned is "" yet the current {@link JAXBContext}
     * includes classes that use the empty namespace URI. This allows
     * the JAXB RI to reserve the "" prefix for the empty namespace URI,
     * which is the only possible prefix for the URI.
     * This restriction is also to simplify the marshaller.
     * </ol>
     *
     * @param namespaceUri
     *      The namespace URI for which the prefix needs to be found.
     *      Never be null. "" is used to denote the default namespace.
     * @param suggestion
     *      When the content tree has a suggestion for the prefix
     *      to the given namespaceUri, that suggestion is passed as a
     *      parameter. Typicall this value comes from the QName.getPrefix
     *      to show the preference of the content tree. This parameter
     *      may be null, and this parameter may represent an already
     *      occupied prefix. 
     * @param requirePrefix
     *      If this method is expected to return non-empty prefix.
     *      When this flag is true, it means that the given namespace URI
     *      cannot be set as the default namespace.
     * 
     * @return
     *      null if there's no prefered prefix for the namespace URI.
     *      In this case, the system will generate a prefix for you.
     * 
     *      Otherwise the system will try to use the returned prefix,
     *      but generally there's no guarantee if the prefix will be
     *      actually used or not.
     * 
     *      return "" to map this namespace URI to the default namespace.
     *      Again, there's no guarantee that this preference will be
     *      honored.
     * 
     *      If this method returns "" when requirePrefix=true, the return
     *      value will be ignored and the system will generate one.
     * 
     * @since JAXB 1.0.1
     */
    public abstract String getPreferredPrefix(String namespaceUri, String suggestion, boolean requirePrefix);

    /**
     * Returns a list of namespace URIs that should be declared
     * at the root element.
     *
     * <p>
     * By default, the JAXB RI 1.0.x produces namespace declarations only when
     * they are necessary, only at where they are used. Because of this
     * lack of look-ahead, sometimes the marshaller produces a lot of
     * namespace declarations that look redundant to human eyes. For example,
     * <pre><xmp>
     * <?xml version="1.0"?>
     * <root>
     *   <ns1:child xmlns:ns1="urn:foo"> ... </ns1:child>
     *   <ns2:child xmlns:ns2="urn:foo"> ... </ns2:child>
     *   <ns3:child xmlns:ns3="urn:foo"> ... </ns3:child>
     *   ...
     * </root>
     * </xmp></pre>
     *
     * <p>
     * The JAXB RI 2.x mostly doesn't exhibit this behavior any more,
     * as it declares all statically known namespace URIs (those URIs
     * that are used as element/attribute names in JAXB annotations),
     * but it may still declare additional namespaces in the middle of
     * a document, for example when (i) a QName as an attribute/element value
     * requires a new namespace URI, or (ii) DOM nodes as a portion of an object
     * tree requires a new namespace URI.
     *
     * <p>
     * If you know in advance that you are going to use a certain set of
     * namespace URIs, you can override this method and have the marshaller
     * declare those namespace URIs at the root element.
     *
     * <p>
     * For example, by returning <code>new String[]{"urn:foo"}</code>,
     * the marshaller will produce:
     * <pre><xmp>
     * <?xml version="1.0"?>
     * <root xmlns:ns1="urn:foo">
     *   <ns1:child> ... </ns1:child>
     *   <ns1:child> ... </ns1:child>
     *   <ns1:child> ... </ns1:child>
     *   ...
     * </root>
     * </xmp></pre>
     * <p>
     * To control prefixes assigned to those namespace URIs, use the
     * {@link #getPreferredPrefix(String, String, boolean)} method. 
     * 
     * @return
     *      A list of namespace URIs as an array of {@link String}s.
     *      This method can return a length-zero array but not null.
     *      None of the array component can be null. To represent
     *      the empty namespace, use the empty string <code>""</code>.
     * 
     * @since
     *      JAXB RI 1.0.2 
     */
    public String[] getPreDeclaredNamespaceUris() {
        return EMPTY_STRING;
    }

    /**
     * Similar to {@link #getPreDeclaredNamespaceUris()} but allows the
     * (prefix,nsUri) pairs to be returned.
     *
     * <p>
     * With {@link #getPreDeclaredNamespaceUris()}, applications who wish to control
     * the prefixes as well as the namespaces needed to implement both
     * {@link #getPreDeclaredNamespaceUris()} and {@link #getPreferredPrefix(String, String, boolean)}.
     *
     * <p>
     * This version eliminates the needs by returning an array of pairs.
     *
     * @return
     *      always return a non-null (but possibly empty) array. The array stores
     *      data like (prefix1,nsUri1,prefix2,nsUri2,...) Use an empty string to represent
     *      the empty namespace URI and the default prefix. Null is not allowed as a value
     *      in the array.
     *
     * @since
     *      JAXB RI 2.0 beta
     */
    public String[] getPreDeclaredNamespaceUris2() {
        return EMPTY_STRING;
    }

    /**
     * Returns a list of (prefix,namespace URI) pairs that represents
     * namespace bindings available on ancestor elements (that need not be repeated
     * by the JAXB RI.)
     *
     * <p>
     * Sometimes JAXB is used to marshal an XML document, which will be
     * used as a subtree of a bigger document. When this happens, it's nice
     * for a JAXB marshaller to be able to use in-scope namespace bindings
     * of the larger document and avoid declaring redundant namespace URIs.
     *
     * <p>
     * This is automatically done when you are marshalling to {@link XMLStreamWriter},
     * {@link XMLEventWriter}, {@link DOMResult}, or {@link Node}, because
     * those output format allows us to inspect what's currently available
     * as in-scope namespace binding. However, with other output format,
     * such as {@link OutputStream}, the JAXB RI cannot do this automatically.
     * That's when this method comes into play.
     *
     * <p>
     * Namespace bindings returned by this method will be used by the JAXB RI,
     * but will not be re-declared. They are assumed to be available when you insert
     * this subtree into a bigger document.
     *
     * <p>
     * It is <b>NOT</b> OK to return  the same binding, or give
     * the receiver a conflicting binding information.
     * It's a responsibility of the caller to make sure that this doesn't happen
     * even if the ancestor elements look like:
     * <pre><xmp>
     *   <foo:abc xmlns:foo="abc">
     *     <foo:abc xmlns:foo="def">
     *       <foo:abc xmlns:foo="abc">
     *         ... JAXB marshalling into here.
     *       </foo:abc>
     *     </foo:abc>
     *   </foo:abc>
     * </xmp></pre>
     *
     * @return
     *      always return a non-null (but possibly empty) array. The array stores
     *      data like (prefix1,nsUri1,prefix2,nsUri2,...) Use an empty string to represent
     *      the empty namespace URI and the default prefix. Null is not allowed as a value
     *      in the array.
     *
     * @since JAXB RI 2.0 beta
     */
    public String[] getContextualNamespaceDecls() {
        return EMPTY_STRING;
    }
}

See the Sample Apps sample application for a detailed example.

2.1.3. Indentation

Property name:com.sun.xml.bind.indentString
Type:java.lang.String
Default value:

" " (four whitespaces)

This property controls the string used for the indentation of XML. An element of depth k will be indented by printing this string k times. Note that the "jaxb.formatted.output" property needs to be set to "true" for the formatting/indentation of the output to occur. See the API documentation for javax.xml.bind.Marshaller interface for details of this property.

2.1.4. Character Escaping Control

Property name:com.sun.xml.bind.characterEscapeHandler
Type:com.sun.xml.bind.marshaller.CharacterEscapeHandler
Default value:

null

By default, the marshaller implementation of the JAXB RI tries to escape characters so they can be safely represented in the output encoding (by using Unicode numeric character references of the form &#dddd;)

Unfortunately, due to various technical reasons, the default behavior may not meet your expectations. If you need to handle escaping more adroitly than the default manner, you can do so by doing the following:

  1. Write a class that implements the com.sun.xml.bind.marshaller.CharacterEscapeHandler interface.

  2. Create a new instance of it.

  3. Set that instance to the Marshaller by using this property.

See the Sample Apps sample application for more details.

2.1.5. XML Declaration Control

Property name:com.sun.xml.bind.xmlDeclaration
Type:boolean
Default value:

true

This experimental JAXB RI 1.0.x property has been adopted as a standard in JAXB 2.0. The 2.0 RI will continue to support this property, but client code should be using the Marshaller.JAXB_FRAGMENT property instead. Please refer to the Marshaller javadoc for a complete description of the behavior.

In JAXB 2.0, calling:

marshaller.setProperty("com.sun.xml.bind.xmlDeclaration", true);

is equivalent to calling:

marshaller.setProperty(Marshaller.JAXB_FRAGMENT, true);

JAXB 1.0 generated code and clients will continue to work exactly the same on the JAXB 2.0 runtime as they did on the JAXB 1.0 runtime.

Enabling fragment marshalling could be useful if you are inserting the output of the XML into another XML.

2.1.6. XML Preamble Control

Property name:com.sun.xml.bind.xmlHeaders
Type:java.lang.String
Default value:

null

This property allows you to specify an XML preamble (<?xml ...> declaration) and any additional PIs, comments, DOCTYPE declaration that follows it. This property takes effect only when you are marshalling to OutputStream, Writer, or StreamResult. Note that this property interacts with the Marshaller.JAXB_FRAGMENT property. If that property is untouched or set to false, then JAXB would always write its XML preamble, so this property can be only used to write PIs, comments, DOCTYPE, etc. On the other hand, if it is set to true, then JAXB will not write its own XML preamble, so this property may contain custom XML preamble.

2.1.7. Jaxb Annotation Control

Property name:com.sun.xml.bind.XmlAccessorFactory
Type:boolean
Default value:

false

This property provides support for a custom com.sun.xml.bind.v2.runtime.reflect.Accessor implementation.  It allows the user to control the access to class fields and properties.

In JAXB 2.1, set the property to enable:

marshaller.setProperty("com.sun.xml.bind.XmlAccessorFactory", true);

3. XJC Customizations

3.1. Customizations

The JAXB RI provides additional customizations that are not defined by the JAXB specification. Note the following:

  • These features may only be used when the JAXB XJC binding compiler is run in the -extension mode.

  • All of the JAXB RI vendor extensions are defined in the "http://java.sun.com/xml/ns/jaxb/xjc" namespace.

  • The namespaces containing extension binding declarations are specified to a JAXB processor by the occurrence of the global attribute @jaxb:extensionBindingPrefixes within an instance of <xs:schema> element. The value of this attribute is a whitespace-separated list of namespace prefixes. For more information, please refer to section 6.1.1 of the JAXB Specification.

3.1.1. Index of Customizations

3.1.2. SCD Support

The JAXB RI supports the use of schema component designator as a means of specifying the customization target (of all standard JAXB customizations as well as vendor extensions explained below.) To use this feature, use the scd attribute on <bindings> element instead of the schemaLocation and node attributes.

<bindings xmlns:tns="http://example.com/myns"
          xmlns="http://java.sun.com/xml/ns/jaxb" version="2.1">
    <bindings
            ...
            scd="tns:foo">
        <!-- this customization applies to the global element declaration -->
        <!-- 'foo' in the http://example.com/myns namespace -->
        <class name="FooElement"/>
    </bindings>
    <bindings
            ...
            scd="~tns:bar">
        <!-- this customization applies to the global type declaration -->
        <!-- 'bar' in the http://example.com/myns namespace -->
        <class name="BarType"/>
    </bindings>
</bindings>  

Compared to the standard XPath based approach, SCD allows more robust and concise way of identifying a target of a customization. For more about SCD, refer to the scd example. Note that SCD is a W3C working draft, and may change in the future.

3.1.3. Extending a Common Super Class

The <xjc:superClass> customization allows you to specify the fully qualified name of the Java class that is to be used as the super class of all the generated implementation classes. The <xjc:superClass> customization can only occur within your <jaxb:globalBindings> customization on the <xs:schema> element:

<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema"
           xmlns:jaxb="http://java.sun.com/xml/ns/jaxb"
           xmlns:xjc="http://java.sun.com/xml/ns/jaxb/xjc"
           jaxb:extensionBindingPrefixes="xjc"
           jaxb:version="2.0">

    <xs:annotation>
        <xs:appinfo>
            <jaxb:globalBindings>
                <xjc:superClass
                        name="org.acme.RocketBooster"/>
            </jaxb:globalBindings>
        </xs:appinfo>
    </xs:annotation>
    
    ...
    
</xs:schema>

In the sample above, the <xjc:superClass> customization will cause all of the generated implementation classes to extend the named class, org.acme.RocketBooster.

3.1.4. Extending a Common Super Interface

The <xjc:superInterface> customization allows you to specify the fully qualified name of the Java interface that is to be used as the root interface of all the generated interfaces. The <xjc:superInterface> customization can only occur within your <jaxb:globalBindings> customization on the <xs:schema> element:

<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema"
           xmlns:jaxb="http://java.sun.com/xml/ns/jaxb"
           xmlns:xjc="http://java.sun.com/xml/ns/jaxb/xjc"
           jaxb:extensionBindingPrefixes="xjc"
           jaxb:version="2.0">

    <xs:annotation>
        <xs:appinfo>
            <jaxb:globalBindings>
                <xjc:superInterface
                        name="org.acme.RocketBooster"/>
            </jaxb:globalBindings>
        </xs:appinfo>
    </xs:annotation>

    ...

</xs:schema>

In the sample above, the <xjc:superInterface> customization will cause all of the generated interfaces to extend the named interface, org.acme.RocketBooster.

3.1.5. Enhanced <jaxb:javaType>

The <xjc:javaType> customization can be used just like the standard <jaxb:javaType> customization, except that it allows you to specify an XmlAdapter-derived class, instead of parse&print method pair.

This customization can be used in all the places <jaxb:javaType> is used, but nowhere else:

<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema"
           xmlns:jaxb="http://java.sun.com/xml/ns/jaxb"
           xmlns:xjc="http://java.sun.com/xml/ns/jaxb/xjc"
           jaxb:extensionBindingPrefixes="xjc"
           jaxb:version="2.0">
    
    ...
    
    <xsd:simpleType name="LayerRate_T">
        <xsd:annotation>
            <xsd:appinfo>
                <xjc:javaType name="org.acme.foo.LayerRate"
                              adapter="org.acme.foo.LayerRateAdapter"/>
            </xsd:appinfo>
        </xsd:annotation>
        
        ... gory simple type definition here ...
        
    </xsd:simpleType>
</xsd:schema>

In the above example, LayerRate_T simple type is adapted by org.acme.foo.LayerRateAdapter, which extends from XmlAdapter.

<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema"
           xmlns:jaxb="http://java.sun.com/xml/ns/jaxb"
           xmlns:xjc="http://java.sun.com/xml/ns/jaxb/xjc"
           jaxb:extensionBindingPrefixes="xjc"
           jaxb:version="2.0">
    
    <xsd:annotation>
        <xsd:appinfo>
            <jaxb:globalBindings>
                <xjc:javaType name="org.acme.foo.MyDateType"
                              xmlType="xsd:dateTime"
                              adapter="org.acme.foo.MyAdapterImpl"/>
            </jaxb:globalBindings>
        </xsd:appinfo>
    </xsd:annotation>

    ...
    
</xsd:schema>

In the above example, all the use of xsd:dateTime type is adapter by org.acme.foo.MyAdapterImpl to org.acme.foo.MyDateType

3.1.6. Experimental simpler & better binding mode

This experimental binding mode can be enabled as a part of the global binding. See below:

<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema"
           xmlns:jaxb="http://java.sun.com/xml/ns/jaxb"
           xmlns:xjc="http://java.sun.com/xml/ns/jaxb/xjc"
           jaxb:extensionBindingPrefixes="xjc"
           jaxb:version="2.0">

    <xs:annotation>
        <xs:appinfo>
            <jaxb:globalBindings generateValueClass="false">
                <xjc:simple/>
            </jaxb:globalBindings>
        </xs:appinfo>
    </xs:annotation>

    ...

</xs:schema>

When enabled, XJC produces Java source code that are more concise and easier to use. Improvements include:

  1. Some content model definitions, such as A,B,A, which used to cause an XJC compilation error and required manual intervention, now compile out of the box without any customization.

  2. Some content model definitions that used to bind to a non-intuitive Java class now binds to a much better Java class:

    <!-- schema -->
    <xs:complexType name="foo">
        <xs:choice>
            <xs:sequence>
                <xs:element name="a" type="xs:int"/>
                <xs:element name="b" type="xs:int"/>
            </xs:sequence>
            <xs:sequence>
                <xs:element name="b" type="xs:int"/>
                <xs:element name="c" type="xs:int"/>
            </xs:sequence>
        </xs:choice>
    </xs:complexType>
    
    // before
    class Foo {
        List<JAXBElement<Integer>> content;
    }
    
    // in <xjc:simple> binding
    class Foo {
        Integer a;
        int b; // notice that b is effectively mandatory, hence primitive
        Integer c;
    }
  3. When repetable elements are bound, the method name will become plural.

    <!-- schema -->
    <xs:complexType name="person">
        <xs:sequence>
            <xs:element name="child" type="xs:string"
                        maxOccurs="unbounded"/>
            <xs:element name="parent" type="xs:string"
                        maxOccurs="unbounded"/>
        </xs:sequence>
    </xs:complexType>
    
    // before
    public class Person {
        protected List<String> child;
        protected List<String> parent;
    }
    
    // in <xjc:simple> binding
    public class Person {
        protected List<String> children;
        protected List<String> parents;
    }     

Once again, readers are warned that this is an experimental binding mode, and therefore the binding is subject to change in future versions of the JAXB RI without notice. Please send feedbacks on this binding to users@jaxb.java.net

3.1.7. Alternative Derivation-by-restriction Binding Mode

Normally, the JAXB specification requires that a derivation-by-restriction be mapped to an inheritance betwee n two Java classes. This is necessary to preserve the type hierarchy, but one of the downsides is that the derived class does not really provide easy-to-use properties that reflect the restricted content model.

This experimental <xjc:treatRestrictionLikeNewType> changes this behavior by not preserving the type inheritance to Java. Instead, it generates two unrelated Java classes, both with proper properties. For example, given the following schema:

<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema"
           xmlns:xjc="http://java.sun.com/xml/ns/jaxb/xjc"
           jaxb:extensionBindingPrefixes="xjc"
           xmlns:jaxb="http://java.sun.com/xml/ns/jaxb"
           jaxb:version="2.0"
           elementFormDefault="qualified">

    <xs:annotation>
        <xs:appinfo>
            <jaxb:globalBindings>
                <xjc:treatRestrictionLikeNewType/>
            </jaxb:globalBindings>
        </xs:appinfo>
    </xs:annotation>

    <xs:complexType name="DerivedType">
        <xs:complexContent>
            <xs:restriction base="ResponseOptionType">
                <xs:sequence>
                    <xs:element name="foo" type="xs:string"/>
                </xs:sequence>
            </xs:restriction>
        </xs:complexContent>
    </xs:complexType>

    <xs:complexType name="ResponseOptionType">
        <xs:sequence>
            <xs:element name="foo" type="xs:string"
                        maxOccurs="unbounded"/>
        </xs:sequence>
    </xs:complexType>

</xs:schema>

The generated Derived class will look like this (comment and annotations removed for brevity):

public class DerivedType {
    protected String foo;

    public String getFoo() { return foo; }
    public void setFoo(String value) { this.foo = value; }
} 

In contrast, without this customization the Derived class would look like the following:

public class DerivedType extends ResponseOptionType {

    // it simply inherits List<String> ResponseOptionType.getFoo()

}

3.1.8. Allow separate compilations to perform element substitutions

In an attempt to make the generated code easier to use, the JAXB specification sometimes choose bindings based on how certain feature is used. One of them is element substitution feature. If no actual element substitution happens in the schema, JAXB assumes that the element is not used for substitution, and generates code that assumes it.

Most of the time this is fine, but when you expect other "extension" schemas to be compiled later on top of your base schema, and if those extension schemas do element substitutions, this binding causes a problem ( see example.)

<xjc:substitutable> customization is a work around for this issue. It explicitly tells XJC that a certain element is used for element substitution head, even though no actual substitution might be present in the current compilation. This customization should be attached in the element declaration itself, like this:

<xs:element name="Model" type="Model">
    <xs:annotation>
        <xs:appinfo>
            <xjc:substitutable/>
        </xs:appinfo>
    </xs:annotation>
</xs:element>

4. DTD

4.1. DTD

The JAXB RI is shipped with experimental DTD support, which lets you compile XML DTDs.

To compile a DTD test.dtd, run the XJC binding compiler as follows:

$ xjc.sh -dtd test.dtd

All the other command-line options of the XJC binding compiler can be applied. Similarly, the xjc ant task supports DTD. The generated code will be no different from what is generated from W3C XML Schema. You'll use the same JAXB API to access the generated code, and it is portable in the sense that it will run on any JAXB 2.0 implementation.

4.1.1. Customization

The customization syntax for DTD is roughly based on the ver.0.21 working draft of the JAXB specification, which is available at xml.coverpages.org. The deviations from this document are:

  • The whitespace attribute of the conversion element takes " preserve", " replace", and " collapse" instead of " preserve"," normalize", and " collapse" as specified in the document.

  • The interface customization just generates marker interfaces with no method.

5. Develop Plugins

This document describes how to write an XJC plugin to extend the code generation of XJC.

5.1. What Can A Plugin Do?

An XJC plugin participates in the code generation from a schema. It can define its own customizations that users can use to control it, it can access the code that the JAXB RI generates, it can generate additional classes/methods/fields/annotations/comments, and it can also replace some of the pluggability points in the compilation process, such as XML name -> Java name conversion.

As a show case of what a plugin can do, take a look at plugins hosted at JAXB2-commons.

5.1.1. Quick Start

To write a plugin, do the following simple steps.

  1. Write a class, say, org.acme.MyPlugin by extending com.sun.tools.xjc.Plugin. See javadoc for how to implement methods.

  2. Write the name of your plugin class in a text file and put it as /META-INF/services/com.sun.tools.xjc.Plugin in your jar file.

Users can then use your plugins by declaring an XJC ant task with your jar files.

<taskdef name="xjc" classname="com.sun.tools.xjc.XJCTask">
    <classpath>
        <fileset dir="jaxb-ri/lib" includes="*.jar"/>
        <fileset dir="your-plugin" includes="*.jar"/>
    </classpath>
</taskdef>

5.1.2. Resources

See this page for more detailed, up-to-date information.

Although we will do our best to maintain the compatibility of the interfaces, it is still subject to change at this point.

Frequently Asked Questions

1. JAXB 2.0
Q: Which version of J2SE does JAXB 2.0 require?
Q: Can I run my existing JAXB 1.0.x applications on the JAXB 2.0 runtime?
Q: What if I want to port my JAXB 1.0.x application to JAXB 2.0?
Q: Are the JAXB runtime API's thread safe?
Q: Why can't I cast the unmarshalled object into the generated type.
Q: Which jar files do I need to distribute with my application that uses the JAXB RI?
Q: How can I cause the Marshaller to generate CDATA blocks?
Q: Can I access <xs:any/> as a DOM node?
Q: How do I find out which version of the JAXB RI I'm using?

1. JAXB 2.0

Q:

Which version of J2SE does JAXB 2.0 require?

A:

Java SE 6.0 or higher.

Q:

Can I run my existing JAXB 1.0.x applications on the JAXB 2.0 runtime?

A:

This is no longer supported. However, you should be able to deploy http://search.maven.org/remotecontent?filepath=com/sun/xml/bind/jaxb1-impl/2.2.5-1/jaxb1-impl-2.2.5-1.jar with your with your application application.

Q:

What if I want to port my JAXB 1.0.x application to JAXB 2.0?

A:

You need to recompile your schema with the newer JAXB 2.0 xjc and modify your application code to work with the new bindings.

Q:

Are the JAXB runtime API's thread safe?

A:

The JAXB Specification currently does not address the thread safety of any of the runtime classes. In the case of the Oracle JAXB RI, the JAXBContext class is thread safe, but the Marshaller, Unmarshaller, and Validator classes are not thread safe.

For example, suppose you have a multi-thread server application that processes incoming XML documents by JAXB. In this case, for the best performance you should have just one instance of JAXBContext in your whole application like this:

class MyServlet extends HttpServlet {
    static final JAXBContext context = initContext();

    private static JAXBContext initContext() {
        return JAXBContext.newInstance("....", MyServlet.class.getClassLoader());
    }
}

And each time you need to unmarshal/marshal/validate a document. Just create a new Unmarshaller/Marshaller/Validator from this context, like this:

public void doGet(HttpServletRequest req, HttpServletResponse resp) {
    Unmarshaller u = context.createUnmarshaller();
    u.unmarshal(...);
}

This is the simplest safe way to use the JAXB RI from multi-threaded applications.

If you really care about the performance, and/or your application is going to read a lot of small documents, then creating Unmarshaller could be relatively an expensive operation. In that case, consider pooling Unmarshaller objects. Different threads may reuse one Unmarshaller instance, as long as you don't use one instance from two threads at the same time.

Q:

Why can't I cast the unmarshalled object into the generated type.

A:

When you invoke JAXBContext.newInstance("aaa.bbb.ccc"), it tries to load classes and resources using the same classloader used to load the JAXBContext class itself. This classloader may be different from the classloader which was used to load your application (see the picture Parent/Child classloader). In this case, you'll see the above error. This problem is often seen with application servers, J2EE containers, Ant, JUnit, and other applications that use sophisticated class loading mechanisms.

Figure 1. Parent/Child classloader

Parent/Child classloader

With some applications, things get even more complicated when the JAXB-generated code can be loaded by either classloader. In this case, JAXBContext.newInstance("aaa.bbb.ccc") will work but the JVM ends up loading two copies of the generated classes for each class loader. As a result, unmarshalling works but an attempt to cast the returned object into the expected type will fail, even though its getClass().getName() returns the expected name.

The solution for both situations is to pass your curent class loader like this:

JAXBContext.newInstance("aaa.bbb.ccc", this.getClass().getClassLoader());

In general, if you are writing code that uses JAXB, it is always better to explicitly pass in a class loader, so that your code will work no matter where it is deployed.

Q:

Which jar files do I need to distribute with my application that uses the JAXB RI?

A:

For JAXB 2.2.x:

$JAXB_HOME/lib/jaxb-api.jar
$JAXB_HOME/lib/jaxb-impl.jar

Q:

How can I cause the Marshaller to generate CDATA blocks?

A:

This functionality is not available from JAXB directly, but you can configure an Apache Xerces-J XMLSerializer to produce CDATA blocks. Please review the JaxbCDATASample.java sample app for more detail.

Q:

Can I access <xs:any/> as a DOM node?

A:

In JAXB 2.0, <xs:any/> is handled correctly without any customization.

  1. If it's strict, it will map to Object or List<Object> and when you unmarshal documents, you'll get objects that map to elements (such as JAXBElements or classes that are annotated with XmlRootElement).

  2. If it's skip, it will map to org.w3c.dom.Element or List<Element> and when you unmarshal documents, you'll get DOM elements.

  3. If it's lax, it will map to the same as with strict, and when you unmarshal documents, you'll get either:

    1. JAXBElements

    2. classes that are annotated with XmlRootElement

    3. DOM elements

Q:

How do I find out which version of the JAXB RI I'm using?

A:

Run the following command

$ java -jar jaxb-xjc.jar -version

Alternatively, each JAXB jar has version information in its META-INF/MANIFEST.MF, such as this:

Manifest-Version: 1.0
Ant-Version: Apache Ant 1.8.2
Created-By: 1.6.0_29-b11 (Sun Microsystems Inc.)
Specification-Title: Java Architecture for XML Binding
Specification-Version: 2.2.6
Specification-Vendor: Oracle Corporation
Implementation-Title: JAXB Reference Implementation 
Implementation-Version: 2.2.5-SNAPSHOT
Implementation-Vendor: Oracle Corporation
Implementation-Vendor-Id: com.sun
Extension-Name: com.sun.xml.bind
Build-Id: 02/09/2012 01:42PM (hudson)
Class-Path: jaxb-api.jar
Terms of Use; Privacy Policy; Copyright ©2013-2014 (revision 20131025.e7cbc9d)
 
 
Close
loading
Please Confirm
Close