Flex bindable - ythy/blog GitHub Wiki

定义

[Bindable] is a one of several meta tags that you can use in flex ActionScript code. It can be applied to properties, or methods that are marked in any scope. It cannot be used with static class members.

The key to using the [Bindable] meta tag is understanding what is going on under the hood when you use it. Essentially using data binding is a type of shorthand for adding event listeners and dispatching events.

详细概念

There are two basic forms of the [Bindable] tag. The first is just [Bindable] followed by a var/property declaration. The Second is [Bindable(event="eventname")] followed by either a var/property declaration, a function/method declaration or one half of a getter/setter declaration.

I'll explain the longer notation first since the other builds on the same concept but with even more shorthand.

When you use [Bindable(event="eventname")] you are essentially telling the compiler that this var/property/function/method (call this the instance member) is 'available' to be used as the source for data binding. You are also telling it that when the value of the instance member has been invalidated/changed and it needs to be re-read that the "eventname" event will be dispatched. In this longer form this all you are doing. You the developer are responsible for actually dispatching the "eventname" event whenever the value needs to be updated in the binding subscribers.

The real efficiency of using data binding comes on the subscribing side. The typical notation you will see in MXML is value="{instance.propertyName}" When you use the notation { } you are telling the compiler to do the following:

  1. Create an event listener that listens to the event named in the bindable meta tag
  2. In that event listener re-read the instance.propertyName and update this value

If you use the shorter form [Bindable], and you add the tag before a property/var, the compiler fills in the blanks and adds some additional functionality to make the property bindable. Essentially you are telling the compiler "add the events and methods you need to make this property bindable" Now the way to think of what the compiler will do under the hood is this.

  1. make a private version of your var
  2. create an "event" to trigger the binding
  3. create a getter function with scope and name of your original var that returns the private verson of the var when called.
  4. create a setter function with scope and name of your original var that sets the private version of the var when called AND dispatches the triggering event.

In essence the compiler will do much of the work for you.

    [Bindable]
    public var xyz

is equivalent

    private var _xyz:String;

    [Bindable(event="updateXYZValue")]
    public function get xyz():String{
        return _xyz;
    }

    public function set xyz(newxyz:String):void{
        _xyz = newxyz;
        dispatchEvent(new Event("updateXYZValue"));
    }

conclusion

  1. [Bindable] XXX 元数据标签定义了 XXX 是可绑定的 即 XXX 的值发生变化时 会派发事件eventB
  2. BindingUtils 或者 text={XXX} 会增加针对eventB的事件监听， 处理目标值相应的更新

参考资料

官方文档