Patterns: Implementing Self-Service in an SOA Environment

Chapter 7. Application and system design guidelines 149

Figure 7-4 Result bean and view bean design pattern

Advantages of result beans

Some reasons to use result beans are:

򐂰 Result beans clearly define what the controller expects back from the

business logic.

򐂰 Once a result bean is defined, the developers of the Controller and the Model

can develop their components independently. This simplifies and optimizes

the development process and allows for parallel development.

򐂰 Since result beans can be serialized, they can be sent to remote servers or

received from remote servers, such as EJB-based distributed applications. In

addition, they can be stored in a file for persistence purposes.

򐂰 The result bean data structure can be reused by multiple business logic

objects and interaction controller objects.

7.2.3 View bean design pattern

A view bean defines the contract between the Controller and the View. It lists all

the attributes the JSP can display. The main benefit of defining such a view bean

HTML

JavaScript

Browser Client

Web Application Server

Command Bean

Model

Page

Construction

View

Also holds the

result data

Business Logic

View

Bean

Controller

Interaction

Controller

Result

Bean

150 Patterns: Implementing Self-Service in an SOA Environment

is to make it easy for the JSP page designer to get all the required data in one

place. The display page often contains the data from the following sources:

򐂰 Result bean properties (returned by the business logic)

򐂰 HTTP request data (including attributes, parameters, cookies, URL string)

򐂰 Session state

򐂰 Servlet context

Figure 7-4 on page 149 demonstrates the relationship between MVC

components, view beans, and result beans. See 7.2.2, “Result bean design

pattern” on page 148, for our discussion on result beans.

The Controller is responsible for instantiating the view bean and initializing all of

the properties of the bean. View beans can be designed to be responsible for

view-specific transformations. For example, a view bean can be responsible for

converting the monetary values into the user-preferred currency. Such a view

bean can have two properties: The monetary value in a base currency and the

currency display type. The Controller can initialize both of these properties. The

view bean can use this information to call a reusable currency conversion library

and get the monetary value in the appropriate format.

Usually, view beans are tightly coupled with a JSP, because the primary purpose

is to provide all the properties the JSP designer would need in one place.

However, under special circumstances one can reuse the view beans by

inheritance.

Both result beans and view beans are implemented using simple Java beans. In

simple Web interactions, both a result bean and a view bean could be

implemented by the same Java bean.

Advantages and disadvantages of view beans

Some advantages of view beans are:

򐂰 Clearly defines all the fields a View (JSP) can display.

򐂰 Once a view bean is clearly defined, the developers of the Controller and the

View can develop their components independently. This simplifies and

optimizes the development process and allows for parallel development.

򐂰 The View (JSP) designer can get all the dynamic data from one view bean

and use <jsp:useBean> and <jsp:getProperty> tags to insert these values.

This allows the JSP designer to concentrate on the look and feel of the page

rather than worry about gathering data from various sources (for example,

sessions, cookies, result beans, and so on) and coding complex View-specific

Java code. View beans effectively hide these complexities from the display

page designer.