Techniques for WCAG 2.0

Skip to Content (Press Enter)

-

SL6: Defining a UI Automation Peer for a Custom Silverlight Control

Applicability

This technique relates to:

User Agent and Assistive Technology Support Notes

See User Agent Support Notes for SL6.

Description

The objective of this technique is to create an AutomationPeer class for a custom Silverlight control. The AutomationPeer exposes accessibility properties of the control in a way that abstracts the Silverlight technology specifics of the control model and maps information to UI Automation concepts, so that these properties can be consumed by the UI Automation accessibility framework.

The AutomationPeer concept is part of the overall architecture design of the UI Automation system. The peer represents a deliberate abstraction of the control, such that a client can obtain pattern-based information about the specific purpose and capability of a control without knowing its implementation-specific object model or having to resort to using a framework-specific object model API. Also, the peers run in a different process than the controls they represent, which has performance and security advantages. For more information on UI Automation architecture, see UI Automation Overview on MSDN.

Creating a custom Silverlight control is one way that Silverlight application authors can create user interface components either for their own application, or as a packaged redistributable that provides the control UI for third parties. Creating an automation peer for a custom control reports control-specific information to the UI Automation accessibility framework, and enables a custom control to participate in all of the same techniques involving UI Automation that can be used for a control that is distributed in the core Silverlight run time. Assistive technologies can use the UI Automation accessibility framework to discover the name and role of the user interface component, and can get and set values by accessing UI Automation patterns. UI Automation thus supports extensibility, while maintaining a discovery system for names, roles and values of UI components.

Control authors associate a peer with a class by implementing a method override for the class implementation. Control authors declare name and role through properties that are general to any UI Automation peer. Control authors expose the means to get and set values by choosing to support one or more patterns that are usually associated with a role. For example, a control in the role of "Button" would typically support an "Invoke" pattern. A consumer of UI Automation could check whether the pattern was supported and then call the pattern-based method Invoke, which would activate the button without any device input events being produced or required.

By convention, controls and their automation peers share a naming pattern. For example, if a control is named Spinner, its automation peer is named SpinnerPeer. However, the actual wiring for the class-peer association is made in the control code by overriding OnCreateAutomationPeer. Thus it is necessary to have access to the control code in order to associate a new peer class implementation with that control.

In addition to properties, automation peers can also expose methods as part of the implemented UI Automation control pattern. For example, a peer implementing the Value pattern can provide an implementation of the SetValue method. The SetValue method can be called by a UI Automation client in order to programmatically set the value of the owner control. The functionality exposed by the implementation of a control pattern can be accessed either by automation-based testing, or by assistive technologies.

Examples

Example 1: SimpleNumericUpDown control and its peer

The example implements a very simple Silverlight custom control named SimpleNumericUpDown. The control is a templateable control, meaning that the UI is defined in a XAML file that serves as the default UI, but any consumer of the control can change the visual appearance by applying a new template. Nevertheless, the basic accessibility characteristics of the control can be shaped by the control author, and can apply even for cases where the visible UI is noticably different. This separation between design-implementation and code-behavior is one reason for the peer-owner design in UI Automation. The majority of the example shows the C# code, including the following :

Control definition class:

   public class SimpleNumericUpDown : Control
   {
       public SimpleNumericUpDown()
       {
           this.DefaultStyleKey = typeof(SimpleNumericUpDown);
       protected override System.Windows.Automation.Peers.AutomationPeer OnCreateAutomationPeer()
       {
           return new SimpleNumericUpDownAutomationPeer(this);
       }
 // templating and event handlers omitted
       public static DependencyProperty NumericValueProperty = DependencyProperty.Register(
           "NumericValue",
           typeof(Int32),
           typeof(SimpleNumericUpDown),
           new PropertyMetadata(0)
           );
       public Int32 NumericValue
       {
           get { return (Int32)this.GetValue(NumericValueProperty); }
           set {this.SetValue(NumericValueProperty,value);}
       }
   }
   

Automation peer definition:

   public class SimpleNumericUpDownAutomationPeer : FrameworkElementAutomationPeer, IValueProvider
   {
       private SimpleNumericUpDown OwnerControl { get { return (SimpleNumericUpDown)Owner; } }
       public SimpleNumericUpDownAutomationPeer(SimpleNumericUpDown owner)
           : base(owner) {}
       //peer overrides
       protected override string GetClassNameCore()
       {
           return "SimpleNumericUpDown";
       }
       protected override AutomationControlType GetAutomationControlTypeCore()
       {
           return AutomationControlType.Spinner;
       }
       public override object GetPattern(PatternInterface patternInterface) {
           if (patternInterface == PatternInterface.Value)
           {
               return this;
           }
           return base.GetPattern(patternInterface);
       }
       // Value pattern implementation
       String IValueProvider.Value
       {
           get { return OwnerControl.NumericValue.ToString(); }
       }
       bool IValueProvider.IsReadOnly {get{return false;}}
       void IValueProvider.SetValue(string value)
       {
           OwnerControl.NumericValue = Convert.ToInt32(value);
       }

This example is shown in operation in the working example of Simple Numeric UpDown control.

Resources

Resources are for information purposes only, no endorsement implied.

Tests

Procedure

  1. Using a browser that supports Silverlight, open an HTML page that references a Silverlight application through an object tag. To see UI Automation, use Microsoft Windows as platform.

  2. Use a verification tool that is capable of showing the full automation tree, and an object’s UI Automation properties and patterns as part of the tree. (For example, use UIAVerify or Silverlight Spy; see Resources links.) Select the item in the automation tree that is accessing the relevant custom automation peer implementation.

  3. Examine the set of properties exposed in the tree. Check that name is reported by Name, that the class name is reported as ClassName, and that there is a role as reported by the value of ControlType.

  4. If the control is expected to report a value, check that the value is reported in the tree somehow. (Exactly which property reports the value varies depending on the control function and pattern; for more information, see Windows Automation API).

  5. Check whether a control pattern is reported in the tree. If a control pattern is reported, test the methods of that pattern using facilities in the verification tool. Verify that invoking the methods has changed the corresponding read only property values in the tree.

Expected Results

#3, #4, and #5 (if applicable) are true.

If this is a sufficient technique for a success criterion, failing this test procedure does not necessarily mean that the success criterion has not been satisfied in some other way, only that this technique has not been successfully implemented and can not be used to claim conformance.

Techniques are Informative

Techniques are informative—that means they are not required. The basis for determining conformance to WCAG 2.0 is the success criteria from the WCAG 2.0 standard—not the techniques. For important information about techniques, please see the Understanding Techniques for WCAG Success Criteria section of Understanding WCAG 2.0.