CS411 JSF Notes1

Introducing JavaServer Faces

Over the last few years, Java has established itself as the leading technology for web application development. Developers are using technologies like servlets and JSP to develop scalable and robust browser-based user interfaces for countless applications with great success. But as web applications become more complex, some developers are longing for the good ol' days of traditional graphical user interface (GUI) frameworks with rich, powerful user interface widgets and event-driven development models. Servlets and JSP have served us well, but HTTP's stateless nature and simple, coarse-grained request/response model forces application developers using these technologies to struggle with details that are handled behind the scenes by GUI frameworks like AWT/Swing, the standard GUI framework for Java.

1.1. What Is JavaServer Faces?

JavaServer Faces (JSF) simplifies development of sophisticated web application user interfaces, primarily by defining a user interface component model tied to a well-defined request processing lifecycle. This allows:

  • A device-independent GUI control framework.
  • JSF can be used to generate graphics in formats other than HTML,using protocols other than HTTP.
  • A better Struts?
  • Like Apache Struts, JSF can be viewed as an MVC framework for building HTML forms, validating their values, invoking business logic, and displaying results.

Creating and Rendering Components

JSF is a very powerful framework with lots of different features and options you can tweak to satisfy your needs. Trying to grasp everything at once would be overwhelming, so we'll take it one step at a time, looking at each feature in the context of the sample expense report application. This chapter focuses on how JSF components are created and rendered, using the sample application filtering criteria and menu areas as concrete examples.

The Basics

At the core, JSF is a Java API built on top of the Servlet API. In addition, it defines JSP custom tag libraries that hide the API layer to make it easy for Page Authors to include JSF components in a JSP page.The figure below shows both these layers and how JSF applications can be developed on top of either one or both.

JSF layers

This picture shows where JSF fits in with other Web-tier technologies. First of all, just like all the other Web-tier technologies, JSF is built over Servlet. In fact, most of JSF APIs are built over Servlet directly. JSF also leverages JSP as well. And JSF custom tags which we will talk about in detail later on are based on JSP custom tag technology.

JSF Architecture - JavaServer Faces' Implementation of MVC

One of the key advantages of JSF is that it is both a Java Web user-interface standard as well as a framework that firmly follows the Model-View-Controller(MVC) design pattern. This makes JSF applications much more manageable because the user-interface code (View) is cleanly separated from the application data and logic (Model). To prepare the JSF context, which provides application data access to the pages, and to guard against unauthorized or improper access of the pages, all user interactions with the application are handled by a front-end "Faces" servlet (Controller).

Figure : JavaServer Faces Implementation of MVC

The JSF Lifecycle

The life cycle of a JavaServer Faces page is somewhat similar to that of a JSP page: The client makes an HTTP request for the page, and the server responds with the page translated to HTML. However, the JavaServer Faces life cycle differs from the JSP life cycle in that it is split up into multiple phases in order to support the sophisticated UI component model. This model requires that component data be converted and validated, component events be handled, and component data be propagated to beans in an orderly fashion.

A JavaServer Faces page is also different from a JSP page in that it is represented by a tree of UI components, called a view. During the life cycle, the JavaServer Faces implementation must build the view while considering state saved from a previous submission of the page. When the client submits a page, the JavaServer Faces implementation performs several tasks, such as validating the data input of components in the view and converting input data to types specified on the server side.

The JavaServer Faces implementation performs all these tasks as a series of steps in the JavaServer Faces request-response life cycle. Figure below illustrates these steps.

The life cycle handles both kinds of requests: initial requests and postbacks. When a user makes an initial request for a page, he or she is requesting the page for the first time. When a user executes a postback, he or she submits the form contained on a page that was previously loaded into the browser as a result of executing an initial request. When the life cycle handles an initial request, it only executes the restore view and render response phases because there is no user input or actions to process. Conversely, when the life cycle handles a postback, it executes all of the phases.

Usually, the first request for a JavaServer Faces pages comes in as a result of clicking a hyperlink on an HTML page that links to the JavaServer Faces page. To render a response that is another JavaServer Faces page, the application creates a new view and stores it in the FacesContext instance, which represents all of the contextual information associated with processing an incoming request and creating a response. The application then acquires object references needed by the view and calls FacesContext.renderResponse, which forces immediate rendering of the view by skipping to the Render Response Phase of the life cycle, as is shown by the arrows labelled Render Response in the diagram.

Sometimes, an application might need to redirect to a different web application resource, such as a web service, or generate a response that does not contain JavaServer Faces components. In these situations, the developer must skip the rendering phase (Render Response Phase) by calling FacesContext.responseComplete. This situation is also shown in the diagram, this time with the arrows labelled Response Complete.

The most common situation is that a JavaServer Faces component submits a request for another JavaServer Faces page. In this case, the JavaServer Faces implementation handles the request and automatically goes through the phases in the life cycle to perform any necessary conversions, validations, and model updates, and to generate the response.

The details of the life cycle explained in this section are primarily intended for developers who need to know information such as when validations, conversions, and events are usually handled and what they can do to change how and when they are handled. Page authors don't necessarily need to know the details of the life cycle.

Restore View Phase

When a request for a JavaServer Faces page is made, such as when a link or a button is clicked, the JavaServer Faces implementation begins the restore view phase.

During this phase, the JavaServer Faces implementation builds the view of the page, wires event handlers and validators to components in the view, and saves the view in the FacesContext instance, which contains all the information needed to process a single request. All the application's component tags, event handlers, converters, and validators have access to the FacesContext instance.

If the request for the page is an initial request, the JavaServer Faces implementation creates an empty view during this phase and the life cycle advances to the render response phase. The empty view will be populated when the page is processed during a postback.

If the request for the page is a postback, a view corresponding to this page already exists. During this phase, the JavaServer Faces implementation restores the view by using the state information saved on the client or the server.

Apply Request Values Phase

After the component tree is restored, each component in the tree extracts its new value from the request parameters by using its decode method. The value is then stored locally on the component. If the conversion of the value fails, an error message associated with the component is generated and queued on FacesContext. This message will be displayed during the render response phase, along with any validation errors resulting from the process validations phase.

If any decode methods or event listeners called renderResponse on the current FacesContext instance, the JavaServer Faces implementation skips to the render response phase.

If events have been queued during this phase, the JavaServer Faces implementation broadcasts the events to interested listeners.

If some components on the page have their immediate attributes set to true, then the validation, conversion, and events associated with these components will be processed during this phase.

At this point, if the application needs to redirect to a different web application resource or generate a response that does not contain any JavaServer Faces components, it can call FacesContext.responseComplete.

At the end of this phase, the components are set to their new values, and messages and events have been queued.

Process Validations Phase

During this phase, the JavaServer Faces implementation processes all validators registered on the components in the tree. It examines the component attributes that specify the rules for the validation and compares these rules to the local value stored for the component.

If the local value is invalid, the JavaServer Faces implementation adds an error message to the FacesContext instance, and the life cycle advances directly to the render response phase so that the page is rendered again with the error messages displayed. If there were conversion errors from the apply request values phase, the messages for these errors are also displayed.

If any validate methods or event listeners called renderResponse on the current FacesContext, the JavaServer Faces implementation skips to the render response phase.

At this point, if the application needs to redirect to a different web application resource or generate a response that does not contain any JavaServer Faces components, it can call FacesContext.responseComplete.

If events have been queued during this phase, the JavaServer Faces implementation broadcasts them to interested listeners.

Update Model Values Phase

After the JavaServer Faces implementation determines that the data is valid, it can walk the component tree and set the corresponding server-side object properties to the components' local values. The JavaServer Faces implementation will update only the bean properties pointed at by an input component's value attribute. If the local data cannot be converted to the types specified by the bean properties, the life cycle advances directly to the render response phase so that the page is rerendered with errors displayed. This is similar to what happens with validation errors.

If any updateModels methods or any listeners called renderResponse on the current FacesContext instance, the JavaServer Faces implementation skips to the render response phase.

At this point, if the application needs to redirect to a different web application resource or generate a response that does not contain any JavaServer Faces components, it can call FacesContext.responseComplete.

If events have been queued during this phase, the JavaServer Faces implementation broadcasts them to interested listeners.

Invoke Application Phase

During this phase, the JavaServer Faces implementation handles any application-level events, such as submitting a form or linking to another page.

At this point, if the application needs to redirect to a different web application resource or generate a response that does not contain any JavaServer Faces components, it can call FacesContext.responseComplete.

If the view being processed was reconstructed from state information from a previous request and if a component has fired an event, these events are broadcast to interested listeners.

Render Response Phase

During this phase, the JavaServer Faces implementation delegates authority for rendering the page to the JSP container if the application is using JSP pages. If this is an initial request, the components represented on the page will be added to the component tree as the JSP container executes the page. If this is not an initial request, the components are already added to the tree so they needn't be added again. In either case, the components will render themselves as the JSP container traverses the tags in the page.

If the request is a postback and errors were encountered during the apply request values phase, process validations phase, or update model values phase, the original page is rendered during this phase. If the pages contain message or messages tags, any queued error messages are displayed on the page.

Advantage of using jsf (vs. MVC with RequestDispatcher)

  • Custom GUI controls
  • JSF provides a set of APIs and associated custom tags to create HTML forms that have complex interfaces.
  • Event Handling
  • JSF makes it easy to designate Java code that is invoked when forms are submitted. The code can respond to particular buttons, changes in particular values, certain user selections, and so on.
  • Managed Beans
  • In JSP, you can use property="*" with jsp:setProperty to automatically populate a bean based on request parameters. JSF extends this capability and adds in several utilities, all of which serve to greatly simplify request parameter processing.
  • Expression Language
  • JSF provides a concise and powerful language for accessing bean properties and collection elements.
  • Format field conversion and validation
  • JSF has builtin capabilities for checking that form values are in the required format and for converting from strings to various other data types. If values are missing or in an improper format, the form can be automatically redisplayed with error messages and with the previously entered values maintained.
  • Centralized file-based configuration
  • Rather then hard-coding information into Java programs, many JSF values are represented in XML or property files. This loose coupling means that many changes can be made without modifying or recompiling Java code, and that wholesale changes can be made by editing a single file. This approach also lets Java and Web developers focus on their specific tasks without needing to know about the overall system layout.
  • Consistent approach
  • JSF encourages consistent use of MVC throughout your application.

Disadvantages of using JSF (vs. MVC with RequestDispatcher)

  • Bigger learning curve
  • To use MVC with the standard RequestDispatcher, you need to be comfortable with the standard JSP and servlet APIs. To use MVC with JSF, you have to be comfortable with the standard JSP and servlet APIs and a large and elaborate framework that is almost equal in size to the core system. This drawback is especially significant with smaller projects, near-term deadlines, and less experienced developers; you could spend as much time learning JSF as building your actual system.
  • Worse documentation
  • Compared to the standard servlet and JSP APIs, JSF has fewer online resources, and many first-time users find the online JSF documentation confusing and poorly organized.
  • Less transparent
  • With JSF applications, there is a lot more going on behind the scenes than with normal Java-based Web applications. As a result, JSF applications are:
  • Harder to understand
  • Harder to benchmark and optimize
  • Undeveloped tool support
  • There are many IDEs with strong support for standard servlet and JSP technology. Support for JSF is only beginning to emerge, and final level is yet to be determined.
  • Rigid approach
  • The flip side of the benefit that JSF encourages a consistent approach to MVC is that JSF makes it difficult use other approaches.

Advantages of JSF (vs. Struts)

  • Custom components
  • JSF makes it relatively easy to combine complex GUIs into a single manageable component; Struts does not.
  • Support for other display technologies
  • JSF is not limited to HTML and HTTP; Struts is
  • Access to beans by name
  • JSF lets you assign names to beans, then you refer to them by name in the forms. Struts have a complex process with several levels of indirection where you have to remember which form is the input for which action.
  • Expression language
  • The JSF expression language is more concise and powerful than the Struts bean:write tag.
  • This is less advantageous if using JSP 2.0 anyhow.
  • Simpler controller and bean definitions
  • JSF does not require your controller and bean classes to extend any particular parent class (e.g., Action) or use any particular method (e.g., execute). Struts does.
  • Simpler config file and overall structure
  • The faces-config.xml file is much easier to use than is the struts-config.xml file. In general, JSF is simpler.
  • More powerful potential tool support
  • The orientation around GUI controls and their handlers opens possibility of simple to use, drag-and-drop IDEs

Disadvantage of using JSF (vs. Struts)