1 Introduction

Wicket core developers have done a wonderful job with this framework and they continue to improve it release after release. However Wicket never provided a freely available documentation and even if you can find on Internet many live examples and many technical articles on it (most of them at Wicket Library and at Wicket in Action ), the lack of an organized and freely available documentation has always been a sore point for this framework.

That's quite an issue because many other popular frameworks (like Spring, Hibernate or Struts) offer a vast and very good documentation which substantially contributed to their success.

This document is not intended to be a complete reference for Wicket but it simply aims to be a straightforward introduction to the framework that should significantly reduce its learning curve. What you will find here reflects my experience with Wicket and it's strictly focused on the framework. The various Wicket-related topics are gradually introduced using pragmatic examples of code that you can find in the according repository on Github.

However remember that Wicket is a vast and powerful tool, so you should feel confident with the topics exposed in this document before starting to code your real applications!

For those who need further documentation on Wicket, there are many good books available for this framework.

Hope you'll find this guide helpful. Have fun with Wicket!

Andrea Del Bene, andrea.del-bene@comsysto.com

PS: this guide is based on Wicket 6. However if you are using an older version you should find this guide useful as well, but it's likely that the code and the snippets won't work with your version.

PPS: although I've tried to do my best working on this tutorial, this document is a work in progress and may contain errors and/or omissions. That's why any feedback of any kind is REALLY appreciated!

2 Why should I learn Wicket?

But before starting to learn the last “coolest” framework we should always ask ourself if it is the right tool for us and how it can improve our everyday job. Java's ecosystem is already full of many well-known web frameworks, so why should we spend our time learning Wicket?

This chapter will show you how Wicket is different from other web frameworks you may know and it will explain also how it can improve your life as web developer.

2.1 We all like spaghetti :-) ...

Some of theme (like Struts, Spring MVC, Velocity, etc...) have become widely adopted and they made the MVC pattern very popular among developers. However, none of these frameworks offers a real object-oriented (OO) abstraction for web pages and we still have to take care of web-related tasks such as HTTP request/response handling, URLs mapping, storing data into user session and so on.

The biggest limit of MVC frameworks is that they don't do much to overcome the impedance mismatch between the stateless nature of HTTP protocol and the need of our web applications of handling a (very complex) state.

To overcome these limits developers have started to adopt a new generation of component oriented web frameworks designed to provide a completely different approach to web development.

2.2 Component oriented frameworks - an overview

1. is built on server-side
2. is built before HTML is sent to client
3. HTML code is generated using this model and not vice versa.

General schema of page request handling for a component oriented framework

With this kind of framework our web pages and their HTML components (forms, input controls, links, etc...), are pure class instances. Since pages are class instances they live inside the JVM heap and we can handle them as we do with any other Java class. This approach is very similar to what GUI frameworks (like Swing or SWT) do with desktop windows and their components. Wicket and the other component oriented frameworks bring to web development the same kind of abstraction that GUI frameworks offer when we build a desktop application. Most of those kind of frameworks hide the details of the HTTP protocol and naturally solve the problem of its stateless nature.

2.3 Benefits of component oriented frameworks for web development

• Web pages are objects: web pages are not just text files sent back to the client. They are object instances and we can harness OOP to design web pages and their components. With Wicket we can also apply inheritance to HTML markup in order to build a consistent graphic layout for our applications (we will see markup inheritance in chapter 4.2).
• We don't have to worry about application's state: pages and components can be considered stateful entities. They are Java objects and they can keep a state inside them and reference other objects. We can stop worrying about keeping track of user data stored inside the HttpSession and we can start managing them in a natural and transparent way.
• Testing web applications is much easier: since pages and components are pure objects, you can use JUnit to test their behavior and to ensure that they render as expected. Wicket has a set of utility classes for unit testing that simulate user interaction with web pages, hence we can write acceptance tests using just JUnit without any other test framework (unit testing is covered in chapter 20).

2.4 Wicket vs the other component oriented frameworks

• Wicket is 100% open source: Wicket is a top Apache project and it doesn't depend on any private company. You don't have to worry about future licensing changes, Wicket will always be released under Apache license 2.0 and freely available.
• Wicket is a community driven project: The Wicket team supports and promotes the dialogue with the framework's users through two mailing lists (one for users and another one for framework developers) and an Apache JIRA (the issue tracking system). Moreover, as any other Apache project, Wicket is developed paying great attention to user feedbacks and to suggested features.
• Wicket is just about Java and good old HTML: almost all web frameworks force users to adopt special tags or to use server side code inside HTML markup. This is clearly in contrast with the concept of separation between presentation and business logic and it leads to a more confusing code in our pages. In Wicket we don't have to take care of generating HTML inside the page itself, and we won't need to use any tag other than standard HTML tags. All we have to do is to attach our components (Java instances) to the HTML tags using a simple tag attribute called wicket:id (we will shortly see how to use it).
• With Wicket we can easily use JavaBeans and POJO in our web tier: one of the most annoying and error-prone task in web development is collecting user input through a form and keeping form fields updated with previously inserted values. This usually requires a huge amount of code to extract input from request parameters (which are strings), parse them to Java types and store them into some kind of variable. And this is just half of the work we have to do as we must implement the inverse path (load data from Java to the web form).Moreover, most of the times our forms will use a JavaBean or a POJO as backing object, meaning that we must manually map form fields with the corresponding object fields and vice versa. Wicket comes with an intuitive and flexible mechanism that does this mapping for us without any configuration overhead (using a convention over configuration approach) and in a transparent way. Chapter 10 will introduce the concept of Wicket model and we will learn how to harness this entity with forms.
• No complex XML needed: Wicket was designed to minimize the amount of configuration files needed to run our applications. No XML file is required except for the standard deployment descriptor web.xml.

3 Wicket says “Hello world!”

In both frameworks we find a base class for GUI components called Component. Wicket pages can be composed (and usually are) by many components, just like AWT windows are composed by Swing/AWT components. Both frameworks promote the reuse of presentation code and GUI elements building custom components. Even if Wicket already comes with a rich set of ready-to-use components, building custom components is a common practice when working with this framework. We'll learn more about custom components in the next chapters.

3.1 Wicket distribution and modules

Please note that the core module depends on the utility and request modules, hence it cannot be used without them.

3.2 Configuration of Wicket applications

All the example projects presented in this document have been generated using Maven and the utility page at http://wicket.apache.org/start/quickstart.html . Appendix A contains the instructions needed to use these projects and build a quickstart application using Apache Maven. All the artifacts used in the next example (files web.xml, HomePage.class and HomePage.html) are automatically generated by Maven.

Wicket application structure

A Wicket application is a standard Java EE web application, hence it is deployed through a web.xml file placed inside folder WEB-INF:

Illustration : The standard directory structure of a Wicket application

The content of web.xml declares a servlet filter (class org.apache.wicket.Protocol.http.WicketFilter) which dispatches web requests to our Wicket application:

<?xml version="1.0" encoding="UTF-8"?>
<web-app>
    <display-name>Wicket Test</display-name>
    <filter>
        <filter-name>TestApplication</filter-name>
        <filter-class>org.apache.wicket.protocol.http.WicketFilter</filter-class>
        <init-param>
          <param-name>applicationClassName</param-name>
          <param-value>org.wicketTutorial.WicketApplication</param-value>
        </init-param>
    </filter>
    <filter-mapping>
        <filter-name>TestApplication</filter-name>
        <url-pattern>/*</url-pattern>
    </filter-mapping>
</web-app>

Since this is a standard servlet filter we must map it to a specific set of URLs through the <filter-mapping> tag). In the xml above we have mapped every URL to our Wicket filter.

Wicket can be started in two modes named respectively DEVELOPMENT and DEPLOYMENT. The first mode activates some extra features which help application development, like resources monitoring and reloading, full stack trace rendering of exceptions, an AJAX debugger window, etc… The DEPLOYMENT mode turns off all these features optimizing performances and resource consumption. In our example projects we will use the default mode which is DEVELOPMENT. Chapter 23.1 contains the chapter “Switching Wicket to DEPLOYMENT mode“ where we can find further details about these two modes as well as the possible ways we have to set the desired one. In any case, DO NOT deploy your applications in a production environment without switching to DEPLOYMENT mode!

The application class

If we look back at web.xml we can see that we have provided the Wicket filter with a parameter called applicationClassName. This value must be the fully qualified class name of a subclass of org.apache.wicket.Application. This subclass represents our web application built upon Wicket and it's responsible for configuring it when the server is starting up. Most of the times our custom application class won't inherit directly from class Application, but rather from class org.apache.wicket.protocol.http.WebApplication which provides a closer integration with servlet infrastructure. Class Application comes with a set of configuration methods that we can override to customize our application's settings. One of these methods is getHomePage() that must be overridden as it is declared abstract:

public abstract Class<? extends Page> getHomePage()

As you may guess from its name, this method specifies which page to use as homepage for our application. Another important method is init():

protected void init()

This method is called when our application is loaded by the web server (Tomcat, Jetty, etc...) and is the ideal place to put our configuration code. The Application class exposes its settings grouping them into interfaces (you can find them in package org.apache.wicket.settings). We can access these interfaces through getter methods that will be gradually introduced in the next chapters when we will cover the related settings.

The current application's instance can be retrieved at any time calling static method Application.get() in our code. We will give more details about this method in chapter 8.3. The content of the application class from project HelloWorldExample is the following:

public class WicketApplication extends WebApplication
{    	
	@Override
	public Class<? extends WebPage> getHomePage()
	{
		return HomePage.class;
	}
	@Override
	public void init()
	{
		super.init();
		// add your configuration here
	}
}

Since this is a very basic example of a Wicket application, we don't need to specify anything inside the init method. The home page of the application is the HomePage class. In the next paragraph we will see how this page is implemented and which conventions we have to follow to create a page in Wicket.

Declaring a WicketFilter inside web.xml descriptor is not the only way we have to kickstart our application. If we prefer to use a servlet instead of a filter, we can use class org.apache.wicket.protocol.http.WicketServlet. See the JavaDoc for further details.

3.3 The HomePage class

To complete our first Wicket application we must explore the home page class that is returned by the Application's method getHomePage() seen above. In Wicket a web page is a subclass of org.apache.wicket.WebPage. This subclass must have a corresponding HTML file which will be used by the framework as template to generate its HTML markup. This file is a regular plain HTML file (its extension must be html).

By default this HTML file must have the same name of the related page class and must be in the same package:

Illustration :Page class and its related HTML file

If you don't like to put class and html side by side (let's say you want all your HTML files in a separated folder) you can use Wicket settings to specify where HTML files can be found. We will cover this topic later in chapter 14.9.

The Java code for the HomePage class is the following:

package org.wicketTutorial;
import org.apache.wicket.request.mapper.parameter.PageParameters;
import org.apache.wicket.markup.html.basic.Label;
import org.apache.wicket.markup.html.WebPage;
public class HomePage extends WebPage {	
    public HomePage() {
	add(new Label("helloMessage", "Hello WicketWorld!"));
    }
}

Apart from subclassing WebPage, HomePage defines a constructor that adds a Label component to itself. Method add(Component component) is inherited from ancestor class org.apache.wicket. MarkupContainer and is used to add children components to a web page. We'll see more about MarkupContainer later in chapter 4.2. Class org.apache.wicket.markup.html.basic.Label is the simplest component shipped with Wicket. It just inserts a string (the second argument of its constructor) inside the corresponding HTML tag. Just like any other Wicket component, Label needs a textual id ('helloMessage' in our example) to be instantiated. At runtime Wicket will use this value to find the HTML tag we want to bind to the component. This tag must have a special attribute called wicket:id and its value must be identical to the component id (comparison is case-sensitive!).

Here is the HTML markup for HomePage (file HomePage.html):

<!DOCTYPE html>
<html>
	<head>
		<meta charset="utf-8" />
		<title>Apache Wicket HelloWorld</title>
	</head>
	<body>
		<div wicket:id="helloMessage">
		[Label's message goes here]
		</div>
	</body>
</html>

We can see that the wicket:id attribute is set according to the value of the component id. If we run this example we will see the text Hello WicketWorld! Inside a <div> tag.

Label replaces the original content of its tag (in our example Label's message goes here) with the string passed as value (Hello WicketWorld! in our example).

If we specify a wicket:id attribute for a tag without adding the corresponding component in our Java code, Wicket will throw a ComponentNotFound Exception. On the contrary if we add a component in our Java code without specifying a corresponding wicket:id attribute in our markup, Wicket will throw a WicketRuntimeException.

3.4 Wicket Links

That said, you shouldn't be surprised to find an abstract method called onClick() inside the Link class. In the following example we have a page with a Link containing an empty implementation of onClick:

public class HomePage extends WebPage {
	public HomePage(){
		add(new Link("id"){
			@Override
			public void onClick() {
				//link code goes here
		    }			
		});
	}
}

By default after onClick has been executed, Wicket will send back to the current page to the client web browser. If we want to navigate to another page we must use method setResponsePage of class Component:

public class HomePage extends WebPage {
	public HomePage(){
		add(new Link("id"){
			@Override
			public void onClick() {			   
                         //we redirect browser to another page.
                         setResponsePage(AnotherPage.class);
			}			
		});
	}
}

In the example above we used a version of setResponsePage which takes as input the class of the target page. In this way a new instance of AnotherPage will be create each time we click on the link. The other version of setResponsePage takes in input a page instance instead of a page class:

@Override
public void onClick() {			   
	//we redirect browser to another page.
	AnotherPage anotherPage = new AnotherPage();
	setResponsePage(anotherPage);
}

The difference between using the first version of setResponsePage rather than the second one will be illustrated in chapter 7, when we will introduce the topic of stateful and stateless pages. For now, we can consider them as equivalent.

Wicket comes with a rich set of link components suited for every need (links to static URL, Ajax-enhanced links, links to a file to download, links to external pages and so on). We will see them in chapter 9.

3.5 Summary

4 Wicket as page layout manager

4.1 Header, footer, left menu, content, etc...

With the adoption of server side technologies like JSP, ASP or PHP the tag <frame> has been replaced by a template-based approach where we divide our page layout into some common areas that will be present in each page of our web application. Then, we manually insert these areas in every page including the appropriate markup fragments.

In this chapter we will see how to use Wicket to build a site layout. The sample layout we will use is a typical page layout consisting of the following areas:

• a header which could contain site title, some logos, a navigation bar, etc...
• a left menu with a bunch of links to different areas/functionalities of the site.
• a footer with generic informations like web master's email, the company address, etc...
• a content area which usually contains the functional part of the page.

The following picture summarises the layout structure:

Once we have chosen a page layout, our web designer can start building up the site theme. The result is a beautiful mock of our future web pages. Over this mock we can map the original layout areas:

Now in order to have a consistent layout across all the site, we must ensure that each page will include the layout areas seen above. With an old template-based approach we must manually put them inside every page. If we were using JSP we would probably end up using include directive to add layout areas in our pages. We would have one include for each of the areas (except for the content):

For the sake of simplicity we can consider each included area as a static HTML fragment.

Now let's see how we can handle the layout of our web application using Wicket.

4.2 Here comes the inheritance!

Markup inheritance

As we have seen in the previous chapter, Wicket pages are pure Java classes, so we can easily write a page which is a subclass of another parent page. But in Wicket inheritance is not limited to the classic object-oriented code inheritance. When a class subclasses a WebPage it also inherits the HTML file of to the parent class. This type of inheritance is called markup inheritance. To better illustrate this concept let's consider the following example where we have a page class called GenericSitePage with the corresponding HTML file GenericSitePage.html. Now let's create a specific page called OrderCheckOutPage where users can check out their orders on our our web site. This class extends GenericSitePage but we don't provide it with any corresponding HTML file. In this scenario OrderCheckOutPage will use GenericSitePage.html as markup file:

Markup inheritance comes in handy for page layout management as it helps us avoid the burden of checking that each page conforms to the site layout. However to fully take advantage of markup inheritance we must first learn how to use another important component of the framework that supports this feature: the panel.

If no markup is found (nor directly assigned to the class, neither inherited from an ancestor) a MarkupNotFoundException is thrown.

Panel class

Class org.apache.wicket.markup.html.panel.Panel is a special component which lets us reuse GUI code and HTML markup across different pages and different web applications. It shares a common ancestor class with WebPage class, which is org.apache.wicket.MarkupContainer:

Illustration: Hierarchy of WebPage and Panel classes

Subclasses of MarkupContainer can contain children components that can be added with method add(Component...) (seen in chapter 3.3). MarkupContainer implements a full set of methods to manage children components. The basic operations we can do on them are:

• add one or more children components (with method add).
• remove a specific child component (with method remove).
• retrieve a specific child component with method get(String). The string parameter is the id of the component or its relative path if the component is nested inside other MarkupContainers. This path is a colon-separated string containing also the ids of the intermediate containers traversed to get to the child component. To illustrate an example of component path, let's consider the code of the following page:

MyPanel myPanel = new MyPanel ("innerContainer");
add(myPanel);

Component MyPanel is a custom panel containing only a label having "name" as id. Under those conditions we could retrieve this label from the container page using the following path expression:

Label name = (Label)get("innerContainer:name");

• replace a specific child component with a new component having the same id (with method replace).
• iterate thought children components with the iterator returned by method iterator or using visitor pattern1 with methods visitChildren.

Both Panel and WebPage have their own associated markup file which is used to render the corresponding component. If such file is not provided, Wicket will apply markup inheritance looking for a markup file through their ancestor classes. When a panel is attached to a container, the content of its markup file is inserted into its related tag.

While panels and pages have much in common, there are some notable differences between these two components that we should keep in mind. The main difference between them is that pages can be rendered as standalone entities while panels must be placed inside a page to be rendered. Another important difference is the content of their markup file: for both WebPage and Panel this is a standard HTML file, but Panel uses a special tag to indicate which part of the whole file will be considered as markup source. This tag is <wicket:panel>. A markup file for a panel will typically look like this:

<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
…
</head>
<body>
   <wicket:panel>
      <!-- Your markup goes here -->
       </wicket:panel>
</body>
</html>

The HTML outside tag <wicket:panel> will be removed during rendering phase. The space outside this tag can be used by both web developers and web designers to place some mock HTML to show how the final panel should look like.

4.3 Divide et impera!

Panels and layout areas

First, let's build a custom panel for each layout area (except for 'content' area). For example given the header area

we can build a panel called HeaderPanel with a related markup file called HeaderPanel.html containing the HTML for this area:

<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
…
</head>
<body>
   <wicket:panel>
      <table width="100%" style="border: 0px none;">
      <tbody>
    <tr>
    <td>
       <img alt="Jug4Tenda" src="wicketLayout_files/logo_jug4tenda.gif">
     </td>
      <td>
    <h1>Gestione Anagrafica</h1>
   </td>   
      </tr>
      </tbody>
      </table>   
   </wicket:panel>
</body>
<html>

The class for this panel simply extends base class Panel:

package helloWorld.layoutTenda;
import org.apache.wicket.markup.html.panel.Panel;
public class HeaderPanel extends Panel {
	public HeaderPanel(String id) {
		super(id);		
	}
}

For each layout area we will build a panel like the one above that holds the appropriate HTML markup. In the end we will have the following set of panels:

• HeaderPanel
• FooterPanel
• MenuPanel

Content area will change from page to page, so we don't need a reusable panel for it.

Template page

Now we can build a generic template page using our brand new panels. Its markup is quite straightforward :

<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8"> 
…
<!--Include CSS-->
…
</head>
<body>
<div id="header" wicket:id="headerPanel">header</div>
<div id="body">
	<div id="menu" wicket:id="menuPanel">menu</div>
	<div id="content" wicket:id="content">content</div>
</div>
<div id="footer" wicket:id="footerPanel">footer</div>
</body>
</html>

The HTML code for this page implements the generic left-menu layout of our site. You can note the 4 <div> tags used as containers for the corresponding areas. The page class contains the code to physically assemble the page and panels:

package helloWorld.layoutTenda;
import org.apache.wicket.markup.html.WebPage;
public class JugTemplate extends WebPage {
	public static final String CONTENT_ID = "contentComponent";
	private Component headerPanel;
	private Component menuPanel;
	private Component footerPanel;
              public JugTemplate(){
		add(headerPanel = new HeaderPanel("headerPanel"));
		add(menuPanel = new MenuPanel("menuPanel"));
		add(footerPanel = new FooterPanel("footerPanel"));
		add(new Label(CONTENT_ID, "Put your content here"));
	}
             //getters for layout areas
       //… 
}

Done! Our template page is ready to be used. Now all the pages of our site will be subclasses of this parent page and they will inherit the layout and the HTML markup. They will only substitute the Label inserted as content area with their custom content.

Final example

As final example we will build the login page for our site. We will call it SimpleLoginPage. First, we need a panel containing the login form. This will be the content area of our page. We will call it LoginFormPanel and the markup is the following:

<html>
<head>
</head>
<body>
   <wicket:panel>
    <div style="margin: auto; width: 40%;">
       <form  id="loginForm" method="get">
         <fieldset id="login" class="center">
            <legend >Login</legend>               
            <span >Username: </span><input type="text" id="username"/><br/>                                                                  
            <span >Password: </span><input type="password" id="password" />
            <p>
               <input type="submit" name="login" value="login"/>
            </p>
         </fieldset>
      </form>
    </div>   
   </wicket:panel>
</body>
</html>

The class for this panel just extends Panel class so we won't see the relative code. The form of this panel is for illustrative purpose only. We will see how to work with Wicket forms in chapters 10 and 11. Since this is a login page we don't want it to display the left menu area. That's not a big deal as Component class exposes a method called setVisible which sets whether the component and its children should be displayed.

The resulting Java code for the login page is the following:

package helloWorld.layoutTenda;
import helloWorld.LoginPanel;
import org.apache.wicket.event.Broadcast;
import org.apache.wicket.event.IEventSink;
public class SimpleLoginPage extends JugTemplate {
	public SimpleLoginPage(){
		super();		
		replace(new LoginPanel(CONTENT_ID));
		getMenuPanel().setVisible(false);
	}
}

Obviously this page doesn't come with a related markup file. You can see the final page in the following picture:

4.4 Markup inheritance with the tag

<html>
<head>
	<meta http-equiv="Content-Type" content="text/html; charset=UTF-8"> 
</head>
<body>
	This is parent body!
	<wicket:child/>
</body>
</html>

The markup of a child page/panel must be placed inside the tag <wicket:extend>. Only the markup inside <wicket:extend> will be included in final markup. Here is an example of child page markup:

<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8"> 
</head>
<body>
    <wicket:extend>
          This is child body!
	</wicket:extend>
</body>
</html>

Considering the two pages seen above, the final markup generated for child page will be the following:

<html>
<head>
	<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body>
	This is parent body!
	<wicket:child>
       <wicket:extend>
           This is child body!
	   </wicket:extend>
    </wicket:child>
</body>
</html>

Our example revisited

Applying <wicket:child> tag to our layout example, we obtain the following markup for the main template page:

<html>
<head>
	<meta http-equiv="Content-Type" content="text/html; charset=UTF-8"> 
</head>
<body>
<div id="header" wicket:id="headerPanel">header</div>
<div id="body">
	<div id="menu" wicket:id="menuPanel">menu</div>
	<wicket:child/>
</div>
<div id="footer" wicket:id="footerPanel">footer</div>
</body>
</html>

We have replaced the <div> tag of the content area with the tag <wicket:child>. Going forward with our example we can build a login page creating class SimpleLoginPage which extends the JugTemplate page, but with a related markup file like this:

<html>
<head>
</head>
<body>
   <wicket:extend>
    <div style="margin: auto; width: 40%;">
       <form  id="loginForm" method="get">
         <fieldset id="login" class="center">
            <legend >Login</legend>               
            <span >Username: </span><input type="text" id="username"/><br/>                                                                  
            <span >Password: </span><input type="password" id="password" />
            <p>
               <input type="submit" name="login" value="login"/>
            </p>
         </fieldset>
      </form>
    </div>   
   </wicket:extend>
</body>
</html>

As we can see this approach doesn't require to create custom panels to use as content area and it can be useful if we don't have to handle a GUI with a high degree of complexity.

4.5 Summary

5 Keeping control over HTML

This chapter will prevent you from having any initial misleading feeling about Wicket showing you how to control and manipulate the generated HTML with the built-in tools shipped with the framework.

5.1 Hiding or disabling a component

Class Component provides two getter methods to determinate if a component is visible or enabled: isVisible and isEnabled.

Even if nothing prevents us from overriding these two methods to implement a custom logic to determinate the state of a component, we should keep in mind that methods isVisible and isEnabled are called multiple times before a component is fully rendered. Hence, if we place non-trivial code inside these two methods, we can sensibly deteriorate the responsiveness of our pages.

As we will see in the next chapter, class Component provides method onConfigure which is more suited to contain code that contributes to determinate component states because it is called just once during rendering phase.

5.2 Modifing tag attributes

As first example of attribute manipulation let's consider a Label component bound to the following markup:

<span wicket:id="simpleLabel"></span>

Suppose we want to add some style to label content making it red and bolded. We can add to the label an AttributeModifier which creates the tag attribute style with value "color:red;font-weight:bold":

label.add(new AttributeModifier("style", "color:red;font-weight:bold"));

If attribute style already exists in the original markup, it will be replaced with the value specified by AttributeModifier. If we don't want to overwrite the existing value of an attribute we can use subclass AttributeAppender which will append its value to the existing one:

label.add(new AttributeAppender("style", "color:red;font-weight:bold"));

We can also create attribute modifiers using factory methods provided by class AttributeModifier and it's also possible to prepend a given value to an existing attribute:

//replaces existing value with the given one
label.add(AttributeModifier.replace("style", "color:red;font-weight:bold"));
//appends the given value to the existing one
label.add(AttributeModifier.append("style", "color:red;font-weight:bold"));
//prepends the given value to the existing one
label.add(AttributeModifier.prepend("style", "color:red;font-weight:bold"));

5.3 Generating tag attribute 'id'

5.4 Creating in-line panels with WebMarkupContainer

In situations like these component org.apache.wicket.markup.html.WebMarkupContainer is better suited than custom panels because it can be directly attached to a tag in the parent markup without needing a corresponding html file (hence it is less reusable). Let's consider for example the main page of a mail service where users can see a list of received mails. Suppose that this page shows a notification box where user can see if new messages have arrived. This box must be hidden if there are no messages to display and it would be nice if we could handle it as if it was a Wicket component.

Suppose also that this information box is a <div> tag like this inside the page:

<div wicket:id="informationBox">
   //here's the body
   You've got <span wicket:id="messagesNumber"></span> new messages.
</div>

Under those conditions we can consider using a WebMarkupContainer component rather than implementing a new panel. The code needed to handle the information box inside the page could be the following:

//Page initialization code
WebMarkupContainer informationBox = new WebMarkupContainer ("informationBox");
informationBox.add(new Label("messagesNumber", messagesNumber));
add(informationBox);
//If there are no new messages, hide informationBox
informationBox.setVisible(false);

As you can see in the snippet above we can handle our information box from Java code as we do with any other Wicket component.

5.5 Working with markup fragments

To better cope with situations like this, Wicket defines component Fragment in package org. apache.wicket.markup.html.panel. Just like its parent component WebMarkupContainer, Fragment doesn't have its own markup file but it uses a markup fragment defined in the markup file of its parent container, which can be a page or a panel. The fragment must be delimited with tag <wicket:fragment> and must be identified by a wicket:id attribute. In addition to the component id, Fragment's constructor takes as input also the id of the fragment and a reference to its container.

In the following example we have defined a fragment in a page and we used it as content area:

Page markup:

<html>
  …
<body>
…
	<div wicket:id="contentArea"></div>
	<wicket:fragment wicket:id="fragmentId">
	   <!-- Fragment markup goes here -->
	</wicket:fragment>
</body>
</html>

Java code:

Fragment fragment = new  Fragment ("contentArea", "fragmentId", this);
add(fragment);

Fragments can be very helpful with complex pages or components. For example let's say that we have a page where users can register to our forum. This page should first display a form where user must insert his/her personal data (name, username, password, email and so on), then, once the user has submitted the form, the page should display a message like “Your registration is complete! Please check your mail to activate your user profile.”.

Instead of displaying this message with a new component or in a new page, we can define two fragments: one for the initial form and one to display the confirmation message. The second fragment will replace the first one after the form has been submitted:

Page markup:

<html>
<body>
	<div wicket:id="contentArea"></div>
	<wicket:fragment wicket:id="formFrag">
	   <!-- Form markup goes here -->
	</wicket:fragment>
	<wicket:fragment wicket:id="messageFrag">
	   <!-- Message markup goes here -->
	</wicket:fragment>
</body>
</html>

Java code:

Fragment fragment = new  Fragment ("contentArea", "formFrag", this);
add(fragment);
//form has been submitted
Fragment fragment = new  Fragment ("contentArea", "messageFrag", this);
replace(fragment);

5.6 Adding header contents to the final page

Considering the markup of a generic panel, we can use <wicket:head> tag in this way:

{html} <wicket:head> <script type="text/javascript"> function myPanelFunction(){ } </script>

<style> .myPanelClass{ font-weight: bold; color: red; } </style> </wicket:head> <body> <wicket:panel>

</wicket:panel> </body> {html}

Wicket will take care of placing the content of <wicket:head> inside the <head> tag of the final page.

The <wicket:head> tag can also be used with children pages/panels which extend parent markup using tag <wicket:extend>.

The content of the <wicket:head> tag is added to the header section once per component class. In other words, if we add multiple instances of the same panel to a page, the <head> tag will be populated just once with the content of <wicket:head>.

The <wicket:head> tag is ideal if we want to define small in-line blocks of CSS or JavaScript. However Wicket provides also a more sophisticated technique to let components contribute to header section with in-line blocks and resource files like CSS or JavaScript files. We will see this technique later in chapter 14.

5.7 Using stub markup in our pages/panels

<html>
<head>
</head>
<body>
	<wicket:remove>
	   <!-- Stub markup goes here -->
	</wicket:remove>
</body>
</html>

5.8 How to render component body only

For example given the following markup and code:

HTML markup:

<html>
<head>
  <title>Hello world page</title>
</head>
<body>
<div wicket:id="helloWorld">[helloWorld]</div>
</body>
</html>

Java code:

Label label = new Label("helloWorld", “Hello World!”);
label.setRenderBodyOnly(true);
add(label);

the output will be:

<html>
<head>
  <title>Hello world page</title>
</head>
<body>
 Hello World!
</body>
</html>

As you can see the <div> tag used for component Label is not present in the final markup.

5.9 Hiding decorating elements with the tag

<label>Total amount: </label><span wicket:id="totalAmount"></span>

Wicket comes with a nice utility tag called <wicket:enclosure> that automatically hides those decorating elements if the related data value is not visible. All we have to do is to put the involved markup inside this tag. Applying <wicket:enclosure> to the previous example we get the following markup:

<wicket:enclosure> 
    <label>Total amount: </label><span wicket:id="totalAmount"></span>
</wicket:enclosure>

Now if component totalAmount is not visible, its description (Total amount:) will be automatically hidden. If we have more then a Wicket component inside <wicket:enclosure> we can use child attribute to specify which component will control the overall visibility:

<wicket:enclosure child="totalAmount"> 
    <label>Total amount: </label><span wicket:id="totalAmount"></span><br/>
	<label>Expected delivery date: </label><span wicket:id="delivDate"></span>
</wicket:enclosure>

child attribute supports also nested components with a colon-separated path:

<wicket:enclosure child="totalAmountContainer:totalAmount"> 
    <div wicket:id="totalAmountContainer">
		<label>Total amount: </label><span wicket:id="totalAmount"></span>
    </div>
    <label>Expected delivery date: </label><span wicket:id="delivDate"></span>
</wicket:enclosure>

5.10 Surrounding existing markup with Border

<?xml version="1.0" encoding="UTF-8"?>
<html xmlns="http://www.w3.org/1999/xhtml" xmlns:wicket="http://wicket.apache.org">
<head></head>
<body>
    <!--  everything above <wicket:border> tag will be discarded...-->
    <wicket:border>
	 <div>
	    foo<br />
    <wicket:body/><br />
           buz <br />
  </div>
    </wicket:border>
    <!--  everything below </wicket:border> tag will be discarded...-->
</body>
</html>

The <wicket:body/> tag used in the example above is used to indicate where the body of the tag will be placed inside border markup. Now if we attached this border to the following tag

<span wicket:id="myBorder">
  bar
</span>

we would obtain the following resulting HTML:

<span wicket:id="myBorder">
	<div>
  		foo<br />
  		bar<br />
  		buz <br />
	</div>
</span>

Border can also contain children components which can be placed either inside its markup file or inside its corresponding HTML tag. In the first case children must be added to the border component with method addToBorder(Component...), while in the second case we must use the add(Component...) method.

The following example illustrates both use cases:

Border class:

public class MyBorder extends Border {
	public MyBorder(String id) {
		super(id);		
	}
}

Border Markup:

<?xml version="1.0" encoding="UTF-8"?>
<html xmlns="http://www.w3.org/1999/xhtml" xmlns:wicket="http://wicket.apache.org">
<head></head>
<body>
    <wicket:border>
	 <div>
	    <div wicket:id="childMarkup"></div>
    <wicket:body/><br />
         </div>
    </wicket:border>
</body>
</html>

Border tag:

<div wicket:id="myBorder">
  <span wicket:id="childTag"></span>
</div>

Initialization code for border:

MyBorder myBorder = new MyBorder("myBorder");
myBorder.addToBorder(new Label("childMarkup", "Child inside markup."));
myBorder.add(new Label("childTag", "Child inside tag."));
add(myBorder);

5.11 Summary

In this chapter we have seen the tools provided by Wicket to gain complete control over the generated HTML. However we didn't see yet how we can repeat a portion of HTML with Wicket. With classic server-side technologies like PHP or JSP we use loops (like while or for) inside our pages to achieve this result. To perform this task Wicket provides a special-purpose family of components called repeaters and designed to repeat their markup body to display a set of items.

But to fully understand how these components work, we must first learn more of Wicket's basics. That's why repeaters will be introduced later in chapter 12.

6 Components lifecycle

6.1 Lifecycle stages of a component

During its life a Wicket component goes through three basic stages:

1. Initialization: a component is instantiated by Wicket and prepared for the rendering phase.
2. Rendering: in this stage Wicket generates component markup. If a component contains children (i.e. is a subclass of MarkupContainer) it must first wait for them to be rendered before starting its own rendering.
3. Removing: this stage is triggered when a component is explicitly removed from its component hierarchy, i.e. when its parent invokes remove(component) on it. This stage is facultative and is never triggered for pages.

The following picture shows the state diagram of component lifecycle:

Once a component has been removed it can be added again to a container, but the initialization stage won't be executed again.

If you read the JavaDoc of class Component you will find a more detailed description of component lifecycle. However this description introduces some advanced topics we didn't covered yet hence, to avoid confusion, in this chapter some details have been omitted and they will be covered later in the next chapters.

For now you can consider just the simplified version of the lifecycle described above.

6.2 Hook methods for component lifecycle

Class Component comes with a number of hook methods that can be overridden in order to customize component behavior during its lifecycle. In the following table these methods are grouped according to the stage in which they are invoked (and they are sorted by execution order):

Now let's take a closer look at each stage and to at hook methods.

6.3 Initialization stage

This stage is performed at the beginning of the component lifecycle. During initialization, the component has already been inserted into its component hierarchy so we can safely access to its parent container or to its page with methods getParent() or getPage(). The only method triggered during this stage is onInitialize(). This method is a sort of “special” constructor where we can execute a custom initialization of our component.

Since onInitialize is similar to a regular constructor, when we override this method we have to call super.onInitialize inside its body, usually as first instruction.

6.4 Rendering stage

Method onConfigure

Method onConfigure() has been introduced in order to provide a good point to manage the component states such as its visibility or enabled state. This method is called before the render phase starts. As stated in chapter 5.1, isVisible and isEnabled are called multiple times when a page or a component is rendered, so it's highly recommended not to directly override these method, but rather to use onConfigure to change component states. On the contrary method onBeforeRender (see the next paragraph) is not indicated for this task because it will not be invoked if component visibility is set to false.

Method onBeforeRender

The most important hook method of this stage is probably onBeforeRender(). This method is called before a component starts its rendering phase and it is our last chance to change its children hierarchy.

If we want add/remove children components this is the right place to do it. In the next example (project LifeCycleStages) we will create a page which alternately displays two different labels, swapping between them each time it is rendered:

public class HomePage extends WebPage
{
	private Label firstLabel;
	private Label secondLabel;
	public HomePage(){
		firstLabel = new Label("label", "First label");
		secondLabel = new Label("label", "Second label");
		add(firstLabel);
		add(new Link("reload"){
			@Override
			public void onClick() {
			}
		});
	}
	@Override
	protected void onBeforeRender() {
		if(contains(firstLabel, true))
			replace(secondLabel);
		else
			replace(firstLabel);
		super.onBeforeRender();
	}
}

The code inside onBeforeRender() is quite trivial as it just checks which label among firstLabel and secondLabel is currently inserted into the component hierarchy and it replaces the inserted label with the other one.

This method is also responsible for invoking children onBeforeRender() so if we decide to override it we have to call super.onBeforeRender(). However, unlike onInitialize(), the call to superclass method should be placed at the end of method's body in order to affect children's rendering with our custom code.

Please note that in the example above we can trigger the rendering stage pressing F5 key or clicking on link “reload”.

If we forget to call superclass version of methods onInitialize() or onBeforeRender(), Wicket will throw an IllegalStateException with the following message:

java.lang.IllegalStateException: org.apache.wicket.Component has not been properly initialized. Something in the hierarchy of <page class name> has not called super.onInitialize()/onBeforeRender() in the override of onInitialize()/ onBeforeRender() method

Method onComponentTag

Method onComponentTag(ComponentTag) is called to process component tag, which can be freely manipulated through its argument of type org.apache.wicket.markup.ComponentTag. For example we can add/remove tag attributes with methods put(String key, String value) and remove(String key), or we can even decide to change the tag or rename it with method setName(String) (the following code is taken from project OnComponentTagExample):

Markup code:

<head>
  <meta charset="utf-8" />
  <title></title>
</head>
<body>		
  <h1 wicket:id="helloMessage"></h1>		
</body>

Java code:

public class HomePage extends WebPage {
   public HomePage() {
      add(new Label("helloMessage", "Hello World"){
         @Override
         protected void onComponentTag(ComponentTag tag) {            
            super.onComponentTag(tag);
            //Turn the h1 tag to a span
            tag.setName("span");
            //Add formatting style
            tag.put("style", "font-weight:bold");
         }
      });
    }
}

Generated markup:

<head>
  <meta charset="utf-8" />
  <title></title>
</head>
<body>		
  <span wicket:id="helloMessage" style="font-weight:bold">Hello World</span>		
</body>

Just like we do with onInitialize, if we decide to override onComponentTag we must remember to call the same method of the super class because also this class may also customize the tag. Overriding onComponentTag is perfectly fine if we have to customize the tag of a specific component, but if we wanted to reuse the code across different components we should consider to use a behavior in place of this hook method.

We have already seen in chapter 5.2 how to use behavior AttributeModifier to manipulate the tag's attribute. In chapter 16.1 we will see that base class Behavior offers also a callback method named onComponentTag(ComponentTag, Component) that can be used in place of the hook method onComponentTag(ComponentTag).

Methods onComponentTagBody

Method onComponentTagBody(MarkupStream, ComponentTag) is called to process the component tag's body. Just like onComponentTag it takes as input a ComponentTag parameter representing the component tag. In addition, we also find a MarkupStream parameter which represents the page markup stream that will be sent back to the client as response.

onComponentTagBody can be used in combination with the Component's method replaceComponentTagBody to render a custom body under specific conditions. For example (taken from project OnComponentTagExample) we can display a brief description instead of the body if the label component is disabled:

public class HomePage extends WebPage {
   public HomePage() {
      add(new Label("helloMessage", "Hello World"){
         @Override
         protected void onComponentTagBody(MarkupStream markupStream, ComponentTag tag) {
           if(!isEnabled())
               replaceComponentTagBody(markupStream, tag, "(the component is disabled)"); 
          else    
               super.onComponentTagBody(markupStream, tag);
         }
      });   
    }
}

Note that the original version of onComponentTagBody is invoked only when we want to preserve the standard rendering mechanism for the tag's body (in our example this happens when the component is enabled).

6.5 Removing stage

This stage is triggered when a component is removed from its component hierarchy. The only hook method for this phase is onRemove(). If our component still holds some resources needed during rendering phase, we can override this method to release them.

Once a component has been removed we are free to add it again to the same container or to a different one.

6.6 Summary

In this chapter we have seen which stages compose the lifecycle of Wicket components and which hook methods they provide. Overriding these methods we can dynamically modify the component hierarchy and we can enrich the behavior of our custom components.

7 Page versioning and caching

7.1 Stateful pages vs stateless

Wicket pages can be divided into two categories: stateful and stateless pages. Stateful pages are those which rely on user session to store they internal state and to keep track of user interaction. On the contrary stateless pages are those which don't change their internal state during their lifecycle and they don't need to occupy space into user session.

From Wicket's point of view the biggest difference between these two types of page is that stateful pages are versioned, meaning that they will be saved into user session every time their internal state has changed. Wicket automatically assigns a session to the user the first time a stateful page is requested. Page versions are stored into user session using Java Serialization mechanism. Stateless pages are never versioned and that's why they don't require a valid user session. If we want to know whether a page is stateless or not, we can call the isPageStateless() method of class Page.

In order to build a stateless page we must comply with some rules to ensure that the page won't need to use user session. These rules are illustrated in paragraph 6.3 but before talking about stateless pages we must first understand how stateful pages are handled and why they are versioned.

7.2 Stateful pages

Stateful pages are versioned in order to support browser's back button: when this button is pressed Wicket must respond by rendering the same page instance previously used.

A new page version is created when a stateful page is requested for the first time or when an existing instance is modified (for example changing its component hierarchy). To identify each page version Wicket uses a session-relative identifier called page id. This is a unique number and it is increased every time a new page version is created.

In the final example of the previous chapter (project LifeCycleStages), you may have noticed the number appended at the and of URL. This number is the page id we are talking about:

In this chapter we will use a revised version of this example project where the component hierarchy is modified inside the Link's onClick()method. This is necessary because Wicket creates a new page version only if the page is modified before its method onBeforeRender() is invoked. The code of the new home page is the following:

public class HomePage extends WebPage
{
	private static final long serialVersionUID = 1L;
	private Label firstLabel;
	private Label secondLabel;
	public HomePage(){
		firstLabel = new Label("label", "First label");
		secondLabel = new Label("label", "Second label");
		add(firstLabel);
		add(new Link("reload"){
			@Override
			public void onClick() {				
				if(getPage().contains(firstLabel, true))
					getPage().replace(secondLabel);
				else
					getPage().replace(firstLabel);		
			}
		});
	}	
}

Now if we run the new example (project LifeCycleStagesRevisited) and we click on the “Reload” button, a new page version is created and the page id is increased by one:

!reload-page-png!

If we press the back button the page version previously rendered (and serialized) will be retrieved (i.e. deserialized) and it will be used again to respond to our request (and page id is decremented):

For more details about page storing you can visit the wiki page at https://cwiki.apache.org/confluence/display/WICKET/Page+Storage . On this page you can find which classes are involved into page storing mechanism and how they work together.

As we have stated at the beginning of this chapter, page versions are stored using Java serialization, therefore every object referenced inside a page must be serializable1. In paragraph 9.6 we will see how to overcome this limit and work with non-serializable objects in our components using detachable Wicket models.

Using a specific page version with PageReference

To retrieve a specific page version in our code we can use class org.apache.wicket. PageReference by providing its constructor with the corresponding page id:

//load page version with page id = 3
PageReference pageReference = new PageReference(3);
//load the related page instance
Page page = pageReference.getPage();

To get the related page instance we must use method getPage.

Turning off page versioning

If for any reason we need to switch off versioning for a given page, we can call its method setVersioned(false).

Pluggable serialization

Starting from version 1.5 it is possible to choose which implementation of Java serialization will be used by Wicket to store page versions. Wicket serializes pages using an implementation of interface org.apache.wicket.serialize.ISerializer. The default implementation is org.apache.wicket.serialize.java.JavaSerializer and it uses the standard Java serialization mechanism based on classes ObjectOutputStream and ObjectInputStream. However on Internet we can find other interesting serialization libraries like Kryo1 which performs faster then the standard implementation.The serializer in use can be customized with the setSerializer(ISerializer) method defined by setting interface org.apache.wicket.settings.IFrameworkSettings.

We can access this interface inside the method init of the class Application using the getFrameworkSettings() method :

@Override
public void init()
{
	super.init();
	getFrameworkSettings().setSerializer(yourSerializer);
}

A serializer based on Kryo library is provided by the WicketStuff project. You can find more information on this project, as well as the instructions to use its modules, in Appendix B.

Page caching

By default Wicket persists versions of pages into a session-relative file on disk, but it uses a two-levels cache to speed up this process. The first level of the cache uses a http session attribute called “wicket:persistentPageManagerData-<APPLICATION_NAME>” to store pages. The second level cache stores pages into application-scoped variables which are identified by a session id and a page id.

The following picture is an overview of these two caching levels:

The session-scoped cache is faster then the other memory levels but it contains only the pages used to serve the last request. Wicket allows us to set the maximum amount of memory allowed for the application-scoped cache and for the page store file. Both parameters can be configured via setting interface org.apache .wicket.settings.IStoreSettings.

This interface provides the setMaxSizePerSession(Bytes bytes) method to set the size for page store file. The Bytes parameter is the maximum size allowed for this file:

@Override
public void init()
{
	super.init();
	getStoreSettings().setMaxSizePerSession(Bytes.kilobytes(500));
}

Class org.apache.wicket.util.lang.Bytes is an utility class provided by Wicket to express size in bytes (for further details refer to the JavaDoc). For the second level cache we can use the setInmemoryCacheSize(int inmemoryCacheSize) method. The integer parameter is the maximum number of page instances that will be saved into application-scoped cache:

@Override
public void init()
{
	super.init();
	getStoreSettings().setInmemoryCacheSize(50);
}

Page expiration

Page instances are not kept in the user session forever. They can be discarded when the limit set with the setMaxSizePerSession method is reached or (more often) when user session expires. When we ask Wicket for a page id corresponding to a page instance removed from the session, we bump into a PageExpiredException and we get the following default error page:

This error page can be customized with the setPageExpiredErrorPage method of the org.apache.wicket.settings.IApplicationSettings interface:

@Override
public void init()
{
	super.init();
	getApplicationSettings().setPageExpiredErrorPage(CustomExpiredErrorPage.class);
}

The page class provided as custom error page must have a public constructor with no argument or a constructor that takes as input a single PageParameters argument (the page must be bookmarkable as described in paragraph 8.1.1).

7.3 Stateless pages

Wicket makes it very easy to build stateful pages, but sometimes we might want to use an “old school” stateless page that doesn't keep memory of its state in the user session. For example consider the public area of a site or a login page: in those cases a stateful page would be a waste of resources or even a security threat, as we will see in paragraph 10.9.

In Wicket a page can be stateless only if it satisfies the following requirements:

1. it has been instantiated by Wicket (i.e. we don't create it with operator new) using a constructor with no argument or a constructor that takes as input a single PageParameters argument (class PageParameters will be covered in chapter 8).
2. All its children components (and behaviors) are in turn stateless, which means that their method isStateless must return true.

The first requirement implies that, rather than creating a page by hand, we should rely on Wicket's capability of resolving page instances, like we do when we use method setResponsePage(Class page).

In order to comply with the second requirement it could be helpful to check if all children components of a page are stateless. To do this we can leverage method visitChildren and the visitor pattern to iterate over components and test if their method isStateless actually returns true:

@Override
protected void onInitialize() {
		super.onInitialize();
		visitChildren(new IVisitor<Component, Void>() {
			@Override
			public void component(Component component, IVisit<Void> arg1) {
				if(!component.isStateless())
		  			System.out.println("Component " + component.getId() + " is not stateless");
			}
		});
	}

Alternatively, we could use the StatelessComponent utility annotation along with the StatelessChecker class (they are both in package org.apache.wicket.devutils. stateless). StatelessChecker will throw an IllegalArgumentException if a component annotated with StatelessComponent doesn't respect the requirements for being stateless. To use StatelessComponent annotation we must first add the StatelessChecker to our application as a component render listener:

@Override
public void init()
{
	super.init();
	getComponentPostOnBeforeRenderListeners().add(new StatelessChecker());
}

Most of the Wicket's built-in components are stateful, hence they can not be used with a stateless page. However some of them have also a stateless version which can be adopted when we need to keep a page stateless. In the rest of the guide we will point out when a built-in component comes also with a stateless version.

A page can be also explicitly declared as stateless setting the appropriate flag to true with the setStatelessHint(true) method. This method will not prevent us from violating the requirements for a stateless page, but if we do so we will get the following warning log message:

Page '<page class>' is not stateless because of component with path '<component path>'

7.4 Summary

In this chapter we have seen how page instances are managed by Wicket. We have learnt that pages can be divided into two families: stateless and stateful pages. Knowing the difference between the two types of pages is important to build the right page for a given task.

However, to complete the discussion about stateless pages we still have to deal with two topics we have just outlined in this chapter: class PageParameters and bookmarkable pages. The first part of chapter 8 will cover these missing topics.

8 Under the hood of the request processing

Wicket provides wrapper classes that allow us to easily access to web entities without the burden of using the low-level APIs of Java Servlet Specification. However it will always be possible to access standard classes (like HttpSession, HttpServletRequest, etc...) that lay under our Wicket application. This chapter will introduce these wrapper classes and it will explain how Wicket uses them to handle the web requests initiated by the user's browser.

8.1 Class Application and request processing

Beside configuring and initializing our application, the Application class is responsible for creating the internal entities used by Wicket to process a request. These entities are instances of the following classes: RequestCycle, Request, Response and Session.

The next paragraphs will illustrate each of these classes, explaining how they are involved into request processing.

8.2 Request and Response classes

The Request and Response classes are located in package org.apache.wicket.request and they provide an abstraction of the concrete request and response used by our web application.

Both classes are declared as abstract but if our application class inherits from WebApplication it will use their sub classes ServletWebRequest and ServletWebResponse, both of them located inside the package org.apache.wicket.protocol.http.servlet.ServletWebRequest and ServletWebResponse wrap respectively a HttpServletRequest and a HttpServletResponse object. If we need to access to these low-level objects we can call Request's method getContainerRequest() and Response's method getContainer Response().

8.3 The “director” of request processing - RequestCycle

Class org.apache.wicket.request.cycle.RequestCycle is the entity in charge of serving a web request. Our application class creates a new RequestCycle on every request with its method createRequestCycle(request, response).

Method createRequestCycle is declared as final, so we can't override it to return a custom subclass of RequestCycle. Instead, we must build a request cycle provider implementing interface org.apache.wicket.IRequestCycleProvider, and then we must tell our application class to use it via the setRequestCycleProvider method.

The current running request cycle can be retrieved at any time by calling its static method RequestCycle.get(). Strictly speaking this method returns the request cycle associated with the current (or local) thread, which is the thread that is serving the current request. A similar get() method is also implemented in classes org.apache.wicket.Application (as we have seen in paragraph 2.2.2) and org.apache.wicket.Session in order to get the application and the session in use by the current thread.

The implementation of the get method takes advantage of the standard class java.lang.ThreadLocal. See its JavaDoc for an introduction to local-thread variables.

Class org.apache.wicket.Component provides the getRequestCycle() method which is a convenience method that internally invokes RequestCycle.get():

public final RequestCycle getRequestCycle() {
	return RequestCycle.get();
}

RequestCycle and request processing

This paragraph will provide just the basic informations about what happens behind the scenes of request processing. When you work with Wicket it's unlikely to have a need for customizing this process, so we won't cover this topic in detail.

In order to process a request, RequestCycle delegates the task to another entity which implements interface org.apache.wicket.request.IRequestHandler. There are different implementations of this interface, each suited for a particular type of requested resource (a page to render, an AJAX request, an URL to an external page, etc.).

To resolve the right handler for a given HTTP request, the RequestCycle uses a set of objects implementing the org.apache.wicket.request.IRequestMapper interface. The mapping interface defines the getCompatibilityScore(Request request) method which returns a score indicating how compatible the request mapper is for the current request. RequestCycle will choose the mapper with the highest score and it will call its mapRequest(Request request) method to get the proper handler for the given request. Once RequestCycle has resolved a request handler, it invokes its method respond(IRequestCycle requestCycle) to start request processing.

The following sequence diagram recaps how a request handler is resolved by the RequestCycle:

Developers can create additional implementations of IRequestMapper and add them to their application via the mount(IRequestMapper mapper) method of the WebApplication class. In paragraph 8.6 we will see how Wicket uses this method to add built-in mappers for mounted pages.

Generating URL with the urlFor and mapUrlFor methods

The RequestCycle is also responsible for generating the URL value (as CharSequence) for the following entities:

• a page class, via the urlFor(Class<C> pageClass, PageParameters parameters) method
• an IRequestHandler via the urlFor(IRequestHandler handler) method
• a ResourceReference via the urlFor(ResourceReference reference, PageParameters params) method (resource entities will be introduced in chapter 13).

The overloaded urlFor method from above also has a corresponding version that returns an instance of org.apache.wicket.request.Url instead of a CharSequence. This version has the prefix 'map' in its name (i.e. it has mapUrlFor as full name).

Method setResponsePage

The RequestCycle class contains the implementation of the setResponsePage method we use to redirect a user to a specific page (see paragraph 2.4). The namesake method of class org.apache.wicket. Component is just a convenience method that internally invokes the actual implementation on current request cycle:

public final void setResponsePage(final Page page) {
	getRequestCycle().setResponsePage(page);
}

RequestCycle's hook methods and listeners

The RequestCycle comes with some hook methods which can be overridden to perform custom actions when request handling reaches a specific stage. These methods are:

• onBeginRequest(): called when the RequestCycle is about to start handling the request.
• onEndRequest(): called when the RequestCycle has finished to handle the request
• onDetach(): called after the request handling has completed and the RequestCycle is about to be detached from its thread. The default implementation of this method invokes detach() on the current session (the Session class will be shortly discussed in paragraph 7.4).

Methods onBeforeRequest and onEndRequest can be used if we need to execute custom actions before and after business code is executed, such as opening a Hibernate/JPA session and closing it when code has terminated.

A more flexible way to interact with the request processing is to use the listener interface org.apache.wicket.request.cycle.IRequestCycleListener. In addition to the three methods already seen for RequestCycle, this interface offers further hooks into request processing:

• onBeginRequest(RequestCycle cycle): (see the description above)
• onEndRequest(RequestCycle cycle): (see the description above)
• onDetach(RequestCycle cycle): (see the description above)
• onRequestHandlerResolved(RequestCycle cycle, IRequestHandler handler): called when an IRequestHandler has been resolved.
• onRequestHandlerScheduled(RequestCycle cycle, IRequestHandler handler): called when an IRequestHandler has been scheduled for execution.
• onRequestHandlerExecuted(RequestCycle cycle, IRequestHandler handler): called when an IRequestHandler has been executed.
• onException(RequestCycle cycle, Exception ex): called when an exception has been thrown during request processing.
• onExceptionRequestHandlerResolved(RequestCycle rc, IRequestHandler rh, Exception ex): called when an IRequestHandler has been resolved and will be used to handle an exception.
• onUrlMapped(RequestCycle cycle, IRequestHandler handler, Url url): called when an URL has been generated for an IRequestHandler object.

To use the request cycle listeners we must add them to our application which in turn will pass them to the new RequestCycle's instances created with createRequestCycle method:

@Override
public void init() {
	super.init();
	IRequestCycleListener myListener;
	//listener initialization…
	getRequestCycleListeners().add(myListener)		
}

The getRequestCycleListeners method returns an instance of class org.apache.wicket. request.cycle.RequestCycleListenerCollection. This class is a sort of typed collection for IRequestCycleListener and it also implements the Composite pattern .

8.4 Session Class

In Wicket we use class org.apache.wicket.Session to handle session-relative informations such as client informations, session attributes, session-level cache (seen in paragraph 6.2.4), etc...

In addition, we know from paragraph 6.1 that Wicket creates a user session to store versions of stateful pages. Similarly to what happens with RequestCycle, the new Session's instances are generated by the Application class with the newSession(Request request, Response response) method. This method is not declared as final, hence it can be overridden if we need to use a custom implementation of the Session class.

By default if our custom application class is a subclass of WebApplication, method newSession will return an instance of class org.apache.wicket.protocol.http.WebSession. As we have mentioned talking about RequestCycle, also class Session provides a static get() method which returns the session associated to the current thread.

Session and listeners

Similar to the RequestCycle, class org.apache.wicket.Session also offers support for listener entities. With Session these entities must implement the callback interface org.apache.wicket. ISessionListener which exposes only the onCreated(Session session) method. As you might guess from its name, this method is called when a new session is created. Session listeners must be added to our application using a typed collection, just like we have done before with request cycle listeners:

@Override
public void init(){
	super.init();
	//listener initialization…
	ISessionListener myListener;
	//add a custom session listener
	getSessionListeners().add(myListener)
}

Handling session attributes

The Session class handles session attributes in much the same way as the standard interface javax.servlet.http.HttpSession. The following methods are provided to create, read and remove session attributes:

• setAttribute(String name, Serializable value): creates an attribute identified by the given name. If the session already contains an attribute with the same name, the new value will replace the existing one. The value must be a serializable object.
• getAttribute(String name): returns the value of the attribute identified by the given name, or null if the name does not correspond to any attribute.
• removeAttribute(String name): removes the attribute identified by the given name.

By default class WebSession will use the underlying http session to store attributes. Wicket will automatically add a prefix to the name of the attributes. This prefix is returned by the WebApplication's method getSessionAttributePrefix().

Accessing to the HTTP session

If for any reason we need to directly access to the underlying HttpSession object, we can retrieve it from the current request with the following code:

HttpSession session = ((ServletWebRequest)RequestCycle.get().getRequest()).getContainerRequest().getSession();

Using the raw session object might be necessary if we have to set a session attribute with a particular name without the prefix added by Wicket. Let's say for example that we are working with Tomcat as web server. One of the administrative tools provided by Tomcat is a page listing all the active user sessions of a given web application:

Tomcat allows us to set the values that will be displayed in columns “Guessed locale” and “Guessed User name”. One possible way to do this is to use session attributes named “Locale” and “userName” but we can't create them via Wicket's Session class because they would not have exactly the name required by Tomcat. Instead, we must use the raw HttpSession and set our attributes on it:

HttpSession session = ((ServletWebRequest)RequestCycle.get().getRequest()).getContainerRequest().getSession();
session.setAttribute("Locale", "ENGLISH");
session.setAttribute("userName", "Mr BadGuy");

Temporary and permanent sessions

Wicket doesn't need to store data into user session as long as the user visits only stateless pages. Nonetheless, even under these conditions, a temporary session object is created to process each request but it is discarded at the end of the current request. To know if the current session is temporary, we can use the isTemporary() method:

Session.get().isTemporary();

If a session is not temporary (i.e. it is permanent), it's identified by an unique id which can be read calling the getId() method. This value will be null if the session is temporary.

Although Wicket is able to automatically recognize when it needs to replace a temporary session with a permanent one, sometimes we may need to manually control this process to make our initially temporary session permanent.

To illustrate this possible scenario let's consider project BindSessionExample where we have a stateless home page which sets a session attribute inside its constructor and then it redirects the user to another page which displays with a label the session attribute previously created. The code of the two pages is as follows:

Home page:

public class HomePage extends WebPage {
    public HomePage(final PageParameters parameters) {
    	Session.get().setAttribute("username", "tommy");
	Session.get().bind();
	setResponsePage(DisplaySessionParameter.class);
    }   
}

Target page:

public class DisplaySessionParameter extends WebPage {
	public DisplaySessionParameter() {
	   super();
	   add(new Label("username", (String) Session.get().getAttribute("username")));
	}
}

Again, we kept page logic very simple to not over-bloat the example with unnecessary code. In the snippet above we have also bolded Session's bind() method which converts temporary session into a permanent one. If the home page has not invoked this method, the session with its attribute would have been discarded at the end of the request and the page DisplaySessionParameter would have displayed an empty value in its label.

Discarding session data

Once a user has finished using our web application, she must be able to log out and clean any session data. To be sure that a permanent session will be discarded at the end of the current request, class Session provides the invalidate() method. If we want to immediately invalidate a given session without waiting for the current request to complete, we can invoke the invalidateNow() method.

Remember that invalidateNow() will immediately remove any instance of components (and pages) from the session, meaning that once we have called this method we won't be able to work with them for the rest of the request process.

Storing arbitrary objects with metadata

JavaServer Pages Specification1 defines 4 scopes in which a page can create and access a variable. These scopes are:

• request: variables declared in this scope can be seen only by pages processing the same request. The lifespan of these variables is (at most) equal to the one of the related request. They are discarded when the full response has been generated or when the request is forwarded somewhere else.
• page: variables declared in this scope can be seen only by the page that has created them.
• session: variables in session scope can be created and accessed by every page used in the same session where they are defined.
• application: this is the widest scope. Variables declared in this scope can be used by any page of a given web application.

Although Wicket doesn't implement the JSP Specification (it is rather an alternative to it), it offers a feature called metadata which resembles scoped variables but is much more powerful. Metadata is quite similar to a Java Map in that it stores pairs of key-value objects where the key must be unique. In Wicket each of the following classes has its own metadata store: RequestCycle, Session, Application and Component.

The key used for metadata is an instance of class org.apache.wicket.MetaDataKey<T>. To put an arbitrary object into metadata we must use the setMetaData method which takes two parameters as input: the key used to store the value and the value itself. If we are using metadata with classes Session or Component, data object must be serializable because Wicket serializes both session and component instances. This constraint is not applied to metadata of classes Application and RequestCycle which can contain a generic object. In any case, the type of data object must be compatible with the type parameter T specified by the key.

To retrieve a previously inserted object we must use the getMetaData(MetaDataKey<T> key) method. In the following example we set a java.sql.Connection object in the application's metadata so it can be used by any page of the application:

Application class code:

public static MetaDataApp extends WebApplication{
	//Do some stuff…
	/**
	* Metadata key definition
	*/
	public static MetaDataKey<Connection> connectionKey = new MetaDataKey<Connection> (){};
	/**
	 * Application's initialization
	 */
	@Override
	public void init(){
		super.init();
		Connection connection;
		//connection initialization…
		setMetaData(connectionKey, connection);
		//Do some other stuff..
	}
}

Code to get the object from the metadata:

Connection connection = Application.get().getMetaData(MetaDataApp.connectionKey);

Since MetaDataKey<T> class is declared as abstract, we must implement it with a subclass or with an anonymous class (like we did in the example above).

8.5 Summary

In this chapter we had a look at how Wicket internally handles a web request. Even if most of the time we won't need to customize this internal process, knowing how it works is essential to use the framework at 100%.

Entities like Application and Session will come in handy again when we will tackle the topic of security in chapter 18.

9 Wicket Links and URL generation

However this component alone is not enough to build all possible kinds of links we may need in our pages. Therefore, Wicket offers other link components suited for those tasks which can not be accomplished with a basic Link.

Besides learning new link components, in this chapter we will also see how to customize the page URL generated by Wicket using the encoding facility provided by the framework and the page parameters that can be passed to a target page.

9.1 PageParameters

A common practice in web development is to pass data to a page using query string parameters (like ?paramName1=paramValu1&paramName2=paramValue2...). Wicket offers a more flexible and object oriented way to do this with models (we will see them in the next chapter). However, even if we are using Wicket, we still need to use query string parameters to exchange data with other Internet-based services. Consider for example a classic confirmation page which is linked inside an email to let users confirm important actions like password changing or the subscription to a mailing list. This kind of page usually expects to receive a query string parameter containing the id of the action to confirm.

Query string parameters can also be referred to as named parameters. In Wicket they are handled with class org.apache.wicket.request.mapper.parameter.PageParameters. Since named parameters are basically name-value pairs, PageParameters works in much the same way as Java Map providing two methods to create/modify a parameter (add(String name, Object value) and set(String name, Object value)), one method to remove an existing parameter (remove(String name)) and one to retrieve the value of a given parameter (get(String name)) . Here is a snippet to illustrate the usage of PageParameters:

PageParameters pageParameters = new PageParameters(); 
//add a couple of parameters
pageParameters.add("name", "John");
pageParameters.add("age", 28);
//retrieve the value of 'age' parameter
pageParameters.get("age");

Now that we have seen how to work with page parameters, let's see how to use them with our pages.

PageParameters and bookmarkable pages

Base class Page comes with a constructor which takes as input a PageParameters instance. If we use this superclass constructor in our page, PageParameters will be used to build the page URL and it can be retrieved at a later time with the Page's getPageParameters() method.

In the following example taken from the PageParametersExample project we have a home page with a link to a second page that uses a version of setResponsePage method that takes as input also a PageParameters to build the target page (named PageWithParameters). The code for the link and for the target page is the following:

Link code:

add(new Link("pageWithIndexParam") {
	@Override
	public void onClick() {
		PageParameters pageParameters = new PageParameters();
		pageParameters.add("foo", "foo");
		pageParameters.add("bar", "bar");
		setResponsePage(PageWithParameters.class, pageParameters);
	}
});

Target page code:

public class PageWithParameters extends WebPage {
	//Override superclass constructor
	public PageWithParameters(PageParameters parameters) {
		super(parameters);
	}
 }

The code is quite straightforward and it’s more interesting to look at the URL generated for the target page:

<app root>/PageParametersExample/wicket/bookmarkable/org.wicketTutorial.PageWithParameters?foo=foo&bar=bar

At first glance the URL above could seem a little weird, except for the last part which contains the two named parameters used to build the target page.

The reason for this “strange” URL is that, as we explained in paragraph 6.2.5, when a page is instantiated using a constructor with no argument or using a constructor that accepts only a PageParameters, Wicket will try to generate a static URL for it, with no session-relative informations. This kind of URL is called bookmarkable because it can be saved by the users as a bookmark and accessed at a later time.

A bookmarkable URL is composed by a fixed prefix (which by default is bookmarkable) and the qualified name of the page class (org.wicketTutorial.PageWithParameters in our example). Segment wicket is another fixed prefix added by default during URL generation. In paragraph 8.6.4 we will see how to customize fixed prefixes with a custom implementation of IMapperContext interface.

Indexed parameters

Besides named parameters, Wicket also supports indexed parameters. These kinds of parameters are rendered as URL segments placed before named parameters. Let's consider for example the following URL:

<application path>/foo/bar?1&baz=baz

The URL above contains two indexed parameters (foo and bar) and a query string consisting of the page id and a named parameter (baz). Just like named parameters also indexed parameters are handled by the PageParameters class. The methods provided by PageParameters for indexed parameters are set(int index, Object object) (to add/modify a parameter), remove(int index)(to remove a parameter) and get(int index) (to read a parameter).

As their name suggests, indexed parameters are identified by a numeric index and they are rendered following the order in which they have been added to the PageParameters. The following is an example of indexed parameters:

PageParameters pageParameters = new PageParameters(); 
//add a couple of parameters
pageParameters.set(0, "foo");
pageParameters.set(1, "bar");
//retrieve the value of the second parameter ("bar")
pageParameters.get(1);

Project PageParametersExample comes also with a link to a page with both indexed parameters and a named parameter:

add(new Link("pageWithNamedIndexParam") {
	@Override
 	public void onClick() {
		PageParameters pageParameters = new PageParameters();
		pageParameters.set(0, "foo");
		pageParameters.set(1, "bar");
		pageParameters.add("baz", "baz");
		setResponsePage(PageWithParameters.class, pageParameters);
	}
});

The URL generated for the linked page (PageWithParameters) is the one seen at the beginning of the paragraph.

9.2 Bookmarkable links

A link to a bookmarkable page can be built with the link component org.apache.wicket.markup.html.link.BookmarkablePageLink:

BookmarkablePageLink bpl=new BookmarkablePageLink(PageWithParameters.class, pageParameters);

The specific purpose of this component is to provide an anchor to a bookmarkable page, hence we don't have to implement any abstract method like we do with Link component.

9.3 Automatically creating bookmarkable links with tag

Bookmarkable pages can be linked directly inside markup files without writing any Java code. Using <wicket:link> tag we ask Wicket to automatically add bookmarkable links for the anchors wrapped inside it. Here is an example of usage of <wicket:link> tag taken from the home page of the project BookmarkablePageAutoLink:

<!DOCTYPE html>
<html xmlns:wicket="http://wicket.apache.org">
	<head>
		<meta charset="utf-8" />
		<title>Apache Wicket Quickstart</title>
	</head>
	<body>		
	   <div id="bd">
	      <wicket:link>
			<a href="HomePage.html">HomePage</a><br/>
			<a href="anotherPackage/SubPackagePage.html">SubPackagePage</a>	
	      </wicket:link>
	   </div>		
	</body>
</html>

The key part of the markup above is the href attribute which must contain the package-relative path to a page. The home page is inside package org.wicketTutorial which in turns contains the sub package anotherPackage. This package hierarchy is reflected by the href attributes: in the first anchor we have a link to the home page itself while the second anchor points to page SubPackagePage which is placed into sub package anotherPackage. Absolute paths are supported as well and we can use them if we want to specify the full package of a given page. For example the link to SubPackagePage could have been written in the following (more verbose) way:

<a href="/org/wicketTutorial/anotherPackage/SubPackagePage.html">SubPackagePage</a>

If we take a look also at the markup of SubPackagePage we can see that it contains a link to the home page which uses the parent directory selector (relative path):

<!DOCTYPE html>
<html xmlns:wicket="http://wicket.apache.org">
	<head>
		<meta charset="utf-8" />
		<title>Apache Wicket Quickstart</title>
	</head>
	<body>		
		<div id="bd">
			<wicket:link>
				<a href="../HomePage.html">HomePage</a><br/>
				<a href="SubPackagePage.html">SubPackagePage</a>			
			</wicket:link>
		</div>		
	</body>
</html>

Please note that any link to the current page (aka self link) is disabled. For example in the home page the self link is rendered like this:

<span><em>HomePage</em></span>

The markup used to render disabled links can be customized using the markup settings (interface IMarkupSettings) available in the application class:

@Override
public void init()
{
	super.init();
	//wrap disabled links with <b> tag
	getMarkupSettings().setDefaultBeforeDisabledLink("<b>");
	getMarkupSettings().setDefaultAfterDisabledLink("</b>");		
}

The purpose of <wicket:link> tag is not limited to just simplifying the usage of bookmarkable pages. As we we will see in chapter 13, this tag can also be adopted to manage web resources like pictures, CSS files, JavaScript files and so on.

9.4 External links

Since Wicket uses plain HTML markup files as templates, we can place an anchor to an external page directly inside the markup file. When we need to dynamically generate external anchors, we can use link component org.apache.wicket.markup.html.link.ExternalLink. In order to build an external link we must specify the value of the href attribute using a model or a plain string. In the next snippet, given an instance of Person, we generate a Google search query for its full name:

Html:

<a wicket:id="externalSite">Search me on Google!</a>

Java code:

Person person = new Person("John", "Smith"); 
String fullName = person.getFullName();
//Space characters must be replaced by character '+'
String googleQuery = "http://www.google.com/search?q=" + fullName.replace(" ", "+");
add(new ExternalLink("externalSite", googleQuery));

Generated anchor:

<a href="http://www.google.com/search?q=John+Smith">Search me on Google!</a>

If we need to specify a dynamic value for the text inside the anchor, we can pass it as an additional constructor parameter:

Html:

<a wicket:id="externalSite">Label goes here...</a>

Java code:

Person person = new Person("John", "Smith"); 
String fullName = person.getFullName();
String googleQuery = "http://www.google.com/search?q=" + fullName.replace(" ", "+");
String linkLabel = "Search '" + fullName + "' on Google.";
add(new ExternalLink("externalSite", googleQuery, linkLabel));

Generated anchor:

<a href="http://www.google.com/search?q=John+Smith">Search 'John Smith' on Google.</a>

9.5 Stateless links

Component Link has a stateful nature, hence it cannot be used with stateless pages. To use links with these kinds of pages Wicket provides the convenience org.apache.wicket.markup.html. link.StatelessLink component which is basically a subtype of Link with the stateless hint set to true.

Please keep in mind that Wicket generates a new instance of a stateless page also to serve stateless links, so the code inside the onClick() method can not depend on instance variables. To illustrate this potential issue let's consider the following code (from the project StatelessPage) where the value of the variable index is used inside onclick():

public class StatelessPage extends WebPage {
	private int index = 0;
	public StatelessPage(PageParameters parameters) {
		super(parameters);
	}
	@Override
	protected void onInitialize() {
		super.onInitialize();
		setStatelessHint(true);
		add(new StatelessLink("statelessLink") {
			@Override
			public void onClick() {
				//It will always print zero
				System.out.println(index++);
			}
		});
	}	
}

The printed value will always be zero because a new instance of the page is used every time the user clicks on the statelessLink link.

9.6 Generating structured and clear URLs

Having structured URLs in our site is a basic requirement if we want to build an efficient SEO1 strategy, but it also contributes to improve user experience with more intuitive URLs. Wicket provides two different ways to control URL generation. The first (and simplest) is to “mount” one or more pages to an arbitrary path, while a more powerful technique is to use custom implementations of IMapperContext and IPageParametersEncoder interfaces. In the next paragraphs we will learn both of these two techniques.

Mounting a single page

With Wicket we can mount a page to a given path in much the same way as we map a servlet filer to a desired path inside file web.xml (see page 9). Using mountPage(String path, Class <T> pageClass) method of the WepApplication class we tell Wicket to respond with a new instance of pageClass whenever a user navigates to the given path. In the application class of the project MountedPagesExample we mount MountedPage to the "/pageMount" path:

@Override
public void init()
{
	super.init();
	mountPage("/pageMount", MountedPage.class);
	//Other initialization code…
}

The path provided to mountPage will be used to generate the URL for any page of the specified class:

//it will return "/pageMount"
RequestCycle.get().urlFor(MountedPage.class);

Under the hood the mountPage method mounts an instance of the request mapper org.apache. wicket.request.mapper.MountedMapper configured for the given path:

public final <T extends Page> void mountPage(final String path,final Class<T> pageClass) {
	mount(new MountedMapper(path, pageClass));
}

Request mappers and the Application's method mount have been introduced in the previous chapter (paragraph 7.3.1).

Using parameter placeholders with mounted pages

The path specified for mounted pages can contain dynamic segments which are populated with the values of the named parameters used to build the page. These segments are declared using special segments called parameter placeholders. Consider the path used in the following example:

mountPage("/pageMount/${foo}/otherSegm", MountedPageWithPlaceholder.class);

The path used above is composed by three segments: the first and the last are fixed while the second will be replaced by the value of the named parameter foo that must be provided when the page MountedPageWithPlaceholder is instantiated:

Java code:

PageParameters pageParameters = new PageParameters();
pageParameters.add("foo", "foo");
setResponsePage(MountedPageWithPlaceholder.class, pageParameters)

Generated URL:

<Application path>/pageMount/foo/otherSegm

On the contrary if we manually insert an URL like '<web app path>/pageMount/bar/otherSegm', we can read value 'bar' retrieving the named parameter foo inside our page.

Place holders can be declared as optional using the '#' character in place of '$':

mountPage("/pageMount/#{foo}/otherSegm", MountedPageOptionalPlaceholder.class);

If the named parameter for an optional placeholder is missing, the corresponding segment is removed from the final URL:

Java code:

PageParameters pageParameters = new PageParameters();
setResponsePage(MountedPageWithPlaceholder.class, pageParameters);

Generated URL:

<Application path>/pageMount/otherSegm

Mounting a package

In addition to mounting a single page, Wicket allows to mount all of the pages inside a package to a given path. Method mountPackage(String path, Class<T> pageClass) of class WepApplication will mount every page inside pageClass's package to the specified path.

The resulting URL for package-mounted pages will have the following structure:

<Application path>/mountedPath/<PageClassName>[optional query string]

For example in the MountedPagesExample project we have mounted all pages inside the subpackage org.tutorialWicket.subPackage with this line of code:

mountPackage("/mountPackage", StatefulPackageMount.class);

StatefulPackageMount is one of the pages placed into the desired package and its URL will be:

<Application path>/mountPackage/StatefulPackageMount?1

Similarly to what is done by the mountPage method, the implementation of the mountPackage method mounts an instance of org.apache.wicket.request.mapper.PackageMapper to the given path.

Providing custom mapper context to request mappers

Interface org.apache.wicket.request.mapper.IMapperContext is used by request mappers to create new page instances and to retrieve static URL segments used to build and parse page URLs. Here is the list of these segments:

• Namespace: it's the first URL segment of non-mounted pages. By default its value is wicket.
• Identifier for non-bookmarkable URLs: it's the segment that identifies non bookmarkable pages. By default its value is page.
• Identifier for bookmarkable URLs: it's the segment that identifies bookmarkable pages. By default its value is bookmarkable (as we have seen before in paragraph 8.1.1).
• Identifier for resources: it's the segment that identifies Wicket resources. Its default value is resources. The topic of resource management will be covered in chapter 13.

IMapperContext provides a getter method for any segment listed above. By default Wicket uses class org.apache.wicket.DefaultMapperContext as mapper context.

Project CustomMapperContext is an example of customization of mapper context where we use index as identifier for non-bookmarkable pages and staticURL as identifier for bookmarkable pages. In this project, instead of implementing our mapper context from scratch, we used DefaultMapperContext as base class overriding just the two methods we need to achieve the desired result (getBookmarkableIdentifier() and getPageIdentifier()). The final implementation is the following:

public class CustomMapperContext extends DefaultMapperContext{
	@Override
	public String getBookmarkableIdentifier() {
		return "staticURL";
	}
	@Override
	public String getPageIdentifier() {
		return "index";
	}
}

Now to use a custom mapper context in our application we must override the newMapperContext() method declared in the Application class and make it return our custom implementation of IMapperContext:

@Override
protected IMapperContext newMapperContext() {
	return new CustomMapperContext();
}

Controlling how page parameters are encoded with IPageParametersEncoder

Some request mappers (like MountedMapper and PackageMapper) can delegate page parameters encoding/decoding to interface org.apache.wicket.request.mapper.parameter.IPage ParametersEncoder. This entity exposes two methods: encodePageParameters() and decodePageParameters(): the first one is invoked to encode page parameters into an URL while the second one extracts parameters from the URL.

Wicket comes with a built-in implementation of this interface which encodes named page parameters as URL segments using the following patter: /paramName1/paramValue1/paramName2/param Value2...

This built-in encoder is org.apache.wicket.request.mapper.parameter.UrlPathPage ParametersEncoder class. In the PageParametersEncoderExample project we have manually mounted a MountedMapper that takes as input also an UrlPathPageParametersEncoder:

@Override
public void init() {
	super.init();
	mount(new MountedMapper("/mountedPath", MountedPage.class, new UrlPathPageParametersEncoder()));
}

The home page of the project contains just a link to the MountedPage web page. The code of the link and the resulting page URL are:

Link code:

add(new Link("mountedPage") {
	@Override
	public void onClick() {
		PageParameters pageParameters = new PageParameters();
		pageParameters.add("foo", "foo");
		pageParameters.add("bar", "bar");
		setResponsePage(MountedPage.class, pageParameters);
	}
});

Generated URL:

<Application path>/mountedPath/foo/foo/bar/bar?1

Encrypting page URLs

Sometimes URLs are a double–edged sword for our site because they can expose too many details about the internal structure of our web application and malicious users could exploit them to perform a cross-site request forgery .

To avoid this kind of security threat we can use the CryptoMapper request mapper which wraps an existing mapper and encrypts the original URL producing a single encrypted segment:

Typically, CryptoMapper is registered into a Wicket application as the root request mapper wrapping the default one:

@Override
public void init() {
	super.init();
	setRootRequestMapper(new CryptoMapper(getRootRequestMapper(), this)); 
	//pages and resources must be mounted after we have set CryptoMapper
	mountPage("/foo/", HomePage.class);

As pointed out in the code above, pages and resources must be mounted after having set CryptoMapper as root mapper, otherwise the mounted paths will not work.

9.7 Summary

Links and URLs are not trivial topics as they may seem and in Wicket they are strictly interconnected. Developers must choose the right trade-off between producing structured URLs and avoiding to make them verbose and vulnerable.

In this chapter we have explored the tools provided by Wicket to control how URLs are generated. We have started with static URLs for bookmarkable pages and we have seen how to pass parameters to target pages with PageParameters. In the second part of the chapter we focused on mounting pages to a specific path and on controlling how parameters are encoded by Wicket. Finally, we have also seen how to encrypt URLs to prevent security vulnerabilities.

10 Wicket models and forms

10.1 What is a model?

Model is essentially a facade interface which allows components to access and modify their data without knowing any detail about how they are managed or persisted. Every component has at most one related model, while a model can be shared among different components. In Wicket a model is any implementation of the interface org.apache.wicket.model.IModel:

The IModel interface defines just the methods needed to get and set a data object (getObject() and setObject()), decoupling components from concrete details about the persistence strategy adopted for data. In addition, the level of indirection introduced by models allows access data object only when it is really needed (for example during the rendering phase) and not earlier when it may not be ready to be used.

Any component can get/set its model as well as its data object using the 4 public shortcut methods listed in the class diagram above. The two methods onModelChanged() and onModelChanging() are triggered by Wicket each time a model is modified: the first one is called after the model has been changed, the second one just before the change occurs. In the examples seen so far we have worked with Label component using its constructor which takes as input two string parameters, the component id and the text to display:

add(new Label("helloMessage", "Hello WicketWorld!"));

This constructor internally builds a model which wraps the second string parameter. That's why we didn't mention label model in the previous examples. Here is the code of this constructor:

public Label(final String id, String label) {
	this(id, new Model<String>(label));
}

Class org.apache.wicket.model.Model is a basic implementation of IModel. It can wrap any object that implements the interface java.io.Serializable. The reason of this constraint over data object is that this model is stored in the web session, and we know from chapter 6 that data are stored into session using serialization.

In general, Wicket models support a detaching capability that allows us to work also with non-serializable objects as data model. We will see the detaching mechanism later in this chapter.

Just like any other Wicket components, Label provides a constructor that takes as input the component id and the model to use with the component. Using this constructor the previous example becomes:

add(new Label("helloMessage", new Model<String>("Hello WicketWorld!")));

The Model class comes with a bunch of factory methods that makes it easier to build new model instances. For example the of(T object) method creates a new instance of Model which wraps any Object instance inside it. So instead of writing

new Model<String>("Hello WicketWorld!")

we can write

Model.of("Hello WicketWorld!")

If the data object is a List, a Map or a Set we can use similar methods called ofList, ofMap and ofSet. From now on we will use these factory methods in our examples.

It's quite clear that if our Label must display a static text it doesn't make much sense to build a model by hand like we did in the last code example. However is not unusual to have a Label that must display a dynamic value, like the input provided by a user or a value read from a database. Wicket models are designed to solve these kinds of problems.

Let's say we need a label to display the current time stamp each time a page is rendered. We can implement a custom model which returns a new Date instance when the getObject() method is called:

IModel timeStampModel = new Model<String>(){
	@Override
	public String getObject() {
		return new Date().toString();
	}
};
add(new Label("timeStamp", timeStampModel));

Even if sometimes writing a custom model could be a good choice to solve a specific problem, Wicket already provides a set of IModel implementations which should fit most of our needs. In the next paragraph we will see a couple of models that allow us to easily integrate JavaBeans with our web applications and in particular with our forms.

By default class Component escapes HTML sensitive characters (like '<', '>' or '&') from the textual representation of its model object. The term 'escape' means that these characters will be replaced with their corresponding HTML entity (for example '<' becomes '&lt; '). This is done for security reasons as a malicious user could attempt to inject markup or JavaScript into our pages. If we want to display the raw content stored inside a model, we can tell the Component class not to escape characters by calling the setEscape ModelStrings(false) method.

10.2 Models and JavaBeans

One of the main goals of Wicket is to use JavaBeans and POJO as data model, overcoming the impedance mismatch between web technologies and OO paradigm. In order to make this task as easy as possible, Wicket offers two special model classes: org.apache.wicket.model.PropertyModel and org.apache.wicket.model.CompoundPropertyModel. We will see how to use them in the next two examples, using the following JavaBean as the data object:

public class Person implements Serializable {
	private String name;
	private String surname;
	private String address;
	private String email;
	private String passportCode;
	private Person spouse;
	private List<Person> children;
	public Person(String name, String surname) {
		this.name = name;
		this.surname = surname;
	}
	public String getFullName(){
   		return name + " " + surname;
	}
	/* 	 
	 * Getters and setters for private fields
     */
}

PropertyModel

Let's say we want to display the name field of a Person instance with a label. We could, of course, use the Model class like we did in the previous example, obtaining something like this:

Person person = new Person();		
//load person's data...
Label label = new Label("name", new Model(person.getName()));

However this solution has a huge drawback: the text displayed by the label will be static and if we change the value of the field, the label won't update its content. Instead, to always display the current value of a class field, we should use the org.apache.wicket.model.PropertyModel model class:

Person person = new Person();		
//load person's data...
Label label = new Label("name", new PropertyModel(person, "name"));

PropertyModel has just one constructor with two parameters: the model object (person in our example) and the name of the property we want to read/write ("name" in our example). This last parameter is called property expression. Internally, methods getObject/setObject use property expression to get/set property's value. To resolve class properties PropertyModel uses class org.apache.wicket.util.lang.Property Resolver which can access any kind of property, private fields included.

Just like the Java language, property expressions support dotted notation to select sub properties. So if we want to display the name of the Person's spouse we can write:

Label label = new Label("spouseName", new PropertyModel(person, "spouse.name"));

PropertyModel is null-safe, which means we don't have to worry if property expression includes a null value in its path. If such a value is encountered, an empty string will be returned.

If property is an array or a List, we can specify an index after its name. For example, to display the name of the first child of a Person we can write the following property expression:

Label label = new Label("firstChildName", new PropertyModel(person, "children.0.name"));

Indexes and map keys can be also specified using squared brackets like “children0.name” or “mapFieldkey.subfield”.

CompoundPropertyModel and model inheritance

Class org.apache.wicket.model.CompoundPropertyModel is a particular kind of model which is usually used in conjunction with another Wicket feature called model inheritance. With this feature, when a component needs to use a model but none has been assigned to it, it will search through the whole container hierarchy for a parent with an inheritable model. Inheritable models are those which implement interface org.apache.wicket.model.IComponentInheritedModel and CompoundPropertyModel is one of them. Once a CompoundPropertyModel has been inherited by a component, it will behave just like a PropertyModel using the id of the component as property expression. As a consequence, to make the most of CompoundPropertyModel we must assign it to one of the containers of a given component, rather than directly to the component itself.

For example if we use CompoundPropertyModel with the previous example (display spouse's name), the code would become like this:

//set CompoundPropertyModel as model for the container of the label
setDefaultModel(new CompoundPropertyModel(person));
Label label = new Label("spouse.name");
add(label);

Note that now the id of the label is equal to the property expression previously used with PropertyModel. Now as a further example let's say we want to extend the code above to display all of the main informations of a person (name, surname, address and email). All we have to do is to add one label for every additional information using the relative property expression as component id:

//Create a person named 'John Smith'
Person person = new Person("John", "Smith");
setDefaultModel(new CompoundPropertyModel(person));
add(new Label("name"));
add(new Label("surname"));
add(new Label("address"));
add(new Label("email"));
add(new Label("spouse.name"));

CompoundPropertyModel can save us a lot of boring coding if we choose the id of components according to properties name. However it's also possible to use this type of model even if the id of a component does not correspond to a valid property expression. The method bind(String property) allows to create a property model from a given CompoundPropertyModel using the provided parameter as property expression. For example if we want to display the spouse's name in a label having "xyz" as id, we can write the following code:

//Create a person named 'John Smith'
Person person = new Person("John", "Smith");
CompoundPropertyModel compoundModel;
setDefaultModel(compoundModel = new CompoundPropertyModel(person));
add(new Label("xyz", compoundModel.bind("spouse.name")));

CompoundPropertyModel are particularly useful when used in combination with Wicket forms, as we will see in the next paragraph.

Model is referred to as static model because the result of its method getObject is fixed an it is not dynamically evaluated each time the method is called. In contrast, models like PropertyModel and CompoundProperty Model are called dynamic models.

10.3 Wicket forms

Web applications use HTML forms to collect user input and send it to the server. Wicket provides org.apache.wicket.markup.html.form.Form class to handle web forms. This component must be bound to <form> tag. The following snippet shows how to create a very basic Wicket form in a page:

Html:

<form wicket:id="form">
    <input type="submit" value="submit"/>
</form>

Java code:

Form form = new Form("form"){
    @Override
    protected void onSubmit() {
    	System.out.println("Form submitted.");
    }
};
add(form);

Method onSubmit is called whenever a form has been submitted and it can be overridden to perform custom actions. Please note that a Wicket form can be submitted using a standard HTML submit button which is not mapped to any component (i.e. it does not have a wicket:id attribute). In the next chapter we will continue to explore Wicket forms and we will see how to submit forms using special components which implement interface org.apache.wicket.markup.html.form.IFormSubmitter.

Form and models

A form should contain some input fields (like text fields, check boxes, radio buttons, drop-down lists, text areas, etc.) to interact with users. Wicket provides an abstraction for all these kinds of elements with component org.apache.wicket.markup.html.form.FormComponent:

The purpose of FormComponent is to store the corresponding user input into its model when the form is submitted. The form is responsible for mapping input values to the corresponding components, avoiding us the burden of manually synchronizing models with input fields and vice versa.

Login form

As first example of interaction between the form and its models, we will build a classic login form which asks for username and password (project LoginForm).

The topic of security will be discussed later in chapter 18. The following form is for example purposes only and is not suited for a real application. If you need to use a login form you should consider to use component org.apache.wicket.authroles.authentication.panel.SignInPanel shipped with Wicket.

This form needs two text fields, one of which must be a password field. We should also use a label to display the result of login process1. For the sake of simplicity, the login logic is all inside onSubmit and is quite trivial.

The following is a possible implementation of our form:

public class LoginForm extends Form {
	private TextField usernameField;
	private PasswordTextField passwordField;
	private Label loginStatus;
	public LoginForm(String id) {
		super(id);
		usernameField = new TextField("username", Model.of(""));
		passwordField = new PasswordTextField("password", Model.of(""));			
		loginStatus = new Label("loginStatus", Model.of(""));
		add(usernameField);
		add(passwordField);
		add(loginStatus);
	}
	public final void onSubmit() {
		String username = (String)usernameField.getDefaultModelObject();
		String password = (String)passwordField.getDefaultModelObject();
		if(username.equals("test") && password.equals("test"))
			loginStatus.setDefaultModelObject("Congratulations!");
		else
			loginStatus.setDefaultModelObject("Wrong username or password!");			
	}
}

Inside form's constructor we build the three components used in the form and we assign them a model containing an empty string:

usernameField = new TextField("username", Model.of(""));
passwordField = new PasswordTextField("password", Model.of(""));			
loginStatus = new Label("loginStatus", Model.of(""));

If we don't provide a model to a form component, we will get the following exception on form submission:

java.lang.IllegalStateException: Attempt to set model object on null model of component:

Component TextField corresponds to the standard text field, without any particular behavior or restriction on the allowed values. We must bind this component to the <input> tag with the attribute type set to "text". PasswordTextField is a subtype of TextFiled and it must be used with an <input> tag with the attribute type set to"password". For security reasons component PasswordTextField cleans its value at each request, so it wil be always empty after the form has been rendered. By default PasswordTextField fields are required, meaning that if we left them empty, the form won't be submitted (i.e. onSubmit won't be called). Class FormComponent provides method setRequired(boolean required) to change this behavior. Inside onSubmit, to get/set model objects we have used shortcut methods setDefaultModelObject and getDefaultModelObject. Both methods are defined in class Component (see class diagram from Illustration 9.1).

The following are the possible markup and code for the login page:

Html:

<html>
	<head>
  		<title>Login page</title>
	</head>
	<body>
		<form id="loginForm" method="get" wicket:id="loginForm">
  			<fieldset>
    			<legend style="color: #F90">Login</legend>
    				<p wicket:id="loginStatus"></p>
    				<span>Username: </span><input wicket:id="username" type="text" id="username" /><br/>
    				<span>Password: </span><input wicket:id="password" type="password" id="password" />
    				<p>
    					<input type="submit" name="Login" value="Login"/>
    				</p>
  	   	    </fieldset>
		</form>
	</body>
</html>

Java code:

public class HomePage extends WebPage {
   public HomePage(final PageParameters parameters) {
		super(parameters);
    	add(new LoginForm("loginForm"));
    }
}

The example shows how Wicket form components can be used to store user input inside their model. However we can dramatically improve the form code using CompoundPropertyModel and its ability to access the properties of its model object. The revisited code is the following (the LoginFormRevisited project):

public class LoginForm extends Form{
		private String username;
		private String password;
		private String loginStatus;
		public LoginForm(String id) {
			super(id);			
			setDefaultModel(new CompoundPropertyModel(this));
			add(new TextField("username"));
			add(new PasswordTextField("password"));
			add(new Label("loginStatus"));
		}
		public final void onSubmit() {			
			if(username.equals("test") && password.equals("test"))
				loginStatus = "Congratulations!";
			else
				loginStatus = "Wrong username or password !";			
		}
	}

In this version the form itself is used as model object for its CompoundPropertyModel. This allows children components to have direct access to form fields and use them as backing objects, without explicitly creating a model for themselves.

Keep in mind that when CompoundPropertyModel is inherited, it does not consider the ids of traversed containers for the final property expression, but it will always use the id of the visited child. To understand this potential pitfall, let's consider the following initialization code of a page:

//Create a person named 'John Smith'
Person person = new Person("John", "Smith");
//Create a person named 'Jill Smith'
Person spouse = new Person("Jill", "Smith");
//Set Jill as John's spouse
person.setSpouse(spouse);
setDefaultModel(new CompoundPropertyModel(person));
WebMarkupContainer spouse = new WebMarkupContainer("spouse");
Label name;
spouse.add(name = new Label("name"));
add(spouse);

The value displayed by label "name" will be "John" and not the spouse's name "Jill" as you may expect. In this example the label doesn't own a model, so it must search up its container hierarchy for an inheritable model. However, its container (WebMarkup Container with id 'spouse') doesn't own a model, hence the request for a model is forwarded to the parent container, which in this case is the page. In the end the label inherits CompoundPropertyModel from page but only its own id is used for the property expression. The containers in between are never taken into account for the final property expression.

10.4 Component DropDownChoice

Class org.apache.wicket.markup.html.form.DropDownChoice is the form component needed to display a list of possible options as a drop-down list where users can select one of the proposed options. This component must be used with <select> tag:

Html:

<form wicket:id="form">
	Select a fruit: <select wicket:id="fruits"></select>
<div><input type="submit" value="submit"/></div>
</form>

Java code:

List<String> fruits = Arrays.asList("apple", "strawberry", "watermelon"); 
form.add(new DropDownChoice<String>("fruits", new Model(), fruits));

Screenshot of generated page:

In addition to the component id, in order to build a DropDownChoice we need to provide to its constructor two further parameters:

• a model containing the current selected item. This parameter is not required if we are going to inherit a CompoundPropertyModel for this component.
• a list of options to display which can be supplied as a model or as a regular java.util.List.

In the example above the possible options are provided as a list of String objects. Now let's take a look at the markup generated for them:

<select name="fruits" wicket:id="fruits">
	<option value="" selected="selected">Choose One</option>
	<option value="0">apple</option>
	<option value="1">strawberry</option>
	<option value="2">watermelon</option>
</select>

The first option is a placeholder item corresponding to a null model value. By default DropDownChoice cannot have a null value so users are forced to select a not-null option. If we want to change this behavior we can set the nullValid flag to true via the setNullValid method. Please note that the placeholder text (“Chose one”) can be localized, as we will see in chapter 12. The other options are identified by the attribute value. By default the value of this attribute is the index of the single option inside the provided list of choices, while the text displayed to the user is obtained by calling toString()on the choice object. This default behavior works fine as long as our options are simple objects like strings, but when we move to more complex objects we may need to implement a more sophisticated algorithm to generate the value to use as the option id and the one to display to user. Wicket has solved this problem with org.apache.wicket.markup.html.form.IChoiceRender interface. This interface defines method getDisplayValue(T object) that is called to generate the value to display for the given choice object, and method getIdValue(T object, int index) that is called to generate the option id. The built-in implementation of this interface is class org.apache.wicket.markup.html.Form. ChoiceRenderer which renders the two values using property expressions.

In the following code we want to show a list of Person objects using their full name as value to display and using their passport code as option id:

Java code:

List<Person> persons; 
//Initialize the list of persons here…
ChoiceRenderer personRenderer = new ChoiceRenderer("fullName", "passportCode");
form.add(new DropDownChoice<String>("persons", new Model<Person>(), persons, personRenderer));

The choice renderer can be assigned to the DropDownChoice using one of its constructor that accepts this type of parameter (like we did in the example above) or after its creation invoking setChoiceRenderer method.

10.5 Model chaining

Models that implement the interface org.apache.wicket.model.IChainingModel can be used to build a chain of models. These kinds of models are able to recognize whether their model object is itself an implementation of IModel and if so, they will call getObject on the wrapped model and the returned value will be the actual model object. In this way we can combine the action of an arbitrary number of models, making exactly a chain of models. Chaining models allows to combine different data persistence strategies, similarly to what we do with chains of I/O streams. To see model chaining in action we will build a page that implements the List/Detail View pattern, where we have a drop-down list of Person objects and a form to display and edit the data of the current selected Person.

The example page will look like this:

What we want to do in this example is to chain the model of the DropDownChoice (which contains the selected Person) with the model of the Form. In this way the Form will work with the selected Person as backing object. The DropDownChoice component can be configured to automatically update its model each time we change the selected item on the client side. All we have to do is to override method wantOn SelectionChangedNotifications to make it return true. In practice, when this method returns true, DropDownChoice will submit its value every time JavaScript event onChange occurs, and its model will be consequently updated. To leverage this functionality, DropDownChoice doesn't need to be inside a form.

The following is the resulting markup of the example page:

…
<body>
	List of persons <select wicket:id="persons"></select> <br/>
	<br/>
	<form wicket:id="form">		
		<div style="display: table;">
			<div style="display: table-row;">
				<div style="display: table-cell;">Name: </div>
				<div style="display: table-cell;">
					<input type="text" wicket:id="name"/> 
				</div>	
			</div>
			<div style="display: table-row;">
				<div style="display: table-cell;">Surname: </div>
				<div style="display: table-cell;">
									<input type="text" wicket:id="surname"/>
								</div>	
							</div>
							<div style="display: table-row;">
								<div style="display: table-cell;">Address: </div>
								<div style="display: table-cell;">
									<input type="text" wicket:id="address"/>
								</div>	
							</div>
							<div style="display: table-row;">
								<div style="display: table-cell;">Email: </div>
								<div style="display: table-cell;">
									<input type="text" wicket:id="email"/>
								</div>
							</div>
						</div>	
						<input type="submit" value="Save"/>
					</form>
				</body>

The initialization code for DropDownChoice is the following:

Model<Person> listModel = new Model<Person>();
ChoiceRenderer<Person> personRender = new ChoiceRenderer<Person>("fullName");
personsList = new DropDownChoice<Person>("persons", listModel, loadPersons(), personRender){
		@Override
		protected boolean wantOnSelectionChangedNotifications() {
			return true;
		}
};

As choice render we have used the basic implementation provided with the org.apache.wicket .markup.html.form.ChoiceRenderer class that we have seen in the previous paragraph. loadPersons() is just an utility method which generates a list of Person instances. The model for DropDownChoice is a simple instance of the Model class.

Here is the whole code of the page (except for the loadPersons() method):

public class PersonListDetails extends WebPage {
  private Form form;
  private DropDownChoice<Person> personsList;
  public PersonListDetails(){
    Model<Person> listModel = new Model<Person>();
    ChoiceRenderer<Person> personRender = new ChoiceRenderer<Person>("fullName");
    personsList = new DropDownChoice<Person>("persons", listModel, loadPersons(),
                                                         personRender){
      @Override
      protected boolean wantOnSelectionChangedNotifications() {
        return true;
      }
	    };
	    add(personsList);
	    form = new Form("form", new CompoundPropertyModel<Person>(listModel));    
	    form.add(new TextField("name"));
	    form.add(new TextField("surname"));
	    form.add(new TextField("address"));
	    form.add(new TextField("email"));
	    add(form);
	  }
	       //loadPersons()
	       //…
	}

The two models work together as a pipeline where the output of method getObject of Model is the model object of CompoundPropertyModel. As we have seen, model chaining allows us to combine the actions of two or more models without creating new custom implementations.

10.6 Detachable models

In chapter 6 we have seen how Wicket uses serialization to store page instances. When an object is serialized, all its referenced objects are recursively serialized. For a page this means that all its children components, their related models as well as the model objects inside them will be serialized. For model objects this could be a serious issue for (at least) two main reasons:

1. The model object could be a very large instance, hence serialization would become very expensive in terms of time and memory.
2. We simply may not be able to use a serializable object as model object. In paragraphs 1.4 and 9.2 we stated that Wicket allows us to use a POJO as backing object, but POJOs are ordinary objects with no prespecified interface, annotation or superclass, hence they are not required to implement the standard Serializable interface.

To cope with these problems IModel extends another interface called IDetachable.

This interface provides a method called detach() which is invoked by Wicket at the end of web request processing when data model is no more needed but before serialization occurs. Overriding this method we can clean any reference to data object keeping just the information needed to retrieve it later (for example the id of the table row where our data are stored). In this way we can avoid the serialization of the object wrapped into the model overcoming both the problem with non-serializable objects and the one with large data objects.

Since IModel inherits from IDetachable, every model of Wicket is “detachable”, although not all of them implement a detaching policy (like the Model class). Usually detaching operations are strictly dependent on the persistence technology adopted for model objects (like a relational db, a NoSQL db, a queue, etc), so it's not unusual to write a custom detachable model suited for the persistence technology chosen for a given project. To ease this task Wicket provides abstract model LoadableDetachableModel. This class internally holds a transient reference to a model object which is initialized the first time getObject()is called to precess a request. The concrete data loading is delegated to abstract method T load(). The reference to a model object is automatically set to null at the end of the request by the detach() method.

The following class diagram summarizes the methods defined inside LoadableDetachableModel.

onDetach and onAttach can be overridden in order to obtain further control over the detaching procedure.

Now as example of a possible use of LoadableDetachableModel, we will build a model designed to work with entities managed via JPA. To understand the following code a basic knowledge of JPA is required even if we won't go into the detail of this standard.

The following model is provided for example purposes only and is not intended to be used in production environment. Important aspects such as transaction management are not taken into account and you should rework the code before considering to use it.

public class JpaLoadableModel<T> extends LoadableDetachableModel<T> {
  private EntityManagerFactory entityManagerFactory;
  private Class<T> entityClass;
  private Serializable identifier;
  private List<Object> constructorParams;
  public JpaLoadableModel(EntityManagerFactory entityManagerFactory, T entity) {
	super();
	PersistenceUnitUtil util = entityManagerFactory.getPersistenceUnitUtil();
		this.entityManagerFactory = entityManagerFactory;
	    this.entityClass = (Class<T>) entity.getClass();
	    this.identifier = (Serializable) util.getIdentifier(entity);
	    setObject(entity);
	}
	@Override
	protected T load() {
	   T entity = null;
	   if(identifier != null) {  
	       EntityManager entityManager = entityManagerFactory.createEntityManager();
	       entity = entityManager.find(entityClass, identifier);
	     }
	     return entity;
	   }
	@Override
	protected void onDetach() {
	   super.onDetach();
	     T entity = getObject();
	     PersistenceUnitUtil persistenceUtil = entityManagerFactory.getPersistenceUnitUtil();
	     if(entity == null) return;
	     identifier = (Serializable) persistenceUtil.getIdentifier(entity);    
	  }
	}

The constructor of the model takes as input two parameters: an implementation of the JPA interface javax.persistence.EntityManagerFactory to manage JPA entities and the entity that must be handled by this model. Inside its constructor the model saves the class of the entity and its id (which could be null if the entity has not been persisted yet). These two informations are required to retrieve the entity at a later time and are used by the load method.

onDetach is responsible for updating the entity id before detachment occurs. The id can change the first time an entity is persisted (JPA generates a new id and assigns it to the entity). Please note that this model is not responsible for saving any changes occurred to the entity object before it is detached. If we don't want to loose these changes we must explicitly persist the entity before the detaching phase occurs.

Since the model of this example holds a reference to the EntityManager Factory, the implementation in use must be serializable.

10.7 Using more than one model in a component

Sometimes our custom components may need to use more than a single model to work properly. In such a case we must manually detach the additional models used by our components. In order to do this we can overwrite the Component's onDetach method that is called at the end of the current request. The following is the generic code of a component that uses two models:

/**
 * 
 * fooModel is used as main model while beeModel must be manually detached
 *
 */
public class ComponetTwoModels extends Component{
	private IModel<Bee> beeModel;
	public ComponetTwoModels(String id, IModel<Foo> fooModel, IModel<Bee> beeModel) {
		super(id, fooModel);
		this.beeModel = beeModel;
	}
	@Override
	public void onDetach() {
               if(beeModel != null)
	   beeModel.detach();
              super.onDetach();
	}
}

When we overwrite onDetach we must call the super class implementation of this method, usually as last line in our custom implementation.

10.8 Use models!

Like many people new to Wicket, you may need a little time to fully understand the power and the advantages of using models. Taking your first steps with Wicket you may be tempted to pass row objects to your components instead of using models:

/**
 * 
 * NOT TO DO: passing row objects to components instead of using models!
 *
 */
public class CustomComponent extends Component{
	private FooBean fooBean;
	public CustomComponent(String id, FooBean fooBean) {
		super(id);
		this.fooBean = fooBean;
	}
	//...some other ugly code :)…
}

That's a bad practice and you must avoid it. Using models we do not only decouple our components from the data source, but we can also relay on them (if they are dynamic) to work with the most up-to-date version of our model object. If we decide to bypass models we lose all these advantages and we force model objects to be serialized.

10.9 Summary

Models are at the core of Wicket and they are the basic ingredient needed to taste the real power of the framework. In this chapter we have seen how to use models to bring data to our components without littering their code with technical details about their persistence strategy. We have also introduced Wicket forms as complementary topic. With forms and models we are able to bring our applications to life allowing them to interact with users. But what we have seen in this chapter about Wicket forms is just the tip of the iceberg. That's why the next chapter is entirely dedicated to them.

11 Wicket forms in detail

One of such features is the ability to work with nested forms (they will be discussed in paragraph 10.5).

In this chapter we will continue to explore Wicket forms learning how to master them and how to build effective and user-proof forms for our web applications.

11.1 Default form processing

In paragraph 9.3 we have seen a very basic usage of the Form component and we didn't pay much attention to what happens behind the scenes of form submission. In Wicket when we submit a form we trigger the following steps on server side:

1. Form validation: user input is checked to see if it satisfies the validation rules set on the form. If validation fails, step number 2 is skipped and the form should display a feedback message to explain to user what went wrong. During this step input values (which are simple strings sent with a web request) are converted into Java objects. In the next paragraphs we will explore the infrastructures provided by Wicket for the three sub-tasks involved with form validation, which are: conversion of user input into objects, validation of user input, and visualization of feedback messages.
2. Updating of models: if validation succeeds, the form updates the model of its children components with the converted values obtained in the previous step.
3. Invoking callback methods onSubmit() or onError(): if we didn't have any validation error, method onSubmit() is called, otherwise onError() will be called. The default implementation of both these methods is left empty and we can override them to perform custom actions.

Please note that the model of form components is updated only if no validation error occurred (i.e. step two is performed only if validation succeeds).

Without going into too much detail, we can say that the first two steps of form processing correspond to the invocation of one or more Form's internal methods (which are declared protected and final). Some examples of these methods are validate(), which is invoked during validation step, and updateFormComponentModels(), which is used at the step that updates the form field models.

The whole form processing is started invoking public method process(IFormSubmitter) (Later in paragraph 10.4 we will introduce interface IFormSubmitter).

11.2 Form validation and feedback messages

A basic example of a validation rule is to make a field required. In paragraph 9.3.2 we have already seen how this can be done calling setRequired(true) on a field. However, to set a validation rule on a FormComponent we must add the corresponding validator to it.

A validator is an implementation of the org.apache.wicket.validation.IValidator interface and the FormComponent has a version of method add which takes as input a reference of this interface.

For example if we want to use a text field to insert an email address, we could use the built-in validator EmailAddressValidator to ensure that the inserted input will respect the email format local-part@domain :

TextField email = new TextField("email");
email.add(new EmailAddressValidator());

Wicket comes with a set of built-in validators that should suit most of our needs. We will see them in paragraph 10.2.3.

Feedback messages and localization

Wicket generates a feedback message for each field that doesn't satisfy one of its validation rules. For example the message generated when a required field is left empty is the following

Field '<label>' is required.

<label> is the value of the label model set on a FormComponent with method setLabel(IModel <String> model). If such model is not provided, component id will be used as the default value.

The entire infrastructure of feedback messages is built on top of the Java internationalization (I18N) support and it uses resource bundles to store messages.

The topics of internationalization will be covered in chapter 12. For now we will give just few notions needed to understand the examples from this chapter.

By default resource bundles are stored into properties files but we can easily configure other sources as described later in paragraph 12.4.5.

Default feedback messages (like the one above for required fields) are stored in the file Application. properties placed inside Wicket the org.apache.wicket package. Opening this file we can find the key and the localized value of the message:

Required=Field '${label}' is required.

We can note the key (Required in our case) and the label parameter written in the expression language (${label}). Scrolling down this file we can also find the message used by the Email AddressValidator:

EmailAddressValidator=The value of '${label}' is not a valid email address.

By default FormComponent provides 3 parameters for feedback message: input (the value that failed validation), label and name (this later is the id of the component).

Remember that component model is updated with the user input only if validation succeeds! As a consequence, we can't retrieve the wrong value inserted for a field from its model. Instead, we should use getValue() method of FormComponent class. (This method will be introduced in the example used in paragraph 10.2.5)

Displaying feedback messages and filtering them

To display feedback messages we must use component org.apache.wicket.markup.html. panel.FeedbackPanel. This component automatically reads all the feedback messages generated during form validation and displays them with an unordered list:

<ul class="feedbackPanel"> 
	<li class="feedbackPanelERROR"> 
		<span class="feedbackPanelERROR">Field 'Username' is required.</span> 
	</li> 
</ul>

CSS classes "feedbackPanel" and "feedbackPanelERROR" can be used in order to customize the style of the message list:

The component can be freely placed inside the page and we can set the maximum amount of displayed messages with the setMaxMessages() method.

Error messages can be filtered using three built-in filters:

• ComponentFeedbackMessageFilter: shows only messages coming from a specific component.
• ContainerFeedbackMessageFilter: shows only messages coming from a specific container or from any of its children components.
• ErrorLevelFeedbackMessageFilter: shows only messages with a level of severity equals or greater than a given lower bound. Class FeedbackMessage defines a set of static constants to express different levels of severity: DEBUG, ERROR, WARNING, INFO, SUCCESS, etc.... Levels of severity for feedback messages are discussed in paragraph 10.2.6.

These filters are intended to be used when there are more than one feedback panel (or more than one form) in the same page. We can pass a filter to a feedback panel via its constructor or using the setFilter method. Custom filters can be created implementing the IFeedbackMessageFilter interface. An example of custom filter is illustrated on page 89.

Built-in validators

Wicket already provides a number of built-in validators ready to be used. The following table is a short reference where validators are listed along with a brief description of what they do. The default feedback message used by each of them is reported as well:

EmailAddressValidator

Checks if input respects the format local-part@domain.

Message:

The value of '${label}' is not a valid email address.

UrlValidator

Checks if input is a valid URL. We can specify in the constructor which protocols are allowed (http://, https://, and ftp://).

Message:

The value of '${label}' is not a valid URL.

DateValidator

Validator class that can be extended or used as a factory class to get date validators to check if a date is bigger than a lower bound (method minimum(Date min)), smaller than a upper bound (method maximum(Date max)) or inside a range (method range(Date min, Date max)).

Messages:

The value of '${label}' is less than the minimum of ${minimum}.

The value of '${label}' is larger than the maximum of ${maximum}.

The value of '${label}' is not between ${minimum} and ${maximum}.

RangeValidator

Validator class that can be extended or used as a factory class to get validators to check if a value is bigger than a given lower bound (method minimum(T min)), smaller than a upper bound (method maximum(T max)) or inside a range (method range(T min,T max)).

The type of the value is a generic subtype of java.lang.Comparable and must implement Serializable interface.

Messages:

The value of '${label}' must be at least ${minimum}.

The value of '${label}' must be at most ${maximum}.

The value of '${label}' must be between ${minimum} and ${maximum}.

StringValidator

Validator class that can be extended or used as a factory class to get validators to check if the length of a string value is bigger then a given lower bound (method minimumLength (int min)), smaller then a given upper bound (method maximumLength (int max)) or within a given range (method lengthBetween(int min, int max)).

To accept only string values consisting of exactly n characters, we must use method exactLength(int length).

Messages:

The value of '${label}' is shorter than the minimum of ${minimum} characters.

The value of '${label}' is longer than the maximum of ${maximum} characters.

The value of '${label}' is not between ${minimum} and ${maximum} characters long.

The value of '${label}' is not exactly ${exact} characters long.

CreditCardValidator

Checks if input is a valid credit card number. This validator supports some of the most popular credit cards (like “American Express", "MasterCard", “Visa” or “Diners Club”).

Message:

The credit card number is invalid.

EqualPasswordInputValidator

This validator checks if two password fields have the same value.

Message:

${label0} and ${label1} must be equal.

Overriding standard feedback messages with custom bundles

If we don't like the default validation feedback messages, we can override them providing custom properties files. In these files we can write our custom messages using the same keys of the messages we want to override. For example if we wanted to override the default message for invalid email addresses, our properties file would contain a line like this:

EmailAddressValidator=Man, your email address is not good!

As we will see in the next chapter, Wicket searches for custom properties files in various positions inside the application's class path, but for now we will consider just the properties file placed next to our application class. The name of this file must be equal to the name of our application class:

The example project OverrideMailMessage overrides email validator's message with a new one which also reports the value that failed validation:

EmailAddressValidator=The value '${input}' inserted for field '${label}' is not a valid email address.

Creating custom validators

If our web application requires a complex validation logic and built-in validators are not enough, we can implement our own custom validators. For example (project UsernameCustomValidator) suppose we are working on the registration page of our site where users can create their profile choosing their username. Our registration form should validate the new username checking if it was already chosen by another user. In a situation like this we may need to implement a custom validator that queries a specific data source to check if a username is already in use.

For the sake of simplicity, the validator of our example will check the given username against a fixed list of three existing usernames.

A custom validator must simply implement interface IValidator:

public class UsernameValidator implements IValidator<String> {
	List<String> existingUsernames = Arrays.asList("bigJack", "anonymous", "mrSmith");
	public void validate(IValidatable<String> validatable) {
		String chosenUserName = validatable.getValue();
		if(existingUsernames.contains(chosenUserName)){
			ValidationError error = new ValidationError(this);
			Random random = new Random();
			error.setVariable("suggestedUserName", 
					validatable.getValue() + random.nextInt());
			validatable.error(error);
		}
	}	
}

The only method defined inside IValidator is validate(IValidatable<T> validatable) and is invoked during validation's step. Interface IValidatable represents the component being validated and it can be used to retrieve the component model (getModel()) or the value to validate (getValue()).

The custom validation logic is all inside IValidator's method validate. When validation fails a validator must use IValidatable's method error(IValidationError error) to generate the appropriate feedback message. In the code above we used the ValidationError class as convenience implementation of the IValidationError interface which represents the validation error that must be displayed to the user. This class provides a constructor that uses the class name of the validator in input as key for the resource to use as feedback message (i.e. 'UsernameValidator' in the example). If we want to specify more then one key to use to locate the error message, we can use method addKey(String key) of ValidationError class.

In our example when validation fails, we suggest a possible username concatenating the given input with a pseudo-random integer. This value is passed to the feedback message with a variable named suggestedUserName. The message is inside application's properties file:

UsernameValidator=The username '${input}' is already in use. Try with '${suggestedUserName}'

To provide further variables to our feedback message we can use method setVariable(String name, Object value) of class ValidationError as we did in our example.

The code of the home page of the project will be examined in the next paragraph after we have introduced the topic of flash messages.

Using flash messages

So far we have considered just the error messages generated during validation step. However Wicket's Component class provides a set of methods to explicitly generate feedback messages called flash messages. These methods are:

• debug(Serializable message)
• info(Serializable message)
• success(Serializable message)
• warn(Serializable message)
• error(Serializable message)
• fatal(Serializable message)

Each of these methods corresponds to a level of severity for the message. The list above is sorted by increasing level of severity.

In the example seen in the previous paragraph we have a form which uses success method to notify user when the inserted username is valid. Inside this form there are two FeedbackPanel components: one to display the error message produced by custom validator and the other one to display the success message. The code of the example page is the following:

HTML:

<body>
	<form wicket:id="form">
		Username: <input type="text" wicket:id="username"/>
		<br/>
		<input type="submit"/>
	</form>
	<div style="color:green" wicket:id="succesMessage">
	</div>
	<div style="color:red" wicket:id="feedbackMessage">
	</div>
</body>

Java code:

public class HomePage extends WebPage {
    public HomePage(final PageParameters parameters) {	
	Form form = new Form("form"){
		@Override
		protected void onSubmit() {
			super.onSubmit();
			success("Username is good!");
		}
	};
	TextField mail;
	form.add(mail = new TextField("username", Model.of("")));
	mail.add(new UsernameValidator());
	add(new FeedbackPanel("feedbackMessage", 
		new ExactErrorLevelFilter(FeedbackMessage.ERROR)));
	add(new FeedbackPanel("succesMessage", 
		new ExactErrorLevelFilter(FeedbackMessage.SUCCESS)));
	add(form);
    }
    class ExactErrorLevelFilter implements IFeedbackMessageFilter{
    	private int errorLevel;
		public ExactErrorLevelFilter(int errorLevel){
			this.errorLevel = errorLevel;
		}
		public boolean accept(FeedbackMessage message) {
			return message.getLevel() == errorLevel;
		}
    }
    //UsernameValidator definition
    //…
}

The two feedback panels must be filtered in order to display just the messages with a given level of severity (ERROR for validator message and SUCCESS for form's flash message). Unfortunately the built-in message filter ErrorLevelFeedbackMessageFilter is not suitable for this task because its filter condition does not check for an exact error level (the given level is used as lower bound value). As a consequence, we had to build a custom filter (inner class ExactErrorLevelFilter) to accept only the desired severity level (see method accept of interface IFeedbackMessageFilter).

11.3 Input value conversion

Working with Wicket we will rarely need to worry about conversion between input values (which are strings because the underlying HTTP protocol) and Java types because in most cases the default conversion mechanism will be smart enough to infer the type of the model object and perform the proper conversion. However, sometimes we may need to work under the hood of this mechanism to make it properly work or to perform custom conversions. That's why this paragraph will illustrate how to control input value conversion.

The component that is responsible for converting input is the FormComponent itself with its convertInput() method. In order to convert its input a FormComponent must know the type of its model object. This parameter can be explicitly set with method setType(Class<?> type):

//this field must receive an integer value
TextField integerField = new TextField("number", new Model()).setType(Integer.class));

If no type has been provided, FormComponent will try to ask its model for this information. The PropertyModel and CompoundPropertyModel models can use reflection to get the type of object model. By default, if FormComponent can not obtain the type of its model object in any way, it will consider it as a simple String.

Once FormComponent has determined the type of model object, it can look up for a converter, which is the entity in charge of converting input to Java object and vice versa. Converters are instances of org.apache.wicket.util.convert.IConverter interface and are registered by our application class on start up.

To get a converter for a specific type we must call method getConverter(Class<C> type) on the interface IConverterLocator returned by Application's method getConverterLocator():

//retrieve converter for Boolean type
Application.get().getConverterLocator().getConverter(Boolean.class);

Components which are subclasses of AbstractSingleSelectChoice don't follow the schema illustrated above to convert user input.

These kinds of components (like DropDownChoice and RadioChoice1) use their choice render and their collection of possible choices to perform input conversion.

Creating custom application-scoped converters

The default converter locator used by Wicket is org.apache.wicket.ConverterLocator. This class provides converters for the most common Java types. Here we can see the converters registered inside its constructor:

public ConverterLocator()
{
	set(Boolean.TYPE, BooleanConverter.INSTANCE);
	set(Boolean.class, BooleanConverter.INSTANCE);
	set(Byte.TYPE, ByteConverter.INSTANCE);
	set(Byte.class, ByteConverter.INSTANCE);
	set(Character.TYPE, CharacterConverter.INSTANCE);
	set(Character.class, CharacterConverter.INSTANCE);
	set(Double.TYPE, DoubleConverter.INSTANCE);
	set(Double.class, DoubleConverter.INSTANCE);
	set(Float.TYPE, FloatConverter.INSTANCE);
	set(Float.class, FloatConverter.INSTANCE);
	set(Integer.TYPE, IntegerConverter.INSTANCE);
	set(Integer.class, IntegerConverter.INSTANCE);
	set(Long.TYPE, LongConverter.INSTANCE);
	set(Long.class, LongConverter.INSTANCE);
	set(Short.TYPE, ShortConverter.INSTANCE);
	set(Short.class, ShortConverter.INSTANCE);
	set(Date.class, new DateConverter());
	set(Calendar.class, new CalendarConverter());
	set(java.sql.Date.class, new SqlDateConverter());
	set(java.sql.Time.class, new SqlTimeConverter());
	set(java.sql.Timestamp.class, new SqlTimestampConverter());
	set(BigDecimal.class, new BigDecimalConverter());
}

If we want to add more converters to our application, we can override Application's method newConverterLocator which is used by application class to build its converter locator.

To illustrate how to implement custom converters and use them in our application, we will build a form with two text field: one to input a regular expression pattern and another one to input a string value that will be split with the given pattern.

The first text field will have an instance of class java.util.regex.Pattern as model object. The final page will look like this (the code of this example is from the CustomConverter project):

The conversion between Pattern and String is quite straightforward. The code of our custom converter is the following:

public class RegExpPatternConverter implements IConverter<Pattern> {
	@Override
	public Pattern convertToObject(String value, Locale locale) {
		return Pattern.compile(value);
	}
	@Override
	public String convertToString(Pattern value, Locale locale) {
		return value.toString();
	}
}

Methods declared by interface IConverter take as input a Locale parameter in order to deal with locale-sensitive data and conversions. We will learn more about locales and internationalization in chapter 12.

Once we have implemented our custom converter, we must override method newConverterLocator() inside our application class and tell it to add our new converter to the default set:

@Override
	protected IConverterLocator newConverterLocator() {
		ConverterLocator defaultLocator = new ConverterLocator();
		defaultLocator.set(Pattern.class, new RegExpPatternConverter());
		return defaultLocator;
	}

Finally, in the home page of the project we build the form which displays (with a flash message) the tokens obtained splitting the string with the given pattern:

public class HomePage extends WebPage {
    private Pattern regExpPatter;
    private String stringToSplit;
    public HomePage(final PageParameters parameters) {		
    	TextField mail;
	TextField stringToSplitTxt;
    	Form form = new Form("form"){
			@Override
			protected void onSubmit() {
				super.onSubmit();
				String messageResult = "Tokens for the given string and pattern:<br/>";
				String[] tokens = regExpPatter.split(stringToSplit);
				for (String token : tokens) {
					messageResult += "- " + token + "<br/>";
				}				
				success(messageResult);
		}
	};
		form.setDefaultModel(new CompoundPropertyModel(this));
		form.add(mail = new TextField("regExpPatter"));
		form.add(stringToSplitTxt = new TextField("stringToSplit"));
		add(new FeedbackPanel("feedbackMessage").setEscapeModelStrings(false));
		add(form);
    }
}

If the user input can not be converted to the target type, FormComponent will generate the default error message “The value of '${label}' is not a valid ${type}.”. The bundle key for this message is IConverter.

11.4 Submit form with an IFormSubmittingComponent

Besides submitting forms with a standard HTML submit button, Wicket allows us to use special components which implement interface IFormSubmittingComponent. This entity is a subinterface of IFormSubmitter:

At the beginning of this chapter we have seen that form processing is started by process method which takes as input an instance of IFormSubmitter. This parameter corresponds to the IFormSubmittingComponent clicked by a user to submit the form and it is null if we have used a standard HTML submit button (like we have done so far).

A submitting component is added to a form just like any other child component using method add(Component...).

A form can have any number of submitting components and we can specify which one among them is the default one by calling the Form's method setDefaultButton(IFormSubmittingComponent  component). The default submitter is the one that will be used when user presses 'Enter' key in a field of the form. In order to make the default button work, Wicket will add to our form a hidden <div> tag containing a text field and a submit button with some JavaScript code to trigger it:

<div style="width:0px;height:0px;position:absolute;left:-100px;top:-100px;overflow:hidden">
	<input type="text" autocomplete="off"/>
	<input type="submit" name="submit2" onclick=" var b=document...."/>
</div>

Just like Wicket forms, interface IFormSubmitter defines methods onSubmit and onError. These two methods have the priority over the namesake methods of the form, meaning that when a form is submitted with an IFormSubmitter, the onSubmit of the submitter is called before the one of the form. Similarly, if validation errors occurs during the first step of form processing, submitter's method onError is called before the form's one.

Starting with Wicket version 6.0 interface IFormSubmitter defines a further callback method called onAfterSubmit(). This method is called after form's method onSubmit() has been executed.

Components Button and SubmitLink

Component org.apache.wicket.markup.html.form.Button is a basic implementation of a form submitter. It can be used with either the <input> or <button> tags. The string model received as input by its constructor is used as button label and it will be the value of the markup attribute value.

In the following snippet we have a form with two submit buttons bound to an <input> tag. One of them is set as default button and both have a string model for the label:

HTML:

<body>
		<form wicket:id="form">
			Username: <input type="text" wicket:id="username"/>
			<br/>
			<input type="submit" wicket:id="submit1"/>
			<input type="submit" wicket:id="submit2"/>
		</form>
</body>

Java code:

public class HomePage extends WebPage {
    public HomePage(final PageParameters parameters) {		
		Form form = new Form("form");
     	form.add(new TextField("username", Model.of("")));
    	form.add(new Button("submit1", Model.of("First submitter")));
		Button secondSubmitter;
		form.add(secondSubmitter = new Button("submit2", Model.of("Second submitter")));
    	form.setDefaultButton(secondSubmitter);
		add(form);
    }
}

Generated markup:

<form wicket:id="form" id="form1" method="post" action="?0-1.IFormSubmitListener-form">
   <div>
      …
      <!-- Code generated by Wicket to handle the default button -->
      …
   </div>			
   Username: <input type="text" wicket:id="username" value="" name="username"/>
   <br/>
   <input type="submit" wicket:id="submit1" name="submit1" id="submit13" value="First submitter"/>
   <input type="submit" wicket:id="submit2" name="submit2" id="submit22" value="Second submitter"/>
</form>

Another component that can be used to submit a form is org.apache.wicket.markup. html.form.SubmitLink. This component uses JavaScript to submit the form. Like the name suggests, the component can be used with the <a> tag but it can be also bound to any other tag that supports the event handler onclick. When used with the <a> tag, the JavaScript code needed to submit the form will be placed inside href attribute while with other tags the script will go inside the event handler onclick.

A notable difference between this component and Button is that SubmitLink can be placed outside the form it must submit. In this case we must specify the form to submit in its constructor:

HTML:

<html xmlns:wicket="http://wicket.apache.org">
<head>
	</head>
	<body>
	<form wicket:id="form">
			Password: <input type="password" wicket:id="password"/>
			<br/>					
		</form>
		<button wicket:id="externalSubmitter">
			Submit
		</button>
	</body>
</html>

Java code:

public class HomePage extends WebPage {
    public HomePage(final PageParameters parameters) {		
		Form form = new Form("form");
		form.add(new PasswordTextField("password", Model.of("")));
		//specify the form to submit
		add(new SubmitLink("externalSubmitter", form));
		add(form);
    }
}

Disabling default form processing

With an IFormSubmittingComponent we can choose to skip the default form submission process by setting the appropriate flag to false with the setDefaultFormProcessing method. When the default form processing is disabled only the submitter's onSubmit is called while form's validation and models updating are skipped.

This can be useful if we want to implement a “Cancel” button on our form which redirects user to another page without validating his/her input.

When we set this flag to false we can decide to manually invoke the form processing by calling the process(IFormSubmittingComponent) method.

11.5 Nested forms

As you might already known, HTLM doesn't allow to have nested forms1. However with Wicket we can overcome this limitation by adding one or more form components to a parent form.

This can be useful if we want to split a big form into smaller ones in order to reuse them and to better distribute responsibilities among different components. Forms can be nested to an arbitrary level:

<form wicket:id="outerForm"> 
	…
	<form wicket:id="innerForm"> 
		…
		<form wicket:id="veryInnerForm">
			…
		</form> 
	</form> 
</form>

When a form is submitted also its nested forms are submitted and they participate in the validation step. This means that if a nested form contains invalid input values, the outer form won't be submitted. On the contrary, nested forms can be singularly submitted without depending on the status of their outer form.

To submit a parent form when one of its children forms is submitted, we must override its method wantSubmitOnNestedFormSubmit and make it return true.

11.6 Multi-line text input

HTML provides a multi-line text input control with <textarea> tag. The Wicket counterpart for this kind of control is org.apache.wicket.markup.html.form.TextArea component:

HTML:

<textarea wicket:id="description" rows="5" cols="40"></textarea>

Java code:

form.add(new TextArea("description", Model.of("")));

Component TextArea is used just like any other single-line text field. To specify the size of the text area we can write attributes rows and cols directly in the markup file or we can create new attribute modifiers and add them to our TextArea component.

11.7 File upload

Wicket supports file uploading with the FileUploadField component which must be used with the <input> tag whose type attribute must be set to "file". In order to send a file on form submission we must enable multipart mode calling MultiPart(true)on our form.

In the next example (project UploadSingleFile) we will see a form which allows users to upload a file into the temporary directory of the server (path /tmp on Unix/Linux systems):

HTML:

<html>
	<head>
	</head>
	<body>
	<h1>Upload your file here!</h1>
	<form wicket:id="form">
		<input type="file" wicket:id="fileUploadField"/> 
		<input type="submit" value="Upload"/>
	</form>
	<div wicket:id="feedbackPanel">
	</div>
	</body>
</html>

Java code:

public class HomePage extends WebPage {
   private FileUploadField fileUploadField;
    public HomePage(final PageParameters parameters) {
    	fileUploadField = new FileUploadField("fileUploadField");
    	Form form = new Form("form"){
    		@Override
    		protected void onSubmit() {
    		  super.onSubmit();
    		  FileUpload fileUpload = fileUploadField.getFileUpload();
    		    try {
			File file = new File(System.getProperty("java.io.tmpdir") + "/" +
    						fileUpload.getClientFileName());
			    fileUpload.writeTo(file);
		        } catch (IOException e) {
			   e.printStackTrace();
			 }
    		}
    	};
	form.setMultiPart(true);
	//set a limit for uploaded file's size
	form.setMaxSize(Bytes.kilobytes(100));
	form.add(fileUploadField);
	add(new FeedbackPanel("feedbackPanel"));
	add(form);
    }
}

The code that copies the uploaded file to the temporary directory is inside the onSubmit method of the Form class. The uploaded file is handled with an instance of class FileUpload returned by the getFileUpload() method of the FileUploadField class. This class provides a set of methods to perform some common tasks like getting the name of the uploaded file (getClientFileName()), coping the file into a directory (writeTo(destinationFile)), calculating file digest (getDigest (digestAlgorithm)) and so on.

Form component can limit the size for uploaded files using its setMaxSize(size) method. In the example we have set this limit to 100 kb to prevent users from uploading files bigger than this size.

The maximum size for uploaded files can also be set at application's level using the setDefaultMaximumUploadSize(Bytes maxSize) method of the IApplicationSettings interface:

@Override
public void init()
{
 getApplicationSettings().setDefaultMaximumUploadSize(Bytes.kilobytes(100));  
}

Upload multiple files

If we need to upload multiple files at once, we can use the MultiFileUploadField component which allows the user to select an arbitrary number of files to send on form submission.

An example showing how to use this component can be found in Wicket module wicket-examples in file MultiUploadPage.java. The live example is hosted at http://www.wicket-library.com/wicket-examples-6.0.x/upload/multi .

11.8 Creating complex form components with FormComponentPanel

In chapter 3.2.2 we have seen how to use class Panel to create custom components with their own markup and with an arbitrary number of children components.

While it's perfectly legal to use Panel also to group form components, the resulting component won't be itself a form component and it won't participate in the form's submission workflow.

This could be a strong limitation if the custom component needs to coordinate its children during sub-tasks like input conversion or model updating. That's why in Wicket we have the org.apache. wicket.markup.html.form.FormComponentPanel component which combines the features of a Panel (it has its own markup file) and a FormComponent (it is a subclass of FormComponent).

A typical scenario in which we may need to implement a custom FormComponentPanel is when our web application and its users work with different units of measurement for the same data.

To illustrate this possible scenario, let's consider a form where a user can insert a temperature that will be recorded after being converted to Kelvin degrees (see the example project CustomForm ComponentPanel).

The Kelvin scale is wildly adopted among the scientific community and it is one of the seven base units of the International System of Units , so it makes perfect sense to store temperatures expressed with this unit of measurement.

However, in our everyday life we still use other temperature scales like Celsius or Fahrenheit, so it would be nice to have a component which internally works with Kelvin degrees and automatically applies conversion between Kelvin temperature scale and the one adopted by the user.

In order to implement such a component, we can make a subclass of FormComponentPanel and leverage the convertInput and onBeforeRender methods: in the implementation of the convertInput method we will convert input value to Kelvin degrees while in the implementation of onBeforeRender method we will take care of converting the Kelvin value to the temperature scale adopted by the user.

Our custom component will contain two children components: a text field to let user insert and edit a temperature value and a label to display the letter corresponding to user's temperature scale (F for Fahrenheit and C for Celsius). The resulting markup file is the following:

<html>
<head>
</head>
<body>
	<wicket:panel>
		Registered temperature: <input size="3" maxlength="3"         
                             wicket:id="registeredTemperature"/> 
		<label wicket:id="mesuramentUnit"></label> 
	</wicket:panel>
</body>
</html>

As shown in the markup above FormComponentPanel uses the same <wicket:panel> tag used by Panel to define its markup. Now let's see the Java code of the new form component starting with the onInitialize() method:

public class TemperatureDegreeField extends FormComponentPanel<Double> {
	private TextField<Double> userDegree;
	public TemperatureDegreeField(String id) {
		super(id);		
	}
	public TemperatureDegreeField(String id, IModel<Double> model) {
		super(id, model);		
	}
	@Override
	protected void onInitialize() {
		super.onInitialize();
	     AbstractReadOnlyModel<String> labelModel=new AbstractReadOnlyModel<String>(){
			@Override
			public String getObject() {
				if(getLocale().equals(Locale.US))
					return "°F";
				return "°C";
			}
		};
		add(new Label("mesuramentUnit", labelModel));
		add(userDegree=new TextField<Double>("registeredTemperature", new 
                           Model<Double>()));
		userDegree.setType(Double.class);
	}

Inside the onInitialize method we have created a read-only model for the label that displays the letter corresponding to the user's temperature scale. To determinate which temperature scale is in use, we retrieve the Locale from the session by calling Component's getLocale() method (we will talk more about this method in chapter 12). Then, if locale is the one corresponding to the United States, the chosen scale will be Fahrenheit, otherwise it will be considered as Celsius.

In the final part of onInitialize() we add the two components to our custom form component. You may have noticed that we have explicitly set the type of model object for the text field to double. This is necessary as the starting model object is a null reference and this prevents the component from automatically determining the type of its model object.

Now we can look at the rest of the code containing the convertInput and onBeforeRender methods:

// continued example
	@Override
	protected void convertInput() {
		Double userDegreeVal = userDegree.getConvertedInput();
		Double kelvinDegree;
		if(getLocale().equals(Locale.US)){
			kelvinDegree = userDegreeVal +  459.67;
			BigDecimal bdKelvin = new BigDecimal(kelvinDegree);
			BigDecimal fraction = new BigDecimal(5).divide(new BigDecimal(9));
			kelvinDegree = bdKelvin.multiply(fraction).doubleValue();
		}else{
			kelvinDegree = userDegreeVal + 273.15;
		}
		setConvertedInput(kelvinDegree);
	}
	@Override
	protected void onBeforeRender() {
		super.onBeforeRender();
		Double kelvinDegree = (Double) getDefaultModelObject();		
		Double userDegreeVal = null;
		if(kelvinDegree == null) return;
		if(getLocale().equals(Locale.US)){
			BigDecimal bdKelvin = new BigDecimal(kelvinDegree);
			BigDecimal fraction = new BigDecimal(9).divide(new BigDecimal(5));
			kelvinDegree = bdKelvin.multiply(fraction).doubleValue();
			userDegreeVal = kelvinDegree - 459.67;
		}else{
			userDegreeVal = kelvinDegree - 273.15;
		}
		userDegree.setModelObject(userDegreeVal);
	}
}

Since our component does not directly receive the user input, convertInput() must read this value from the inner text field using FormComponent's getConvertedInput() method which returns the input value already converted to the type specified for the component (Double in our case). Once we have the user input we convert it to kelvin degrees and we use the resulting value to set the converted input for our custom component (using method setConvertedInput(T convertedInput)).

Method onBeforeRender() is responsible for synchronizing the model of the inner textfield with the model of our custom component. To do this we retrieve the model object of the custom component with the getDefaultModelObject() method, then we convert it to the temperature scale adopted by the user and finally we use this value to set the model object of the text field.

11.9 Stateless form

In chapter 6 we have seen how Wicket pages can be divided into two categories: stateful and stateless. Pages that are stateless don't need to be stored in the user session and they should be used when we don't need to save any user data in the user session (for example in the public area of a site).

Besides saving resources on server-side, stateless pages can be adopted to improve user experience and to avoid security weaknesses. A typical situation where a stateless page can bring these benefits is when we have to implement a login page.

For this kind of page we might encounter two potential problems if we chose to use a stateful page. The first problem occurs when the user tries to login without a valid session assigned to him. This could happen if the user leaves the login page opened for a period of time bigger than the session's timeout and then he decides to log in. Under these conditions the user will be redirected to a 'Page expired' error page, which is not exactly a nice thing for user experience.

The second problem occurs when a malicious user or a web crawler program attempts to login into our web application, generating a huge number of page versions and consequently increasing the size of the user session.

To avoid these kinds of problems we should build a stateless login page which does not depend on a user session. Wicket provides a special version of the Form component called StatelessForm which is stateless by default (i.e its method getStatelessHint() returns true), hence it's an ideal solution when we want to build a stateless page with a form. A possible implementation of our login form is the following (example project StatelessLoginForm):

HTML:

<html>
   <head>
      <meta charset="utf-8" />
   </head>
   <body>
      <div>Session is <b wicket:id="sessionType"></b></div>
      <br/>
      <div>Type 'user' as correct credentials</div>
      <form wicket:id="form">
         <fieldset>
            Username: <input type="text" wicket:id="username"/> <br/>
            Password: <input type="password" wicket:id="password"/><br/>
            <input type="submit"/>
         </fieldset>
      </form>
      <br/>
      <div wicket:id="feedbackPanel"></div>
   </body>
</html>

Java code:

public class HomePage extends WebPage {
    private Label sessionType;
    private String password;
    private String username;
    public HomePage(final PageParameters parameters) {
       StatelessForm form = new StatelessForm("form"){
         @Override
         protected void onSubmit() {
            //sign in if username and password are “user”
            if("user".equals(username) && username.equals(password))
               info("Username and password are correct!");
            else
               error("Wrong username or password");
         }
      };
      form.add(new PasswordTextField("password"));
      form.add(new TextField("username"));
      add(form.setDefaultModel(new CompoundPropertyModel(this)));
      add(sessionType = new Label("sessionType", Model.of("")));
      add(new FeedbackPanel("feedbackPanel"));
    }
    @Override
    protected void onBeforeRender() {
       super.onBeforeRender();
       if(getSession().isTemporary())
          sessionType.setDefaultModelObject("temporary");
       else
          sessionType.setDefaultModelObject("permanent");
    }
}

Label sessionType shows if current session is temporary or not and is set inside onBeforeRender(): if our page is really stateless the session will be always temporary. We have also inserted a feedback panel in the home page that shows if the credentials are correct. This was done to make the example form more interactive.

11.10 Working with radio buttons and checkboxes

In this paragraph we will see which components can be used to handle HTML radio buttons and checkboxes. Both these input elements are usually grouped together to display a list of possible choices:

A check box can be used as single component to set a boolean property. For this purpose Wicket provides the org.apache.wicket.markup.html.form.CheckBox component which must be attached to <input type="checkbox".../> tag. In the next example (project SingleCheckBox) we will consider a form similar to the one used in paragraph 9.5 to edit a Person object, but with an additional checkbox to let the user decide if she wants to subscribe to our mailing list or not. The form uses the following bean as backing object:

public class RegistrationInfo implements Serializable {
	private String name;
	private String surname;
	private String address;
	private String email;
	private boolean subscribeList;
	/*Getters and setters*/
}

The markup and the code for this example are the following:

HTML:

<form wicket:id="form">		
		<div style="display: table;">
			<div style="display: table-row;">
				<div style="display: table-cell;">Name: </div>
				<div style="display: table-cell;">
					<input type="text" wicket:id="name"/> 
				</div>	
			</div>
			<div style="display: table-row;">
				<div style="display: table-cell;">Surname: </div>
				<div style="display: table-cell;">
					<input type="text" wicket:id="surname"/>
				</div>	
			</div>
			<div style="display: table-row;">
				<div style="display: table-cell;">Address: </div>
				<div style="display: table-cell;">
					<input type="text" wicket:id="address"/>
				</div>	
			</div>
			<div style="display: table-row;">
				<div style="display: table-cell;">Email: </div>
				<div style="display: table-cell;">
					<input type="text" wicket:id="email"/>
				</div>
			</div>
			<div style="display: table-row;">
				<div style="display: table-cell;">Subscribe list:</div>
				<div style="display: table-cell;">
					<input type="checkbox" wicket:id="subscribeList"/>
				</div>
			</div>
		</div>	
	<input type="submit" value="Save"/>
</form>

Java code:

public HomePage(final PageParameters parameters) {
    	RegistrationInfo registrtionInfo = new RegistrationInfo();
    	registrtionInfo.setSubscribeList(true);
    	Form form = new Form("form", 
    			new CompoundPropertyModel<RegistrationInfo>(registrtionInfo));
    	form.add(new TextField("name"));
	form.add(new TextField("surname"));
	form.add(new TextField("address"));
	form.add(new TextField("email"));
	form.add(new CheckBox("subscribeList"));
	add(form);
}

Please note that the checkbox will be initially selected because we have set to true the subscribe flag during the model object creation (with instruction registrtionInfo.setSubscribeList(true)):

Working with grouped checkboxes

When we need to display a given number of options with checkboxes, we can use the org.apache. wicket.markup.html.form.CheckBoxMultipleChoice component. For example, If our options are a list of strings, we can display them in this way:

HTML:

<div wicket:id="checkGroup">
		<input type="checkbox"/>It will be replaced by the actual checkboxes…
</div>

Java code:

List<String> fruits = Arrays.asList("apple", "strawberry", "watermelon"); 
form.add(new CheckBoxMultipleChoice("checkGroup", new ListModel<String>(new  
								 ArrayList<String>()), fruits));

Screenshot:

This component can be attached to a <div> tag or to a <span> tag. No specific content is required for this tag as it will be populated with the actual checkboxes. Since this component allows multiple selection, its model object is a list. In the example above we have used model class org.apache. wicket.model.util.ListModel which is specifically designed to wrap a List object.

By default CheckBoxMultipleChoice inserts a <br/> tag as suffix after each option. We can configure both the suffix and the prefix used by the component with the setPrefix and setSuffix methods.

When our options are more complex objects than simple strings, we can render them using an IChoiceRender, as we did for DropDownChoice in paragraph 9.4:

HTML:

<div wicket:id="checkGroup">
		<input type="checkbox"/>It will be replaced by actual checkboxes…
</div>

Java code:

Person john = new Person("John", "Smith");
Person bob = new Person("Bob", "Smith");
Person jill = new Person("Jill", "Smith");
List<Person> theSmiths = Arrays.asList(john, bob, jill); 
ChoiceRenderer render = new ChoiceRenderer("name");
     	form.add(new CheckBoxMultipleChoice("checkGroup", new ListModel<String>(ArrayList<String>()),   
                                    theSmiths, render));

Screenshot:

How to implement a "Select all" checkbox

A nice feature we can offer to users when we have a group of checkboxes is a “special” checkbox which selects/unselects all the other options of the group:

Wicket comes with a couple of utility components that make it easy to implement such a feature. They are CheckboxMultipleChoiceSelector and CheckBoxSelector classes, both inside org. apache.wicket.markup.html.form package. The difference between these two components is that the first works with an instance of CheckBoxMultipleChoice while the second takes in input a list of CheckBox objects:

/* CheckboxMultipleChoiceSelector usage: */
CheckBoxMultipleChoice checkGroup;
//checkGroup initialization…
CheckboxMultipleChoiceSelector cbmcs = new CheckboxMultipleChoiceSelector("id", checkGroup);
/* CheckBoxSelector usage: */
CheckBox checkBox1, checkBox2, checkBox3;
//checks initialization…
CheckBoxSelector cbmcs = new CheckBoxSelector("id", checkBox1, checkBox2, checkBox3);

Working with grouped radio buttons

For groups of radio buttons we can use the org.apache.wicket.markup.html.form. RadioChoice component which works in much the same way as CheckBoxMultipleChoice:

HTML:

<div wicket:id="radioGroup">
	<input type="radio"/>It will be replaced by actual radio buttons…
</div>

Java code:

List<String> fruits = Arrays.asList("apple", "strawberry", "watermelon"); 
form.add(new RadioChoice("radioGroup", Model.of(""), fruits));

Screenshot:

Just like CheckBoxMultipleChoice, this component provides the setPrefix and setSuffix methods to configure the prefix and suffix for our options and it supports IChoiceRender as well. In addition, RadioChoice provides the wantOnSelectionChangedNotifications() method to notify the web server when the selected option changes (this is the same method seen for DropDownChoice in paragraph 9.4).

11.11 Selecting multiple values with ListMultipleChoices and Palette

Checkboxes work well when we have a small amount of options to display, but they quickly become chaotic as the number of options increases. To overcome this limit we can use the <select> tag switching it to multiple-choice mode with attribute multiple="multiple":

Now the user can select multiple options by holding down Ctrl key (or Command key for Mac) and selecting them.

To work with multiple choice list Wicket provides the org.apache.wicket.markup.html.form. ListMultipleChoice component:

HTML:

<select wicket:id="fruits">
	<option>choice 1</option>
	<option>choice 2</option>
</select>

Java code:

List<String> fruits = Arrays.asList("apple", "strawberry", "watermelon"); 
form.add(new ListMultipleChoice("fruits", new ListModel<String>(new ArrayList<String>()), fruits));

Screenshot:

This component must be bound to a <select> tag but the attribute multiple="multiple" is not required as it will automatically be added by the component.

The number of visible rows can be set with the setMaxRows(int maxRows) method.

Component Palette

While multiple choice list solves the problem of handling a big number of multiple choices, it is not much intuitive for end users. That's why desktop GUIs have introduced a more complex component which can be generally referred to as multi select transfer component (it doesn't have an actual official name):

This kind of component is composed by two multiple-choice lists, one on the left displaying the available options and the other one on the right displaying the selected options. User can move options from a list to another by double clicking on them or using the buttons placed between the two list.

Built-in org.apache.wicket.extensions.markup.html.form.palette.Palette component provides an out-of-the-box implementation of a multi select transfer component. It works in a similar way to ListMultipleChoice:

HTML:

<div wicket:id="palette">
   Select will be replaced by the actual content…
	   <select multiple="multiple">
     <option>option1</option>
     <option>option2</option>
     <option>option3</option>
</div>

Java code:

Person john = new Person("John", "Smith");
Person bob = new Person("Bob", "Smith");
Person jill = new Person("Jill", "Smith");
Person andrea = new Person("Andrea", "Smith");
List<Person> theSmiths = Arrays.asList(john, bob, jill, andrea); 
ChoiceRenderer render = new ChoiceRenderer("name");
form.add(new Palette("palette", Model.of(new ArrayList<String>()), new ListModel<String> (theSmiths), render, 5, true));

Screenshot:

The last two parameters of the Palette's constructor (an integer value and a boolean value) are, respectively, the number of visible rows for the two lists and a flag to choose if we want to display the two optional buttons which move selected options up and down. The descriptions of the two lists (“Available” and “Selected”) can be customized providing two resources with keys palette.available and palette.selected.

The markup of this component uses a number of CSS classes which can be extended/overriden to customize the style of the component. We can find these classes and see which tags they decorate in the default markup file of the component:

<table cellspacing="0" cellpadding="2" class="palette">
<tr>
	<td class="header headerAvailable"><span wicket:id="availableHeader">[available header]</span></td>
	<td>&#160;</td>
	<td class="header headerSelected"><span wicket:id="selectedHeader">[selected header]</span>                                           
        </td> 
</tr>
<tr>
	<td class="pane choices">
		<select wicket:id="choices" class="choicesSelect">[choices]</select>	
	</td>
	<td class="buttons">
		<button type="button" wicket:id="addButton" class="button add"><div/> 
               </button><br/>
		<button type="button" wicket:id="removeButton" class="button remove"><div/> 
               </button><br/>
		<button type="button" wicket:id="moveUpButton" class="button up"><div/>  
               </button><br/>
		<button type="button" wicket:id="moveDownButton" class="button down"><div/>  
               </button><br/>
	</td>
	<td class="pane selection">
		<select class="selectionSelect" wicket:id="selection">[selection]</select>	
	</td>
</tr>
</table>

11.12 Summary

In the second part of the chapter we learnt how to build reusable form components and how to implement a stateless form. We have ended the chapter with an overview of the built-in form components needed to handle standard input form elements like checkboxes, radio buttons and multiple selections lists.

12 Displaying multiple items with repeaters

<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<title>Insert title here</title>
</head>
<body>
  <%
    for(int i = 12; i<=32; i++) {
      %>
      <div>Hello! I'm index n°<%= %></div>
  <% 
    }
  %>
</body>

To ease this task Wicket provides a number of special-purpose components called repeaters which are designed to use their related markup to display the items of a given set in a more natural and less chaotic way.

In this chapter we will see some of the built-in repeaters that come with Wicket.

12.1 Component RepeatingView

Component org.apache.wicket.markup.repeater.RepeatingView is a container which renders its children components using the tag it is bound to. It can contain an arbitrary number of children elements and we can obtain a new valid id for a new child calling its method newChildId(). This component is particularly suited when we have to repeat a simple markup fragment, for example when we want to display some items as a HTML list:

HTML:

<ul>
    <li wicket:id="listItems"></li>
</ul>

Java Code:

RepeatingView listItems = new RepeatingView("listItems");
listItems.add(new Label(listItems.newChildId(), "green");
listItems.add(new Label(listItems.newChildId(), "blue");
listItems.add(new Label(listItems.newChildId(), "red");

Generated markup:

<ul>
    <li>green</li>
    <li>blue</li>
    <li>red</li>
</ul>

As we can see in this example, each child component has been rendered using the parent markup as if it was its own.

12.2 Component ListView

As its name suggests, component org.apache.wicket.markup.html.list.ListView is designed to display a given list of objects which can be provided as a standard Java List or as a model containing the concrete List. ListView iterates over the list and creates a child component of class org.apache.wicket.markup.html.list.ListItem for every encountered item.

Unlike RepeatingView this component is intended to be used with complex markup fragments containing nested components.

To generate its children, ListView calls its abstract method populateItem(ListItem<T> item) for each item in the list, so we must provide an implementation of this method to tell the component how to create its children components. In the following example we use a ListView to display a list of Person objects:

HTML:

…
	<body>
		<div id="bd" style="display: table;">
			<div wicket:id="persons" style="display: table-row;">
				<div style="display: table-cell;"><b>Full name: </b></div>
				<div wicket:id="fullName" style="display: table-cell;"></div>
			</div>
		</div>
	</body>
...

Java Code (Page Constructor):

public HomePage(final PageParameters parameters) {
	   List<Person> persons = Arrays.asList(new Person("John", "Smith"), 
                                        new Person("Dan", "Wong"));
   add(new ListView<Person>("persons", persons) {
	@Override
	protected void populateItem(ListItem<Person> item) {
	   item.add(new Label("fullName", new PropertyModel(item.getModel(), "fullName")));
	}			
   });
}

Screenshot of generated page:

In this example we have displayed the full name of two Person's instances. The most interesting part of the code is the implementation of method populateItem where parameter item is the current child component created by ListView and its model contains the corresponding element of the list. Please note that inside populateItem we must add nested components to item object and not directly to the ListView.

ListView and Form

By default ListView replaces its children components with new instances every time is rendered. Unfortunately this behavior is a problem if ListView is inside a form and it contains form components. The problem is caused by the fact that children components are replaced by new ones before form is rendered, hence they can't keep their input value if validation fails and, furthermore, their feedback messages can not be displayed.

To avoid this kind of problem we can force ListView to reuse its children components using its method setReuseItems and passing true as parameter. If for any reason we need to refresh children components after we have invoked setReuseItems(true), we can use MarkupContainer's method removeAll() to force ListView to rebuild them.

12.3 Component RefreshingView

Component org.apache.wicket.markup.repeater.RefreshingView is a subclass of RepeatingView that comes with a customizable rendering strategy for its children components.

RefreshingView defines abstract methods populateItem(Item) and getItemModels(). The first method is similar to the namesake method seen for ListView, but it takes in input an instance of class org.apache.wicket.markup.repeater.Item which is a subclass of ListItem. RefreshingView is designed to display a collection of models containing the actual items. An iterator over these models is returned by the other abstract method getItemModels.

The following code is a version of the previous example that uses RefreshingView in place of ListView:

HTML:

…
	<body>
		<div id="bd" style="display: table;">
			<div wicket:id="persons" style="display: table-row;">
				<div style="display: table-cell;"><b>Full name: </b></div>
				<div wicket:id="fullName" style="display: table-cell;"></div>
			</div>
		</div>
	</body>
...

Java Code (Page Constructor):

public HomePage(final PageParameters parameters) {
   //define the list of models to use
   final List<IModel<Person>> persons = new ArrayList<IModel<Person>>();
   persons.add(Model.of(new Person("John", "Smith")); 
   persons.add(Model.of(new Person("Dan", "Wong"));
   add(new RefreshingView<Person>("persons") {
	@Override
	protected void populateItem(Item<Person> item) {
	   item.add(new Label("fullName", new PropertyModel(item.getModel(), "fullName")));
	}
@Override
	protected Iterator<IModel<Person>> getItemModels() {
	   return persons.iterator();
}			
   });
}

Item reuse strategy

By default, just like ListView, RefreshingView replaces its children with new instances every time is rendered. The strategy that decides if and how children components must be refreshed is returned by method getItemReuseStrategy. This strategy is an implementation of interface IItemReuseStrategy. The default implementation used by RefreshingView is class DefaultItemReuseStrategy but Wicket provides also strategy ReuseIfModelsEqualStrategy which reuses an item if its model has been returned by the iterator obtained with method getItemModels.

To set a custom strategy we must use method setItemReuseStrategy.

12.4 Pageable repeaters

Wicket offers a number of components that should be used when we have to display a big number of items (for example the results of a select SQL query).

All these components implements interface org.apache.wicket.markup.html.navigation. paging.IPageable and use interface IDataProvider (placed in package org.apache.wicket. markup.repeater.data) as data source. This interface is designed to support data paging. We will see an example of data paging later in paragraph 11.4.2.

The methods defined by IDataProvider are the following:

• iterator(long first, long count): returns an iterator over a subset of the entire dataset. The subset starts from the item at position first and includes all the next count items (i.e. it's the closed interval first,first+count).
• size(): gets the size of the entire dataset.
• model(T object): this method is used to wrap an item returned by the iterator with a model. This can be necessary if, for example, we need to wrap items with a detachable model to prevent them from being serialized.

Wicket already provides implementations of IDataProvider to work with a List as data source (ListDataProvider) and to support data sorting (SortableDataProvider).

Component DataView

Class org.apache.wicket.markup.repeater.data.DataView is the simplest pageable repeater shipped with Wicket. DataView comes with abstract method populateItem(Item) that must be implemented to configure children components. In the following example we use a DataView to display a list of Person objects in a HTML table:

HTML:

<table>
	<tr>
	   <th>Name</th><th>Surename</th><th>Address</th><th>Email</th>
	</tr>
	<tr wicket:id="rows">
	   <td wicket:id="dataRow"></td>
	</tr>
</table>

Java Code:

//method loadPersons is defined elsewhere
List<Person> persons = loadPersons();
ListDataProvider<Person> listDataProvider = new ListDataProvider<Person>(persons);
DataView<Person> dataView = new DataView<Person>("row", listDataProvider) {
  @Override
  protected void populateItem(Item<Person> item) {
    Person person = item.getModelObject();
    RepeatingView repeatingView = new RepeatingView("dataRow");
    repeatingView.add(new Label(repeatingView.newChildId(), person.getName()));
    repeatingView.add(new Label(repeatingView.newChildId(), person.getSurename()));
    repeatingView.add(new Label(repeatingView.newChildId(), person.getAddress()));    
    repeatingView.add(new Label(repeatingView.newChildId(), person.getEmail()));
    item.add(repeatingView); 
  }
};
add(dataView);

Please note that in the code above we have used also a RepeatingView component to populate the rows of the table.

In the next paragraph we will see a similar example that adds support for data paging.

Data paging

To enable data paging on a pageable repeaters, we must first set the number of items to display per page with method setItemsPerPage(long items). Then, we must attach the repeater to panel PagingNavigator (placed in package org.apache.wicket.markup.html.navigation .paging) which is responsible for rendering a navigation bar containing the links illustrated in the following picture:

Project PageDataViewExample mixes a DataView component with a PagingNavigator to display the list of all countries of the world sorted by alphabetical order1. Here is the initialization code of the project home page:

HTML:

<table>
  <tr>
    <th>ISO 3166-1</th><th>Name</th><th>Long name</th><th>Capital</th><th>Population</th>
  </tr>
  <tr wicket:id="rows">
    <td wicket:id="dataRow"></td>
  </tr>
</table>

Java Code:

public HomePage(final PageParameters parameters) {
  super(parameters);
  //method loadCountriesFromCsv is defined elsewhere in the class.
  //It reads countries data from a csv file and returns each row as an array of Strings.
  List<String[]> countries = loadCountriesFromCsv();
  ListDataProvider<String[]> listDataProvider = new ListDataProvider<String[]>(countries);
  DataView<String[]> dataView = new DataView<String[]>("rows", listDataProvider) {
    @Override
    protected void populateItem(Item<String[]> item) {
      String[] countriesArr = item.getModelObject();
      RepeatingView repeatingView = new RepeatingView("dataRow");
      for (int i = 0; i < countriesArr.length; i++){
         repeatingView.add(new Label(repeatingView.newChildId(), countriesArr[i]));
      }
      item.add(repeatingView);
    }
  };
  dataView.setItemsPerPage(15);
  add(dataView);
  add(new PagingNavigator("pagingNavigator", dataView));
}

The data of a single country (ISO code, name, long name, capital and population) are handled with an array of strings. The usage of PagingNavigator it's quite straightforward as we need to simply pass the pageable repeater to its constructor.

To explore the other pageable repeaters shipped with Wicket you can visit the page at http://www.wicket-library.com/wicket-examples/repeater/ where you can find live examples of these components.

Wicket provides also component PageableListView which is a sublcass of ListView that implements interface IPageable, hence it can be considered a pageable repeaters even if it doesn't use interface IDataProvider as data source.

12.5 Summary

In this chapter we have explored the built-in set of components called repeaters which are designed to repeat their own markup in output to display a set of items. We have started with component RepeatingView which can be used to repeat a simple markup fragment.

Then, we have seen components ListView and RefreshingView which should be used when the markup to repeat contains nested components to populate.

Finally, we have discussed those repeaters that support data paging and that are called pageable repeaters. We ended the chapter looking at an example where a pageable repeater is used with panel PagingNavigator to make its dataset navigable by the user.

13 Internationalization with Wicket

13.1 Localization

As we have seen in chapter 10, the infrastructure of feedback messages is built on top of Java internationalization (i18n) support, so it should not be surprising that the same infrastructure is used also for localization purpose. However, while so far we have used only the <ApplicationClassName>.properties file to store our custom messages, in this chapter we will see that also pages, components, validators and even Java packages can have their own resource bundles. This allows us to split bundles into multiple files keeping them close to where they are used. But before diving into the details of internationalization with Wicket, it's worthwhile to quickly review how i18n works under Java, see what classes are involved and how they are integrated into Wicket.

Providing a full description of Java support for i18n is clearly out of the scope of this document. If you need more informations about this topic you can find them in the JavaDocs and in the official i18n tutorial .

Class Locale and ResourceBundle

Class java.util.Locale represents a specific country or language of the world and is used in Java to retrieve other locale-dependent informations like numeric and date formats, the currency in use in a country and so on. Such kind of informations are accessed through special entities called resource bundles which are implemented by class java.util.ResourceBundle. Every resource bundle is identified by a full name which is built using four parameters: a base name (which is required), a language code, a country code and a variant (which are all optional). These three optional parameters are provided by an instance of Locale with its three corresponding getter methods: getLanguage(), getCountry() and getVariant(). Parameter language code is a lowercase ISO 639 2-letter code (like zh for Chinese, de for German and so on) while country code is an uppercase ISO 3166 2-letter code (like CN for China, DE for Germany and so on). The final full name will have the following structure (NOTE: tokens inside squared brackets are optional):

<base name>[ <language code>[ <COUNTRY_CODE>[_<variant code>]]]

For example a bundle with MyBundle as base name and localized for Mandarin Chinese (language code zh, country code CH, variant cmn) will have MyBundle_zh_CH_cmn as full name. A base name can be a fully qualified class name, meaning that it can include a package name before the actual base name. The specified package will be the container of the given bundle. For example if we use org.foo.MyBundle as base name, the bundle named MyBundle will be searched inside package org.foo. The actual base name (MyBundle in our example) will be used to build the full name of the bundle following the same rules seen above. ResourceBundle is an abstract factory class, hence it exposes a number of factory methods named getBundle to load a concrete bundle. Without going into too much details we can say that a bundle corresponds to a file in the classpath. To find a file for a given bundle, getBundle needs first to generate an ordered list of candidate bundle names. These names are the set of all possible full names for a given bundle. For example if we have org.foo.MyBundle as base name and the current locale is the one seen before for Mandarin Chinese, the candidate names will be:

1. org.foo.MyBundle_zh_CH_cmn
2. org.foo.MyBundle_zh_CH
3. org.foo.MyBundle_zh
4. org.foo.MyBundle

The list of these candidate names is generated starting from the most specific one and subtracting an optional parameter at each step. The last name of the list corresponds to the default resource bundle which is the most general name and is equal to the base name. Once that getBundle has generated the list of candidate names, it will iterate over them to find the first one for which is possible to load a class or a properties file. The class must be a subclass of ResourceBundle having as class name the full name used in the current iteration. If such a class is not found, getBundle will try to locate a properties file having a file name equals to the current full name (Java will automatically append extension .properties to the full name). For example given the resource bundle of the previous example, Java will search first for class org.foo.MyBundle_zh_CH_cmn and then for file MyBundle_zh_CH_cmn.properties inside package org.foo. If no file is found for any of the candidate names, a MissingResourceException will be thrown. Bundles contains local-dependent string resources identified by a key that is unique in the given bundle. So once we have obtained a valid bundle we can access these objects with method getString (String key).

As we have seen before working with feedback messages, in Wicket most of the times we will work with properties files rather than with bundle classes. In chapter 10 we used a properties file having as base name the class name of the application class and without any information about the locale. This file is the default resource bundle for a Wicket application. In paragraph 12.4 we will explore the algorithm used in Wicket to locate the available bundles for a given component. Once we have learnt how to leverage this algorithm, we will be able to split our bundles into more files organized in a logical hierarchy.

13.2 Localization in Wicket

A component can get the current locale in use calling its method getLocale(). By default this method will be recursively called on component's parent containers until one of them returns a valid locale. If no one of them returns a locale, this method will get the one associated with the current user session. This locale is automatically generated by Wicket in accordance with the language settings of the browser.

Developers can change the locale of the current session with Session's method setLocale (Locale locale):

Session.get().setLocale(locale)

Style and variation parameters for bundles

In addition to locale's informations, Wicket supports two further parameters to identify a resource bundle: style and variation. Parameter style is a string value and is defined at session-level. To set/get the style for the current session we can use the corresponding setter and getter of class Session:

Session.get().setStyle("myStyle");
Session.get().getStyle();

If set, style's value contributes to the final full name of the bundle and it is placed between the base name and the locale's informations:

<base name>[ style][ <language code>[ <COUNTRY_CODE>[ <variant code>]]]

Wicket gives the priority to candidate names containing the style information (if available). The other parameter we can use for localization is variation. Just like style also variation is a string value, but it is defined at component-level. The value of variation is returned by Component's method getVariation(). By default this method returns the variation of the parent component or a null value if a component hasn't a parent (i.e. it's a page). If we want to customize this parameter we must overwrite method getVariation and make it return the desired value.

Variation's value contributes to the final full name of the bundle and is placed before style parameter:

<base name>[ variation][_style][ <language code>[ <COUNTRY_CODE>[ <variant code>]]]

Using XML files as resource bundles

Java uses the standard character set ISO 8859-11 to encode text files like properties files. Unfortunately ISO 8859-1 does not support most of the extra-European languages like Chinese or Japanese. The only way to use properties files with such languages is to use escaped Unicode characters, but this leads to not human-readable files. For example if we wanted to write the word 'website' in simplified Chinese (the ideograms are 网站) we should write the Unicode characters u7F51u7AD9. That's why starting from version 1.5, Java introduced the support for XML files as resource bundles. XML files are generally encoded with character sets UTF-8 or UTF-16 which support every symbol of the Unicode standard. In order to be a valid resource bundle the XML file must conform to the DTD available at http://java.sun.com/dtd/properties.dtd .

Here is an example of XML resource bundle taken from project LocalizedGreetings (file WicketApplication_zh.properties.xml) containing the translation in simplified Chinese of the greeting message “Welcome to the website!”:

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE properties SYSTEM "http://java.sun.com/dtd/properties.dtd">
<properties>
	<entry key="greetingMessage">欢迎光临本网站！</entry>
</properties>

To use XML bundles in Wicket we don't need to put in place any additional configuration. The only rule we have to respect with these files is to use properties.xml as extension while their base name follows the same rules seen so far for bundle names.

Reading bundles from code

Class Component makes reading bundles very easy with method getString(String key). This method searches for a resource with the given key looking into the resource bundles visited by the lookup algorithm illustrated in paragraph 12.4. For example if we have a greeting message with key greetingMessage in our application's resource bundle, we can read it from our component code with this instruction:

getString("greetingMessage");

Localization of bundles in Wicket

In chapter 10 we have used as resource bundle the properties file placed next to our application class. This file is the default resource bundle for the entire application and it is used by the lookup algorithm if it doesn't find any better match for a given component and locale. If we want to provide localized versions of this file we must simply follow the rules of Java i18n and put our translated resources into another properties file with a name corresponding to the desired locale. For example project LocalizedGreetings comes with the default application's properties file ( WicketApplication.properties) containing a greeting message:

greetingMessage=Welcome to the site!

Along with this file we can also find a bundle for German (WicketApplication_de.properties) and another one in XML format for simplified Chinese (WicketApplication_zh.properties.xml). The example project consists of a single page (HomePage.java) displaying the greeting message. The current locale can be changed with a drop-down list and the possible options are English (the default one), German and simplified Chinese:

The label displaying the greeting message has a custom read-only model which returns the message with method getString. The initialization code for this label is this:

AbstractReadOnlyModel<String> model = new AbstractReadOnlyModel<String>() {			
			@Override
			public String getObject() {
				return getString("greetingMessage");			
			}
};
add(new Label("greetingMessage", model));

Class org.apache.wicket.model.AbstractReadOnlyModel is a convenience class for implementing read-only models. In this project we have implemented a custom read-only model for illustrative purposes only because Wicket already provides built-in models for the same task. We will see them in paragraph 12.6.

The rest of the code of the home page builds the stateless form and the drop-down menu used to change the locale.

List<Locale> locales = Arrays.asList(Locale.ENGLISH, Locale.CHINESE, Locale.GERMAN);
final DropDownChoice<Locale> changeLocale = 
             new DropDownChoice<Locale>("changeLocale", new Model<Locale>(), locales);
StatelessForm form = new StatelessForm("form"){
	@Override
	protected void onSubmit() {
		Session.get().setLocale(changeLocale.getModelObject());
	}
};
setStatelessHint(true);
add(form.add(changeLocale))

Localization of markup files

Although resource bundles exist to extract local-dependent elements from our code and from UI components, in Wicket we can decide to provide different markup files for different locale settings. Just like standard markup files, by default localized markup files must be placed next to component's class and their file name must contain the locale's informations. In the following picture, CustomPanel comes with a standard (or default) markup file and with another one localized for German:

When the current locale corresponds to German country (language code de), markup file CustomPanel_de.html will be used in place of the default one.

Reading bundles with tag <wicket:message>

String resources can be also retrieved directly from markup code using tag <wicket:message>. The key of the desired resource is specified with attribute key:

<wicket:message key="greetingMessage">message goes here</wicket:message>

wicket:message can be adopted also to localize the attributes of a tag. The name of the attribute and the resource key are expressed as a colon-separated value. In the following markup the content of attribute value will be replaced with the localized resource having 'key4value' as key:

<input type="submit" value="Preview value" wicket:message="value:key4value"/>

If we want to specify multiple attributes at once, we can separate them with a coma:

<input type="submit" value="Preview value" wicket:message="value:key4value,title:key4title"/>

13.3 Bundles lookup algorithm

As we hinted at the beginning of this chapter, by default Wicket provides a very flexible algorithm to locate the resource bundles available for a given component. In this paragraph we will learn how this default lookup algorithm works and which options it offers to manage our bundle files.

Localizing pages and panels

Similarly to application class, also component classes can have their own bundle files having as base name the class name of the related component and placed in the same package. So for example if class CustomPanel is a custom panel we created, we can provide it with a default bundle file called CustomPanel.properties containing the textual resources used by this panel. This rule applies to page classes as well:

One fundamental thing to keep in mind when we work with these kinds of bundles is that the lookup algorithm gives priority to the bundles of the containers of the component that is requesting a localized resource. The more a container is higher in the hierarchy, the bigger is its priority over the other components. This mechanism was made to allow containers to overwrite resources used by children components. As a consequence the values inside the resource bundle of a page will have the priority over the other values with the same key defined in the bundles of children components.

To better grasp this concept let's consider the component hierarchy depicted in the following picture:

If CustomPanel tries to retrieve the string resource having 'message' as key, it will get the value 'Wellcome!' and not the one defined inside its own bundle file.

The default message-lookup algorithm is not limited to component hierarchy but it also includes the class hierarchy of every component visited in the search strategy described so far. This makes bundle files inheritable, just like markup files. When the hierarchy of a container component is explored, any ancestor has the priority over children components. Consider for example the hierarchy in the following picture:

Similarly to the previous example, the bundle owned by CustomPanel is overwritten by the bundle of page class BasePage (which has been inherited by CustomPage).

Component-specific resources

In order to make a resource specific for a given child component, we can prefix the message key with the id of the desired component. Consider for example the following code and bundle of a generic page:

Page code:

add(new Label("label",new ResourceModel("labelValue")));
add(new Label("anotherLabel",new ResourceModel("labelValue")));

Page bundle:

labelValue=Default value
anotherLabel.labelValue=Value for anotherLabel

Label with id anotherLabel will display the value 'Value for anotherLabel' while label label will display 'Default value'. In a similar fashion, parent containers can specify a resource for a nested child component prepending also its relative path (the path is dot-separated):

Page code:

Form form = new Form("form");
form.add(new Label("anotherLabel",new ResourceModel("labelValue")));
add(form);

Page bundle:

labelValue=Default value
anotherLabel.labelValue=Value for anotherLabel
form.anotherLabel.labelValue=Value for anotherLabel inside form

With the code and the bundle above, the label inside the form will display the value 'Value for anotherLabel inside form'.

Package bundles

If no one of the previous steps can find a resource for the given key, the algorithm will look for package bundles. These bundles have package as base name and they can be placed in one of the package of our application:

Packages are traversed starting from the one containing the component requesting for a resource and going up to the root package.

Bundles for feedback messages

The algorithm described so far applies to feedback messages as well. In case of validation errors, the component that has caused the error will be considered as the component which the string resource is relative to. Furthermore, just like application class and components, validators can have their own bundles placed next to their class and having as base name their class name. This allows us to distribute validators along with the messages they use to report errors:

Validator's resource bundles have the lowest priority in the lookup algorithm. They can be overwritten by resource bundles of components, packages and application class.

Extending the default lookup algorithm

Wicket implements the default lookup algorithm using the strategy pattern1. The concrete strategies are abstracted with the interface org.apache.wicket.resource.loader.IStringResource Loader. By default Wicket uses the following implementations of IStringResourceLoader (sorted by execution order):

1. ComponentStringResourceLoader: implements most of the default algorithm. It searches for a given resource across bundles from the container hierarchy, from class hierarchy and from the given component.
2. PackageStringResourceLoader: searches into package bundles.
3. ClassStringResourceLoader: searches into bundles of a given class. By default the target class is the application class.
4. ValidatorStringResourceLoader: searches for resources into validator's bundles. A list of validators is provided by the form component that failed validation.
5. InitializerStringResourceLoader: this resource allows internationalization to interact with the initialization mechanism of the framework that will be illustrated in paragraph 15.4.

Developer can customize lookup algorithm removing default resource loaders or adding custom implementations to the list of the resource loaders in use. This task can be accomplished using method getStringResourceLoaders of setting interface org.apache.wicket.settings. IResourceSettings:

@Override
public void init()
{
  super.init();
  //retrieve IResourceSettings and then the list of resource loaders
  List<IStringResourceLoader> resourceLoaders= getResourceSettings(). 
                                                                 getStringResourceLoaders();
  //customize the list...

13.4 Localization of component's choices

Components that inherit from AbstractChoice (such as DropDownChoice, CheckBoxMultiple Choice and RadioChoice) must override method localizeDisplayValues and make it return true to localize the values displayed for their choices. By default this method return false so values are displayed as they are. Once localization is activated we can use display values as key for our localized string resources. In project LocalizedChoicesExample we have a drop-down list that displays four colors (green, red, blue, and yellow) which are localized in three languages (English, German and Italian). The current locale can be changed with another drop-down menu (in a similar fashion to project LocalizedGreetings). The code of the home page and the relative bundles are the following:

Java code:

public HomePage(final PageParameters parameters) {
	super(parameters);
	List<Locale> locales = Arrays.asList(Locale.ENGLISH, Locale.ITALIAN, Locale.GERMAN);
	List<String> colors = Arrays.asList("green", "red", "blue", "yellow");
	final DropDownChoice<Locale> changeLocale = new DropDownChoice<Locale>("changeLocale", 
                                                    new Model<Locale>(), locales);
	StatelessForm form = new StatelessForm("form"){
		@Override
		protected void onSubmit() {
			Session.get().setLocale(changeLocale.getModelObject());
		}
	};
	DropDownChoice<String> selectColor = new DropDownChoice<String>("selectColor", new 
                                                            Model<String>(), colors){
		@Override
		protected boolean localizeDisplayValues() {
			return true;
		}
	};
	form.add(selectColor);
	add(form.add(changeLocale));
    }

Default bundle (English):

selectColor.null=Select a color
green=Green
red=Red
blue=Blue
yellow=Yellow

German bundle:

selectColor.null=Wahlen sie eine farbe
green=Grun
red=Rot
blue=Blau
yellow=Gelb

Italian bundle:

selectColor.null=Scegli un colore
green=Verde
red=Rosso
blue=Blu
yellow=Giallo

Along with the localized versions of colors names, in the bundles above we can also find a custom value for the placeholder text (“Chose one ”) used for null value. The resource key for this resource is 'null' or '<component id>.null' if we want to make it component-specific.

13.5 Internationalization and Models

Internationalization is another good chance to taste the power of models. Wicket provides two built-in models to better integrate our components with string resources: they are ResourceModel and StringResourceModel.

ResourceModel

Model org.apache.wicket.model.ResourceModel acts just like the read-only model we have implemented in paragraph 12.3.4. It simply retrieves a string resource corresponding to a given key:

//build a ResourceModel for key 'greetingMessage'
new ResourceModel("greetingMessage");

We can also specify a default value to use if the requested resource is not found:

//build a ResourceModel with a default value
new ResourceModel("notExistingResource", "Resource not found.");

StringResourceModel

Model org.apache.wicket.model.StringResourceModel allows to work with complex and dynamic string resources containing parameters and property expressions. The basic constructor of this model takes in input a resource key and another model. This further model can be used by both the key and the related resource to specify dynamic values with property expressions. For example let's say that we are working on an e-commerce site which has a page where users can see an overview of their orders. To handle the state of user's orders we will use the following bean and enum (the code is from project StringResourceModelExample):

Bean:

public class Order implements Serializable {
	private Date orderDate;
	private ORDER_STATUS status;
	public Order(Date orderDate, ORDER_STATUS status) {
		super();
		this.orderDate = orderDate;
		this.status = status;
	}
	//Getters and setters for private fields
}

Enum:

public enum ORDER_STATUS {
	PAYMENT_ACCEPTED(0),
	IN_PROGRESS(1),
	SHIPPING(2),
	DELIVERED(3);
	private int code;
	//Getters and setters for private fields	
}

Now what we want to do in this page is to print a simple label which displays the status of an order and the date on which the order has been submitted. All the informations about the order will be passed to a StringResourceModel with a model containing the bean Order. The bundle in use contains the following key/value pairs:

orderStatus.0=Your payment submitted on ${orderDate} has been accepted.
orderStatus.1=Your order submitted on ${orderDate} is in progress.
orderStatus.2=Your order submitted on ${orderDate} has been shipped.
orderStatus.3=Your order submitted on ${orderDate} has been delivered.

The values above contain a property expression (${orderDate}) that will be evaluated on the data object of the model. The same technique can be applied to the resource key in order to load the right resource according to the state of the order:

Order order = new Order(new Date(), ORDER_STATUS.IN_PROGRESS);
add(new Label("orderStatus", new StringResourceModel("orderStatus.${status.code}", Model.of(order))));

As we can see in the code above also the key contains a property expression (${status.code}) which makes its value dynamic. In this way the state of an object (an Order in our example) can determinate which resource will be loaded by StringResourceModel. If we don't use properties expressions we can provide a null value as model and in this case StringResourceModel will behave exactly as a ResourceModel. StringResourceModel supports also the same parameter substitution used by standard class java.text.MessageFormat. Parameters can be generic objects but if we use a model as parameter, StringResourceModel will use the data object inside it as actual value (it will call getObject on the model). Parameters are passed to constructor as a vararg argument. Here is an example of usage of parameter substitution:

Java code:

PropertyModel propertyModel = new PropertyModel<Order>(order, "orderDate");
//build a string model with two parameters: a property model and an integer value
StringResourceModel srm = new StringResourceModel("orderStatus.delay", null, propertyModel, 3);

Bundle:

orderStatus.delay=Your order submitted on ${0} has been delayed by {1} days.

One further parameter we can specify when we build a StringResourceModel is the component that must be used by the lookup algorithm. Normally this parameter is not relevant, but if we need to use a particular bundle owned by a component not considered by the algorithm, we can specify this component as second parameter. If we pass all possible parameters to StringResourceModel's constructor we obtain something like this:

new StringResourceModel("myKey", myComponent, myModel, param1, param2, param3,...);

13.6 Summary

Internationalization is a mandatory step if we want to take our applications (and our business!) abroad. Choosing the right strategy to manage our localized resources is fundamental to avoid to make a mess of them. In this chapter we have explored the built-in support for localization provided by Wicket, and we have learnt which solutions it offers to manage resource bundles. In the final part of the chapter we have seen how to localize the options displayed by a component (such as DropDownChoice or RadioChoice) and we also introduced two new models specifically designed to localize our components without introducing in their code any detail about internationalization.

14 Resource management with Wicket

In paragraph 4.6 we have already seen how to add CSS and JavaScript contents to the header section of the page. In the first half of this chapter we will learn a more sophisticated technique that allows us to manage static resources directly from code and “pack” them with our custom components.

Then, in the second part of the chapter we will see how to implement custom resources to enrich our web application with more complex and dynamic functionalities.

14.1 Static vs dynamic resources

In Wicket a resource is an entity that can interact with the current request and response and It must implement interface org.apache.wicket.request.resource.IResource. This interface defines just method respond(IResource.Attributes attributes) where the nested class IResource. Attributes provides access to request, response and page parameters objects.

Resources can be static or dynamic. Static resources don't entail any computational effort to be generated and they generally correspond to a resource on the filesystem. On the contrary dynamic resources are generated on the fly when they are requested, following a specific logic coded inside them.

An example of dynamic resource is the built-in class CaptchaImageResource in package org. apache.wicket.extensions.markup.html.captcha which generates a captcha image each time is rendered.

As we will see in paragraph 13.6, developers can build custom resources extending base class org.apache.wicket.request.resource.AbstractResource.

14.2 Resource references

Most of the times in Wicket we won't directly instantiate a resource but rather we will use a reference to it. Resource references are represented by abstract class org.apache.wicket.request.resource .ResourceReference which returns a concrete resource with factory method getResource(). In this way we can lazy-initialize resources loading them only the first time they are requested.

14.3 Package resources

With HTML we use to include static resources in our pages using tags like <script>, <link> or <img>. This is what we have done so far writing our custom panels and pages. However, when we work with a component-oriented framework like Wicket, this classic approach becomes inadequate because it makes custom components hardly reusable. This happens when a component depends on a big number of resources. In such a case, if somebody wanted to use our custom component in his application, he would be forced to know which resources it depends on and make them available.

To solve this problem Wicket allows us to place static resource files into component package (like we do with markup and properties files) and load them from component code.

These kinds of resources are called package resources (a CSS and a JavaScript file in this screenshot):

With package resources custom components become independent and self-contained and client code can use them without worrying about their dependencies.

To load package resources Wicket provides class org.apache.wicket.request.resource. PackageResourceReference.

To identify a package resource we need to specify a class inside the target package and the name of the desired resource (most of the times this will be a file name).

In the following example taken from project ImageAsPackageRes, CustomPanel loads a picture file available as package resource and it displays it in a <img> tag using the built-in component org. apache.wicket.markup.html.image.Image:

HTML:

<html>
<head>...</head>
<body>
<wicket:panel>
	Package resource image: <img wicket:id="packageResPicture"/>
</wicket:panel>
</body>
</html>

Jave Code:

public class CustomPanel extends Panel {
	public CustomPanel(String id) {
		super(id);
		PackageResourceReference resourceReference = 
	            new PackageResourceReference(getClass(), "calendar.jpg");
		add(new Image("packageResPicture", resourceReference));
	}
}

Wicket will take care of generating a valid URL for file calendar.jpg. URLs for package resources have the following structure:

<path to application root>/wicket/resource/<fully qualified class name>/<resource file name> -<ver-<id>>.

In our example the URL for our picture file calendar.jpg is the following:

./wicket/resource/org.wicketTutorial.CustomPanel/calendar-ver-1297887542000.jpg

The first part of the URL is the relative path to the application root. In our example our page is already at the application's root so we have only a single-dotted segment. The next two segments, wicket and resource, are respectively the namespace and the identifier for resources seen in paragraph 8.6.4.

The fourth segment is the fully qualified name of the class used to locate the resource and it is the scope of the package resource. In the last segment of the URL we can find the name of the resource (the file name).

As you can see Wicket has automatically appended to the file name a version identifier (ver-1297887542000). When Wicket runs in DEVELOPMENT mode this identifier contains the timestamp in millisecond indicating the last time the resource file was modified. This can be useful when we are developing our application and resource files are frequently modified. Appending the timestamp to the original name we are sure that our browser will use always the last version of the file and not an old, out of date, cached version.

When instead Wicket is running in DEPLOYMENT mode, the version identifier will contain the MD5 digest of the file instead of the timestamp. The digest is computed only the first time the resource is requested. This perfectly makes sense as static resources don't change so often when our application runs into production environment and when this appends the application is redeployed.

Package resources can be localized following the same rules seen for resource bundles and markup files:

In the example illustrated in the picture above, if we try to retrieve package resource calendar.jpg when the current locale is set to French, the actual file returned will be calendar_fr.jpg.

Using package resources with tag <wicket:link>

In paragraph 8.3 we have used tag <wicket:link> to automatically create links to bookmarkable pages. The same technique can be used also for package resources in order to use them directly from markup file. Let's assume for example that we have a picture file called icon.png placed in the same package of the current page. Under these conditions we can display the picture file using the following markup fragment:

<wicket:link>
   <img src="icon.png"/>
</wicket:link>

In the example above Wicket will populate the attribute src with the URL corresponding to the package resource icon.png. <wicket:link> supports also tag <link> for CSS files and tag <script> for JavaScript files.

14.4 Adding resources to page header section

Wicket comes with interface org.apache.wicket.markup.html.IHeaderContributor which allows components and behaviors (which will be introduced later in paragraph 15.1) to contribute to the header section of their page. The only method defined in this interface is renderHead (IHeaderResponse response) where IHeaderResponse is an interface which defines method render(HeaderItem item) to write static resources or free-form text into the header section of the page.

Header entries are instances of abstract class org.apache.wicket.markup.head.HeaderItem. Wicket provides a set of built-in implementations of this class suited for the most common types of resources. With the exception of PriorityHeaderItem, every implementation of HeaderItem is an abstract factory class:

• CssHeaderItem: represents a CSS resource. Factory methods provided by this class are forReference which takes in input a resource reference, forUrl which creates an CSS item from a given URL and forCSS which takes in input an arbitrary CSS string and an optional id value to identify the resource.
• JavaScriptHeaderItem: represents a JavaScript resource. Just like CssHeaderItem it provides factory methods forReference and forUrl along with method forScript which takes in input an arbitrary string representing the script and an optional id value to identify the resource.
• OnDomReadyHeaderItem: it adds JavaScript code that will be executed after the DOM has been built, but before external files (such as picture, CSS, etc...) have been loaded. The class provides a factory method forScript which takes in input an arbitrary string representing the script to execute.
• OnEventHeaderItem: the JavaScript code added with this class is executed when a specific JavaScript event is triggered on a given DOM element. The factory method is forScript(String target, String event, CharSequence javaScript), where target is the id of a DOM element (or the element itself), event is the event that must trigger our code and javaScript is the code to execute.
• OnLoadHeaderItem: the JavaScript code added with this class is executed after the whole page is loaded, external files included. The factory method is forScript(CharSequence javaScript).
• PriorityHeaderItem: it wraps another header item and ensures that it will have the priority over the other items during rendering phase.
• StringHeaderItem: with this class we can add an arbitrary text to the header section. Factory method is forString(CharSequence string).

In the following example our custom component loads a CSS file as a package resource (placed in the same package) and it adds it to header section.

public class MyComponent extends Component{
  @Override
  public void renderHead(IHeaderResponse response) {
      PackageResourceReference cssFile = 
                            new PackageResourceReference(this.getClass(), "style.css");
    CssHeaderItem cssItem = CssHeaderItem.forReference(cssFile);
    response.render(cssItem);
  }
}

14.5 Resource dependencies

Class ResourceReference allows to specify the resources it depends on overriding method getDependencies(). The method returns an iterator over the set of HeaderItems that must be rendered before the resource referenced by ResourceReference can be used. This can be really helpful when our resources are JavaScript or CSS libraries that in turn depend on other libraries.

For example we can use this method to ensure that a custom reference to JQueryUI library will find JQuery already loaded in the page:

Url jqueyuiUrl = Url.parse("https://ajax.googleapis.com/ajax/libs/jqueryui/" + 
                                                                 "1.10.2/jquery-ui.min.js");
UrlResourceReference jqueryuiRef = new UrlResourceReference(jqueyuiUrl){
	@Override
	public Iterable<? extends HeaderItem> getDependencies() {
		Application application = Application.get();
		ResourceReference jqueryRef = application.getJavaScriptLibrarySettings(). 
                                             getJQueryReference();
		return Arrays.asList(JavaScriptHeaderItem.forReference(jqueryRef));
	}
};

Please note that in the code above we have built a resource reference using a URL to the desired library instead of a package resource holding the physical file.

The same method getDependencies() is defined also for class HeaderItem.

14.6 Custom resources

In Wicket the best way to add dynamic functionalities to our application (such as csv export, a pdf generated on the fly, etc...) is implementing a custom resource. In this paragraph as example of custom resource we will build a basic RSS feeds generator which can be used to publish feeds on our site (project CustomResourceMounting). Instead of generating a RSS feed by hand we will use Rome framework and its utility classes.

As hinted above in paragraph 13.1, class AbstractResource can be used as base class to implement new resources. This class defines abstract method newResourceResponse which is invoked when the resource is requested. The following is the code of our RSS feeds generator:

public class RSSProducerResource extends AbstractResource {
  @Override
  protected ResourceResponse newResourceResponse(Attributes attributes) {
    ResourceResponse resourceResponse = new ResourceResponse();
    resourceResponse.setContentType("text/xml");
    resourceResponse.setTextEncoding("utf-8");
    resourceResponse.setWriteCallback(new WriteCallback()
    {
      @Override
      public void writeData(Attributes attributes) throws IOException
      {
        OutputStream outputStream = attributes.getResponse().getOutputStream();
        Writer writer = new OutputStreamWriter(outputStream);
        SyndFeedOutput output = new SyndFeedOutput();
            try {
          output.output(getFeed(), writer);
        } catch (FeedException e) {
          throw new WicketRuntimeException("Problems writing feed to response...");
        }
      }      
    });
    return resourceResponse;
  }
  // method getFeed()…
}

Method newResourceResponse returns an instance of ResourceResponse representing the response generated by the custom resource. Since RSS feeds are based on XML, in the code above we have set the type of the response to text/xml and the text encoding to utf-8.

To specify the content that will be returned by our resource we must also provide an implementation of inner class WriteCallback which is responsible for writing content data to response's output stream. In our project we used class SyndFeedOutput from Rome framework to write our feed to response. Method getFeed() is just an utility method that generates a sample RSS feed (which is an instance of interface com.sun.syndication.feed.synd.SyndFeed).

Now that we have our custom resource in place, we can use it in the home page of the project. The easiest way to make a resource available to users is to expose it with link component ResourceLink:

add(new ResourceLink("rssLink", new RSSProducerResource()));

In the next paragraphs we will see how to register a resource at application-level and how to mount it to an arbitrary URL.

14.7 Mounting resources

Just like pages also resources can be mounted to a specific path. Class WebApplication provides method mountResource which is almost identical to mountPage seen in paragraph 8.6.1:

@Override
public void init() {
  super.init();
  //resource mounted to path /foo/bar
  ResourceReference resourceReference = new ResourceReference("rssProducer"){
     RSSReaderResource rssResource = new RSSReaderResource();
     @Override
     public IResource getResource() {
	return rssResource;
  }};
  mountResource("/foo/bar", resourceReference);
}

With the configuration above (taken from project CustomResourceMounting) every request to /foo/bar will be served by the custom resource built in the previous paragraph.

Parameter placeholders are supported as well:

@Override
public void init() {
  super.init();
  //resource mounted to path /foo with a required indexed parameter
  ResourceReference resourceReference = new ResourceReference("rssProducer"){
     RSSReaderResource rssResource = new RSSReaderResource();
     @Override
     public IResource getResource() {
	return rssResource;
  }};
  mountResource("/bar/${baz}", resourceReference);
}

14.8 Shared resources

Resources can be added to a global registry in order to share them at application-level. Shared resources are identified by an application-scoped key and they can be easily retrieved at a later time using reference class SharedResourceReference. The global registry can be accessed with Application's method getSharedResources. In the following excerpt of code (taken again from project CustomResourceMounting) we register an instance of our custom RSS feeds producer as application-shared resource:

//init application's method
  @Override
  public void init(){
    RSSProducerResource rssResource = new RSSProducerResource();
    // …
    getSharedResources().add("globalRSSProducer", rssResource);    
  }

Now to use an application-shared resource we can simply retrieve it using class SharedResource Reference and providing the key previously used to register the resource:

add(new ResourceLink("globalRssLink", new SharedResourceReference("globalRSSProducer")));

The URL generated for application shared resources follows the same pattern seen for package resources:

./wicket/resource/org.apache.wicket.Application/globalRSSProducer

The last segment of the URL is the key of the resource while the previous segment contains the scope of the resource. For application-scoped resources the scope is always the fully qualified name of class Application. This should not be surprising since global resources are visible at application level (i.e. the scope is the application).

Package resources are also application-shared resources but they don't need to be explicitly registered.

Remember that we can get the URL of a resource reference using method urlFor(ResourceReference resourceRef, PageParameters params ) available with both class RequestCycle and class Component.

14.9 Customizing resource loading

Wicket loads application's resources delegating this task to a resource locator represented by interface org.apache.wicket.core.util.resource.locator.IResourceStreamLocator. To retrieve or modify the current resource locator we can use the getter and setter methods defined by setting interface IResourceSettings:

//init application's method
  @Override
  public void init(){   
    //get the resource locator 
    getResourceSettings().getResourceStreamLocator();
    //set the resource locator    
    getResourceSettings().setResourceStreamLocator(myLocator);
  }

The default locator used by Wicket is class ResourceStreamLocator which in turn tries to load a requested resource using a set of implementations of interface IResourceFinder. This interface defines method find(Class class, String pathname) which tries to resolve a resource corresponding to the given class and path.

The default implementation of IResourceFinder used by Wicket is ClassPathResourceFinder which searches for resources into the application class path. This is the implementation we have used so far in our examples. However some developers may prefer storing markup files and other resources in a separate folder rather than placing them side by side with Java classes.

To customize resource loading we can add further resource finders to our application in order to extend the resource-lookup algorithm to different locations. Wicket already comes with two other implementations of IResourceFinder designed to search for resources into a specific folder on the file system. The first is class Path and it's defined in package org.apache.wicket.util.file. The constructor of this class takes in input an arbitrary folder that can be expressed as a string path or as an instance of Wicket utility class Folder (in package org.apache.wicket.util.file). The second implementation of interface IResourceFinder is class WebApplicationPath which looks into a folder placed inside webapp's root path (but not inside folder WEB-INF).

Project CustomFolder4MarkupExample uses WebApplicationPath to load the markup file and the resource bundle for its home page from a custom folder. The folder is called markupFolder and it is placed in the root path of the webapp. The following picture illustrates the file structure of the project:

As we can see in the picture above, we must preserve the package structure also in the custom folder used as resource container. The code used inside application class to configure WebApplicationPath is the following:

@Override
public void init()
{
	getResourceSettings().getResourceFinders().add(
			new WebApplicationPath(getServletContext(), "markupFolder"));
}

Method getResourceFinders() defined by setting interface IResourceSettings returns the list of resource finders defined in our application. The constructor of WebApplicationPath takes in input also an instance of standard interface javax.servlet.ServletContext which can be retrieved with WebApplication's method getServletContext().

By default, if resource files can not be found inside application classpath, Wicket will search for them inside “resources” folder. You may have noted this folder in the previous picture. It is placed next to the folder “java” containing our source files:

This folder can be used to store resource files without writing any configuration code.

14.10 Summary

In this chapter we have learnt how to manage resources with the built-in mechanism provided by Wicket. With this mechanism we handle resources from Java code and Wicket will automatically take care of generating a valid URL for them. We have also seen how resources can be bundled as package resources with a component that depends on them to make it self-contained.

Then, in the second part of the chapter, we have built a custom resource and we have learnt how to mount it to an arbitrary URL and how to make it globally available as shared resource.

Finally, in the last part of the paragraph we took a peek at the mechanism provided by the framework to customize the locations where the resource-lookup algorithm searches for resources.

15 An example of integration with JavaScript

15.1 What we want to do...

For end-users a datepicker is one of the most appreciated widget. It allows to simply edit a date value with the help of a user-friendly pop-up calendar. That's why nearly all UI frameworks provide a version of this widget.

Popular JavaScript libraries like YUI and JQuery come with a ready-to-use datepicker to enrich the user experience of our web applications. Wicket already provides a component which integrates a text field with a calendar widget from YUI library, but there is no built-in component that uses a datepicker based on JQuery library.

As both JQuery and its child project JQueryUI have gained a huge popularity in the last years, it's quite interesting to see how to integrate them in Wicket building a custom component. In this chapter we will create a custom datepicker based on the corresponding widget from JQueryUI project:

On Internet you can find different libraries that already offer a strong integration between Wicket and JQuery. The goal of this chapter is to see how to integrate Wicket with a JavaScript framework building a simple homemade datepicker which is not intended to provide every feature of the original JavaScript widget.

What features we want to implement

Before starting to write code, we must clearly define what features we want to implement for our component. The new component should:

• Be self-contained: we must be able to distribute it and use it in other projects without requiring any kind of additional configuration.
• Have a customizable date format: developer must be able to decide the date format used to display date value and to parse user input.
• Be localizable: the pop-up calendar must be localizable in order to support different languages.

That's what we'd like to have with our custom datepicker. In the rest of the chapter we will see how to implement the features listed above and which resources must be packaged with our component.

15.2 ...and how we will do it

Our new component will extend the a built-in text field org.apache.wicket.extensions .markup.html.form.DateTextField which already uses a java.util.Date as model object and already performs conversion and validation for input values. Since the component must be self-contained, we must ensure that the JavaScript libraries it relies on (JQuery and JQuery UI) will be always available.

Starting from version 6.0 Wicket has adopted JQuery as backing JavaScript library so we can use the version bundled with Wicket for our custom datepicker.

To make JQuery UI available we should instead go to its official site, download the required artifacts and use them as package resources of our component.

Component package resources

JQuery UI needs the following static resources in order to work properly:

• jquery-ui.min.js: the minified version of the library.
• jquery-ui.css: the CSS containing the style used by JQuery UI widgets.
• jquery-ui-i18n.min.js: the minified JavaScript containing the built-in support for localization.
• Folder 'images': the folder containing picture files used by JQuery UI widgets.

In the following picture we can see these package resources with our component class (named JQueryDateField):

Along with the four static resources listed above, we can find also file calendar.jpg, which is the calendar icon used to open the pop up calendar, and file JQDatePicker.js which contains the following custom JavaScript code that binds our component to a JQuery UI datepicker:

function initJQDatepicker(inputId, countryIsoCode, dateFormat,  calendarIcon) {
	var localizedArray = $.datepicker.regional[countryIsoCode];
	localizedArray['buttonImage'] = calendarIcon;
	localizedArray['dateFormat'] = dateFormat;
	initCalendar(localizedArray);
	$("#" + inputId).datepicker(localizedArray);	
};
function initCalendar(localizedArray){
	 localizedArray['changeMonth']= true;
	 localizedArray['changeYear']= true;
	 localizedArray['showOn'] = 'button';
	 localizedArray['buttonImageOnly'] = true;
};

Function initJQDatepicker takes in input the following parameters:

• inputId: the id of the HTML text field corresponding to our custom component instance.
• countryIsoCode: a two-letter low-case ISO language code. It can contain also the two-letter upper-case ISO country code separated with a minus sign (for example en-GB)
• dateFormat: the date format to use for parsing and displaying date values.
• calendarIcon: the relative URL of the icon used as calendar icon.

As we will see in the next paragraphs, its up to our component to generate this parameters and invoke the initJQDatepicker function.

Function initCalendar is a simple utility function that sets the initialization array for datepicker widget. For more details on JQuery UI datepicker usage see the documentation at http://jqueryui.com/ datepicker.

Initialization code

The initialization code for our component is contained inside its method onInitialize and is the following:

@Override
protected void onInitialize() {
	super.onInitialize();
	setOutputMarkupId(true);
	datePattern =  new ResourceModel("jqueryDateField.shortDatePattern", "mm/dd/yy")
                                          .getObject();		
	countryIsoCode = new ResourceModel("jqueryDateField.countryIsoCode", "en-GB")                                                             
                                          .getObject();
	PackageResourceReference resourceReference = 
                            new PackageResourceReference(getClass(), "calendar.jpg");
	urlForIcon = urlFor(resourceReference, new PageParameters());
	dateConverter = new PatternDateConverter(datePattern, false);	
}
@Override
public <Date> IConverter<Date> getConverter(Class<Date> type) {
	return (IConverter<Date>) dateConverter;
}

The first thing to do inside onInitialize is to ensure that our component will have a markup id for its related text field. This is done invoking setOutputMarkupId(true).

Next, JQueryDateField tries to retrieve the date format and the ISO language code that must be used as initialization parameters. This is done using class ResourceModel which searches for a given resource in the available bundles. If no value is found for date format or for ISO language code, default values will be used ('mm/dd/yy' and 'en-GB').

To generate the relative URL for calendar icon, we load it as package resource reference and then we use Component's method urlFor to get the URL value (we have seen this method in paragraph 7.3.2).

The last configuration instruction executed inside onInitialize is the instantiation of the custom converter used by our component. This converter is an instance of the built-in class org.apache. wicket.datetime.PatternDateConvert and must use the previously retrieved date format to perform conversion operations. Now to tell our component to use this converter we must return it overriding FormComponent's method getConverter.

Header contributor code

The rest of the code of our custom component is inside method renderHeader, which is responsible for adding to page header the bundled JQuery library, the three files from JQuery UI distribution, the custom file JQDatePicker.js and the invocation of function initJQDatepicker:

@Override
public void renderHead(IHeaderResponse response) {
	super.renderHead(response);
	//if component is disabled we don't have to load the JQueryUI datepicker
	if(!isEnabledInHierarchy())
		return;
	//add bundled JQuery
	IJavaScriptLibrarySettings javaScriptSettings =          
                      getApplication().getJavaScriptLibrarySettings();
	response.render(JavaScriptHeaderItem.
			forReference(javaScriptSettings.getJQueryReference()));
	//add package resources
	response.render(JavaScriptHeaderItem.
	      forReference(new PackageResourceReference(getClass(), "jquery-ui.min.js")));
	response.render(JavaScriptHeaderItem.
	      forReference(new PackageResourceReference(getClass(), "jquery-ui-i18n.min.js")));
	response.render(CssHeaderItem.
	      forReference(new PackageResourceReference(getClass(), "jquery-ui.css")));
	//add custom file JQDatePicker.js. Reference JQDatePickerRef is a static field
	response.render(JavaScriptHeaderItem.forReference(JQDatePickerRef));
	//add the init script for datepicker
	String jqueryDateFormat = datePattern.replace("yyyy", "yy").toLowerCase();
	String initScript = ";initJQDatepicker('" + getMarkupId() + "', '" + countryIsoCode +
                            "', '" + jqueryDateFormat + "', " + "'" + urlForIcon +"');";
	response.render(OnLoadHeaderItem.forScript(initScript));
}

If component is disabled the calendar icon must be hidden and no datepicker must be displayed. That's why renderHeader is skipped if component is not enabled.

To get a reference to the bundled JQuery library we used the JavaScript setting interface IJava ScriptLibrarySettings and its method getJQueryReference.

In the last part of renderHeader we build the string to invoke function initJQDatepicker using the values obtained inside onInitialize. Unfortunately the date format used by JQuery UI is different from the one adopted in Java so we have to convert it before building the JavaScript code. This init script is rendered into header section using a OnLoadHeaderItem to ensure that it will be executed after all the other scripts have been loaded.

If we add more than one instance of our custom component to a single page, static resources are rendered to the header section just once. Wicket automatically checks if a static resource is already referenced by a page and if so, it will not render it again.

This does not apply to the init script which is dynamically generated and is rendered for every instance of the component.

Our datepicker is not ready yet to be used with AJAX. In chapter 16 we will see how to modify it to make it AJAX-compatible.

15.3 Summary

In this brief chapter we have seen how custom components can be integrated with DHTML technologies. To do so we have used most of what we have learnt in this guide. Now we are able to build complex components with a rich user experience. However this is not enough yet to develop Web 2.0 applications. We still have to cover a fundamental technology like AJAX and some other Wicket-related topics that will help us building our application in more modular and efficient way.

16 Wicket advanced topics

16.1 Enriching components with behaviors

With class org.apache.wicket.behavior.Behavior Wicket provides a very flexible mechanism to share common features across different components and to enrich existing components with further functionalities. As the class name suggests, Behavior adds a generic behavior to a component modifying its markup and/or contributing to the header section of the page (Behavior implements the interface IHeaderContributor).

One or more behaviors can be added to a component with Component's method add(Behavior...), while to remove a behavior we must use method remove(Behavior).

Here is a partial list of methods defined inside class Behavior along with a brief description of what they do:

• beforeRender(Component component): called when a component is about to be rendered.
• afterRender(Component component): called after a component has been rendered.
• onComponentTag(Component component, ComponentTag tag): called when component tag is being rendered.
• getStatelessHint(Component component): returns if a behavior is stateless or not.
• bind(Component component): called after a behavior has been added to a component.
• unbind(Component component): called when a behavior has been removed from a component.
• detach(Component component): overriding this method a behavior can detach its state before being serialized.
• isEnabled(Component component): tells if the current behavior is enabled for a given component. When a behavior is disabled it will be simply ignored and not executed.
• isTemporary(Component component): tells component if the current behavior is temporary. A temporary behavior is discarded at the end of the current request (i.e it's executed only once).
• onConfigure(Component component): called right after the owner component has been configured.
• onRemove(Component component): called when the owner component has been removed from its container.
• renderHead(Component component, IHeaderResponse response): overriding this method behaviors can render resources to the header section of the page.

For example the following behavior prepends a red asterisk to the tag of a form component if this one is required:

public class RedAsteriskBehavior extends Behavior {
  @Override
  public void beforeRender(Component component) {
      Response response = component.getResponse();
      StringBuffer asterisktHtml = new StringBuffer(200);
      if(componet instanceof FormComponent 
            && ((FormComponent)component).isRequired()){
        asteriskHtml.append(" <b style="color:red;font-size:medium">*</b>");
      }  
      response.write(asteriskHtml);
  }
}

Since method beforeRender is called before the coupled component is rendered, we can use it to prepend custom markup to component tag. This can be done writing our markup directly to the current Response object, as we did in the example above.

Please note that we could achieve the same result overriding component method onBeforeRender. However using a behavior we can easily reuse our custom code with any other kind of component without modifying its source code. As general best practice we should always consider to implement a new functionality using a behavior if it can be shared among different kinds of component.

Behaviors play also a strategic role in the built-in AJAX support provided by Wicket, as we will see in the next chapter.

16.2 Generating callback URLs with IRequestListener

With Wicket it's quite easy to build a callback URL that executes a specific method on server side. This method must be defined in a functional interface1 that inherits from built-in org.apache.wicket. IRequestListener and it must be a void method with no parameters in input:

public interface IMyListener extends IRequestListener
{
	/**
	 * Called when the relative callback URL is requested.
	 */
	void myCallbackMethod();
}

To control how the method will be invoked we must use class org.apache.wicket.Request ListenerInterface. In Wicket is a common practice to instantiate this class as a public static field inside the relative callback interface:

public interface IMyListener extends IRequestListener
{
	/**RequestListenerInterface instance*/
	public static final RequestListenerInterface INTERFACE = new 
                                           RequestListenerInterface(IMyListener.class);
	/**
	 * Called when the relative callback URL is requested.
	 */
	void myCallbackMethod();
}

By default RequestListenerInterface will respond rendering the current page after the callback method has been executed (if we have a non-AJAX request). To change this behavior we can use setter method setRenderPageAfterInvocation(boolean).

Now that our callback interface is complete we can generate a callback URL with Component's method urlFor(RequestListenerInterface, PageParameters) or with method urlFor (Behavior, RequestListenerInterface, PageParameters) if we are using a callback interface with a behavior (see the following example).

Project CallbackURLExample contains a behavior (class OnChangeSingleChoiceBehavior) that implements a callback interface to update the model of an AbstractSingleSelectChoice component when user changes the selected option (it provides the same functionality of method want OnSelectionChangedNotifications).

Instead of a custom callback interface, OnChangeSingleChoiceBehavior implements built-in interface org.apache.wicket.behavior.IBehaviorListener which is designed to generate a callback URL for behaviors. The callback method defined in this interface is onRequest() and the following is the implementation provided by OnSelectionChangedNotifications:

@Override
public void onRequest() {	
	Request request = RequestCycle.get().getRequest();
	IRequestParameters requestParameters = request.getRequestParameters();
	StringValue choiceId = requestParameters.getParameterValue("choiceId");
	//boundComponent is the component that the behavior it is bound to.
	boundComponent.setDefaultModelObject(convertChoiceIdToChoice(choiceId.toString()));
}

When invoked via URL, the behavior expects to find a request parameter (choiceId) containing the id of the selected choice. This value is used to obtain the corresponding choice object that must be used to set the model of the component that the behavior is bound to (boundComponent). Method convertChoiceIdToChoice is in charge of retrieving the choice object given its id and it has been copied from class AbstractSingleSelectChoice.

Another interesting part of OnChangeSingleChoiceBehavior is its method onComponentTag where some JavaScript “magic” is used to move user's browser to the callback URL when event “change” occurs on bound component:

@Override
public void onComponentTag(Component component, ComponentTag tag) {
	super.onComponentTag(component, tag);
	CharSequence callBackURL = getCallbackUrl();
	String separatorChar = (callBackURL.toString().indexOf('?') > -1 ? "&" : "?");
	String finalScript = "var isSelect = $(this).is('select');n" +
				 "var component;n" +	
				 "if(isSelect)n" +
				 "	component = $(this);n" +
				 "else n" +
				 "	component = $(this).find('input:radio:checked');n" +
				 "window.location.href='" + callBackURL +  separatorChar + 
				 "choiceId=' + " + "component.val()";
	tag.put("onchange", finalScript);
}

The goal of onComponentTag is to build an onchange handler that forces user's browser to move to the callback URL (modifing standard property window.location.href). Please note that we have appended the expected parameter (choiceId) to the URL retrieving its value with a JQuery selector suited for the current type of component (a drop-down menu or a radio group). Since we are using JQuery in our JavaScript code, the behavior comes also with method renderHead that adds the bundled JQuery library to the current page.

Method getCallbackUrl() is used to generate the callback URL for our custom behavior and it has been copied from built-in class AbstractAjaxBehavior:

public CharSequence getCallbackUrl(){
	if (boundComponent == null){
		throw new IllegalArgumentException(
			"Behavior must be bound to a component to create the URL");
	}
	final RequestListenerInterface rli;
	rli = IBehaviorListener.INTERFACE;
	return boundComponent.urlFor(this, rli, new PageParameters());
}

Static field IBehaviorListener.INTERFACE is the implementation of RequestListener Interface defined inside callback interface IBehaviorListener.

The home page of project CallbackURLExample contains a DropDownChoice and a RadioChoice which use our custom behavior. There are also two labels to display the content of the models of the two components:

Implementing interface IBehaviorListener makes a behavior stateful because its callback URL is specific for a given instance of component.

Wicket events infrastructure

Starting from version 1.5 Wicket offers an event-based infrastructure for inter-component communication. The infrastructure is based on two simple interfaces (both in package org. apache.wicket.event) : IEventSource and IEventSink.

The first interface must be implemented by those entities that want to broadcast en event while the second interface must be implemented by those entities that want to receive a broadcast event.

The following entities already implement both these two interfaces (i.e. they can be either sender or receiver): Component, Session, RequestCycle and Application. IEventSource exposes a single method named send which takes in input three parameters:

• sink: an implementation of IEventSink that will be the receiver of the event.
• broadcast: a Broadcast enum which defines the broadcast method used to dispatch the event to the sink and to other entities such as sink children, sink containers, session object, application object and the current request cycle. It has four possible values:

• payload: a generic object representing the data sent with the event.

Each broadcast mode has its own traversal order for Session, RequestCycle and Application. See JavaDoc of class Broadcast for further details about this order.

Interface IEventSink exposes callback method onEvent(IEvent<?> event) which is triggered when a sink receives an event. The interface IEvent represents the received event and provides getter methods to retrieve the event broadcast type, the source of the event and its payload. Typically the received event is used checking the type of its payload object:

@Override
public void onEvent(IEvent event) {
  //if the type of payload is MyPayloadClass perform some actions 
  if(event.getPayload() instanceof MyPayloadClass) {
     //execute some business code.
  }else{
     //other business code
  }   		
}

Project InterComponetsEventsExample provides a concrete example of sending an event to a component (named 'container in the middle') using all the available broadcast methods:

16.3 Initializers

Some components or resources may need to be configured before being used in our applications. While so far we used Application's init method to initialize these kinds of entities, Wicket offers a more flexible and modular way to configure our classes.

During application's bootstrap Wicket searches for any properties file named wicket.properties placed in one of the classpath roots visible to the application. When one of these files is found, the initializer defined inside it will be executed. An initializer is an implementation of interface org.apache.wicket.IInitializer and is defined inside wicket.properties with a line like this:

initializer=org.wicketTutorial.MyInitializer

The fully qualified class name corresponds to the initializer that must be executed. Interface IInitializer defines method init(Application) which should contain our initialization code, and method destroy(Application) which is invoked when application is terminated:

public class MyInitializer implements IInitializer{
	public void init(Application application) {
		//initialization code 
	}
	public void destroy(Application application) {
		//code to execute when application is terminated
	}	
}

Only one initializer can be defined in a single wicket.properties file. To overcome this limit we can create a main initializer that in turn executes every initializer we need:

public class MainInitializer implements IInitializer{
	public void init(Application application) {
		new AnotherInitializer().init(application);
		new YetAnotherInitializer().init(application);
		//… 
	}
	//destroy… 
}

16.4 Using JMX with Wicket

JMX (Java Management Extensions) is the standard technology adopted in Java for managing and monitoring running applications or Java Virtual Machines. Wicket offers support for JMX through module wicket-jmx. In this paragraph we will see how we can connect to a Wicket application using JMX. In our example we will use JConsole as JMX client. This program is bundled with Java SE since version 5 and we can run it typing jconsole in our OS shell.

Once JConsole has started it will ask us to establish a new connection to a Java process, choosing between a local process or a remote one. In the following picture we have selected the process corresponding to the local instance of Jetty server we used to run one of our example projects:

After we have established a JMX connection, JConsole will show us the following set of tabs:

JMX exposes application-specific informations using special objects called MBeans (Manageable Beans), hence if we want to control our application we must open the corresponding tab. The MBeans containing the application's informations is named org.apache.wicket.app.<filter/servlet name>.

In our example we have used wicket.test as filter name for our application:

As we can see in the picture above, every MBean exposes a node containing its attributes and another node showing the possible operations that can be performed on the object. In the case of a Wicket application the available operations are clearMarkupCache and clearLocalizerCache:

With these two operations we can force Wicket to clear the internal caches used to load components markup and resource bundles. This can be particularly useful if we have our application running in DEPLOYMENT mode and we want to publish minor fixes for markup or bundle files (like spelling or typo corrections) without restarting the entire application. Without cleaning these two caches Wicket would continue to use cached values ignoring any change made to markup or bundle files.

Some of the exposed properties are editable, hence we can tune their values while the application is running. For example if we look at the properties of ApplicationSettings we can set the maximum size allowed for an upload modifying the attribute DefaultMaximumUploadSize:

16.5 Generating HTML markup from code

So far, as markup source for our pages/panels we have used a static markup file, no matter if it was inherited or directly associated to the component. Now we want to investigate a more complex use case where we want to dynamical generate the markup directly inside component code.

To become a markup producer, a component must simply implement interface org.apache. wicket.markup.IMarkupResourceStreamProvider. The only method defined in this interface is getMarkupResourceStream(MarkupContainer, Class<?>) which returns an utility interface called IResourceStream representing the actual markup.

In the following example we have a custom panel without a related markup file that generates a simple <div> tag as markup:

public class AutoMarkupGenPanel extends Panel implements IMarkupResourceStreamProvider {
	public AutoMarkupGenPanel(String id, IModel<?> model) {
		super(id, model);		
	}
	@Override
	public IResourceStream getMarkupResourceStream(MarkupContainer container,
			Class<?> containerClass) {
		String markup = "<div>Panel markup</div>";
		StringResourceStream resourceStream = new StringResourceStream(markup);
		return resourceStream;
	}
}

Class StringResourceStream is a resource stream that uses a String instance as backing object.

Avoiding markup caching

As we have seen in the previous paragraph, Wicket uses an internal cache for components markup. This can be a problem if our component dynamical generates its markup when it is rendered because once the markup has been cached, Wicket will always use the cached version for the specific component. To overwrite this default caching policy, a component can implement interface IMarkupCacheKeyProvider.

This interface defines method getCacheKey(MarkupContainer, Class<?>) which returns a string value representing the key used by Wicket to retrieve the markup of the component from the cache. If this value is null the markup will not be cached, allowing the component to display the last generated markup each time it is rendered:

public class NoCacheMarkupPanel extends Panel implements IMarkupCacheKeyProvider {
	public NoCacheMarkupPanel(String id, IModel<?> model) {
		super(id, model);		
	}
	/**
	* Generate a dynamic HTML markup that changes every time
	* the component is rendered
	*/
	@Override
	public IResourceStream getMarkupResourceStream(MarkupContainer container,
			Class<?> containerClass) {
		String markup = "<div>Panel with current nanotime: " + System.nanoTime() +
				 "</div>"; 
		StringResourceStream resourceStream = new StringResourceStream(markup);
		return resourceStream;
	}
	/**
	* Avoid markup caching for this component
	*/
	@Override
	public String getCacheKey(MarkupContainer arg0, Class<?> arg1) {
		return null;
	}
}

16.6 Summary

In this chapter we have introduced some advanced topics we didn't have the chance to cover yet. We have started talking about behaviors and we have seen how they can be used to enrich existing components (promoting a component-oriented approach). Behaviors are also fundamental to work with AJAX in Wicket, as we will see in the next chapter.

After behaviors we have learnt how to generate callback URLs to execute a custom method on server side defined inside a specific callback interface.

The third topic of the chapter has been the event infrastructure provided in Wicket for inter-component communication which brings to our components a desktop-like event-driven architecture.

Then, we have introduced a new entity called initializer which can be used to configure resources and component in a modular and self-contained way.

We have also looked at Wicket support for JMX and we have seen how to use this technology for monitoring and managing our running applications.

Finally we have introduced a new technique to generate the markup of a component from its Java code.

17 Working with AJAX

In this chapter we will learn how to leverage the AJAX support provided by Wicket to make our applications fully Web 2.0 compliant.

17.1 How to use AJAX components and behaviors

Wicket support for AJAX is implemented in file wicket-ajax-jquery.js which makes complete transparent to Java code any detail about AJAX communication.

AJAX components and behaviors shipped with Wicket expose one or more callback methods which are executed when they receive an AJAX request. One of the arguments of these methods is an instance of interface org.apache.wicket.ajax.AjaxRequestTarget.

For example component AjaxLink (in package org.apache.wicket.ajax.markup.html) defines abstract method onClick(AjaxRequestTarget target) which is executed when user clicks on the component:

new AjaxLink("ajaxLink"){
	@Override
	public void onClick(AjaxRequestTarget target) {
	    //some server side code…
	}  	
};

Using AjaxRequestTarget we can specify the content that must be sent back to the client as response to the current AJAX request. The most commonly used method of this interface is probably add(Component… components). With this method we tell Wicket to render again the specified components and refresh their markup via AJAX:

new AjaxLink("ajaxLink"){
	@Override
	public void onClick(AjaxRequestTarget target) {
	    //modify the model of a label and refresh it on browser
	    label.setDefaultModelObject("Another value 4 label.");
	    target.add(label);
	}  	
};

Components can be refreshed via Ajax only if they have rendered a markup id for their related tag. As a consequence, we must remember to set a valid id value on every component we want to add to AjaxRequestTarget. This can be done using one of the two methods seen in paragraph 4.3:

final Label label = new Label("labelComponent", "Initial value.");
//autogenerate a markup id
label.setOutputMarkupId(true);
add(label);
//…
new AjaxLink("ajaxLink"){
	@Override
	public void onClick(AjaxRequestTarget target) {
	    //modify the model of a label and refresh it on client side
	    label.setDefaultModelObject("Another value 4 label.");
	    target.add(label);
	}  	
};

Another common use of AjaxRequestTarget is to prepend or append some JavaScript code to the generated response. For example the following AJAX link displays an alert box as response to user's click:

new AjaxLink("ajaxLink"){
	@Override
	public void onClick(AjaxRequestTarget target) {
	    target.appendJavaScript(";alert('Hello!!');");
	}  	
};

Repeaters component that have org.apache.wicket.markup.repeater AbstractRepeater as base class (like ListView, RepeatingView, etc...) can not be directly updated via AJAX.

If we want to refresh their markup via AJAX we must add one of their parent containers to the AjaxRequestTarget.

17.2 Build-in AJAX components

Wicket distribution comes with a number of built-in AJAX components ready to be used. Some of them are the ajaxified version of common components like links and buttons, while others are AJAX-specific components.

AJAX components are not different from any other component seen so far and they don't require any additional configuration to be used. As we will shortly see, switching from a classic link or button to the ajaxified version is just a matter of appending “Ajax” to the component class name.

This paragraph provides an overview of what we can find in Wicket to start writing AJAX-enhanced web applications.

Links and buttons

In the previous paragraph we have already introduced component AjaxLink. Wicket provides also the ajaxified versions of submitting components SubmitLink and Button which are simply called AjaxSubmitLink and AjaxButton. These components come with a version of methods onSubmit, onError and onAfterSubmit that takes in input also an instance of AjaxRequestTarget.

Both components are in package org.apache.wicket.ajax.markup.html.form.

Fallback components

Building an entire site using AJAX can be risky as some clients may not support this technology. In order to provide an usable version of our site also to these clients, we can use components AjaxFallbackLink and AjaxFallbackButton which are able to automatically degrade to a standard link or to a standard button if client doesn't support AJAX.

AJAX Checkbox

Class org.apache.wicket.ajax.markup.html.form.AjaxCheckBox is a checkbox component that updates its model via AJAX when user changes its value. Its AJAX callback method is onUpdate(AjaxRequestTarget target). The component extends standard checkbox component CheckBox adding an AjaxFormComponentUpdatingBehavior to itself (we will see this behavior later in paragraph 16.3.3).

AJAX editable labels

An editable label is a special label that can be edited by the user when she/he clicks on it. Wicket ships three different implementations for this component (all inside package org.apache.wicket. extensions.ajax.markup.html):

• AjaxEditableLabel: it's a basic version of editable label. User can edit the content of the label with a text field. This is also the base class for the other two editable labels.
• AjaxEditableMultiLineLabel: this label supports multi-line values and uses a text area as editor component.
• AjaxEditableChoiceLabel: this label uses a drop-down menu to edit its value.

Base component AjaxEditableLabel exposes the following set of AJAX-aware methods that can be overriden:

• onEdit(AjaxRequestTarget target): called when user clicks on component. The default implementation shows the component used to edit the value of the label.
• onSubmit(AjaxRequestTarget target): called when the value has been successfully updated with the new input.
• onError(AjaxRequestTarget target): called when the new inserted input has failed validation.
• onCancel(AjaxRequestTarget target): called when user has exited from editing mode pressing escape key. The default implementation brings back the label to its initial state hiding the editor component.

Wicket module wicket-examples contains page class EditableLabelPage.java which shows all these three components together. You can see this page in action at http://www.wicket-library.com/wicket-examples-6.0.x/ajax/editable-label :

Autocomplete text field

On Internet we can find many examples of text fields that display a list of suggestions (or options) while the user types a text inside them. This feature is known as autocomplete functionality.

Wicket offers an out-of-the-box implementation of an autocomplete text field with component org.apache.wicket.extensions.ajax.markup.html.autocomplete.AutoCompleteTextField.

When using AutoCompleteTextField we are required to implement its abstract method getChoices(String input) where the input parameter is the current input of the component. This method returns an iterator over the suggestions that will be displayed as a drop-down menu:

Suggestions are rendered using a render which implements interface IAutoCompleteRenderer. The default implementation simply calls toString() on each suggestion object. If we need to work with a custom render we can specify it via component constructor.

AutoCompleteTextField supports a wide range of settings that are passed to its constructor with class AutoCompleteSettings.

One of the most interesting parameter we can specify for AutoCompleteTextField is the throttle delay which is the amount of time (in milliseconds) that must elapse between a change of input value and the transmission of a new Ajax request to display suggestions. This parameter can be set with method setThrottleDelay(int):

AutoCompleteSettings settings = new AutoCompleteSettings();
//set throttle to 400 ms: component will wait 400ms before displaying the options		
settings.setThrottleDelay(400);
//...		
AutoCompleteTextField field = new AutoCompleteTextField<T>("field", model) {
	@Override
	protected Iterator getChoices(String arg0) {
		//return an iterator over the options 
	}
};

Wicket module wicket-examples contains page class AutoCompletePagePage.java which shows an example of autocomplete text field. The running example is available at http://www.wicket-library.com/wicket-examples-6.0.x/ajax/autocomplete .

Modal window

Class org.apache.wicket.extensions.ajax.markup.html.modal.ModalWindow is an implementation of a modal window based on AJAX:

The content of a modal window can be either another component or a page. In the first case the id of the component used as content must be retrieved with method getContentId().

If instead we want to use a page as window content, we must implement the inner interface ModalWindow.PageCreator and pass it to method setPageCreator. The page used as content will be embedded in a <iframe> tag.

To display a modal window we must call its method show(AjaxRequestTarget target). This is usually done inside the AJAX callback method of another component (like an AjaxLink). The following markup and code are taken from project BasicModalWindowExample and illustrate a basic usage of a modal window:

HTML:

<body>
	<h2>Modal Windod example</h2>
	<a wicket:id="openWindow">Open the window!</a>
	<div wicket:id="modalWindow"></div>
</body>

Java Code:

public HomePage(final PageParameters parameters) {
   	super(parameters);
   	final ModalWindow modalWindow = new ModalWindow("modalWindow");
   	Label label = new Label(modalWindow.getContentId(), "I'm a modal window!");
   	modalWindow.setContent(label);
   	modalWindow.setTitle("Modal window");
   	add(modalWindow);
   	add(new AjaxLink("openWindow") {
	  @Override
	  public void onClick(AjaxRequestTarget target) {
		modalWindow.show(target);				
	  }    		
	});
}

Just like any other component also ModalWindow must be added to a markup tag, like we did in our example using a <div> tag. Wicket will automatically hide this tag in the final markup appending the style value display:none. The component provides different setter methods to customize the appearance of the window:

• setTitle(String): specifies the title of the window
• setResizable(boolean): by default the window is resizeable. If we need to make its size fixed we can use this method to turn off this feature.
• setInitialWidth(int) and setInitialHeight(int): set the initial dimensions of the window.
• setMinimalWidth(int) and setMinimalHeight(int): specify the minimal dimensions of the window.
• setCookieName(String): this method can be used to specify the name of the cookie used on client side to store size and position of the window when it is closed. The component will use this cookie to restore these two parameters the next time the window will be opened. If no cookie name is provided, the component will not remember its last position and size.
• setCssClassName(String): specifies the CSS class used for the window.
• setAutoSize(boolean): when this flag is set to true the window will automatically adjust its size to fit content width and height. By default it is false.

The modal window can be closed from code using its method close(AjaxRequestTarget target). The currently opened window can be closed also with the following JavaScript instruction:

Wicket.Window.get().close();

ModalWindow gives the opportunity to perform custom actions when window is closing. Inner interface ModalWindow.WindowClosedCallback can be implemented and passed to window's method setWindowClosedCallback to specify the callback that must be executed after window has been closed:

modalWindow.setWindowClosedCallback(new ModalWindow.WindowClosedCallback() {
	@Override
	public void onClose(AjaxRequestTarget target) {
	  //custom code…
	}			
});

Tree repeaters

Class org.apache.wicket.extensions.markup.html.repeater.tree.AbstractTree is the base class of another family of repeaters called tree repeaters and designed to display a data hierarchy as a tree, resembling the behavior and the look & feel of desktop tree components. A classic example of tree component on desktop is the tree used by nearly all file managers to navigate file system:

Because of their highly interactive nature, tree repeaters are implemented as AJAX components, meaning that they are updated via AJAX when we expand or collapse their nodes.

The basic implementation of a tree repeater shipped with Wicket is component NestedTree. In order to use a tree repeater we must provide an implementation of interface ITreeProvider which is in charge of returning the nodes that compose the tree.

Wicket comes with a built-in implementation of ITreeProvider called TreeModelProvider that works with the same tree model and nodes used by Swing component javax.swing.JTree. These Swing entities should be familiar to you if you have previously worked with the old tree repeaters (components Tree and TreeTable) that have been deprecated with Wicket 6 and that are strongly dependent on Swing-based model and nodes. TreeModelProvider can be used to migrate your code to the new tree repeaters.

In the next example (project CheckBoxAjaxTree) we will build a tree that displays some of the main cities of three European countries: Italy, Germany and France. The cities are sub-nodes of a main node representing the relative county. The nodes of the final tree will be also selectable with a checkbox control. The whole tree will have the classic look & feel of Windows XP. This is how our tree will look like:

We will start to explore the code of this example from the home page. The first portion of code we will see is where we build the nodes and the TreeModelProvider for the three. As tree node we will use Swing class javax.swing.tree.DefaultMutableTreeNode:

public class HomePage extends WebPage {
    public HomePage(final PageParameters parameters) {
     super(parameters);
     DefaultMutableTreeNode root = new DefaultMutableTreeNode("Cities of Europe");
     addNodes(addNodes(root, "Italy"), "Rome", "Venice", "Milan", "Florence");
     addNodes(addNodes(root, "Germany"),"Stuttgart","Munich","Berlin","Dusseldorf", "Dresden");
     addNodes(addNodes(root, "France"), "Paris","Toulouse","Strasbourg","Bordeaux", "Lyon");
     DefaultTreeModel treeModel = new DefaultTreeModel(root);
     TreeModelProvider<DefaultMutableTreeNode> modelProvider = new 
                                  TreeModelProvider<DefaultMutableTreeNode>(treeModel) {
         @Override
         public IModel<DefaultMutableTreeNode> model(DefaultMutableTreeNode object) {
            return Model.of(object);
        }
     };
     //To be continued...

Nodes have been built using simple strings as data objects and invoking custom utility method addNodes which converts string parameters into children nodes for a given parent node. Once we have our tree of DefaultMutableTreeNodes we can build the Swing tree model (DefaultTreeModel) that will be the backing object for a TreeModelProvider. This provider wraps each node in a model invoking its abstract method model. In our example we have used a simple Model as wrapper model.

Scrolling down the code we can see how the tree component is instantiated and configured before being added to the home page:

//Continued from previous snippet…
 NestedTree<DefaultMutableTreeNode> tree = new NestedTree<DefaultMutableTreeNode>("tree", 
                                                      modelProvider)
  {
   @Override
   protected Component newContentComponent(String id, IModel<DefaultMutableTreeNode>model)
   {
     return new CheckedFolder<DefaultMutableTreeNode>(id, this, model);
   }
  };
  //select Windows theme
  tree.add(new WindowsTheme());
  add(tree);
  }
  //implementation of addNodes
  //…
}

To use tree repeaters we must implement their abstract method newContentComponent which is called internally by base class AbstractTree when a new node must be built. As content component we have used built-in class CheckedFolder which combines a Folder component with a CheckBox form control.

The final step before adding the tree to its page is to apply a theme to it. Wicket comes with two behaviors, WindowsTheme and HumanTheme, which correspond to the classic Windows XP theme and to the Human theme from Ubuntu.

Our checkable tree is finished but our work is not over yet because the component doesn't offer many functionalities as it is. Unfortunately neither NestedTree nor CheckedFolder provide a means for collecting checked nodes and returning them to client code. It's up to us to implement a way to keep track of checked nodes.

Another nice feature we would like to implement for our tree is the following user-friendly behavior that should occur when a user checks/unchecks a node:

• When a node is checked also all its children nodes (if any) must be checked. We must also ensure that all the ancestors of the checked node (root included) are checked, otherwise we would get an inconsistent selection.
• When a node is unchecked also all its children nodes (if any) must be unchecked and we must also ensure that ancestors get unchecked if they have no more checked children.

The first goal (keeping track of checked node) can be accomplished building a custom version of CheckedFolder that uses a shared Java Set to store checked node and to verify if its node has been checked. This kind of solution requires a custom model for checkbox component in order to reflect its checked status when its container node is rendered. This model must implement typed interface IModel<Boolean> and must be returned by CheckedFolder's method newCheckBoxModel.

For the second goal (auto select/unselect children and ancestor nodes) we can use CheckedFolder's callback method onUpdate(AjaxRequestTarget) that is invoked after a checkbox is clicked and its value has been updated. Overriding this method we can handle user click adding/removing nodes to/from the Java Set.

Following this implementation plan we can start coding our custom CheckedFolder (named AutocheckedFolder):

public class AutocheckedFolder<T> extends CheckedFolder<T> {
   private ITreeProvider<T> treeProvider;
   private IModel<Set<T>> checkedNodes;
   private IModel<Boolean> checkboxModel;
   public AutocheckedFolder(String id, AbstractTree<T> tree, 
                        IModel<T> model, IModel<Set<T>> checkedNodes) {
      super(id, tree, model);   
      this.treeProvider = tree.getProvider();
      this.checkedNodes = checkedNodes;            
   }
   @Override
   protected IModel<Boolean> newCheckBoxModel(IModel<T> model) {
      checkboxModel =  new CheckModel();
      return checkboxModel;
   }
   @Override
   protected void onUpdate(AjaxRequestTarget target) {
      super.onUpdate(target);
      T node = getModelObject();
      boolean nodeChecked = checkboxModel.getObject();
      addRemoveSubNodes(node, nodeChecked);            
      addRemoveAncestorNodes(node, nodeChecked);            
   }
  class CheckModel extends AbstractCheckBoxModel{
      @Override
      public boolean isSelected() {
         return checkedNodes.getObject().contains(getModelObject());
      }
      @Override
      public void select() {
         checkedNodes.getObject().add(getModelObject());
      }
      @Override
      public void unselect() {
         checkedNodes.getObject().remove(getModelObject());
      }				
  }
}

The constructor of this new component takes in input a further parameter which is the set containing checked nodes.

Class CheckModel is the custom model we have implemented for checkbox control. As base class for this model we have used AbstractCheckBoxModel which is provided to implement custom models for checkbox controls.

Methods addRemoveSubNodes and addRemoveAncestorNodes are called to automatically add/remove children and ancestor nodes to/from the current Set. Their implementation is mainly focused on the navigation of tree nodes and it heavily depends on the internal implementation of the tree, so we won't dwell on their code.

Now we are just one step away from completing our tree as we still have to find a way to update the checked status of both children and ancestors nodes on client side. Although we could easily accomplish this task by simply refreshing the whole tree via AJAX, we would like to find a better and more performant solution for this task.

When we modify the checked status of a node we don't expand/collapse any node of the three so we can simply update the desired checkboxes rather than updating the entire tree component. This alternative approach could lead to a more responsive interface and to a strong reduction of bandwidth consumption.

With the help of JQuery we can code a couple of JavaScript functions that can be used to check/ uncheck all the children and ancestors of a given node. Then, we can append these functions to the current AjaxRequest at the end of method onUpdate:

@Override
   protected void onUpdate(AjaxRequestTarget target) {
      super.onUpdate(target);
      T node = getModelObject();
      boolean nodeChecked = checkboxModel.getObject();
      addRemoveSubNodes(node, nodeChecked);            
      addRemoveAncestorNodes(node, nodeChecked);    
      updateNodeOnClientSide(target, nodeChecked);		
   }
   protected void updateNodeOnClientSide(AjaxRequestTarget target,
			boolean nodeChecked) {
      target.appendJavaScript(";CheckAncestorsAndChildren.checkChildren('" + getMarkupId() + 
                              "'," + nodeChecked + ");");
      target.appendJavaScript(";CheckAncestorsAndChildren.checkAncestors('" + getMarkupId() + 
                              "'," + nodeChecked + ");");
   }

The JavaScript code can be found inside file autocheckedFolder.js which is added to the header section as package resource:

@Override
public void renderHead(IHeaderResponse response) {
	PackageResourceReference scriptFile = new PackageResourceReference(this.getClass(), 
                                                      "autocheckedFolder.js");
	response.render(JavaScriptHeaderItem.forReference(scriptFile));
}

Working with hidden components

When a component is not visible its markup and the related id attribute are not rendered in the final page, hence it can not be updated via AJAX. To overcome this problem we must use Component's method setOutputMarkupPlaceholderTag(true) which has the effect of rendering a hidden <span> tag containing the markup id of the hidden component:

final Label label = new Label("labelComponent", "Initial value.");
//make label invisible
label.setVisible(false);
//ensure that label will leave a placeholder for its markup id
label.setOutputMarkupPlaceholderTag(true);
add(label);
//…
new AjaxLink("ajaxLink"){
	@Override
	public void onClick(AjaxRequestTarget target) {
	    //turn label to visible
	    label.setVisible(true);
	    target.add(label);
	}  	
};

Please note that in the code above we didn't invoked method setOutputMarkupId(true) as setOutputMarkupPlaceholderTag already does it internally.

17.3 Built-in AJAX behaviors

In addition to specific components, Wicket offers also a set of built in AJAX behaviors that can be used to easily add AJAX functionalities to existing components. As we will see in this paragraph AJAX behaviors can be used also to ajaxify components that weren't initially designed to work with this technology. All the following behaviors are inside package org.apache.wicket.ajax.

AjaxEventBehavior

AjaxEventBehavior allows to handle a JavaScript event (like click, change, etc...) on server side via AJAX. Its constructor takes in input the name of the event that must be handled. Every time this event is fired for a given component on client side, the callback method onEvent(AjaxRequestTarget target) is executed. onEvent is abstract, hence we must implement it to tell AjaxEventBehavior what to do when the specified event occurs.

In project AjaxEventBehaviorExample we used this behavior to build a “clickable” Label component that counts the number of clicks. Here is the code from the home page of the project:

HTML:

<body>
  <div wicket:id="clickCounterLabel"></div>
  User has clicked <span wicket:id="clickCounter"></span> time/s on the label above.
</body>

Java Code:

public class HomePage extends WebPage {
   public HomePage(final PageParameters parameters) {
      super(parameters);
      final ClickCounterLabel clickCounterLabel = 
         new ClickCounterLabel("clickCounterLabel", "Click on me!");
      final Label clickCounter =
         new Label("clickCounter", new PropertyModel(clickCounterLabel, "clickCounter"));
      
      clickCounterLabel.setOutputMarkupId(true);
      clickCounterLabel.add(new AjaxEventBehavior("click"){
         @Override
         protected void onEvent(AjaxRequestTarget target) {
            clickCounterLabel.clickCounter++;
            target.add(clickCounter);
         }         
      });
      add(clickCounterLabel);
      add(clickCounter.setOutputMarkupId(true));      
    }
}
class ClickCounterLabel extends Label{
   public int clickCounter;
   public ClickCounterLabel(String id) {
      super(id);
   }
   public ClickCounterLabel(String id, IModel<?> model) {
      super(id, model);
   }
   public ClickCounterLabel(String id, String label) {
      super(id, label);      
   }
}

In the code above we have declared a custom label class named ClickCounterLabel that exposes a public integer field called clickCounter. Then, in the home page we have attached a AjaxEventBehavior to our custom label to increment clickCounter every time it receives a click event.

The number of clicks is displayed with another standard label named clickCounter.

AjaxFormSubmitBehavior

This behavior allows to send a form via AJAX when the component it is attached to receives the specified event. The component doesn't need to be inside the form if we use the constructor version that, in addition to the name of the event, takes in input also the target form:

Form form = new Form("form");		
Button submitButton = new Button("submitButton");
//submit form when button is clicked		
submitButton.add(new AjaxFormSubmitBehavior(form, "click"){});
add(form);
add(submitButton);

AjaxFormComponentUpdatingBehavior

This behavior updates the model of the form component it is attached to when a given event occurs. The standard form submitting process is skipped and the behavior validates only its form component.

The behavior doesn't work with radio buttons and checkboxes. For these kinds of components we must use AjaxFormChoiceComponentUpdatingBehavior:

Form form = new Form("form");		
TextField textField = new TextField("textField", Model.of(""));
//update the model of the text field each time event "change" occurs
textField.add(new AjaxFormComponentUpdatingBehavior("change"){
	@Override
	protected void onUpdate(AjaxRequestTarget target) {
		//...				
	}
});
add(form.add(textField));

AbstractAjaxTimerBehavior

AbstractAjaxTimerBehavior executes callback method onTimer(AjaxRequestTarget target) at a specified interval. The behavior can be stopped and restarted at a later time with methods stop(AjaxRequestTarget target) and restart(AjaxRequestTarget target):

Label dynamicLabel = new Label("dynamicLabel");
//trigger an AJAX request every three seconds		
dynamicLabel.add(new AbstractAjaxTimerBehavior(Duration.seconds(3)) {			
	@Override
	protected void onTimer(AjaxRequestTarget target) {
		//...				
	}
});
add(dynamicLabel);

17.4 Using an activity indicator

One of the things we must take care of when we use AJAX is to notify user when an AJAX request is already in progress. This is usually done displaying an animated picture as activity indicator while the AJAX request is running.

Wicket comes with a variant of components AjaxButton, AjaxLink and AjaxFallbackLink that display a default activity indicator during AJAX request processing. These components are respectively IndicatingAjaxButton, IndicatingAjaxLink and IndicatingAjaxFallbackLink.

The default activity indicator used in Wicket can be easily integrated in our components using behavior AjaxIndicatorAppender (available in package org.apache.wicket.extensions.ajax. markup.html) and implementing the interface IAjaxIndicatorAware (in package org.apache. wicket.ajax).

IAjaxIndicatorAware declares method getAjaxIndicatorMarkupId() which returns the id of the markup element used to display the activity indicator. This id can be obtained from the AjaxIndicatorAppender behavior that has been added to the current component. The following code snippet summarizes the steps needed to integrate the default activity indicator with an ajaxified component:

//1-Implement interface IAjaxIndicatorAware
public class MyComponent extends Component implements IAjaxIndicatorAware {
	//2-Instantiate an AjaxIndicatorAppender
	private AjaxIndicatorAppender indicatorAppender =
			new AjaxIndicatorAppender();
	public MyComponent(String id, IModel<?> model) {
		super(id, model);
		//3-Add the AjaxIndicatorAppender to the component
		add(indicatorAppender);
	}
	//4-Return the markup id obtained from AjaxIndicatorAppender
	public String getAjaxIndicatorMarkupId() {		
		return indicatorAppender.getMarkupId();
	}
//…
}

If we need to change the default picture used as activity indicator, we can override method getIndicatorUrl() of AjaxIndicatorAppender and return the URL to the desired picture.

17.5 AJAX request attributes and call listeners

Starting from version 6.0 Wicket has introduced two entities which allow us to control how an AJAX request is generated on client side and to specify the custom JavaScript code we want to execute during request handling. These entities are class AjaxRequestAttributes and interface IAjaxCallListener, both placed in package org.apache.wicket.ajax.attributes.

AjaxRequestAttributes exposes the attributes used to generate the JavaScript call invoked on client side to start an AJAX request. Each attribute will be passed as a JSON parameter to the JavaScript function Wicket.Ajax.ajax which is responsible for sending the concrete AJAX request. Every JSON parameter is identified by a short name. Here is a partial list of the available parameters:

A full list of the available request parameters as well as more details on the related JavaScript code can be found at https://cwiki.apache.org/confluence/ display/WICKET/Wicket+Ajax display/WICKET/Wicket+Ajax .

Parameters 'u' (callback URL) and 'c' (the id of the component) are generated by the AJAX behavior that will serve the AJAX call and they are not accessible through AjaxRequestAttributes.

Here is the final AJAX function generate for the behavior used in example project AjaxEventBehavior Example:

Wicket.Ajax.ajax({"u":"./?0-1.IBehaviorListener.0-clickCounterLabel","e":"click",               
                  "c":"clickCounterLabel1"});

Even if most of the times we will let Wicket generate request attributes for us, both AJAX components and behaviors give us the chance to modify them overriding their method updateAjaxAttributes (AjaxRequestAttributes attributes).

One of the attribute we may need to modify is the list of IAjaxCallListeners returned by method getAjaxCallListeners().

IAjaxCallListener defines a set of methods which return the JavaScript code (as a CharSequence) that must be executed on client side when the AJAX request handling reaches a given stage:

• getBeforeHandler(Component): returns the JavaScript code that will be executed before any other handlers returned by IAjaxCallListener. The code is executed in a scope where it can use variable attrs, which is an array containing the JSON parameters passed to Wicket.Ajax.ajax.
• getPrecondition(Component): returns the JavaScript code that will be used as precondition for the AJAX call. If the script returns false then neither the Ajax call nor the other handlers will be executed. The code is executed in a scope where it can use variable attrs, which is the same variable seen for getBeforeHandler.
• getBeforeSendHandler(Component): returns the JavaScript code that will be executed just before the AJAX call is performed. The code is executed in a scope where it can use variables attrs, jqXHR and settings:
• attrs is the same variable seen for getBeforeHandler.
• jqXHR is the the jQuery XMLHttpRequest object used to make the AJAX call.
• settings contains the settings used for calling jQuery.ajax().
• getAfterHandler(Component): returns the JavaScript code that will be executed after the AJAX call. The code is executed in a scope where it can use variable attrs, which is the same variable seen before for getBeforeHandler.
• getSuccessHandler(Component): returns the JavaScript code that will be executed if the AJAX call has successfully returned. The code is executed in a scope where it can use variables attrs, jqXHR, data and textStatus:
• attrs and jqXHR are same variables seen for getBeforeSendHandler:
• data is the data returned by the AJAX call. Its type depends on parameter wr (Wicket AJAX response).
• textStatus it's the status returned as text.
• getFailureHandler(Component): returns the JavaScript code that will be executed if the AJAX call has returned with a failure. The code is executed in a scope where it can use variable attrs, which is the same variable seen for getBeforeHandler.
• getCompleteHandler(Component): returns the JavaScript that will be invoked after success or failure handler has been executed. The code is executed in a scope where it can use variables attrs, jqXHR and textStatus which are the same variables seen for getSuccessHandler.

In the next paragraph we will see an example of custom IAjaxCallListener designed to disable a component during AJAX request processing.

17.6 Creating custom AJAX call listener

Displaying an activity indicator is a nice way to notify user that an AJAX request is already running, but sometimes is not enough. In some situations we may need to completely disable a component during AJAX request processing, for example when we want to avoid that impatient users submit a form multiple times. In this paragraph we will see how to accomplish this goal building a custom and reusable IAjaxCallListener. The code used in this example is from project CustomAjaxListenerExample.

What we want for our listener

The listener should execute some JavaScript code to disable a given component when the component it is attached to is about to make an AJAX call. Then, when the AJAX request has been completed, the listener should bring back the disabled component to an active state.

When a component is disabled it must be clear to user that an AJAX request is running and that he/she must wait for it to complete. To achieve this result we want to disable a given component covering it with a semi-transparent overlay area with an activity indicator in the middle.

The final result will look like this:

How to implement the listener

The listener will implement methods getBeforeHandler and getAfterHandler: the first will return the code needed to place an overlay <div> on the desired component while the second must remove this overlay when the AJAX call has completed.

To move and resize the overlay area we will use another module from JQueryUI library that allows us to position DOM elements on our page relative to another element.

So our listener will depend on four static resources: the JQuery library, the position module of JQuery UI, the custom code used to move the overlay <div> and the picture used as activity indicator. Except for the activity indicator, all these resources must be added to page header section in order to be used.

Ajax call listeners can contribute to header section by simply implementing interface IComponentAwareHeaderContributor. Wicket provides adapter class AjaxCallListener that implements both IAjaxCallListener and IComponentAwareHeaderContributor. We will use this class as base class for our listener.

JavaScript code

Now that we know what to do on the Java side, let's have a look at the custom JavaScript code that must be returned by our listener (file moveHiderAndIndicator.js):

DisableComponentListener = {
   disableElement: function(elementId, activeIconUrl){
      var hiderId = elementId + "-disable-layer";
      var indicatorId = elementId + "-indicator-picture";
      elementId = "#" + elementId;
      //create the overlay <div>
      $(elementId).after('<div id="' + hiderId 
         + '" style="position:absolute;">'
         + '<img id="' + indicatorId +  '" src="' + activeIconUrl + '"/>'
         + '</div>');
      hiderId = "#" + hiderId;
      //set the style properties of the overlay <div>
      $(hiderId).css('opacity', '0.8');               
      $(hiderId).css('text-align', 'center');
      $(hiderId).css('background-color', 'WhiteSmoke');
      $(hiderId).css('border', '1px solid DarkGray');
      //set the dimention of the overlay <div>
      $(hiderId).width($(elementId).outerWidth());
      $(hiderId).height($(elementId).outerHeight());       	 
      //positioning the overlay <div> on the component that must be disabled.     
      $(hiderId).position({of: $(elementId),at: 'top left', my: 'top left'});
      //positioning the activity indicator in the middle of the overlay <div>
      $("#" + indicatorId).position({of: $(hiderId), at: 'center center',
                                     my: 'center center'});
   },
   //function hideComponent

Function DisableComponentListener.disableElement places the overlay <div> an the activity indicator on the desired component. The parameters in input are the markup id of the component we want to disable and the URL of the activity indicator picture. These two parameters must be provided by our custom listener.

The rest of custom JavaScript contains function DisableComponentListener.hideComponent which is just a wrapper around the JQuery function remove():

hideComponent: function(elementId){
	var hiderId = elementId + "-disable-layer";
	$('#' + hiderId).remove();
	}
};

Java class code

The code of our custom listener is the following:

public class DisableComponentListener extends AjaxCallListener {
   private static PackageResourceReference customScriptReference = new   
   PackageResourceReference(DisableComponentListener.class,"moveHiderAndIndicator.js");
   private static PackageResourceReference jqueryUiPositionRef = new    
   PackageResourceReference(DisableComponentListener.class, "jquery-ui-position.min.js");
   private static PackageResourceReference indicatorReference = 
         new PackageResourceReference(DisableComponentListener.class, "ajax-loader.gif");
   private Component targetComponent;
   public DisableComponentListener(Component targetComponent){
      this.targetComponent = targetComponent;
   }
   @Override
   public CharSequence getBeforeHandler(Component component) {   
      CharSequence indicatorUrl = getIndicatorUrl(component);
      return ";DisableComponentListener.disableElement('" + targetComponent.getMarkupId() 
                                                   + "'," + "'" + indicatorUrl + "');";
   }
   @Override
   public CharSequence getCompleteHandler(Component component) {
      return ";DisableComponentListener.hideComponent('" + targetComponent.getMarkupId() + 
                                                                               "');";
   }
   protected CharSequence getIndicatorUrl(Component component) {
      return component.urlFor(indicatorReference, null);
   }
   @Override
   public void renderHead(Component component, IHeaderResponse response) {   
      ResourceReference jqueryReference = Application.get().getJavaScriptLibrarySettings(). 
                                     getJQueryReference();
      response.render(JavaScriptHeaderItem.forReference(jqueryReference));      
      response.render(JavaScriptHeaderItem.forReference(jqueryUiPositionRef));
      response.render(JavaScriptHeaderItem.forReference(customScriptReference));
   }
}

As you can see in the code above we have created a function (getIndicatorUrl) to retrieve the URL of the indicator picture. This was done in order to make the picture customizable by overriding this method.

Once we have our listener in place, we can finally use it in our example overwriting method updateAjaxAttributes of the AJAX button that submits the form:

//…
new AjaxButton("ajaxButton"){
	@Override
	protected void updateAjaxAttributes(AjaxRequestAttributes attributes) {
	  super.updateAjaxAttributes(attributes);
	  attributes.getAjaxCallListeners().add(new DisableComponentListener(form));
	}
}
//...

Global listeners

So far we have seen how to use an AJAX call listener to track the AJAX activity of a single component. In addition to these kinds of listeners, Wicket provides also global listeners which are triggered for any AJAX request sent from a page.

Global AJAX call events are handled with JavaScript. We can register a callback function for a specific event of the AJAX call lifecycle with function Wicket.Event.subscribe('<eventName>', <callback Function>). The first parameter of this function is the name of the event we want to handle. The possible names are:

• '/ajax/call/before': called before any other event handler.
• '/ajax/call/beforeSend': called just before the AJAX call.
• '/ajax/call/after': called after the AJAX request has been sent.
• '/ajax/call/success': called if the AJAX call has successfully returned.
• '/ajax/call/failure': called if the AJAX call has returned with a failure.
• '/ajax/call/complete': called when the AJAX call has completed.
• '/dom/node/removing': called when a component is about to be removed via AJAX. This happens when component markup is updated via AJAX (i.e. the component itself or one of its containers has been added to AjaxRequestTarget)
• '/dom/node/added': called when a component has been added via AJAX. Just like '/dom/node/removing', this event is triggered when a component is added to AjaxRequest Target.

The callback function takes in input the following parameters: attrs, jqXHR, textStatus, jqEvent and errorThrown. The first three parameters are the same seen before with IAjaxCallListener while jqEvent is an event internally fired by Wicket. The last parameter errorThrown indicates if an error has occurred during the AJAX call.

To see a basic example of use of a global AJAX call listener, let's go back to our custom datepicker created in chapter 14. When we built it we didn't think about a possible use of the component with AJAX. When a complex component like our datepicker is refreshed via AJAX, the following two side effects can occur:

• After been refreshed, the component loses every JavaScript handler set on it. This is not a problem for our datepicker as it sets a new JQuery datepicker every time is rendered (inside method renderHead).
• The markup previously created with JavaScript is not removed. For our datepicker this means that the icon used to open the calendar won't be removed while a new one will be added each time the component is refreshed.

To solve the second unwanted side effect we can register a global AJAX call listener that completely removes the datepicker functionality from our component before it is removed due to an AJAX refresh (which fires event '/dom/node/removing').

Project CustomDatepickerAjax contains a new version of our datepicker which adds to its JavaScript file JQDatePicker.js the code needed to register a callback function that gets rid of the JQuery datepicker before the component is removed from the DOM:

Wicket.Event.subscribe('/dom/node/removing', 
    function(jqEvent, attributes, jqXHR, errorThrown, textStatus) {
	var componentId = '#' + attributes['id'];
	if($(componentId).datepicker !== undefined)
	      $(componentId).datepicker('destroy');
     }
);

The code above retrieves the id of the component that is about to be removed using parameter attributes. Then it checks if a JQuery datepicker was defined for the given component and if so, it removes the widget calling function destroy.

17.7 Summary

AJAX is another example of how Wicket can simplify web technologies providing a good component and object oriented abstraction of them.

In this chapter we have seen how to take advantage of the AJAX support provided by Wicket to write AJAX-enhanced applications. Most of the chapter has been dedicated to the built-in components and behaviors that let us adopt AJAX without almost any effort.

In the final part of the chapter we have seen how Wicket physically implements an AJAX call on client side using AJAX request attributes. Then, we have learnt how to use call listeners to execute custom JavaScript during AJAX request lifecycle.

18 Integration with enterprise containers

The first decade of 2000s has seen the rising of new frameworks (like Spring ) and new specifications (like EJB 3.1 ) aimed to simplify the management of enterprise resources and (among other things) their integration with presentation code.

All these new technologies are based on the concepts of container and dependency injection. Container is the environment where our enterprise resources are created and configured while dependency injection is a pattern implemented by containers to inject into an object the resources it depends on.

Wicket can be easily integrated with enterprise containers using component instantiation listeners. These entities are instances of interface org.apache.wicket.application.IComponent InstantiationListener and can be registered during application's initialization. IComponentInstantiationListener defines callback method onInstantiation(Component component) which can be used to provide custom instantiation logic for Wicket components.

Wicket distribution and project WicketStuff already provide a set of built-in listeners to integrate our applications with EJB 3.1 compliant containers (like JBoss Seam) or with some of the most popular enterprise frameworks like Guice or Spring.

In this chapter we will see two basic examples of injecting a container-defined object into a page using first an implementation of the EJB 3.1 specifications (project OpenEJB ) and then using Spring.

18.1 Integrating Wicket with EJB

WicketStuff provides a module called wicketstuff-javaee-inject that contains component instantiation listener JavaEEComponentInjector. If we register this listener in our application we can use standard EJB annotations to inject dependencies into our Wicket components.

To register a component instantiation listener in Wicket we must use Application's method getComponentInstantiationListeners which returns a typed collection of IComponent InstantiationListeners.

The following initialization code is taken from project EjbInjectionExample:

public class WicketApplication extends WebApplication
{    	
       //Constructor...
	@Override
	public void init()
	{
		super.init();
		getComponentInstantiationListeners().add(new JavaEEComponentInjector(this));		
	}	
}

In this example the object that we want to inject is a simple class containing a greeting message:

@ManagedBean
public class EnterpriseMessage {
	public String message = "Welcome to the EJB world!";
}

Please note that we have used annotation ManagedBean to decorate our object. Now to inject it into the home page we must add a field of type EnterpriseMessage and annotate it with annotation EJB:

public class HomePage extends WebPage {
	@EJB
	private EnterpriseMessage enterpriseMessage;
	//getter and setter for enterpriseMessage...
	public HomePage(final PageParameters parameters) {
		super(parameters);
		add(new Label("message", enterpriseMessage.message));
	}
}

That is all. We can point the browser to the home page of the project and see the greeting message injected into the page:

18.2 Integrating Wicket with Spring

If we need to inject dependencies with Spring we can use listener org.apache.wicket.spring. injection.annot.SpringComponentInjector provided by module wicket-spring.

For the sake of simplicity in the example project SpringInjectionExample we have used Spring class AnnotationConfigApplicationContext to avoid any XML file and create a Spring context directly from code:

public class WicketApplication extends WebApplication
{      
  //Constructor...
  @Override
  public void init()
  {
    super.init();
    AnnotationConfigApplicationContext ctx = new AnnotationConfigApplicationContext();
    //Scan package for annotated beans
    ctx.scan("org.wicketTutorial.ejbBean");
    ctx.refresh();
    getComponentInstantiationListeners().add(new SpringComponentInjector(this, ctx));
  }	
}

As we can see in the code above, the constructor of SpringComponentInjector takes in input also an instance of Spring context.

The injected object is the same used in the previous project EjbInjectionExample, it differs only for the greeting message:

@ManagedBean
public class EnterpriseMessage {
	public String message = "Welcome to the Spring world!";
}

In the home page of the project the object is injected using Wicket annotation @SpringBean:

public class HomePage extends WebPage {
  @SpringBean
  private EnterpriseMessage enterpriseMessage;
  //getter and setter for enterpriseMessage...
  public HomePage(final PageParameters parameters) {
  	super(parameters);
	add(new Label("message", enterpriseMessage.message));
  }
}

By default @SpringBean searches into Spring context for a bean having the same type of the annotated field. If we want we can specify also the name of the bean to use as injected object and we can declare if the dependency is required or not. By default dependencies are required and if they can not be resolved to a compatible bean, Wicket will throw an IllegalStateException:

//set the dependency as not required, i.e the field can be left null
  @SpringBean(name="anotherName", required=false)
  private EnterpriseMessage enterpriseMessage;

18.3 JSR-330 annotations

Spring (and Guice) users can use standard JSR-330 annotations to wire their dependencies. This will make their code more interoperable with other containers that support this standard:

//inject a bean specifying its name with JSR-330 annotations
  Inject Named("anotherName")
  private EnterpriseMessage enterpriseMessage;

18.4 Summary

In this chapter we have seen how to integrate Wicket applications with Spring and with an EJB container. Module wicket-examples contains also an example of integration with Guice (see application class org.apache.wicket.examples.guice.GuiceApplication).

19 Security with Wicket

In this chapter we will explore the security infrastructure provided by Wicket and we will learn how to use it to implement authentication and authorizations in our web applications.

19.1 Authentication

The first step in implementing a security policy is assigning a trusted identity to our users, which means that we must authenticate them. Web applications usually adopt a form-based authentication with a login form that asks user for a unique username and the relative password:

Wicket supports form-based authentication with session class AuthenticatedWebSession and application class AuthenticatedWebApplication, both placed inside package org.apache. wicket.authroles.authentication.

AuthenticatedWebSession

Class AuthenticatedWebSession comes with the following set of public methods to manage user authentication:

• authenticate(String username, String password): this is an abstract method that must be implemented by every subclass of AuthenticatedWebSession. It should contain the actual code that checks for user's identity. It returns a boolean value which is true if authentication has succeeded or false otherwise.
• signIn(String username, String password): this method internally calls authenticate and set the flag signedIn to true if authentication succeeds.
• isSignedIn():getter method for flag signedIn.
• signOut(): sets the flag signedIn to false.
• invalidate(): calls signOut and invalidates session.

Remember that signOut does not discard any session-relative data. If we want to get rid of these data, we must invoke method invalidate instead of signOut.

Another abstract method we must implement when we use AuthenticatedWebSession is getRoles which is inherited from parent class AbstractAuthenticatedWebSession. This method can be ignored for now as it will be discussed later when we will talk about role-based authorization.

AuthenticatedWebApplication

Class AuthenticatedWebApplication provides the following methods to support form-based authentication:

• getWebSessionClass(): abstract method that returns the session class to use for this application. The returned class must be a subclass of AbstractAuthenticatedWeb Session.
• getSignInPageClass(): abstract method that returns the page to use as sign in page when a user must be authenticated.
• restartResponseAtSignInPage(): forces the current response to restart at the sign in page. After we have used this method to redirect a user, we can make her/him return to the original page calling Componet's method continueToOriginalDestination().

The other methods implemented inside AuthenticatedWebApplication will be introduced when we will talk about authorizations.

A basic example of authentication

Project BasicAuthenticationExample is a basic example of form-based authentication implemented with classes AuthenticatedWebSession and AuthenticatedWebApplication.

The homepage of the project contains only a link to page AuthenticatedPage which can be accessed only if user is signed in. The code of AuthenticatedPage is this following:

public class AuthenticatedPage extends WebPage {
   @Override
   protected void onConfigure() {
      AuthenticatedWebApplication app = (AuthenticatedWebApplication)Application.get();
      //if user is not signed in, redirect him to sign in page
      if(!AuthenticatedWebSession.get().isSignedIn())
         app.restartResponseAtSignInPage();
   }
   @Override
   protected void onInitialize() {
      super.onInitialize();
      add(new Link("goToHomePage") {
         @Override
         public void onClick() {
            setResponsePage(getApplication().getHomePage());
         }
      });
      add(new Link("logOut") {
         @Override
         public void onClick() {
            AuthenticatedWebSession.get().invalidate();
            setResponsePage(getApplication().getHomePage());
         }
      });
   }
}

Page AuthenticatedPage checks inside onConfigure if user is signed in and if not, it redirects her/him to the sign in page with method restartResponseAtSignInPage. The page contains also a link to the homepage and another link that signs out user.

The sign in page is implemented in class SignInPage and contains the form used to authenticate users:

public class SignInPage extends WebPage {
   private String username;
   private String password;
   @Override
   protected void onInitialize() {
      super.onInitialize();
      StatelessForm form = new StatelessForm("form"){
         @Override
         protected void onSubmit() {
            if(Strings.isEmpty(username))
               return;
            boolean authResult = AuthenticatedWebSession.get().signIn(username, password);
            //if authentication succeeds redirect user to the requested page
            if(authResult)
               continueToOriginalDestination();
         }
      };
      form.setDefaultModel(new CompoundPropertyModel(this));
      form.add(new TextField("username"));
      form.add(new PasswordTextField("password"));
      add(form);
   }
}

The form is responsible for handling user authentication inside its method onSubmit. The username and password are passed to AuthenticatedWebSession's method signIn(username, password) and if authentication succeeds, the user is redirected to the original page with method continueToOriginalDestination.

The session class and the application class used in the project are reported here:

Session class:

public class BasicAuthenticationSession extends AuthenticatedWebSession {
	public BasicAuthenticationSession(Request request) {
		super(request);		
	}
	@Override
	public boolean authenticate(String username, String password) {
	      //user is authenticated if both username and password are equal to 'wicketer'
		return username.equals(password) && username.equals("wicketer");
	}
	@Override
	public Roles getRoles() {
		return null;
	}
}

Application class:

public class WicketApplication extends AuthenticatedWebApplication{    	
	@Override
	public Class<HomePage> getHomePage(){
		return HomePage.class;
	}
	@Override
	protected Class<? extends AbstractAuthenticatedWebSession> getWebSessionClass(){
		return BasicAuthenticationSession.class;
	}
	@Override
	protected Class<? extends WebPage> getSignInPageClass() {
		return SignInPage.class;
	}
}

The authentication logic inside authenticate has been kept quite trivial in order to make the code as clean as possible. Please note also that session class must have a constructor that accepts an instance of class Request.

Redirecting user to an intermediate page

Method restartResponseAtSignInPage is an example of redirecting user to an intermediate page before allowing him to access to the requested page. This method internally throws exception org.apache.wicket.RestartResponseAtInterceptPageException which saves the URL of the requested page into session metadata and then redirects user to the page passed as constructor parameter (the sign in page).

Component's method redirectToInterceptPage(Page) works in much the same way as restartResponseAtSignInPage but it allows us to specify which page to use as intermediate page:

redirectToInterceptPage(intermediatePage);

Since both restartResponseAtSignInPage and redirectToIntercept Page internally throw an exception, the code placed after them will not be executed.

19.2 Authorizations

The authorization support provided by Wicket is built around the concept of authorization strategy which is represented by interface IAuthorizationStrategy (in package org.apache.wicket .authorization):

public interface IAuthorizationStrategy
{
  //interface methods 
 <T extends IRequestableComponent> boolean isInstantiationAuthorized(Class<T> componentClass);
 boolean isActionAuthorized(Component component, Action action);
 //default authorization strategy that allows everything
 public static final IAuthorizationStrategy ALLOW_ALL = new IAuthorizationStrategy()
 {
  @Override
  public <T extends IRequestableComponent> boolean isInstantiationAuthorized(final Class<T> c)
  {
    return true;
  }
  @Override
  public boolean isActionAuthorized(Component c, Action action)
  {
    return true;
  }
 };
}

This interface defines two methods:

• isInstantiationAuthorized checks if user is allowed to instantiate a given component.
• isActionAuthorized checks if user is authorized to perform a given action on a component's instance. The standard actions checked by this method are defined into class Action and are Action.ENABLE and Action.RENDER.

Inside IAuthorizationStrategy we can also find a default implementation of the interface (called ALLOW_ALL) that allows everyone to instantiate every component and perform every possible action on it. This is the default strategy adopted by class Application.

To change the authorization strategy in use we must register the desired implementation into security settings (interface ISecuritySettings) during initialization phase with method setAuthorization Strategy:

//Application class code… 
  @Override
  public void init()
  {
    super.init();
    getSecuritySettings().setAuthorizationStrategy(myAuthorizationStrategy);
  }	
//...

If we want to combine the action of two or more authorization strategies we can chain them with strategy CompoundAuthorizationStrategy which implements composite patter for authorization strategies.

Most of the times we won't need to implement an IAuthorizationStrategy from scratch as Wicket already comes with a set of built-in strategies. In the next paragraphs we will see some of these strategies that can be used to implement an effective and flexible security policy.

SimplePageAuthorizationStrategy

Abstract class SimplePageAuthorizationStrategy (in package org.apache.wicket. authorization.strategies.page) is a strategy that checks user authorizations calling abstract method isAuthorized only for those pages that are subclasses of a given supertype. If isAuthorized returns false, the user is redirected to the sign in page specified as second constructor parameter:

SimplePageAuthorizationStrategy authorizationStrategy = new SimplePageAuthorizationStrategy( 
                                                  PageClassToCheck.class, SignInPage.class)
{
  protected boolean isAuthorized()
  {		                
    //Authentication code…
  }
};

By default SimplePageAuthorizationStrategy checks for permissions only on pages. If we want to change this behavior and check also other kinds of components, we must override method isActionAuthorized and implement our custom logic inside it.

Role-based strategies

At the end of paragraph 18.1.1 we have introduced AbstractAuthenticatedWebSession's method getRoles which is provided to support role-based authorization returning the set of roles granted to the current user.

In Wicket roles are simple strings like “BASIC_USER” or “ADMIN” (they don't need to be capitalized) and they are handled with class org.apache.wicket.authroles.authorization.strategies. role.Roles. This class extends standard HashSet collection adding some functionalities to check whether the set contains one or more roles. Class Roles already defines roles Roles.USER and Roles.ADMIN.

The session class in the following example returns a custom “SIGNED_IN” role for every authenticated user and it adds an Roles.ADMIN role if username is equal to superuser:

class BasicAuthenticationRolesSession extends AuthenticatedWebSession {
	private String userName;
	public BasicAuthenticationRolesSession(Request request) {
		super(request);		
	}
	@Override
	public boolean authenticate(String username, String password) {
		boolean authResult= false;
		authResult = //some authentication logic...
		if(authResult)
			userName = username;
		return authResult;
	}
	@Override
	public Roles getRoles() {
		Roles resultRoles = new Roles();
		if(isSignedIn())
			resultRoles.add("SIGNED_IN");
		if(userName.equals("superuser"))
			resultRoles.add(Roles.ADMIN);
		return resultRoles;
	}
}

Roles can be adopted to apply security restrictions on our pages and components. This can be done using one of the two built-in authorization strategies that extend super class AbstractRole AuthorizationStrategyWicket: MetaDataRoleAuthorizationStrategy and Annotations RoleAuthorizationStrategy

The difference between these two strategies is that MetaDataRoleAuthorizationStrategy handles role-based authorizations with Wicket metadata while AnnotationsRoleAuthorization-Strategy uses Java annotations.

Application class AuthenticatedWebApplication already sets MetaData RoleAuthorizationStrategy and AnnotationsRoleAuthorization Strategy as its own authorization strategies (it uses a compound strategy as we will see in paragraph 18.2.4).

The code that we will see in the next examples is for illustrative purpose only. If our application class inherits from AuthenticatedWebApplication we won't need to configure anything to use these two strategies.

Using roles with metadata

Strategy MetaDataRoleAuthorizationStrategy uses application and components metadata to implement role-based authorizations. The class defines a set of static methods authorize that can be used to specify which roles are allowed to instantiate a component and which roles can perform a given action on a component.

The following code snippet reports both application and session classes from project MetaDataRolesStrategyExample and illustrates how to use MetaDataRoleAuthorizationStrategy to allow access to a given page (AdminOnlyPage) only to ADMIN role:

Application class:

public class WicketApplication extends AuthenticatedWebApplication{    		
   @Override
   public Class<? extends WebPage> getHomePage(){
      return HomePage.class;
   }
   @Override
   protected Class<? extends AbstractAuthenticatedWebSession> getWebSessionClass() {
      return BasicAuthenticationSession.class;
   }
   @Override
   protected Class<? extends WebPage> getSignInPageClass() {
      return SignInPage.class;
   }
   @Override
   public void init(){   
      getSecuritySettings().setAuthorizationStrategy(new MetaDataRoleAuthorizationStrategy(this));
      MetaDataRoleAuthorizationStrategy.authorize(AdminOnlyPage.class, Roles.ADMIN);
   }
}

Session class:

public class BasicAuthenticationSession extends AuthenticatedWebSession {
   private String username;
   public BasicAuthenticationSession(Request request) {
      super(request);      
   }
   @Override
   public boolean authenticate(String username, String password) {
      //user is authenticated if username and password are equal
     boolean authResult = username.equals(password);
      if(authResult)
         this.username = username;
      return authResult;
   }
   public Roles getRoles() {
      Roles resultRoles = new Roles();
      //if user is signed in add the relative role
      if(isSignedIn())
         resultRoles.add("SIGNED_IN");
      //if username is equal to 'superuser' add the ADMIN role
      if(username!= null && username.equals("superuser"))
         resultRoles.add(Roles.ADMIN);
      return resultRoles;
   }
   @Override
   public void signOut() {
      super.signOut();
      username = null;
   }
}

The code that instantiates MetaDataRoleAuthorizationStrategy and set it as application's strategy is inside application class method init.

Any subclass of AbstractRoleAuthorizationStrategyWicket needs an implementation of interface IRoleCheckingStrategy to be instantiated. For this purpose in the code above we used the application class itself because its base class AuthenticatedWebApplication already implements interface IRoleCheckingStrategy. By default AuthenticatedWebApplication checks for authorizations using the roles returned by the current AbstractAuthenticated WebSession. As final step inside init we grant the access to page AdminOnlyPage to ADMIN role calling method authorize.

The code from session class has three interesting methods. The first is authenticate which considers as valid credentials every pair of username and password having the same value. The second notable method is getRoles which returns role SIGNED_IN if user is authenticated and it adds role ADMIN if username is equal to superuser. Finally, we have method signOut which has been overridden in order to clean the username field used internally to generate roles.

Now if we run the project and we try to access to AdminOnlyPage from the home page without having the ADMIN role, we will be redirected to the default access-denied page used by Wicket:

The access-denied page can be customized using method setAccessDeniedPage(Class<? extends Page>) of setting interface IApplicationSettings:

//Application class code…
   @Override
   public void init(){   
      getApplicationSettings().setAccessDeniedPage(MyCustomAccessDeniedPage.class); 
   }

Just like custom “Page expired” page (see chapter 6.2.5), also custom “Access denied” page must be bookmarkable.

Using roles with annotations

Strategy AnnotationsRoleAuthorizationStrategy relies on two built-in annotations to handle role-based authorizations. These annotations are AuthorizeInstantiation and Authorize Action. As their names suggest the first annotation specifies which roles are allowed to instantiate the annotated component while the second must be used to indicate which roles are allowed to perform a specific action on the annotated component.

In the following example we use annotations to make a page accessible only to signed-in users and to enable it only if user has the ADMIN role:

@AuthorizeInstantiation("SIGNED_IN")
@AuthorizeAction(action = "ENABLE", roles = {"ADMIN"})
public class MyPage extends WebPage {
   //Page class code…
}

Remember that when a component is not enabled, user can render it but he can neither click on its links nor interact with its forms.

Example project AnnotationsRolesStrategyExample is a revisited version of MetaDataRolesStrategy Example where we use AnnotationsRoleAuthorizationStrategy as authorization strategy. To ensure that page AdminOnlyPage is accessible only to ADMIN role we have used the following annotation:

@AuthorizeInstantiation("ADMIN")
public class AdminOnlyPage extends WebPage {
    //Page class code…
}

Catching an unauthorized component instantiation

Interface IUnauthorizedComponentInstantiationListener (in package org.apache. wicket.authorization) is provided to give the chance to handle the case in which a user tries to instantiate a component without having the permissions to do it. The method defined inside this interface is onUnauthorizedInstantiation(Component) and it is executed whenever a user attempts to execute an unauthorized instantiation.

This listener must be registered into application's security settings with method setUnauthorized ComponentInstantiationListener defined by setting interface ISecuritySettings. In the following code snippet we register a listener that redirect user to a warning page if he tries to do a not-allowed instantiation:

public class WicketApplication extends AuthenticatedWebApplication{   
     //Application code…
     @Override
     public void init(){    
        getSecuritySettings().setUnauthorizedComponentInstantiationListener(
			new IUnauthorizedComponentInstantiationListener() {
	    @Override
	    public void onUnauthorizedInstantiation(Component component) {
	        component.setResponsePage(AuthWarningPage.class);
	    }
        });
     }
}

In addition to interface IRoleCheckingStrategy, class AuthenticatedWebApplication implements also IUnauthorizedComponentInstantiationListener and registers itself as listener for unauthorized instantiations.

By default AuthenticatedWebApplication redirects users to sign-in page if they are not signed-in and they try to instantiate a restricted component. Otherwise, if users are already signed in but they are not allowed to instantiate a given component, an UnauthorizedInstantiationException will be thrown.

Strategy RoleAuthorizationStrategy

Class RoleAuthorizationStrategy is a compound strategy that combines both MetaData RoleAuthorizationStrategy and AnnotationsRoleAuthorizationStrategy.

This is the strategy used internally by AuthenticatedWebApplication.

19.3 Using HTTPS protocol

HTTPS is the standard technology adopted on Internet to create a secure communication channel between web applications and their users.

In Wicket we can easily protect our pages with HTTPS mounting a special request mapper called HttpsMapper and using annotation RequireHttps with those pages we want to serve over this protocol. Both these two entities are in package org.apache.wicket.protocol.https.

HttpsMapper wraps an existing mapper and redirects incoming requests to HTTPS if the related response must render a page containing annotation RequireHttps. Most of the times the wrapped mapper will be the root one, just like we saw before for CryptoManager in paragraph 8.6.6.

Another parameter needed to build a HttpsMapper is an instance of class HttpsConfig. This class allows us to specify which ports must be used for HTTPS and HTTP. By default the port numbers used by these two protocols are respectively 443 and 80.

The following code is taken from project HttpsProtocolExample and illustrates how to enable HTTPS in our applications:

//Application class code…
@Override
public void init(){   
   setRootRequestMapper(new HttpsMapper(getRootRequestMapper(), 
                                       new HttpsConfig(8080, 443))); 
}

Now we can use annotation RequireHttps to specify which pages must be served using HTTPS:

@RequireHttps
public class HomePage extends WebPage {
    public HomePage(final PageParameters parameters) {
    	super(parameters);	
    }
}

If we want to protect many pages with HTTPS without adding annotation RequireHttps to each of them, we can annotate a marker interface or a base page class and implement/extend it in any page we want to make secure:

// Marker interface:
@RequireHttps
public interface IMarker{
}
// Base class:
@RequireHttps
public class BaseClass extends WebPage{
//Page code…
}
// Secure page inheriting from BaseClass:
public class HttpsPage extends BaseClass{
//Page code…
}
// Secure page implementing IMarker:
public class HttpsPage implements IMarker{
//Page code…
}

19.4 Package Resource Guard

Wicket internally uses an entity called package resource guard to protect package resources from external access. This entity is an implementation of interface org.apache.wicket.markup.html. IPackageResourceGuard.

By default Wicket applications use as package resource guard class SecurePackageResource Guard, which allows to access only to the following file extensions (grouped by type):

To modify the set of allowed files formats we can add one or more patterns with method addPattern (String). The rules to write a pattern are the following:

• patterns start with either a "+" or a "-". In the first case the pattern will add one or more file to the set while starting a pattern with a “-” we exclude all the files matching the given pattern. For example pattern “-web.xml” excludes all web.xml files in all directories.
• wildcard character “*” is supported as placeholder for zero or more characters. For example pattern “+*.mp4” adds all the mp4 files inside all directories.
• subdirectories are supported as well. For example pattern “+documents/*.pdf” adds all pdf files under “documents” directory. Character “*” can be used with directories to specify a nesting level. For example “+documents/*/*.pdf” adds all pdf files placed one level below “documents” directory.
• a double wildcard character “**” indicates zero or more subdirectories. For example patter “+documents/**/*.pdf” adds all pdf files placed inside “documents” directory or inside any of its subdirectories.

Patterns that allow to access to every file with a given extensions (such as “+*.pdf”) should be always avoided in favour of more restrictive expressions that contain a directory structure:

//Application class code…
@Override
public void init()   
{
      IPackageResourceGuard packageResourceGuard = application.getResourceSettings() 
                                                              .getPackageResourceGuard();
      if (packageResourceGuard instanceof SecurePackageResourceGuard)
      {
         SecurePackageResourceGuard guard = (SecurePackageResourceGuard) packageResourceGuard;
         //Allow to access only to pdf files placed in the “public” directory.
         guard.addPattern("+public/*.pdf");
      }
}

19.5 Summary

In this chapter we have seen the components and the mechanisms that allow us to implement security policies in our Wicket-based applications. Wicket comes with an out of the box support for both authorization and authentication.

The central element of authorization mechanism is the interface IAuthorizationStrategy which decouples our components from any detail about security strategy. The implementations of this interface must decide if a user is allowed to instantiate a given page or component and if she/he can perform a given action on it.

Wicket natively supports role-based authorizations with strategies MetaDataRoleAuthorization Strategy and AnnotationsRoleAuthorizationStrategy. The difference between these two strategies is that the first offers a programmatic approach for role handling while the second promotes a declarative approach using built-in annotations.

After having explored how Wicket internally implements authentication and authorization, in the last part of the chapter we have learnt how to configure our applications to support HTTPS and how to specify which pages must be served over this protocol.

In the last paragraph we have seen how Wicket protects package resources with a guard entity that allows us to decide which package resources can be accessed from users.

20 Test Driven Development with Wicket

In this chapter we will see how to write unit tests for our applications and components and we will learn how to use helper classes to simulate user navigation and write acceptance tests without the need of any testing framework other than JUnit.

The JUnit version used in this chapter is 4.x.

20.1 Utility class WicketTester

A good way to start getting confident with Wicket unit testing support is looking at the test case class TestHomePage that is automatically generated by Maven when we use Wicket archetype to create a new project:

Here is the content of TestHomePage:

public class TestHomePage{
	private WicketTester tester;
	@Before
	public void setUp(){
		tester = new WicketTester(new WicketApplication());
	}
	@Test
	public void homepageRendersSuccessfully(){
		//start and render the test page
		tester.startPage(HomePage.class);
		//assert rendered page class
		tester.assertRenderedPage(HomePage.class);
	}
}

The central class in a Wicket testing is org.apache.wicket.util.tester.WicketTester. This utility class provides a set of methods to render a component, click links, check if page contains a given component or a feedback message, and so on.

The basic test case shipped with TestHomePage illustrates how WicketTester is typically instantiated (inside method setUp()). In order to test our components, WicketTester needs to use an instance of WebApplication. Usually, we will use our application class as WebApplication, but we can also decide to build WicketTester invoking its no-argument constructor and letting it automatically build a mock web application (an instance of class org.apache.wicket.mock. MockApplication).

The code from TestHomePage introduces two basic methods to test our pages. The first is method startPage that renders a new instance of the given page class and sets it as current rendered page for WicketTester. The second method is assertRenderedPage which checks if the current rendered page is an instance of the given class. In this way if TestHomePage succeeds we are sure that page HomePage has been rendered without any problem. The last rendered page can be retrieved with method getLastRenderedPage.

That's only a taste of what WicketTester can do. In the next paragraphs we will see how it can be used to test every element that composes a Wicket page (links, models, behaviors, etc...).

Testing links

A click on a Wicket link can be simulated with method clickLink which takes in input the link component or the page-relative path to it.

To see an example of usage of clickLink, let's consider again project LifeCycleStagesRevisited. As we know from chapter 5 the home page of the project alternately displays two different labels (“First label” and “Second label”), swapping between them each time button "reload" is clicked. The code from its test case checks that label has actually changed after button "reload" has been pressed:

//…
@Test
public void switchLabelTest(){
	//start and render the test page
	tester.startPage(HomePage.class);
	//assert rendered page class
	tester.assertRenderedPage(HomePage.class);
	//assert rendered label
	tester.assertLabel("label", "First label");
	//simulate a click on "reload" button
	tester.clickLink("reload");
	//assert rendered label
	tester.assertLabel("label", "Second label");	
}
//...

In the code above we have used clickLink to click on the "reload" button and force page to be rendered again. In addition, we have used also method assertLabel that checks if a given label contains the expected text.

By default clickLink assumes that AJAX is enabled on client side. To switch AJAX off we can use another version of this method that takes in input the path to the link component and a boolean flag that indicates if AJAX must be enabled (true) or not (false).

//…
//simulate a click on a button without AJAX support
tester.clickLink("reload", false);
//...

Testing component status

WicketTester provides also a set of methods to test the states of a component. They are:

• assertEnabled(String path)/assertDisabled(String path): they test if a component is enabled or not.
• assertVisible(String path)/assertInvisible(String path): they test component visibility.
• assertRequired(String path): checks if a form component is required.

In the test case from project CustomDatepickerAjax we used assertEnabled/assertDisabled to check if button "update" really disables our datepicker:

//…
@Test
public void testDisableDatePickerWithButton(){
	//start and render the test page
	tester.startPage(HomePage.class);
	//assert that datepicker is enabled
	tester.assertEnabled("form:datepicker");
	//click on update button to disable datepicker
	tester.clickLink("update");
	//assert that datepicker is disabled
	tester.assertDisabled("form:datepicker");		
}
//...

Testing components in isolation

Method startComponent(Component) can be used to test a component in isolation without having to create a container page for this purpose. The target component is rendered and both its methods onInitialize() and onBeforeRender() are executed. In the test case from project CustomFormComponentPanel we used this method to check if our custom form component correctly renders its internal label:

//…
@Test
public void testCustomPanelContainsLabel(){
	TemperatureDegreeField field = new TemperatureDegreeField("field", Model.of(0.00));
	//Use standard JUnit class Assert	
	Assert.assertNull(field.get("mesuramentUnit"));		
	tester.startComponent(field);		
	Assert.assertNotNull(field.get("mesuramentUnit"));
}
//...

If test requires a page we can use startComponentInPage(Component) which automatically generates a page for our component.

Testing the response

WicketTester allows us to access to the last response generated during testing with method getLastResponse. The returned value is an instance of class MockHttpServletResponse that provides helper methods to extract informations from mocked request.

In the test case from project CustomResourceMounting we extract the text contained in the last response with method getDocument and we check if it is equal to the RSS feed used for the test:

//…
@Test
public void testMountedResourceResponse() throws IOException, FeedException{		tester.startResource(new RSSProducerResource());
	String responseTxt = tester.getLastResponse().getDocument();
	//write the RSS feed used in the test into a ByteArrayOutputStream
	ByteArrayOutputStream outputStream = new ByteArrayOutputStream();
	Writer writer = new OutputStreamWriter(outputStream);
	SyndFeedOutput output = new SyndFeedOutput();
	output.output(RSSProducerResource.getFeed(), writer);
	//the response and the RSS must be equal 
	Assert.assertEquals(responseTxt, outputStream.toString());
}
//...

To simulate a request to the custom resource we used method startResource which can be used also with resource references.

Testing URLs

WicketTester can be pointed to an arbitrary URL with method executeUrl(String url). This can be useful to test mounted pages, resources or request mappers:

//…
//the resource was mapped at '/foo/bar'
tester.executeUrl("./foo/bar");	
//...

Testing AJAX components

If our application uses AJAX to refresh components markup, we can test if AjaxRequestTarget contains a given component with WicketTester's method assertComponentOnAjaxResponse:

//…
//test if AjaxRequestTarget contains a component (using its instance)
tester.assertComponentOnAjaxResponse(amountLabel);	
//…
//test if AjaxRequestTarget contains a component (using its path)
tester.assertComponentOnAjaxResponse("pathToLabel:labelId");

It's also possible to use method isComponentOnAjaxResponse(Component cmp) to know if a component has been added to AjaxRequestTarget:

//…
//test if AjaxRequestTarget does NOT contain amountLabel 
assertFalse(tester.isComponentOnAjaxResponse(amountLabel));	
//...

Testing AJAX events

Behavior AjaxEventBehavior and its subclasses can be tested simulating AJAX events with WicketTester's method executeAjaxEvent(Component cmp, String event). Here is the sample code from project TestAjaxEventsExample:

Home page code:

public class HomePage extends WebPage {
 public static String INIT_VALUE = "Initial value";
 public static String OTHER_VALUE = "Other value";
 public HomePage(final PageParameters parameters) {
	super(parameters);
	Label label;
	add(label = new Label("label", INIT_VALUE));				
	label.add(new AjaxEventBehavior("click") {
		@Override
		protected void onEvent(AjaxRequestTarget target) {
			//change label's data object
			getComponent().setDefaultModelObject(OTHER_VALUE);
			target.add(getComponent());
		}
	}).setOutputMarkupId(true);
	//…
 }
}

Test method:

@Test
public void testAjaxBehavior(){
	//start and render the test page
	tester.startPage(HomePage.class);
	//test if label has the initial expected value
	tester.assertLabel("label", HomePage.INIT_VALUE);		
	//simulate an AJAX "click" event
	tester.executeAjaxEvent("label", "click");
	//test if label has changed as expected
	tester.assertLabel("label", HomePage.OTHER_VALUE);
}

Testing AJAX behaviors

To test a generic AJAX behavior we can simulate a request to it using WicketTester's method executeBehavior(AbstractAjaxBehavior behavior):

//…
AjaxFormComponentUpdatingBehavior ajaxBehavior = new AjaxFormComponentUpdatingBehavior 
                                                                           ("change"){
	@Override
	protected void onUpdate(AjaxRequestTarget target) {
		//...				
	}
};
component.add(ajaxBehavior);
//…
//execute AJAX behavior, i.e. onUpdate will be invoked 
tester.executeBehavior(ajaxBehavior));	
//...

Using a custom servlet context

In paragraph 13.9 we have seen how to configure our application to store resource files into a custom folder placed inside webapp root folder (see project CustomFolder4MarkupExample).

In order to write testing code for applications that use this kind of customization, we must tell WicketTester which folder to use as webapp root. This is necessary as under test environment we don't have any web server, hence it's impossible for WicketTester to retrieve this parameter from servlet context.

Webapp root folder can be passed to WicketTester's constructor as further parameter like we did in the test case of project CustomFolder4MarkupExample:

public class TestHomePage{
   private WicketTester tester;
   @Before
   public void setUp(){
      //build the path to webapp root folder   
      File curDirectory = new File(System.getProperty("user.dir"));
      File webContextDir = new File(curDirectory, "src/main/webapp");
      tester = new WicketTester(new WicketApplication(), webContextDir.getAbsolutePath());
   }
   //test methods…
}

After a test method has been executed, we may need to clear any possible side effect occurred to the Application and Session objects. This can be done invoking WicketTester's method destroy():

@After
public void tearDown(){
	//clear any side effect occurred during test.
	tester.destroy();
}

20.2 Testing Wicket forms

Wicket provides utility class FormTester that is expressly designed to test Wicket forms. A new FormTester is returned by WicketTester's method newFormTester(String, boolean) which takes in input the page-relative path of the form we want to test and a boolean flag indicating if its form components must be filled with a blank string:

//…
//create a new form tester without filling its form components with a blank string
FormTester formTester = tester.newFormTester("form", false);
//...

FormTester can simulate form submission with method submit which takes in input as optional parameter the submitting component to use instead of the default one:

//…
//create a new form tester without filling its form components with a blank string
FormTester formTester = tester.newFormTester("form", false);
//submit form with default submitter
formTester.submit();
//…
//submit form using inner component 'button' as alternate button
formTester.submit("button");

If we want to submit a form with an external link component we can use method submitLink (String path, boolean pageRelative) specifying the path to the link.

In the next paragraphs we will see how to use WicketTester and FormTester to interact with a form and with its children components.

Setting form components input

The purpose of a HTML form is to collect user input. FormTester comes with the following set of methods that simulate input insertion into form's fields:

• setValue(String path, String value): inserts the given textual value into the specified component. It can be used with components TextField and TextArea. A version of this method that accepts a component instance instead of its path is also available.
• setValue(String checkboxId, boolean value): sets the value of a given CheckBox component.
• setFile(String formComponentId, File file, String contentType): sets a File object on a FileUploadField component.
• select(String formComponentId, int index): selects an option among a list of possible options owned by a component. It supports components that are subclasses of AbstractChoice along with RadioGroup and CheckGroup.
• selectMultiple(String formComponentId, int indexes): selects all the options corresponding to the given array of indexes. It can be used with multiple-choice components like CheckGroup or ListMultipleChoice.

setValue is used inside method insertUsernamePassword to set the username and password fields of the form used in project StatelessLoginForm:

protected void insertUsernamePassword(String username, String password) {
	//start and render the test page
	tester.startPage(HomePage.class);
	FormTester formTester = tester.newFormTester("form");
	//set credentials
	formTester.setValue("username", username);
	formTester.setValue("password", password);		
	//submit form
	formTester.submit();
}

Testing feedback messages

To check if a page contains one or more expected feedback messages we can use the following methods provided by WicketTester:

• assertFeedback(String path, String… messages): asserts that a given panel contains the specified messages
• assertInfoMessages(String… expectedInfoMessages): asserts that the expected info messages are rendered in the page.
• assertErrorMessages(String… expectedErrorMessages): asserts that the expected error messages are rendered in the page.

assertInfoMessages and assertErrorMessages are used in the test case from project StatelessLoginForm to check that form generates a feedback message in accordance with the login result:

@Test
public void testMessageForSuccessfulLogin(){
	inserUsernamePassword("user", "user");	
	tester.assertInfoMessages("Username and password are correct!");
}
@Test
public void testMessageForFailedLogin (){
	inserUsernamePassword("wrongCredential", "wrongCredential");		
	tester.assertErrorMessages("Wrong username or password");
}

Testing models

Component model can be tested as well. With method assertModelValue we can test if a specific component has the expected data object inside its model.

This method has been used in the test case of project ModelChainingExample to check if the form and the drop-down menu share the same data object:

@Test
public void testFormSelectSameModelObject(){
	PersonListDetails personListDetails = new PersonListDetails();
	DropDownChoice dropDownChoice = (DropDownChoice) personListDetails.get("persons");
	List choices = dropDownChoice.getChoices();
	//select the second option of the drop-down menu
	dropDownChoice.setModelObject(choices.get(1));
	//start and render the test page
	tester.startPage(personListDetails);		
	//assert that form has the same data object used by drop-down menu
	tester.assertModelValue("form", dropDownChoice.getModelObject());
}

20.3 Testing markup with TagTester

If we need to test component markup at a more fine-grained level, we can use class TagTester from package org.apache.wicket.util.tester.

This test class allows to check if the generated markup contains one or more tags having a given attribute with a given value. TagTester can not be directly instantiated but it comes with three factory methods that return one or more TagTester matching the searching criteria. In the following test case (from project TagTesterExample) we retrieve the first tag of the home page (a <span> tag) having attribute class equal to myClass:

HomePage markup:

<html xmlns:wicket="http://wicket.apache.org">
	<head>
		<meta charset="utf-8" />
		<title></title>
	</head>
	<body>
		<span class="myClass"></span>
		<div class="myClass"></div>
	</body>
</html>

Test method:

@Test
public void homePageMarkupTest()
{
	//start and render the test page
	tester.startPage(HomePage.class);
	//retrieve response's markup
	String responseTxt = tester.getLastResponse().getDocument();
	TagTester tagTester = TagTester.createTagByAttribute(responseTxt, "class", "myClass");
	Assert.assertNotNull(tagTester);
	Assert.assertEquals("span", tagTester.getName());
	List<TagTester> tagTesterList = TagTester.createTagsByAttribute(responseTxt, "class", 
                                                                       "myClass", false);
	Assert.assertEquals(2, tagTesterList.size());
}

The name of the tag found by TagTester can be retrieved with its method getName. Method createTagsByAttribute returns all the tags that have the given value on the given attribute. In the code above we have used this method to test that our markup contains two tags having attribute class equal to myClass.

20.4 Summary

With a component-oriented framework we can test our pages and components as we use to do with any other Java entity. Wicket offers a complete support for writing testing code, offering built-in tools to test nearly all the elements that build up our applications (pages, containers, links, behaviors, etc...).

The main entity discussed in this chapter has been class WicketTester which can be used to write unit tests and acceptance tests for our application, but we have also seen how to test forms with FormTester and how to inspect markup with TagTester.

In addition to learning how to use the utility classes provided by Wicket for testing, we have also experienced the innovative approach of Wicket to web testing that allows to test components in isolation without the need of running our tests with a web server and depending only on JUnit as testing framework.

21 Test Driven Development with Wicket and Spring

21.1 Configuration of the runtime environment

In order to get the Wicket framework up to speed when your server is up and running, you usually configure a WicketFilter instance in your web application deployment descriptor file (web.xml) while passing it a single init parameter called applicationClassName that points to your main implementation class extending org.apache.wicket.protocol.http.WebApplication where all of your application-wide settings and initialization requirements are dealt with:

<filter>
    <filter-name>wicketfilter</filter-name>
    <filter-class>org.apache.wicket.protocol.http.WicketFilter</filter-class>
    <init-param>
        <param-name>applicationClassName</param-name>
        <param-value>com.comsysto.webapp.MyWebApplication</param-value>
    </init-param>
</filter>

In case you want to get Wicket application up and running while leaving the application configuration and dependency injection issues to the Spring container, the configuration to be provided within the deployment descriptor looks slightly different though:

<web-app>
    <filter>
        <filter-name>wicketfilter</filter-name>
        <filter-class>org.apache.wicket.protocol.http.WicketFilter</filter-class>
        <init-param>
            <param-name>applicationFactoryClassName</param-name>
            <param-value>org.apache.wicket.spring.SpringWebApplicationFactory</param-value>
        </init-param>
    </filter>
    <listener>
        <listener-class>org.springframework.web.context.ContextLoaderListener</listener-class>
    </listener>
    <context-param>
        <param-name>contextConfigLocation<param-name>
        <param-value>/WEB-INF/applicationContext.xml<param-value>
    </context-param>
</web-app>

The additional configuration part containing listener and context parameter definition is a usual Spring container related configuration detail. ContextLoaderListener is an implementation of standard Servlet API ServletContextListener interface provided by the Spring framework itself and is responsible for looking up an according bean definition file(s) specified by the context param above and creating an ApplicationContext instance during servlet context initialization accordingly. When integrating an ApplicationContext instance with Wicket, one of the beans defined in the above mentioned Spring bean definition file has to be your own specific extension of org.apache.wicket.protocol.http.WebApplication. You can either define a suitable bean in the bean definition file itself:

<beans>
    <bean id="myWebApp" class="com.comsysto.webapp.MyWebApplication"/>
</beans>

or use powerful classpath scanning feature of the Spring framework and annotate the MyWebApplication implementation with the appropriate @Component annotation accordingly while enabling the Spring container to scan the according package(s) of your application for relevant bean definitions:

<beans>
    <context:component-scan base-package="com.comsysto.webapp" />
    <context:component-scan base-package="com.comsysto.webapp.service" />
    <context:component-scan base-package="com.comsysto.webapp.repository" />
</beans>

Either way, if everything goes well, you'll get a pre-configured ApplicationContext all set up during the startup of your web container. One of the beans in the ApplicationContext will be your own extension of Wicket's WebApplication type. SpringWebApplicationFactory implementation provided by the Wicket framework itself that you have defined as the applicationFactoryClassName in the configuration of your WicketFilter will then be used in order to retrieve that very same WebApplication bean out of your Spring ApplicationContext. The Factory expects one and only one extension of Wicket's very own WebApplication type to be found within the ApplicationContext instance at runtime. If no such bean or more than one bean extending WebApplication is found in the given ApplicationContext an according IllegalStateException will be raised and initialization of your web application will fail:

Map<?,?> beans = BeanFactoryUtils.beansOfTypeIncludingAncestors(ac,WebApplication.class, false, false);
if (beans.size() == 0)
{
	throw new IllegalStateException("bean of type [" + WebApplication.class.getName() +
			"] not found");
}
if (beans.size() > 1)
{
	throw new IllegalStateException("more than one bean of type [" +
			WebApplication.class.getName() + "] found, must have only one");
}

After the WebApplication bean has been successfully retrieved from the ApplicationContext via SpringWebApplicationFactory, WicketFilter will then, as part of its own initialization process, trigger both internalInit() and init() methods of the WebApplication bean. The latter one is the exact spot where the last piece of the runtime configuration puzzle between Wicket and Spring is to be placed :

@Component
public class MyWebApplication extends WebApplication {
    @Override
    protected void init() {
        addComponentInstantiationListener(new SpringComponentInjector(this));
    }
}

SpringComponentInjector provided by the Wicket framework enables you to get dependencies from the ApplicationContext directly injected into your Wicket components by simply annotating these with the according @SpringBean annotation.

21.2 Configuration of the JUnit based integration test environment

One of the main features of Apache Wicket framework is the ability to easily write and run plain unit tests for your Pages and all other kinds of Components that even include the verification of the rendering process itself by using JUnit framework and the WicketTester API only. When using Spring framework for application configuration together with Wicket, as we do, you can even use the same tools to easily write and run full blown integration tests for your web application as well. All you have to do is use Spring's TestContext framework additionally to configure and run your JUnit based integration tests. The Spring Framework provides a set of Spring specific annotations that you can use in your integration tests in conjunction with the TestContext framework itself in order to easily configure an according ApplicationContext instance for your tests as well as for appropriate transaction management before, during and after your test execution. Following code snippet represents a simple JUnit 4 based test case using Spring's specific annotations in order to initialize an ApplicationContext instance prior to executing the test itself:

@RunWith(SpringJUnit4ClassRunner.class)
@ContextConfiguration(locations = {"classpath:WEB-INF/applicationContext.xml"})
@TransactionConfiguration(transactionManager = "txManager", defaultRollback = false)
public class LoginPageTest {
    private WicketTester tester;
    @Autowired
    private ApplicationContext ctx;
    @Autowired
    private MyWebApplication myWebApplication;
    @Before
    public void setUp() {
        tester = new WicketTester(myWebApplication);
    }
    @Test
    @Transactional
    @Rollback(true)
    public void testRenderMyPage() {
        tester.startPage(LoginPage.class);
        tester.assertRenderedPage(LoginPage.class);
        tester.assertComponent("login", LoginComponent.class);
    }
}

By defining three annotations on the class level (see code snippet above) in your test, Spring's TestContext framework takes care of preparing and initializing an ApplicationContext instance having all the beans defined in the according Spring context file as well as the transaction management in case your integration test includes some kind of database access. Fields marked with @Autowired annotation will be automatically dependency injected as well so that you can easily access and use these for your testing purposes. Since MyWebApplication, which extends Wicket's WebApplication type and represents the main class of our web application, is also a bean within the ApplicationContext managed by Spring, it will also be provided to us by the test framework itself and can be easily used in order to initialize a WicketTester instance later on during the execution of the test's setUp() method. With this kind of simple, annotation based test configuration we are able to run an integration test that verifies whether a LoginPage gets started and initialized, whether the rendering of the page runs smoothly and whether the page itself contains a LoginComponent that we possibly need in order to process user's login successfully.

When you run this test though, you'll unfortunately get the following exception raised:

java.lang.IllegalStateException: No WebApplicationContext found: no ContextLoaderListener registered?
	at org.springframework.web.context.support.WebApplicationContextUtils.getRequiredWebApplicationContext(WebApplicationContextUtils.java:84)
	at org.apache.wicket.spring.injection.annot.SpringComponentInjector.<init>(SpringComponentInjector.java:72)
	at com.comsysto.serviceplatform.uiwebapp.MyWebApplication.initializeSpringComponentInjector(MyWebApplication.java:59)
	at com.comsysto.serviceplatform.uiwebapp.MyWebApplication.init(MyWebApplication.java:49)
	at org.apache.wicket.protocol.http.WicketFilter.init(WicketFilter.java:719)
	at org.apache.wicket.protocol.http.MockWebApplication.<init>(MockWebApplication.java:168)
	at org.apache.wicket.util.tester.BaseWicketTester.<init>(BaseWicketTester.java:219)
	at org.apache.wicket.util.tester.WicketTester.<init>(WicketTester.java:325)
	at org.apache.wicket.util.tester.WicketTester.<init>(WicketTester.java:308)

As you can see above, the Exception gets raised during the initialization of the WicketTester instance even before the actual test method gets executed. Even though we have applied rather cool and simple annotation based test configuration already described and passed in perfectly well prepared ApplicationContext instance to the WicketTester instance in the constructor, somewhere down the rabbit hole someone complained that no WebApplicationContext instance could have been found which seems to be required in order to initialize the WicketTester properly.

The problem that we run against here is due to the fact that SpringComponentInjector during its own initialization is trying to get hold of an according Spring's ApplicationContext instance that would normally be there in a runtime environment but does not find any since we are running in a test environment currently. SpringComponentInjector delegates to Spring's own WebApplicationContextUtils class to retrieve the instance of ApplicationContext out of the ServletContext which is perfectly fine for a runtime environment but is unfortunately failing in a test environment:

public static WebApplicationContext getRequiredWebApplicationContext(ServletContext sc)
		throws IllegalStateException {
	WebApplicationContext wac = getWebApplicationContext(sc);
	if (wac == null) {
		throw new IllegalStateException("No WebApplicationContext found: no ContextLoaderListener registered?");
	}
	return wac;
}

If you still remember we defined a ContextLoaderListener in our web.xml file as part of the configuration of our runtime environment that makes sure an according WebApplicationContext instance gets initialized and registered against the ServletContext properly. Luckily, this problem can easily be solved if we slightly change the way we initialize SpringComponentInjector in our main MyWebApplication class. Apart from the constructor that we have used so far, there is another constructor in the SpringComponentInjector class that expects the caller to provide it with an according ApplicationContext instance rather than trying to resolve one on its own:

public SpringComponentInjector(WebApplication webapp, ApplicationContext ctx,
		boolean wrapInProxies)
{
	if (webapp == null)
	{
		throw new IllegalArgumentException("Argument [[webapp]] cannot be null");
	}
	if (ctx == null)
	{
		throw new IllegalArgumentException("Argument [[ctx]] cannot be null");
	}
	// store context in application's metadata …
	webapp.setMetaData(CONTEXT_KEY, new ApplicationContextHolder(ctx));
	// … and create and register the annotation aware injector
	InjectorHolder.setInjector(new AnnotSpringInjector(new ContextLocator(), wrapInProxies));
}

In order to use this constructor instead of the one we used previously, we now obviously need to get hold of the ApplicationContext instance on our own in our MyWebApplication implementation. The easiest way to do this is to use Spring's own concept of lifecycle callbacks provided to the beans managed by the Spring container. Since our MyWebApplication is also a bean managed by the Spring container at runtime (enabled by the classpath scanning and @Component annotation on a type level), we can declare it to implement ApplicationContextAware interface which ensures that it gets provided with the ApplicationContext instance that it runs in by the Spring container itself during startup.

public interface ApplicationContextAware {
	void setApplicationContext(ApplicationContext applicationContext) throws BeansException;
}

So the relevant parts of MyWebApplication type will now look something like the following code snippet:

@Component
public class MyWebApplication extends WebApplication implements ApplicationContextAware {
    @Override
    protected void init() {
        addComponentInstantiationListener(new SpringComponentInjector(this, ctx, true));
    }
    public void setApplicationContext(ApplicationContext applicationContext) throws BeansException {
        this.ctx = applicationContext;
    }
}

For additional clarification of how MyWebApplication now relates to both Wicket and Spring framework here is an according class diagram:

21.3 Summary

With the configuration outlined above, no additional modifications are required to the test itself. It's going to turn green now. This way you can use exactly the same Spring context configuration that you'd use in your runtime environment for running your JUnit based integration tests as well.

22 Wicket Best Practices

22.1 Encapsulate components correctly

A component should be self-contained. The user of a component should neither have to know nor care about its internal structure. She should just be familiar with its external interfaces and its documentation in order to be able to use it. This means in detail: Every component that extends Wicket's own Panel type (thus is a Panel itself) must provide its own HTML template. In contrast, when a component extends the classes WebMarkupContainer or Form, there is no HTML template. This implies that you should add components through composition in WebMarkupContainer or Form.

Listing 1:

// Poor component
public class RegistrationForm extends Form<Registration> {
    public RegistrationForm(String id, IModel<Registration> regModel) {
        super(id, new CompoundPropertyModel<Registration>(regModel))
        // Wrong: RegistrationForm provides its own components
        add(new TextField("username"));
        add(new TextField("firstname"));
        add(new TextField("lastname"));
    }
}

This snippet is an example for a poor component. The user of the RegistrationForm must know the internal structure of the markup and component in order to use it.

Listing 2:

public class RegistrationPage extends Page {
    public RegistrationPage(IModel<Registration> regModel) {
        Form<?> form = new RegistrationForm("form");
        form.add(new SubmitButton("register") {
            public void onSubmit() {
                 // do something
            }
        });
        add(form);
    }
}

<html>
<body>
    <form wicket:id="form">
        <!-- These are internal structure information from RegistrationForm -->
        Username <input type="text" wicket:id="username"/>
        First name <input type="text" wicket:id="firstname"/>
        Last name <input type="text" wicket:id="lastname"/>
        <!-- Above new components from page which the user knows  -->
        <input type="submit" wicket:id="register" value="Register"/>
    </form>
</body>
</html>

The code above shows the usage of the poor component in the RegistrationPage. You can see that the input fields firstname, lastname and username get used, even though these components are not added explicitly to the RegistrationPage. Avoid this, because other developers cannot directly see that the components were added in RegistrationPage class.

Listing 3:

// Good component
public class RegistrationInputPanel extends Panel{
    public RegistrationInputPanel(String id, IModel<Registration> regModel) {
        super(id, regModel);
        IModel<Registration> compound = new CompoundPropertyModel<Registration(regmodel)
        Form<Registration> form = new Form<Registration>("form", compound);
        // Correct: Add components to Form over the instance variable
        form.add(new TextField("username"));
        form.add(new TextField("firstname"));
        form.add(new TextField("lastname"));
        add(form);
    }
}

<html>
<body>
    <wicket:panel>
    <form wicket:id="form">
        Username <input type="text" wicket:id="username"/>
        First name <input type="text" wicket:id="firstname"/>
        Last name <input type="text" wicket:id="lastname"/>
    </form>
    </wicket:panel>
</body>
</html>

Now we have a properly encapsulated input component which provides its own markup. Furthermore you can see the correct usage of a Wicket Form. The components get added by calling form.add(Component) on the instance variable. On the other hand, it is allowed to add behaviours and validators over inheritance, because those do not have markup ids which must be bound.

With that, the usage of RegistrationInputPanel is much more intuitive. There is no markup of other embedded components present anymore, just markup of components which get directly added. The RegistrationPage provides its own form that delegates the submit to all Wicket nested forms which are contained in the component tree.

Listing 4:

public class RegistrationPage extends Page {
    public RegistrationPage(IModel<Registration> regModel) {
        Form<?> form = new Form("form");
        form.add(new RegistrationInputPanel("registration", regModel);
        form.add(new SubmitButton("register") {
            public void onSubmit() {
              // do something
            }
        });
        add(form);
    }
}

<html>
<body>
    <form wicket:id="form">
        <div wicket:id="registration">
           Display the RegistrationInputPanel
        </div>
        <input type=&rdquo;submit&rdquo; wicket:id="register" value="Register"/>
    </form>
</body>
</html>

22.2 Put models and page data in fields

In contrast to Struts, Wicket pages and components are no singletons, they are stateful and session-scoped. This enables us to store user-specific information within pages and components. The information should be stored in fields. This way you can access the information within a class while avoiding long method signatures only for passing the same information around. Instances of components can exist for several requests. For example, a page with a form which gets submitted and produces validation errors uses the same page instance. Furthermore the same page instance gets used when the user presses the back button of the browser and resubmits this formular again. Information which gets passed by the constructor should be assigned to fields (normally this must be models). When storing information in fields you should consider that the information is serializable, because the pages are stored in Wicket's page map. By default the page map stores the pages on the hard disk. A non-serializable object leads to NullPointerExceptions and NonSerializableExceptions. Additionally, big data (like binary stuff) should not be stored directly in fields because this can cause performance losses and memory leaks during serialization and deserialization. In this case, you should use the LoadableDetachableModel which can be assigned to a field because this provides an efficient mechanism to load and detach data.

22.3 Correct naming for Wicket IDs

For many developers, naming is a dispensable thing, but I think it is one of the major topics in software development. With the help of correct naming, you can easily identify the business aspects of a software component. Additionally good naming avoids unneccessary and bad comments.

Bad namings for Wicket-IDs are birthdateTextField, firstnameField and addressPanel. Why? The naming contains two aspects: A technical aspect ( "TextField" ) and the business aspect ( "birthdate" ). Only the the business aspect is relevant because both the HTML template as well as the Java code already contain the technical details ("new TextField("birthdate"))". Additionally, such names add a lot of effort when you do technical refactorings, e.g. if you have to replace a TextField by a DatePicker and the Wicket ID birthdateTextField becomes birthdateDatePicker. Another reason for avoiding technical aspects in Wicket IDs is the CompoundPropertyModel. This model delegates the properties to its child components named by Wicket IDs (see listing 3). For example the TextField username automatically calls setUsername() and getUsername() on the Registration object. A setter like setUsernameTextfield() would be very inconvenient here.

22.4 Avoid changes at the component tree

You should consider Wicket's component tree a constant and fixed skeleton which gets revived when its model is filled with data like a robot without brain. Without brain the robot is not able to do anything and is just a dead and fixed skeleton. However, when you fill it with data, it becomes alive and can act. There is no need for changing hardware when filling him with data. In Wicket, you should manipulate the component tree as little as possible. Consequently, you should avoid calling methods like Component.replace(Component) and Component.remove(Component). Calling these methods indicates missing usage or misusage of Wicket's models. Furthermore the component trees should not be constructed using conditions (see listing 5). This reduces the possibility of reusing the same instance significantly.

Listing 5:

// typical for struts
if(MySession.get().isNotLoggedIn()) {
    add(new LoginBoxPanel("login"))
}
else {
    add(new EmptyPanel("login"))
}

Instead of constructing LoginBoxPanel conditionally, it is recommended to always add the panel and control the visibility by overriding isVisible(). So the component LoginBoxPanel is responsible for displaying itself. We move the responsibility into the same component which executes the login. Brilliant! Cleanly encapsulated business logic. There is no decision from outside, the component handles all the logic. You can see another example in "Implement visibilities of components correctly".

22.5 Implement visibilities of components correctly

Visibility of components is an important topic. In Wicket you control any component's visibility via the methods isVisible() and setVisible(). These methods are within Wicket's base class Component and therefore it is applicable for every component and page. Let's have a look at a concrete example of LoginBoxPanel. The panel just gets displayed when the user is not logged in.

Listing 6:

// Poor implementation
LoginBoxPanel loginBox = new LoginBoxPanel("login");
loginBox.setVisible(MySession.get().isNotLoggedIn());
add(loginBox);

Listing 6 shows a poor implementation, because a decision about the visibility is made while instanciating the component. Again, in Wicket instances of components exist for several requests. To reuse the same instance you have to call loginBox.setVisible(false). This is very unhandy, because we always have to call setVisible() and manage the visibility. Furthermore you are going to duplicate the states, because visible is equal to "not logged in". So we have two saved states, one for the business aspect "not logged in" and one for the technical aspect "visible". Both is always equal. This approach is error-prone and fragile, because we always have to pay attention to setting the correct information every time. But this is often forgotten because the logic is widely spread over the code. The solution is the Hollywood principle: "Don't call us, we'll call you.". Take a look at the following diagram illustrating an application flow with some calls. We avoid three calls through the Hollywood-Principle and we just have to instanciate the LoginBoxPanel.

Listing 7:

public class LoginBoxPanel {
    // constructor etc.
    @Override
    public boolean isVisible() {
        return MySession.get().isNotLoggedIn();
    }
};

Now the control over visibility has been inverted, the LoginBoxPanel decides on its visibility autonomously. For each call of isVisible() there is a refreshed interpretion of the login state. Hence, there is no additional state that might be outdated. The logic is centralized in one line code and not spread throughout the application. Furthermore, you can easily identify that the technical aspect isVisible() correlates to the business aspect "logged in". The same rules can be applied to the method isEnabled(). If isEnabled() returns false the components get displayed in gray. Forms which are within an inactive or invisible component do not get executed.

Note that there are cases in which you cannot avoid to call the methods setVisible() and setEnabled(). An example: The user presses a button to display an inlined registration form. In general, you can apply the following rules: data driven components override these methods and delegates to the data model. User triggered events call the method setVisible(boolean). You can also override these methods with inline implementations:

Listing 8:

new Label("headline", headlineModel) {
    @Override
    public boolean isVisible() {
        // Hidden headline if text starts with "Berlusconi"
        String headline = getModelObject();
        return headline.startWith("Berlusconi");
    }
}

Note: Some people insist on overriding isVisible() being a bad thing . The method isVisible() gets called very often (more than once for each request!), so you have to ensure that the calls within isVisible() are cheap. The main point is that the visibility of a component should be controlled by its own and not be controlled by other components. This avoids a wide-spread logic over the whole application. Another way you can realize this is to override onConfigure() and set the visibility there. This method gets called once during each request.

22.6 Always use models

Always use models - period! Do not pass raw objects directly to components. Instances of pages and components can exist for several requests. If you use raw objects, you cannot replace them later. An example is an entity which gets loaded at each request within a LoadableDetachableModel. The entity manager creates a new object reference, but the page would keep the obsolete instance. Always pass IModel in the constructor of your components:

Listing 9:

public class RegistrationInputPanel extends Panel{
    // Correct: The class Registration gets wrapped by IModel
    public RegistrationInputPanel(String id, IModel<Registration> regModel) {
        // add components
    }
}

This code can use any implementation of IModel, e.g. the class Model, a PropertyModel or a custom implementation of LoadableDetachableModel which loads and persists the values automatically. The model implementations gets very easy to replace. You - as a developer - just need to know: if I call IModel.getObject(), I will get an object of type Registration. Where the object comes from is within the responsibility of the model implementation and the calling component. For example you can pass the model while instanciating the component. If you avoid using models, you will almost certainly have to modify the component tree sooner or later which forces you to duplicate states and thus produce unmaintainable code. Additionally, you should use models due to serialization issues. Objects which get stored in fields of pages and components get serialized and deserialized on each request. This can be inefficient in some cases.

22.7 Do not unwrap models within the constructor hierarchy

Avoid unwrapping models within the constructor hierarchy, i.e. do not call IModel.getObject() within any constructor. As already mentioned, a page instance can exist for several page requests, so you might store obsolete and redundant infomation. It is reasonable to unpack Wicket Models at events (user actions), that are methods like onUpdate(), onClick() or onSubmit()@:

Listing 10:

new Form("register") {
    public void onSubmit() {
        // correct, unwrap model in an event call
        Registration reg = registrationModel.getObject()
        userService.register(reg);
    }
}

An additional possibility to unwrap models is via overriding methods like isVisible(), isEnabled() or onBeforeRender().

22.8 Pass models extended components

Always try to pass models on to the parent component. By that, you ensure that at the end of every request the method IModel.detach() gets called. This method is responsible for a data cleanup. Another example: you have implemented your own model which persists the data in the detach() method. So the call of detach() is neccessary for that your data gets persisted. You can see an exemplary passing to the super constructor here:

Listing 11:

public class RegistrationInputPanel extends Panel{
    public RegistrationInputPanel(String id, IModel<Registration> regModel) {
        super(id, regModel)
        // add components
    }
}

22.9 Validators must not change any data or models

Validators should just validate. Consider a bank account form which has a BankFormValidator. This validator checks the bank data over a webservice and corrects the bank name. Nobody would expect that a validator modifies information. Such logic has to be located in Form.onSubmit() or in the event logic of a button.

22.10 Do not pass components to constructors

Do not pass entire components or pages to constructors of other components.

Listing 12:

// Bad solution
public class SettingsPage extends Page {
    public SettingsPage (IModel<Settings> settingsModel, final Webpage backToPage) {
        Form<?> form = new Form("form");
        // add components
        form.add(new SubmitButton("changeSettings") {
            public void onSubmit() {
               // do something
               setResponsePage(backToPage);
            }
        });
        add(form);
    }
}

The SettingsPage expects the page which should be displayed after a successful submit to be passed to its constructor. This solution works, but is very bad practice. You need to know during the instanciation of SettingsPage where you want to redirect the user. This requires a predetermined order of instanciation. It is better to order the instanciation based on business logic (e.g. the order in the HTML template). Furthermore, you need an unnecessary instance of the next success page which might never be displayed. The solution is once again the Hollywood principle. For this you create an abstract method or a hook:

Listing 13:

// Good solution
public class SettingsPage extends Page {
    public SettingsPage (IModel<Settings> settingsModel) {
        Form<?> form = new Form("form");
        // add components
        form.add(new SubmitButton("changeSettings") {
            public void onSubmit() {
               // do something
               onSettingsChanged();
            }
         });
         add(form);
    }
    // hook
    protected void onSettingsChanged() {
    }
// The usage of the new component
Link<Void> settings = new Link<Void>("settings") {
    public void onClick() {
        setResponsePage(new SettingsPage(settingsModel) {
            @Override
            protected void onSettingsChanged() {
               // reference to the current page
               setResponsePage(this);
            }
        });
    }
}
add(settings);

This solution has more code, but it is more flexible and reuseable. We can see there is an event onSettingsChanged() and this event is called after a successful change. Furthermore, there is the possibility to execute additional code besides setting the next page. For example, you can display messages or persist information.

22.11 Use the Wicket session only for global data

The Wicket session is your own extension of Wicket's base session. It is fully typed. There is no map structure to store information unlike the servlet session. You just should use Wicket's session for global data. Authentication is a good example for global data. The login and user information is required on nearly each page. For a blog application it would be good to know whether the user is an author who is allowed to compose blog entries. So you are able to hide or or show links to edit a blog entry. In general you should store the whole authorization logic in Wicket's session, because it is a global thing and you would expect it there. Data of forms and flows which only span certain pages should not stored in the session. This data can be passed from one page to the next via the constructor (see listing 14). As a consequence of this, the models and data have a clearly defined lifecycle that reflects the corresponding the page flow.

Listing 14:

public class MyPage extends WebPage {
    IModel<MyData> myDataModel;
    public MyPage(IModel<MyData> myDataModel) {
        this.myDataModel = myDataModel;
        Link<Void> next = new Link<Void>("next") {
             public void onClick() {
                  // do something
                  setResponsePage(new NextPage(myDataModel));
             }
        }
        add(next);
    }
}

You should pass concrete information to the page. All models can simply be stored in fields because Wicket pages are user-specific instances and no singletons in contrast to Struts. The big advantage of this approach is that the data gets automatically cleaned up when a user completes or exits the page flow. No manual cleanup anymore! This is basically an automatic garbage collector for your session.

22.12 Do not use factories for components

The factory pattern is useful, but nevertheless not suitable for Wicket components.

Listing 15:

public class CmsFactory {
   public Label getCmsLabel(String markupId, final String url) {
       IModel<String> fragment = new AbstractReadOnlyModel<String>() {
          @Override
          public String getObject() {
             return loadSomeContent(url);
          }
       };
       Label result = new Label(markupId, fragment);
       result.setRenderBodyOnly(true);
       result.setEscapeModelStrings(false);
       return result;
   }
   public String loadContent(String url) {
      // load some content
   }
}
// create the component within the page:
public class MyPage extends WebPage {
   @SpringBean
   CmsFactory cmsFactory;
   public MyPage() {
      add(cmsFactory.getCmsLabel("id", "http://url.to.load.from"));
   }
}

This approach for adding a label from the CmsFactory to a page seems to be okay at first glance, but it comes with some disadvantages. There is no possibility to use inheritance anymore. Furthermore, there is no possibility to override isVisible() and isEnabled(). The factory could also be a Spring service which instanciates the component. A better solution is to create a CmsLabel.

Listing 16:

public class CmsLabel extends Label {
   @SpringBean
   CmsResource cmsResource;
   public CmsLabel(String id, IModel<String> urlModel) {
      super(id, urlModel);
      IModel<String> fragment = new AbstractReadOnlyModel<String>(){
         @Override
         public String getObject() {
            return cmsResource.loadSomeContent(urlModel.getObject());
         }
      };
      setRenderBodyOnly(true);
      setEscapeModelStrings(false);
   }
}
// create the component within a page
public class MyPage extends WebPage {
   public MyPage() {
      add(new CmsLabel("id", Model.of("http://url.to.load.from")));
   }
}

The label in listing 16 is clearly encapsulated in a component without using a factory. Now you can easily create inline implementations and override isVisible() or other stuff. Naturally, you might claim "I need a factory to initialize some values in the component, e.g. a Spring service.". For this you can create a implementation of IComponentInstantiationListener. This listener gets called on the super-constructor of every component. The most popular implementation of this interface is the SpringComponentInjector which injects Spring beans in components when the fields are annotated with @SpringBean. You can easliy write and add your own implementation of IComponentInstantiationListener. So there is no reason for using a factory anymore. More information about the instanciation listener is located in Wicket's JavaDoc.

22.13 Every page and component must be tested

Every page and component should have a test. The simplest test just renders the component and validates its technical correctness. For example, a child component should have a matching wicket id in the markup. If the wicket id is not correctly bound - through a typo or if it was just forgotten - the test will fail. An advanced test could test a form, where a backend call gets executed and validated over a mock. So you can validate your component's behaviour. This is a simple way to detect and fix technical and business logic bugs during the build process. Wicket is very suitable for a test driven development approach. For instance, if you run a unit test which fails and shows a message that the wicket id not bound, you will avoid an unneccessary server startup (a server startup takes longer than running a unit test). This reduces the development turnaround. A disadvantage is the difficult testing possibility of AJAX components. However, the testing features of Wicket are much more sophisticated than in other web frameworks.

22.14 Avoid interactions with other servlet filters

Try to get within the Wicket world whenever possible. Avoid the usage of other servlet filters. For this you can use the RequestCycle and override the methods onBeginRequest() and onEndRequest(). You can apply the same to the HttpSession. The equivalent in Wicket is the WebSession. Just extend the WebSession and override the newSession()-method from the Application class. There are very few reasons to access the servlet interfaces. An example could be to read an external cookie to authenticate a user. Those parts should be properly encapsulated and avoided when possible. For this example, you could do the handling within the Wicket session because this is an authentication.

22.15 Cut small classes and methods

Avoid monolithic classes. Often I have seen that developers put the whole stuff into constructors. These classes are getting very unclear and chaotic because you use inline implementations over serveral levels. It is recommended to group logical units and extract methods with a correct business naming. This enhances the clarity and the understandability of the business aspect. When a developer navigates to a component, he is not interested in the technical aspect at first, however he just need the business aspect. To retrieve technical information of a component you can navigate to the method implementation. In case of doubt you should consider to extract seperate components. Smaller components increase the chances of reuse and make testing easier. Listing 17 shows an example of a possible structuring.

Listing 17:

public class BlogEditPage extends WebPage {
    private IModel<Blog> blogModel;
    public BlogEditPage(IModel<Blog> blogModel) {
        super(new PageParameters());
        this.blogModel = blogModel;
        add(createBlogEditForm());
    }
    private Form<Blog> createBlogEditForm() {
        Form<Blog> form = newBlogEditForm();
        form.add(createHeadlineField());
        form.add(createContentField());
        form.add(createTagField());
        form.add(createViewRightPanel());
        form.add(createCommentRightPanel());
        form.setOutputMarkupId(true);
        return form;
    }
    // more methods here
}

22.16 The argument "Bad documentation"

It is a widespread opinion that Wicket has a bad documentation. This argument is just partly correct. There are a lot of code samples and snippets which can be used as code templates. Furthermore, there is a big community that answers complex questions very quickly. In Wicket it is very hard to document everything, because nearly everything is extensible and replaceable. If a component is not completely suitable, you will extend or replace it. Working with Wicket means permanently navigating through code. For example, just consider validators. How can I find all navigators that exist? Open the interface IValidator (Eclipse: Ctrl + Shift + T) and then open the type hierachy (Crtl + T). Now we can see all the validators existing in Wicket and our project.

22.17 Summary

The best practices presented in this chapter should help you to write better and more maintainable code in Wicket. All described methodologies were already proven in a few Wicket projects. If you follow these advices, your Wicket projects will get future-proof and hopefully successful.

23 Working with Maven (Appendix)

23.1 Switching Wicket to DEPLOYMENT mode

As pointed out in the note at page 9, Wicket can be started in two modes, DEVELOPMENT and DEPLOYMENT. When we are in DEVELOPMENT mode Wicket warns us at application startup with the following message:

********************************************************************
*** WARNING: Wicket is running in DEVELOPMENT mode.              ***
***                               ^^^^^^^^^^^                    ***
*** Do NOT deploy to your live server(s) without changing this.  ***
*** See Application#getConfigurationType() for more information. ***
********************************************************************

As we can read Wicket itself discourages us from using DEVELOPMENT mode into production environment. The running mode of our application can be configured in three different ways. The first one is adding a filter parameter inside deployment descriptor web.xml:

<filter>      
	<filter-name>wicket.MyApp</filter-name>
	<filter-class>org.apache.wicket.protocol.http.WicketFilter</filter-class>
	<init-param>
		<param-name>applicationClassName</param-name>
		<param-value>org.wicketTutorial.WicketApplication</param-value>
	</init-param>
	<init-param>
            <param-name>configuration</param-name>
            <param-value>deployment</param-value>
	</init-param>
</filter>

The additional parameter is written in bold. The same parameter can be also expressed as context parameter:

<context-param>
    <param-name>configuration</param-name>
    <param-value>deployment</param-value>
</context-param>

The third way to set the running mode is using system property wicket.configuration. This parameter can be specified in the command line that starts up the server:

java -Dwicket.configuration=deployment ...

Remember that system properties overwrite other settings, so they are ideal to ensure that on production machine the running mode will be always set to DEPLOYMENT.

23.2 Creating a Wicket project from scratch and importing it into our favourite IDE

In order to follow the instructions of this paragraph you must have Maven installed on your system. The installation of Maven is out of the scope of this guide but you can easily find an extensive documentation about it on Internet. Another requirement is a good Internet connection (a flat ADSL is enough) because Maven needs to connect to its central repository to download the required dependencies.

From Maven to our IDE

Wicket project and its dependencies are managed using Maven1. This tool is very useful also when we want to create a new project based on Wicket from scratch. With a couple of shell commands we can generate a new project properly configured and ready to be imported into our favourite IDE. The main step to create such a project is to run the command which generates project's structure and its artifacts. If we are not familiar with Maven or we simply don't want to type this command by hand, we can use the utility form on Wicket site at http://wicket.apache.org/start/quickstart.html :

Here we have to specify the root package of our project (GroupId), the project name (ArtifactId) and which version of Wicket we want to use (Version). Once we have run the resulting command in the OS shell, we will have a new folder with the same name of the project (i.e the ArtifactId). Inside this folder we can find a file called pom.xml. This is the main file used by Maven to manage our project. For example, using “org.wicketTutorial” as GroupId and “MyProject” as ArtifactId, we would obtain the following artifacts:

.MyProject
        |   pom.xml
        |
        ---src
            +---main
            |   +---java
            |   |   ---org
            |   |       ---wicketTutorial
            |   |               HomePage.html
            |   |               HomePage.java
            |   |               WicketApplication.java
            |   |
            |   +---resources
            |   |       log4j.properties
            |   |
            |   ---webapp
            |       ---WEB-INF
            |               web.xml
            |
            ---test
                ---java
                    ---org
                        ---wicketTutorial
                                TestHomePage.java

Amongst other things, file pom.xml contains a section delimited by tag <dependencies> which declares the dependencies of our project. By default the Maven archetype will add the following Wicket modules as dependencies:

…
<dependencies>
	<!--  WICKET DEPENDENCIES -->
	<dependency>
		<groupId>org.apache.wicket</groupId>
		<artifactId>wicket-core</artifactId>
		<version>${wicket.version}</version>
	</dependency>
	<dependency>
		<groupId>org.apache.wicket</groupId>
		<artifactId>wicket-ioc</artifactId>
		<version>${wicket.version}</version>
	</dependency>
	<!-- OPTIONAL DEPENDENCY
	<dependency>
		<groupId>org.apache.wicket</groupId>
		<artifactId>wicket-extensions</artifactId>
		<version>${wicket.version}</version>
	</dependency>
	--> 
	…
</dependencies>
...

If we need to use more Wicket modules or additional libraries, we can add the appropriate XML fragments here.

Importing a Maven project into our IDE

Maven projects can be easily imported into the most popular Java IDEs. However, the procedure needed to do this differs from IDE to IDE. In this paragraph we can find the instructions to import Maven projects into three of the most popular IDEs among Java developers : NetBeans, JetBrains IDEA and Eclipse.

NetBeans Starting from version 6.7, NetBeans includes Maven support, hence we can start it and directly open the folder containing our project:

Intellj IDEA Intellj IDEA comes with a Maven importing functionality that can be started under “File/New Project/Import from external model/Maven”. Then, we just have to select the pom.xml file of our project:

Eclipse If our IDE is Eclipse the import procedure is a little more complex. Before opening the new project we must generate the Eclipse project artifacts running the following command from project root:

mvn eclipse:eclipse

Now to import our project into Eclipse we must create a classpath variable called M2_REPO that must point to your local Maven repository. This can be done selecting “Window/Preferences” and searching for “Classpath Variables”. The folder containing our local Maven repository is usually under our user folder and is called .m2 (for example under Unix system is /home/<myUserName>/.m2/repository):

Once we have created the classpath variable we can go to “File/Import.../Existing Project into Workspace”, select the directory of the project and press “Finish”:

Once the project has been imported into Eclipse, we are free to use our favourite plug-ins to run it or debug it (like for example "run-jetty-run": http://code.google.com/p/run-jetty-run/ ).

Please note the option “Copy projects into workspace” in the previous illustration. If we select it, the original project generated with Maven won't be affected by the changes made inside Eclipse because we will work on a copy of it under the current workspace.

If we modify the pom.xml file (for example adding further dependencies) we must regenerate project's artifacts and refresh the project (F5 key) to reflect changes into Eclipse.

Speeding up development with plugins.

Now that we have our project loaded into our IDE we could start coding our components directly by hand. However it would be a shame to not leverage the free and good Wicket plugins available for our IDE. The following is a brief overview of the most widely used plugins for each of the three main IDEs considered so far.

NetBeans NetBeans offers Wicket support thought 'NetBeans Plugin for Wicket' hosted at http://java.net/projects/nbwicketsupport/ . This plugin is released under CDDL-1.0 license. You can find a nice introduction guide to this plugin at http://netbeans.org/kb/docs/web/quickstart-webapps-wicket.html .

Intellj IDEA For JetBrain IDEA we can use WicketForge plugin, hosted at Google Code http://code.google.com/p/wicketforge/ . The plugin is released under ASF 2.0 license.

Eclipse With Eclipse we can install one of the plugins that supports Wicket. As of the writing of this document, the most popular is probably Qwickie, available in the Eclipse Marketplace and hosted on Google Code at http://code.google.com/p/qwickie/ . QWickie is released under ASF 2.0 license.

24 Project WicketStuff (Appendix)

24.1 What is project WicketStuff

WicketStuff is an umbrella project that gathers different Wicket-related projects developed and maintained by the community. The project is hosted on GitHub at https://github.com/wicketstuff/core Every module is structured as a parent Maven project containing the actual project that implements the new functionality and an example project that illustrates how to use it in our code. The resulting directory structure of each module is the following:

<module name>-parent
        |
        +---<module name>
        ---<module name>-examples

So far we have introduced only modules Kryo Serializer and JavaEE Inject, but WicketStuff comes with many other modules that can be used in our applications. Some of them come in handy to improve the user experience of our pages with complex components or integrating some popular web services (like Google Maps ) and JavaScript libraries (like TinyMCE ).

This appendix provides a quick overview of what WicketStuff offers to enhance the usability and the visually-appealing of our pages.

Every WicketStuff module can be downloaded as JAR archive at http://mvnrepository.com . This site provides also the XML fragment needed to include it as a dependency into our pom.xml file.

24.2 Module tinymce

Module tinymce offers integration with the namesake JavaScript library that turns our “humble” text-areas into a full-featured HTML WYSIWYG editor:

To “tinyfy” a textarea component we must use behavior TinyMceBehavior:

TextArea textArea = new TextArea("textArea", new Model(""));
textArea.add(new TinyMceBehavior());

By default TinyMceBehavior adds only a basic set of functionalities to our textarea:

To add more functionalities we must use class TinyMCESettings to register additional TinyMCE plugins and to customize the toolbars buttons. The following code is an excerpt from example page FullFeaturedTinyMCEPage:

TinyMCESettings settings = new TinyMCESettings(TinyMCESettings.Theme.advanced);
//…
// first toolbar
//…
settings.add(Button.newdocument, TinyMCESettings.Toolbar.first,
		                             TinyMCESettings.Position.before);
settings.add(Button.separator, TinyMCESettings.Toolbar.first,
			                      TinyMCESettings.Position.before);
settings.add(Button.fontselect, TinyMCESettings.Toolbar.first,
			                      TinyMCESettings.Position.after);
//…
// other settings
settings.setToolbarAlign(TinyMCESettings.Align.left);
settings.setToolbarLocation(TinyMCESettings.Location.top);
settings.setStatusbarLocation(TinyMCESettings.Location.bottom);
settings.setResizing(true);
//…
TextArea textArea = new TextArea("ta", new Model(TEXT));
textArea.add(new TinyMceBehavior(settings));

For more configuration examples see pages inside package wicket.contrib.examples.tinymce in the example project of the module.

24.3 Module wicketstuff-gmap3

Module wicketstuff-gmap3 integrates Google Maps service with Wicket providing component org.wicketstuff.gmap.GMap. If we want to embed Google Maps into one of our pages we just need to add component GMap inside the page. The following snippet is taken from example page SimplePage:

HTML:

…
<body>
  <div wicket:id="map">Map</div>
</body>
...

Java code:

public class SimplePage extends WicketExamplePage
{
    public SimplePage()
    {
        GMap map = new GMap("map");
        map.setStreetViewControlEnabled(false);
        map.setScaleControlEnabled(true);
        map.setScrollWheelZoomEnabled(true);
        map.setCenter(new GLatLng(52.47649, 13.228573));        
        add(map);
    }
}

The component defines a number of setters to customize its behavior and appearance. More info can be found on wiki page https://github.com/wicketstuff/core/wiki/Gmap3 .

24.4 Module wicketstuff-googlecharts

To integrate the Google Chart tool into our pages we can use module wicketstuff-googlecharts. To display a chart we must combine the following entities: component Chart, interface IChartData and class ChartProvider, all inside package org.wicketstuff.googlecharts. The following snippet is taken from example page Home:

HTML:

…
  <h2>Hello World</h2>
  <img wicket:id="helloWorld"/>
...

Java code:

IChartData data = new AbstractChartData(){
    public double[][] getData(){
       return new double[][] { { 34, 22 } };
    }
};
ChartProvider provider = new ChartProvider(new Dimension(250, 100), ChartType.PIE_3D, data);
provider.setPieLabels(new String[] { "Hello", "World" });
add(new Chart("helloWorld", provider));

Displayed chart:

As we can see in the snippet above, component Chart must be used with <img> tag while the input data returned by IChartData must be a two-dimensional array of double values.

24.5 Module wicketstuff-inmethod-grid

Module wicketstuff-inmethod-grid implements a sophisticated grid-component with class com. inmethod.grid.datagrid.DataGrid.

Just like pageable repeaters (seen in paragraph 11.4) DataGrid provides data pagination and uses interface IDataProvider as data source. In addition the component is completely ajaxified:

DataGrid supports also editable cells and row selection:

The following snippet illustrate how to use DataGrid and is taken from wiki page https://github.com/wicketstuff/core/wiki/InMethodGrid :

HTML:

…
  <div wicket:id="grid">Grid</div>
...

Java code:

final List<Person> personList = //load a list of Persons
final ListDataProvider listDataProvider = new ListDataProvider(personList);
//define grid's columns
List<IGridColumn> cols = (List) Arrays.asList(
	     new PropertyColumn(new Model("First Name"), "firstName"),
	     new PropertyColumn(new Model("Last Name"), "lastName"));
DataGrid grid = new DefaultDataGrid("grid", new DataProviderAdapter(listDataProvider), cols);
add(grid);

In the code above we have used convenience class DefaultDataGrid that is a subclass of DataGrid and it already comes with a navigation toolbar.

The example pages are under package com.inmethod.grid.examples.pages in the example project which is hosted at http://www.wicket-library.com/inmethod-grid/data-grid/simple .

25 Lost In Redirection With Apache Wicket (Appendix)

In case of invalid data being submitted with a form instance for example, it seemed like according error messages wouldn’t be presented when the same form page gets displayed again. Sometimes! And sometimes they would! One of those nightmares of rather deterministic programmer’s life. This so called Lost In Redirection problem, even if it looks like a wicket bug at first, is rather a result of a default setting in wicket regarding the processing of form submissions in general. In order to prevent another wide known problem of double form submissions, Wicket uses a so called REDIRECT_TO_BUFFER strategy for dealing with rendering a page after web form’s processing (@see IRequestCycleSettings#RenderStrategy).

What does the default RenderStrategy actually do?

Both logical parts of a single HTTP request, an action and a render part get processed within the same request, but instead of streaming the render result to the browser directly, the result is cached on the server first.

Wicket will create an according BufferedHttpServletResponse instance that will be used to cache the resulting HttpServletResponse within the WebApplication.

After the buffered response is cached the HTTP status code of 302 get’s provided back to the browser resulting in an additional GET request to the redirect URL (which Wicket sets to the URL of the Form itself). There is a special handling code for this case in the WicketFilter instance that then looks up a Map of buffered responses within the WebApplication accordingly. If an appropriate already cached response for the current request is found, it get’s streamed back to the browser immediately. No additional form processing happens now. The following is a code snippet taken from WicketFilter:

// Are we using REDIRECT_TO_BUFFER?
if (webApplication.getRequestCycleSettings().getRenderStrategy() == IRequestCycleSettings.REDIRECT_TO_BUFFER)
{
    // Try to see if there is a redirect stored
    // try get an existing session
    ISessionStore sessionStore = webApplication.getSessionStore();
    String sessionId = sessionStore.getSessionId(request, false);
    if (sessionId != null)
    {
        BufferedHttpServletResponse bufferedResponse = null;
        String queryString = servletRequest.getQueryString();
        // look for buffered response
        if (!Strings.isEmpty(queryString))
        {
            bufferedResponse = webApplication.popBufferedResponse(sessionId,
                queryString);
        }
        else
        {
            bufferedResponse = webApplication.popBufferedResponse(sessionId,
                relativePath);
        }
        // if a buffered response was found
        if (bufferedResponse != null)
        {
            bufferedResponse.writeTo(servletResponse);
            // redirect responses are ignored for the request
            // logger…
            return true;
        }
    }
}

So what happens in case you have 2 server running your application with session replication and load balancing turned on while using the default RenderStrategy described above?

Since a Map of buffered responses is cached within a WebApplication instance that does not get replicated between the nodes obviously, a redirect request that is suppose to pick up the previously cached response (having possibly form violation messages inside) potentially get’s directed to the second node in your cluster by the load balancer. The second node does not have any responses already prepared and cached for your user. The node therefore handles the request as a completely new request for the same form page and displays a fresh new form page instance to the user accordingly.