How to: Implement ICommandSource

This example shows how to create a command source by implementing ICommandSource. A command source is an object that knows how to invoke a command. The ICommandSource interface exposes three members: Command, CommandParameter, and CommandTarget. Command is the command which will be invoked. The CommandParameter is a user-defined data type which is passed from the command source to the method which handles the command. The CommandTarget is the object that the command is being executed on.

In this example, a class is created which subclasses the Slider control and implements ICommandSource. For the complete source code, see the Implement ICommand Source Sample.

Example

WPF provides a number of classes which implement ICommandSource, such as Button, MenuItem, and ListBoxItem. A command source defines how it invokes a command. Button and MenuItem invoke a command when they are clicked. A ListBoxItem invokes a command when it is double clicked. These classes only become a command source when their Command property is set.

For this example we will invoke the command when the slider is moved, or more accurately, when the Value property is changed.

The following is the class definition.

public class CommandSlider : Slider, ICommandSource
{
    public CommandSlider() : base()
    {

    }

The next step is to implement the ICommandSource members. In this example, the properties are implemented as DependencyProperty objects. This enables the properties to use data binding. For more information about the DependencyProperty class, see the Dependency Properties Overview. For more information about data binding, see the Data Binding Overview.

Only the Command property is shown here.

// Make Command a dependency property so it can use databinding.
public static readonly DependencyProperty CommandProperty =
    DependencyProperty.Register(
        "Command",
        typeof(ICommand),
        typeof(CommandSlider),
        new PropertyMetadata((ICommand)null,
        new PropertyChangedCallback(CommandChanged)));

public ICommand Command
{
    get 
    {
        return (ICommand)GetValue(CommandProperty);
    }
    set 
    {
        SetValue(CommandProperty, value);
    }
}

The following is the DependencyProperty change callback.

// Command dependency property change callback.
private static void CommandChanged(DependencyObject d,
    DependencyPropertyChangedEventArgs e)
{
    CommandSlider cs = (CommandSlider)d;
    cs.HookUpCommand((ICommand)e.OldValue,(ICommand)e.NewValue);
}

The next step is to add and remove the command which is associated with the command source. The Command property cannot simply be overwritten when a new command is added, because the event handlers associated with the previous command, if there was one, must be removed first.

// Add a new command to the Command Property.
private void HookUpCommand(ICommand oldCommand, ICommand newCommand)
{
    // If oldCommand is not null, then we need to remove the handlers.
    if (oldCommand != null)
    {
        RemoveCommand(oldCommand, newCommand);
    }
    AddCommand(oldCommand, newCommand);
}

// Remove an old command from the Command Property.
private void RemoveCommand(ICommand oldCommand, ICommand newCommand)
{
    EventHandler handler = CanExecuteChanged;
    oldCommand.CanExecuteChanged -= handler;
}

// Add the command.
private void AddCommand(ICommand oldCommand, ICommand newCommand)
{
    EventHandler handler = new EventHandler(CanExecuteChanged);
    canExecuteChangedHandler = handler;
    if (newCommand != null)
    {
        newCommand.CanExecuteChanged += canExecuteChangedHandler;
    }
}

The last step is to create logic for the CanExecuteChanged handler and the Execute method.

The CanExecuteChanged event notifies the command source that the ability of the command to execute on the current command target may have changed. When a command source receives this event, it typically calls the CanExecute method on the command. If the command cannot execute on the current command target, the command source will typically disable itself. If the command can execute on the current command target, the command source will typically enable itself.

private void CanExecuteChanged(object sender, EventArgs e)
{

    if (this.Command != null)
    {
        RoutedCommand command = this.Command as RoutedCommand;

        // If a RoutedCommand.
        if (command != null)
        {
            if (command.CanExecute(CommandParameter, CommandTarget))
            {
                this.IsEnabled = true;
            }
            else
            {
                this.IsEnabled = false;
            }
        }
        // If a not RoutedCommand.
        else
        {
            if (Command.CanExecute(CommandParameter))
            {
                this.IsEnabled = true;
            }
            else
            {
                this.IsEnabled = false;
            }
        }
    }
}

The last step is the Execute method. If the command is a RoutedCommand, the RoutedCommand Execute method is called; otherwise, the ICommand Execute method is called.

// If Command is defined, moving the slider will invoke the command;
// Otherwise, the slider will behave normally.
protected override void OnValueChanged(double oldValue, double newValue)
{
    base.OnValueChanged(oldValue, newValue);

    if (this.Command != null)
    {
        RoutedCommand command = Command as RoutedCommand;

        if (command != null)
        {
            command.Execute(CommandParameter, CommandTarget);
        }
        else
        {
            ((ICommand)Command).Execute(CommandParameter);
        }
    }
}

The Implement ICommand Source Sample creates an sample application which uses this command source.

See Also

Reference

ICommandSource
ICommand
RoutedCommand

Concepts

Commanding Overview