Introduction to Service Design and Engineering | University of Trento | Webpage
To facilitate the lives of programmers around the world (at least for some), a long time ago we humans created object-oriented programming. But how do we go from objects to serialized representations of them in XML or JSON? And viceversa?. In this module, we will be mapping XML/JSON from and to Objects. What we covered up to here is also the base-ground we need before we start implementing actual services.
Links: PPT slides | PDF slides | Source code
For the purpose of mapping XML (and JSON) to and from Objects (i.e., serializing/deserializing), we will explore three main technologies: Java Annotations (wich will come handy for many other things), JAXB, and (briefly) Dozer.
- Java Annotations provide data about a program that is not part of the program itself.
- They have no direct effect on the operation of the code they annotate.
- Annotations can be applied to a program's declarations of classes, fields, methods, and other program elements.
- They uses include:
- Information for the compiler: to detect errors or suppress warnings.
- Compile-time and deployment-time processing: software tools can process annotation information to generate code, XML files, and so forth.
- Runtime processing: some annotations are available to be examined at runtime
- In this module, we will java annotations to specify how a class must be transformed from and to XML/JSON
- JAXB stands for Java Architecture for XML Binding. Is a Java standard that defines how Java objects are converted from and to XML.
- As opposed to XPATH, it allows us to map XML to Java Classes, allowing our Java program to operate only on plain old java objects (not a document tree)
- Using java annotations, JAXB libraries can
- marshal (i.e., convert) Java objects into XML
- un-marshal XML back into Java objects
- JAXB also includes a compiler that generates Java Classes from an XML Schema
-
To use JAXB, you will need to have it in your classpath (or at least in the build path of your project).
-
To solve this, we will use Apache Ivy, a lightweight dependency manager that easily integrates with ant scripts (whenever possible, we will use Ivy to get libraries we need)
-
Learning ivy is beyond the scope of the lab, but it is actually very simple. You will find a simple way to use it in this modules's "build.xml" file.
-
To use IVY, you need to (1) copy PART 1 of the build script (see comments in build.xml) to your future build scripts; and (2) have an ivy.xml file in your project home (same folder where the build.xml is) where you will specify the dependencies.
-
You can use the Maven Repository website to search for libraries and to get the proper ivy dependency declarations. Below is an example from this module's ivy.xml, which will bring JAXB API and XJC compiler to this module's example
<dependency org="javax.xml.bind" name="jaxb-api" rev="2.2.11"/> <dependency org="com.sun.xml.bind" name="jaxb-xjc" rev="2.2.7"/>
-
Optionally:
- If you want to have ivy already installed in your computer (so that you don't need PART 1 of this session build.xml in your future scripts) you can download it from here, unpack it wherever you want and then copy the ivy jar file into your ant lib directory, ANT_HOME/lib.
- Moreover, in the installation guide for Eclipse we have included the installation of Apache IvyDE plugin, which uses its own ivy installation to manage dependencies in eclipse projects (we will see this later when we create a project for this module)
- Similarly, if you want to have JAXB in your system (including xjc binary), you can download it from here, unpack the it somewhere and then add its "bin" folder to your system PATH.
-
Open your terminal window in your local lab04 folder and run the following:
ant compile ant execute.HPWriter
-
You will see that before executing the "compile" target, ivy.jar will be downloaded into an "ivy" folder, and libraries that are specified in "ivy.xml" will be downloaded to "lib"
-
Moreover, the target generate would have created the folder Examples/src/bookstore/generated with four classes in it (we will get back to this later)
-
And finally, as a result from execute.HPWriter, there is a new people.xml in your lab04 folder.
-
So, what has happened here?
- Open Eclipse and create a project with the location at the lab04 folder
- Open and explore the new HealthProfileReader example to get acquainted with JAXB annotations
- Explore Example/src/model/Person.java
- Explore Example/src/model/HealthProfile.java
- Explore Example/src/dao/PeopleStore.java (dao stands for "data access object", a typical data accessing pattern)
- Explore and run Example/src/HealthProfileWriter.java
- Exercise 04.01: let's try writing ourselves the HealthProfileReader that will use the "people.xml" generated by the writer.
-
Where do those generated classes we saw earlier come from?
-
JAXB comes with an XML Schema binding compilation tool (which was invoked by the target generate in our build script by using a "taskdef" definition).
-
Explore the classes under the newly created "generated folder". They are the result of generating classes based on the XML schema defined in catalog.xsd
-
Now, Marshal these classes into an XML
ant execute.JAXBUnMarshaller
-
Explore catalog.xml
-
UnMarshal them into Java objects
ant execute.JAXBUnMarshaller
-
Exercise 04.02: What should you change in in the bookstore example to add an element year within each article of a journal?.
-
You have a domain model (i.e., the model that is mapped to your database)
public class PersonDB { private String firstName; private String lastName; private String address; private String dbID;
-
What if your client is waiting this, and it does not care about ids?
<person> <fName>Cristhian</fName> <lName>Parra</lName> <address>Povo Trento</address> </person>
-
In other words, we want to keep domain objects separated from the logic that manages the transformation into XML/JSON representations of the resources in our model
-
A design pattern to achieve this is to use Transfer Objects in the middle (i.e., the objects that are actually exchanged between client and server)
-
Dozer is a Java Bean to Java Bean mapper that recursively copies data from one object to another
-
It supports mapping between attribute names and between types.
-
Standard conversions are provided automatically
-
You are allowed to specify custom conversions via XML
-
With Dozer, your internal domain objects are not exposed to external presentation layers or to external consumers.
-
Dozer maps your domain objects to external APIs calls and vice-versa.
-
Dozer can work both with XML and JSON
-
You can install it in your project via ivy
<!-- Ivy dependency declaration for Dozer --> <dependency org="net.sf.dozer" name="dozer" rev="5.5.1"/>
-
If you do not use ivy, you can download it from github, unpack it and add ${dozer.home}/dist/dozer.jar to your classpath.
-
You will also need to add required thirdparty runtime jars to your classpath
-
Chek the package dozerproject in the examples of this module. Exlore the files:
- dozerproject.entity.Person.java
- dozerproject.transfer.PersonBean.java
- dozerproject.delegate.PersonBeanDelegate.java
-
The basic idea is that you have create:
- entity classes to represent your domain model (i.e., your data model)
- transfer classes to manage what and how the domain model is going to be mapped to the presentation layer (xml, json)
-
Additionally, we can create delegate classes to manage the mapping operation (i.e., instantiating the dozer mapper and setting up the operation).
-
Dozer maps domain objects (entities) into (transfer objects) that can be implemented with jaxb annotations, facilitating their serialization
-
The marshalling is then separated from the management of the data objects (and therefore, JAXB annotations are not mixed with the data, but only with presentation objects)
-
Mapping from entities to transfer objects can be defined via an xml mapping file (e.g., Example/dozerMappings.xml) or via its more experimental version with java annotations in the transfer classes.
-
A third mapping option is by using dozer API programmatically (we will not cover this, but you can check it out here)
-
**Exercise 04.03: ** Let's add to our Dozer example the HealthProfile class (using annotations) and in such a way that the BMI calculation is only included in the transfer bean.
-
Below is an example of an XML mapping file for Dozer
<?xml version="1.0" encoding="UTF-8"?> <mappings xmlns="http://dozer.sourceforge.net" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://dozer.sourceforge.net http://dozer.sourceforge.net/schema/beanmapping.xsd"> <mapping> <class-a>dozerproject.PersonDB</class-a> <class-b>lsde.lab4.presentation.PersonUI</class-b> <field> <a>firstName</a> <b>fName</b> </field> <field> <a>lastName</a> <b>lName</b> </field> </mapping> </mappings>
-
Below is an example of an Java program that uses Dozer to map beans
public static void main (String argus[]){ DozerMapper mapper = new DozerMapper(); PersonDB pdb = getPersonFromDB(); // load mapping files List myMappingFiles = new ArrayList(); myMappingFiles.add("File:./dozerMappings.xml"); // prepare DozerMapper DozerBeanMapper mapper = new DozerBeanMapper(); mapper.setMappingFiles(myMappingFiles); // do the mapping PersonUI destObject = (PersonUI) mapper.map(sourceObject, PersonUI.class); // serialize the mapped object to the final representation (e.g., xml) String xmlFile = "person.xml"; File xmlDocument = newFile(xmlFile); try { JAXBContext jaxbContext = JAXBContext.newInstance(PersonUI.class); Marshaller marshaller = jaxbContext.createMarshaller(); marshaller.setProperty("jaxb.formatted.output", new Boolean(true)); marshaller.marshal(destObject, new FileOutputStream(xmlDocument)); } catch (IOException e) { }
- Finally, although XML is one of the cornerstones of service oriented computing, nowadays, JSON is the most common language for message exchange.
- The main reason behind this comes from the advent of mobile computing: JSON is less complex to process, thereby requiring less computation resources, and less battery.
- One way of having both (XML and JSON) to coexist (which they do in most web services implementation frameworks) is reusing annotations.
- A library that supports reusing JAXB annotations for mapping also to and from JSON is Jackson.
- Open and run HealthProfileJson Example to see how it works
- Checkout the details of Assignment 1 and stay tune for the deadline announcement in the hands-on session.