User Guide
Table Of Contents
- Contents
- Working with Flash MX 2004
- Creating Basic Components in Flash MX 2004
- Creating Advanced Components in Flash MX 2004
- Contents
- About Creating components
- Writing the component’s ActionScript code
- Simple example of a class file
- General process for writing a class file
- Selecting a parent class
- Identifying the class, symbol, and owner names
- About the component instantiation life cycle
- Writing the constructor
- Specifying clip parameters
- Implementing the constructObject2() method
- Implementing the init() method
- Implementing the createChildren() method
- Implementing the commitProperties() method
- Implementing the measure() method
- Implementing the layoutChildren() method
- Implementing the draw() method
- Defining getters and setters
- Component metadata
- Defining component parameters
- Handling events
- Using the Event metadata
- About invalidation
- Skinning custom controls
- Adding styles
- Making components accessible
- Improving component usability
- Best practices when designing a component
- Using the ModalText example
- Troubleshooting
- Index
56 Chapter 3: Creating Advanced Components in Flash MX 2004
Flex only calls the commitProperties() method after it calls the invalidateProperties()
method.
For example, the ViewStack container uses the
commitProperties() method to maximize
performance. When you set the
ViewStack.selectedIndex property, the ViewStack container
doesn’t update to a new page right away. Instead, it privately stores a
pendingSelectedIndex
property. When it is time for Flash Player to update the screen, Flex calls the
commitProperties(), measure(), layoutChildren(), and draw() methods. In the
commitProperties() method, the ViewStack container checks to see whether the
pendingSelectedIndex property is set, and it updates the selected index at that time.
The motivation to use the
commitProperties() method is to delay processing until the last
minute, so that Flash Player avoids doing computationally expensive, redundant work. For
example, if a script changes the
ViewStack.selectedIndex property 15 times, you would want
to minimize the number of times the display of the ViewStack container updates when the
selectedIndex property changes. By using the commitProperties() method, you can update
the
pendingSelectedIndex property 15 times, and then only do the rendering once.
This is most useful for properties that are computationally expensive to update. If setting a
property is inexpensive, you can avoid using the
commitProperties() method.
Implementing the measure() method
Generally, the
measure() method is only called once when the component is instantiated. The
component’s container sets the initial size and Flex calculates the preferred minimum and
maximum sizes. You can use the
measure() method to explicitly set the size of the component,
although Macromedia does not recommend doing this when developing components.
You can set the following properties in the
measure() method. Flex calculates them, but you can
override them:
• _measuredMinWidth
• _measuredMaxWidth
• _measuredMinHeight
• _measuredMaxHeight
• _measuredWidthFlex
• _measuredHeightFlex
• _measuredPreferredWidth
• _measuredPreferredHeight
The Flex properties (those that end with the word Flex) define limits for when the object is
resized. These measured properties are used for layout in containers if your component doesn’t
explicitly set a
preferredWidth or preferredHeight attribute.
Controls calculate the values of these based on runtime properties. For example, the Button
control’s
measure() method examines how wide its label is in order to compute the value of the
_measuredPreferredWidth property.