What type of marshalling does JAXB use?

[Luke Murphy]

Hi,
The marshalling for SAX, DOM and StAX are all well documented and it is easy to get a good understanding of them.
While I know JAXB has a nice clean API and cool annotations and makes marshalling very simple, I'd like to know how it marshalls under the hood.
My understanding is that it builds up a Tree in the same way DOM does but it is slightly more performant than DOM.
Is this correct?

Also, if it is building up a tree I presume it suffers from some of the drawbacks DOM does for example a large memory footprint. And it therefore wouldn't be as performant as StAX? Would this be correct?

Is there any way JAXB can be configured to be more performant? for example to use a Pull or PUSH model as SAX StAX does?

Many thanks.

[Karthik Shiraly]

Hi,

It can use whatever you want it to use - both Marshaller.marshal() and Unmarshaller.unmarshal() support SAX, DOM and StAX sources/readers and results.

In one of my apps, I used JAXB unmarshalling with StAX XMLEventReader and Unmarshaller.Listener.afterUnmarshal() to parse an incoming HTTP XML (of 10000+ users) as and when the tags were coming in, without having to wait for the entire HTTP response to arrive.

[Luke Murphy]

Karthik Shiraly wrote:Hi,

It can use whatever you want it to use - both Marshaller.marshal() and Unmarshaller.unmarshal() support SAX, DOM and StAX sources/readers and results.

In one of my apps, I used JAXB unmarshalling with StAX XMLEventReader and Unmarshaller.Listener.afterUnmarshal() to parse an incoming HTTP XML (of 10000+ users) as and when the tags were coming in, without having to wait for the entire HTTP response to arrive.

How do you configure it to use StAX, SAX or DOM then?

[Karthik Shiraly]

Hi,

It's described in the javadoc itself. I quote:

Marshalling to a javax.xml.transform.SAXResult: 

       // assume MyContentHandler instanceof ContentHandler
       SAXResult result = new SAXResult( new MyContentHandler() );

       m.marshal( element, result );
    
Marshalling to a javax.xml.transform.DOMResult: 

       DOMResult result = new DOMResult();
       
       m.marshal( element, result );
    
Marshalling to a javax.xml.transform.StreamResult: 

       StreamResult result = new StreamResult( System.out );
 
       m.marshal( element, result );
    
Marshalling to a javax.xml.stream.XMLStreamWriter: 

       XMLStreamWriter xmlStreamWriter = 
           XMLOutputFactory.newInstance().createXMLStreamWriter( ... );
 
       m.marshal( element, xmlStreamWriter );
    
Marshalling to a javax.xml.stream.XMLEventWriter: 

       XMLEventWriter xmlEventWriter = 
           XMLOutputFactory.newInstance().createXMLEventWriter( ... );
 
       m.marshal( element, xmlEventWriter );
    

Unmarshalling from a org.w3c.dom.Node: 

       JAXBContext jc = JAXBContext.newInstance( "com.acme.foo" );
       Unmarshaller u = jc.createUnmarshaller();
 
       DocumentBuilderFactory dbf = DocumentBuilderFactory.newInstance();
       dbf.setNamespaceAware(true);
       DocumentBuilder db = dbf.newDocumentBuilder();
       Document doc = db.parse(new File( "nosferatu.xml"));

       Object o = u.unmarshal( doc );
    
Unmarshalling from a javax.xml.transform.sax.SAXSource using a client specified validating SAX2.0 parser: 

       // configure a validating SAX2.0 parser (Xerces2)
       static final String JAXP_SCHEMA_LANGUAGE =
           "http://java.sun.com/xml/jaxp/properties/schemaLanguage";
       static final String JAXP_SCHEMA_LOCATION =
           "http://java.sun.com/xml/jaxp/properties/schemaSource";
       static final String W3C_XML_SCHEMA =
           "http://www.w3.org/2001/XMLSchema";

       System.setProperty( "javax.xml.parsers.SAXParserFactory",
                           "org.apache.xerces.jaxp.SAXParserFactoryImpl" );

       SAXParserFactory spf = SAXParserFactory.newInstance();
       spf.setNamespaceAware(true);
       spf.setValidating(true);
       SAXParser saxParser = spf.newSAXParser();
       
       try {
           saxParser.setProperty(JAXP_SCHEMA_LANGUAGE, W3C_XML_SCHEMA);
           saxParser.setProperty(JAXP_SCHEMA_LOCATION, "http://....");
       } catch (SAXNotRecognizedException x) {
           // exception handling omitted
       }

       XMLReader xmlReader = saxParser.getXMLReader();
       SAXSource source = 
           new SAXSource( xmlReader, new InputSource( "http://..." ) );

       // Setup JAXB to unmarshal
       JAXBContext jc = JAXBContext.newInstance( "com.acme.foo" );
       Unmarshaller u = jc.createUnmarshaller();
       ValidationEventCollector vec = new ValidationEventCollector();
       u.setEventHandler( vec );
       
       // turn off the JAXB provider's default validation mechanism to 
       // avoid duplicate validation
       u.setValidating( false )

       // unmarshal
       Object o = u.unmarshal( source );

       // check for events
       if( vec.hasEvents() ) {
          // iterate over events
       }
    
Unmarshalling from a StAX XMLStreamReader: 

       JAXBContext jc = JAXBContext.newInstance( "com.acme.foo" );
       Unmarshaller u = jc.createUnmarshaller();
 
       javax.xml.stream.XMLStreamReader xmlStreamReader = 
           javax.xml.stream.XMLInputFactory().newInstance().createXMLStreamReader( ... );
 
       Object o = u.unmarshal( xmlStreamReader );
    
Unmarshalling from a StAX XMLEventReader: 

       JAXBContext jc = JAXBContext.newInstance( "com.acme.foo" );
       Unmarshaller u = jc.createUnmarshaller();
 
       javax.xml.stream.XMLEventReader xmlEventReader = 
           javax.xml.stream.XMLInputFactory().newInstance().createXMLEventReader( ... );
 
       Object o = u.unmarshal( xmlEventReader );
    

Just call the overloaded version of marshal/unmarshal that suits the approach you want.

[Luke Murphy]

Good answer thanks.