JiBX offers the possibility to create separate input and output bindings. Just use the attribute direction in the root element of the binding file. To create a XML binding just for the marshalling objects, use <binding direction="output">. To create an XML binding just for unmarhshalling XML files, use <binding direction="input">.

With this possibility, you can define for example an input binding, which doesn’t map every element of a XML file. But in the output binding you can define some constants, which get’s mapped, if objects gets marshaled to XML files.

JiBX is one of the fastest XML object mapping frameworks out there (check out this performance test). Furthermore, it is very flexible in binding XML structures to objects. In this blog post I will describe, how a basic JiBX project can be set up with Maven2. The focus of this entry is on getting in touch with JiBX (and a little bit with Maven).
To use JiBX in our project, we just have to add the dependency to our pom.xml. JiBX is fast, but to get it’s high performance, it uses byte code enhancement. For this reason, the byte code need’s to be edited by JiBX after the compilation phase. This is done with a Maven JiBX plugin, which is executed in the build phase “process-classes”.

Besides, I also wanted a JAR file, which contains all dependencies and has a defined main class (to which I will come to later). The creation of the JAR file is done with the Maven Assembly.

Check out the pom.xml for this test run.

After setting up the pom.xml, we can begin to develop a mapping example with Eclipse. JiBX has two main parts on which we have to focus, binding and the runtime part.
In the binding part, JiBX enhances the byte code (based on a binding definition) to enable the mapping between Java objects and a XML files. In the runtim part, the JiBX runtime components are used to perform the marshal (objects to XML) or unmarshal (XML to objects) tasks.

As you have seen in the pom.xml, the JiBX binding files should be located in src/main/jibx. The names of the JiBX files should end with -binding.xml.

Our goal is to map an XML file, based on following schema CustomerSchema.xsd:

<?xml version="1.0" encoding="UTF-8"?>
<schema xmlns="http://www.w3.org/2001/XMLSchema"
  targetNamespace="http://stefanjaeger.ch/CustomerSchema"
  xmlns:sja="http://stefanjaeger.ch/CustomerSchema"
  elementFormDefault="qualified">
  <element name="customer">
    <complexType>
      <sequence>
        <element ref="sja:person" maxOccurs="1" />
        <element ref="sja:city" maxOccurs="1" />
      </sequence>
    </complexType>
  </element>
  <element name="person">
    <complexType>
      <sequence>
        <element ref="sja:first-name" maxOccurs="1" />
        <element ref="sja:last-name" maxOccurs="1" />
      </sequence>
    </complexType>
  </element>
  <element name="first-name" type="string" />
  <element name="last-name" type="string" />
  <element name="city" type="string" />
</schema>

Our XML Customer1.xml file looks like:

<?xml version="1.0" encoding="UTF-8"?>
<customer xmlns="http://stefanjaeger.ch/CustomerSchema"
  xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
  xsi:schemaLocation="http://stefanjaeger.ch/CustomerSchema CustomerSchema.xsd">
  <person>
    <first-name>Stefan</first-name>
    <last-name>Jaeger</last-name>
  </person>
  <city>Davos Dorf</city>
</customer>

We want to map the XML in the two objects Customer and Person (to simplify these lines of code, the getter, setter and toString methods are not described).

public class Customer {
  private Person person;
  private String city;
}
 
public class Person {
  private String firstName;
  private String lastName;
}

To map the XML file to these classes, we create following binding definition. Because we use a XML schema, we have to define the namespace. Also the fully qualified name of the mapped class Customer need’s to be used. Because the class Person is in relation with the Customer class, JiBX recognizes automatically the fully qualified name of Person.

<?xml version="1.0" encoding="UTF-8"?>
<binding>
  <mapping name="customer" class="ch.stefanjaeger.jibx.Customer">
  <namespace uri="http://stefanjaeger.ch/CustomerSchema" default="elements"/>
    <structure name="person" field="person">
      <value name="first-name" field="firstName" />
      <value name="last-name" field="lastName" />
    </structure>
    <value name="city" field="city" />
  </mapping>
</binding>

Now, we have a XML file, some Java classes and a binding between them. What we now need is an application, which does the unmarshalling of the XML file.

public class JiBXDemoApplication {
  public static void main(String[] args) throws Exception {
    IBindingFactory bfact = BindingDirectory.getFactory(Customer.class);
    IUnmarshallingContext uctx = bfact.createUnmarshallingContext();
    Customer customer = (Customer) uctx.unmarshalDocument(
      new FileInputStream("Customer1.xml"), null);
 
    System.out.println(customer.toString());
  }
}

We are finished now. Just build the project with the command mvn package and let’s see, how the test run is working. Remember to place the XML schema and the XML file to the same location as the JAR file (after mvn package, it should be located in /target/).

The command

java -jar jibx_playground-1.0-SNAPSHOT-jar-with-dependencies.jar

results in

ch.stefanjaeger.jibx.Customer@12d03f9
[person=ch.stefanjaeger.jibx.Person@15dfd77
 [firstName=Stefan,lastName=Jaeger]
,city=Davos Dorf]

I have to admit, that I did the same example with ANT. It tooks me a little bit longer… From today on, I’m a Maven fan

Download here the source code of this example. You only have to run mvn package to build the project.

How many times in your lifetime have you overwritten the functions equals(), hashCode() or toString()? I did many times…

The Apache Commons Lang Package offers three basic classes for building the equals(), hashCode() or toString() method.

1. EqualsBuilder
   JavaDoc of the EqualsBuilder
   All available fields can be append to the EqualsBuilder, which checks, if two objects are equals. It works with all primitive datatypes and objects (if they have implemented the equals method, of course).
    
2. HashCodeBuilder
   JavaDoc of the HashCodeBuilder
   If you overwrite the equals() method, you also have to overwrite the hashCode() method. This builder has almost the same methods like the EqualsBuilder, but you should define two randomly chosen odd numbers in the constructor. Ideally the numbers are unique over all classes in the program.
    
3. ToStringBuilder
   JavaDoc of the ToStringBuilder
   Appends all the desired members and prints something like Person@7f54[name=Stephen,age=29,smoker=false]

I have to admit that I have spend in recent years a lot of my free time into developing with PHP. For me, PHP offers a lot of nice features and it is a easy to write script language. Of course, it’s not Java and it lacks in some features (like type safety, maintainability, and so on).

But in PHP there were two functions, which I used a lot. Implode and Explode. These two functions offered a easy way to split or concatenate strings. Let me make two examples.

explode: splits the string “a;b;c;d” into the array (“a”,”b”,”c”,”d”)
implode: concatenate the elements of the array (“a”,”b”,”c”,”d”) into the string “a;b;c;d”

Of course, the delimiter (the ; in the example) can be chosen individually.

Now, I was searching for such functions in Java. Especially the implode method I missed a lot. But there is a open source framework, which holds the solution. It’s the Lang package of the Apache Commons project (http://commons.apache.org/lang/). Inside this framework, there is the StringUtils class, which offers some interesting string operations.

StringUtils.join(Collection collection, String separator) is the same as implode.
StringUtils.split(String str, String separatorChars) is the same as explode.

Of course, StringUtils offers some other nice features. Just try it out!

In team project, CruiseControl is often in use to maintain a continuous build process. It is a Java program and it was first designed for Java development process. But as a successfully project, there is also a .NET implementation, called CruiseControl.NET. With this .NET implementation, Windows users can now benefit from a tiny tool called CCTray.

CCTray monitors the build process and informs a Windows user, if a build is successfully or if it has failed. This is done with a tiny balloon tip:

To use CCTray, download it from the Sourceforge project page. After the installation you can add your CruiseControl instance. Because we use a Java CruiseControl instance, you have to define the XML path. The Java CruiseControl provieds a XML servlet, which is accessible from http:// cruisecontrol/xml. After adding your Server, you can choose the projects, which should be monitored.

CCTray helps you to maintain the build process and is a tiny, but very helpful tool.