Showing posts with label WP7. Show all posts
Showing posts with label WP7. Show all posts

Thursday, September 15, 2011

Design-time support for Caliburn.Micro

Recently me together with Bartek Pampuch have been wondering if it’s possible to fire Caliburn.Micro’s binding process at design-time. Caliburn.Micro is a really great framework (just take a look at some of our multitouch apps based on this framework and BFSharp). You just name the controls appropriately and all magic happens automatically. Controls are bound to the View Model’s properties and methods for you. The problem is that it is all happing at run time not at design time. When you are preparing the form in the Visual Studio or Expression Blend you aren’t able to see how the form will look like with bound data. In some cases such feature would very useful, especially when we want to use the sample data generated by Expression Blend or Visual Studio designers. In this post we will show you how to run Caliburn.Micro’s conventions based binding mechanism at design time.

Firstly, we discovered that we can change the objects tree representing the screen and that change is not persisted back to xaml file. You can read more about this feature in my previous post. Secondly, we tried to run Caliburn.Micro’s binding process to see what happens. What was really amazing is that after first try all just started working smoothly! The framework itself is implemented so well :)

One of the samples provided with Caliburn.Micro release is a very simple application called GameLibrary. AddGameView.xaml file defines the screen that looks like this:

image

After setting up sample data generated by Visual Studio and setting attached property 'DesignTime.Enable=”True” the same form looks like this:

clip_image001

In the simplest scenario all we need to run Caliburn.Micro binding process at design time is setting design time data context and enabling binding via Enable attached property. In sample above we are setting sample data generated by Visual Studio because the AddGameViewModel class doesn’t provide default constructor. Of course we could always add a default constructor in code and define view model instance in xaml file.

I chose this particular form intentionally because it contains the control that hasn’t defined custom binding convention by default. The Rating control is responsible for displaying Rating property value using stars. Rating property value is 0.8 so 4 stars should be displayed. In such cases where we are using controls with custom binding convention we need to register those conventions at design time. Any application using Caliburn.Micro has a type inherited from Bootstrapper type. This type is a starting point of our application and the instance of that type very often is defined as a resource inside App.xaml file. In fact Caliburn.Micro framework is already prepared for design time scenarios because Bootstrapper type contains virtual method called StartDesignTime. Let’s override this method and register appropriate convention:


public class Bootstrapper
{
public Bootstrapper()
{
if (Execute.InDesignMode) StartDesignTime(); else StartRuntime();
}
}

public class Bootstrapper<TRootModel> : Bootstrapper
{
}

public class AppBootstrapper : Bootstrapper<IShell>
{
private static bool isInitializedInDesignTime = false;

protected override void StartDesignTime()
{
base.StartDesignTime();

if (isInitializedInDesignTime)
return;
isInitializedInDesignTime = true;

ConfigureConventions();
}


void ConfigureConventions()
{
ConventionManager.AddElementConvention<Rating>(Rating.ValueProperty, "Value", "ValueChanged");
}
}

The instance of AppBootstapper type can be created many times at design time so boolean flag ensures that the conventions will be registered only once. When we compile the project and reopen the form we will see something like this:

image

This form represents a very simple scenario where the view model type has only properties of simple types like: string, double. boolean. It doesn’t contain any property of collection type or other view model type. In such scenarios Caliburn.Micro can find the appropriate type of view (control type) based on the type of view model. Let’s see another form called ResultsView.xaml to demonstrate this case:

image

This is a typical problem when we work with Caliburn.Micro and we cannot see anything at design time because the control is entirely collapsed :). Let’s see what happen after connecting sample data.

image

The list contains some elements but the font color is white so we aren’t able to read it because of the default Visual Studio background color. We can try to open this form inside Expression Blend where the background is dark or we can use custom design time attributes presented in the previous post.

image

Now we can see that Caliburn.Micro is not able to find view for view model type representing list item. It’s because we are using sample data mechanism from Visual Studio which generates dynamic type _.di0.GameLibrary.ViewModels.IndividualResultViewModel with the shape of the original view model type GameLibrary.ViewModels.IndividualResultViewModel. We need to change the way Caliburn.Micro is searching for view type based on specified view model type.

protected override void StartDesignTime()
{
base.StartDesignTime();

if (isInitializedInDesignTime)
return;
isInitializedInDesignTime = true;

ConfigureConventions();

AssemblySource.Instance.AddRange(new[] { typeof(App).Assembly });

var originalLocateTypeForModelType = ViewLocator.LocateTypeForModelType;
Func<Type, bool> isDesignTimeType = type => type.Assembly.IsDynamic;
ViewLocator.LocateTypeForModelType = (modelType, displayLocation, context) =>
{
var type = originalLocateTypeForModelType(modelType, displayLocation, context);
if (type == null && isDesignTimeType(modelType))
{
if (modelType.Name == "IndividualResultViewModel")
{
type = typeof(IndividualResultView);
}
}
return type;
};

IoC.GetInstance = base.GetInstance;
IoC.GetAllInstances = base.GetAllInstances;
}

Finally the list of items and generated sample data look like this:

image

Attached property is not the most convenient way of extending the designer functionality because we need to write xaml code. I tried to rewrite attached property to custom behavior so we could use dra&drop instead of writing the code. The problem is that behaviors code is not executed at design time. But I have good news. If you are creating UI with Expression Blend you can use our attached property on the property grid like a normal property.

image

It was possible thanks to the usage of attribute called AttachedPropertyBrowsableForTypeAttribute decorating attached property. Everything that was presented so far works both in Silverlight and WPF environments. It’s time to reveal how the magic works.



public static class DesignTime
{
public static DependencyProperty EnableProperty =
DependencyProperty.RegisterAttached(
"Enable",
typeof(bool),
typeof(DesignTime),
new PropertyMetadata(new PropertyChangedCallback(EnableChanged)));

#if !SILVERLIGHT && !WP7
[AttachedPropertyBrowsableForTypeAttribute(typeof(DependencyObject))]
#endif
public static bool GetEnable(DependencyObject dependencyObject)
{
return (bool)dependencyObject.GetValue(EnableProperty);
}


public static void SetEnable(DependencyObject dependencyObject, bool value)
{
dependencyObject.SetValue(EnableProperty, value);
}

static void EnableChanged(DependencyObject d, DependencyPropertyChangedEventArgs e)
{
if (!Execute.InDesignMode)
return;

BindingOperations.SetBinding(d, DataContextProperty, (bool)e.NewValue ? new Binding() : null);
}

private static readonly DependencyProperty DataContextProperty =
DependencyProperty.RegisterAttached(
"DataContext",
typeof(object),
typeof(DesignTime),
new PropertyMetadata(new PropertyChangedCallback(DataContextChanged))
);

private static void DataContextChanged(DependencyObject d, DependencyPropertyChangedEventArgs e)
{
if (!Execute.InDesignMode)
return;

object enable = d.GetValue(EnableProperty);
if (enable == null || ((bool)enable) == false || e.NewValue == null)
return;

var fe = d as FrameworkElement;
if (fe == null)
return;

ViewModelBinder.Bind(e.NewValue, d, string.IsNullOrEmpty(fe.Name) ? fe.GetHashCode().ToString() : fe.Name);
}
}

I hope you find this solution useful.

Download

Custom design-time attributes in Silverlight and WPF designer

We all know the standard design-time attributes in the Silverlight designer like d:DataContext, d:DesignWidth or d:DesignHeight. Let’s see how easy we can add some new attributes.

image
using System.ComponentModel;
using System.Windows;

namespace DesignTimeProperties
{
public static class d
{
static bool? inDesignMode;

/// <summary>
/// Indicates whether or not the framework is in design-time mode. (Caliburn.Micro implementation)
/// </summary>
private static bool InDesignMode
{
get
{
if (inDesignMode == null)
{
var prop = DesignerProperties.IsInDesignModeProperty;
inDesignMode = (bool)DependencyPropertyDescriptor.FromProperty(prop, typeof(FrameworkElement)).Metadata.DefaultValue;

if (!inDesignMode.GetValueOrDefault(false) && System.Diagnostics.Process.GetCurrentProcess()
.ProcessName.StartsWith("devenv", System.StringComparison.Ordinal))
inDesignMode = true;
}

return inDesignMode.GetValueOrDefault(false);
}
}

public static DependencyProperty BackgroundProperty = DependencyProperty.RegisterAttached(
"Background", typeof(System.Windows.Media.Brush), typeof(d),
new PropertyMetadata(new PropertyChangedCallback(BackgroundChanged)));

public static System.Windows.Media.Brush GetBackground(DependencyObject dependencyObject)
{
return (System.Windows.Media.Brush)dependencyObject.GetValue(BackgroundProperty);
}
public static void SetBackground(DependencyObject dependencyObject, System.Windows.Media.Brush value)
{
dependencyObject.SetValue(BackgroundProperty, value);
}
private static void BackgroundChanged(DependencyObject d, DependencyPropertyChangedEventArgs e)
{
if (!InDesignMode)
return;

d.GetType().GetProperty("Background").SetValue(d, e.NewValue, null);
}
}
}

Great, but what about many others very useful properties ? Do we need to write them all manually? No, we don’t. I have prepared T4 template that generates code for all properties of all controls available in WPF and Silverlight assemblies. When you look carefully at the code above you will find that I am using the reflection to set the value of the property. It’s because some properties like Background are defined many times in many different controls so instead of checking the actual control type I assume that the control contains appropriate property. This assumption simplifies the implementation very much. And of course the property can be set only in design time when we are editing the form inside Expression Blend or Visual Studio. T4 template code analyzes all properties from all controls and properties with the same names but different types are skipped during generation process.

Download

Thursday, May 5, 2011

Simplifying XNA game development using Async CTP or F# asynchronous workflow (GameAwaiter)

Last time I have been writing about launching the Async CTP on WP7 platform. The Async CTP Refresh has been released during MIX 2011 and finally WP7 platform is supported, so my last post is quite obsolete right now. But I hope that, after reading it, you better understand how Async CTP works underneath.

This time we will go one step forward. Once we have Tasks on WP7 and two new keywords await and async in C# we can use all that to change the way we write games in XNA.

If you know something about TPL and XNA framework it may seem a bit strange how I am going to mix them both since at first look TPL is all about building multithreaded applications and XNA application uses one single thread in most cases.

I will not not write too much about XNA framework itself  (especially because I am a beginner in game development :) ) but if you try writing even a very simple 2D game you will find out that code for XNA is really hard to understand and maintain. Probably you will have many fields storing state of the game items, many boolean flags and a lot of if-then-else statements changing that state. You can say that the game is actually one big state machine. That’s what I’ve found out after writing my fist game in XNA.

Async CTP can change the structure of the code dramatically so we can easily discover the flow of the game logic and what is most interesting everything is running on single thread.

In this post I will show you three implementations of a very simple 2D game called Smiley. All of those implementations use XNA framework but the first one is a typical XNA approach and the last two are totally different. They are using Async CTP and the F# asynchronous workflow.

image

I took the idea for the game from the Tomas Petricek’s master thesis and it can be presented on the three screens:

image

image

 image

We tap the screen to start the game, then we have 20 seconds to hit moving Smiley image as many times as we can and at the end the number of hits is displayed for 4 seconds. The first screen appears again. The Smiley position is changing in 2 seconds periods or directly after tapping on it.

Originally, that game was implemented in F# using Joinads. A few months ago I have implemented that game also in F# using Reactive Framework and treating the IObservable<T> type as the Monad type. With that I built my own workflow builder. I will try to describe that approach in the next blog post but now let’s come back to the subject of today’s post.

As you can see in the class diagram above, all three implementations have a common base class called SmileyGameBase. That class inherits from Game class which is the main XNA component representing the whole game. And all we need to do when writing an XNA game is to override two methods: Update and Draw. Update method is responsible for recalculating the game state and the Draw method draws all game items like texts, textures and so on. Both methods are called by XNA environment (Update before Draw) many times during each second when the game is running (about 30 times per second on WP7). In our game the Draw method looks the same in all three cases but the Update is specific for each of them.

public abstract class SmileyGameBase : Game
{
    // ... 

    protected override void Draw(GameTime gameTime)
    {
        GraphicsDevice.Clear(Color.Black);

        _spriteBatch.Begin();

        if (_isRunning)
        {
            float fontScale = 2;
            _spriteBatch.DrawString(_spriteFont, "time : " + _time, new Vector2(10, 5), 
                Color.White, 0, Vector2.Zero, fontScale, SpriteEffects.None, 0);
            _spriteBatch.DrawString(_spriteFont, "score : " + _score, new Vector2(500, 5), 
                Color.White, 0, Vector2.Zero, fontScale, SpriteEffects.None, 0);

            _spriteBatch.Draw(_smileyTexture, _smileyPosition, Color.White);
        }
        else
        {
            float fontScale = 3;
            _spriteBatch.DrawString(_spriteFont, _message, new Vector2(10, 100), 
                Color.Red, 0, Vector2.Zero, fontScale, SpriteEffects.None, 0);
        }

        _spriteBatch.End();

        base.Draw(gameTime);
    }
}

The code is pretty simple, we have few fields representing the game state and the Draw method draws Smiley image and prints some text messages. Now let’s look at the XNA and Async CTP implementations to compare them together.

This is the Async CTP implementation.

using System.Threading.Tasks;
using Microsoft.Xna.Framework.Input.Touch;
    
namespace SmileyGame.Common
{
    using AsyncXnaIntegration;
   
    public abstract class AsyncCtpGame : SmileyGameBase
    {
        private GameAwaiter _gameAwaiter;

        protected AsyncCtpGame()
        {
            _gameAwaiter = new GameAwaiter(this);
            Components.Add(_gameAwaiter);
        }

        protected override void LoadContent()
        {
            base.LoadContent();
            StartMenu();
        }

        async private void StartMenu()
        {
            while (true)
            {
                _message = "tap to start the game ... ";
                await _gameAwaiter.Gesture(GestureType.Tap);
                _isRunning = true;
                await StartGame();
                _isRunning = false;
                _message = "your score : " + _score;
                await _gameAwaiter.Delay(4000);
            }
        }

        async private Task StartGame()
        {
            _score = 0;
            var timer = StartTimer(GameDuration);

            while (true)
            {
                var match = await _gameAwaiter.WhenAny(timer,
                    _gameAwaiter.Delay(ChangePositionPeriod), _gameAwaiter.Gesture(IsSmileyClicked));
                switch (match.Index)
                {
                    case 0:
                        return;

                    case 1:
                        _smileyPosition = GetRandomPostion();
                        break;

                    case 2:
                        _smileyPosition = GetRandomPostion();
                        _score++;
                        break;

                    default: break;
                }

            }
        }

        async private Task StartTimer(int n)
        {
            while (n >= 0)
            {
                _time = n;
                await _gameAwaiter.Delay(TimerPeriod);
                n--;
            }
        }
    }
}

And here is the pure XNA implementation.

using System;
using System.Collections.Generic;
using System.Linq;
using Microsoft.Xna.Framework;
using Microsoft.Xna.Framework.Input.Touch;

namespace SmileyGame.Common
{
    public abstract class XnaGame : SmileyGameBase
    {
        protected XnaGame()
        {
            _message = "tap to start the game ... ";
        }

        private TimeSpan _nextTimerTime;
        private TimeSpan _nextChangePostionTime;
        private TimeSpan? _nextDisplayFinalScoreTime;
        
        protected override void Update(GameTime gameTime)
        {
            base.Update(gameTime);

            var gestures = GetGestures();

            if (!_isRunning)
            {
                if (_nextDisplayFinalScoreTime.HasValue)
                {
                    if (gameTime.TotalGameTime > _nextDisplayFinalScoreTime.Value)
                    {
                        _nextDisplayFinalScoreTime = null;
                        _message = "tap to start the game ... ";
                    }   
                }
                else if (gestures.Any(g => IsGestureType(g, GestureType.Tap)))
                {
                    _isRunning = true;
                    _score = 0;
                    _time = GameDuration;
                    _nextTimerTime = gameTime.TotalGameTime + TimeSpan.FromMilliseconds(TimerPeriod);
                    _nextChangePostionTime = gameTime.TotalGameTime + TimeSpan.FromMilliseconds(ChangePositionPeriod);
                }
            }
            else
            {                
                if (gameTime.TotalGameTime > _nextTimerTime )
                {
                    _time--;
                    _nextTimerTime = gameTime.TotalGameTime + TimeSpan.FromMilliseconds(TimerPeriod);
                    if (_time < 0)
                    {
                        _isRunning = false;
                        _nextDisplayFinalScoreTime = gameTime.TotalGameTime + TimeSpan.FromMilliseconds(4000);
                        _message = "your score : " + _score;
                    }                    
                }
                else if (gameTime.TotalGameTime > _nextChangePostionTime )
                {
                    _smileyPosition = GetRandomPostion();
                    _nextChangePostionTime = gameTime.TotalGameTime + TimeSpan.FromMilliseconds(ChangePositionPeriod);
             
                }
                else if (gestures.Any(IsSmileyClicked))
                {
                    _smileyPosition = GetRandomPostion();
                    _nextChangePostionTime = gameTime.TotalGameTime + TimeSpan.FromMilliseconds(ChangePositionPeriod);
                    _score++;
                }
            }
        }

        private static bool IsGestureType(GestureSample gestureSample, GestureType gestureType)
        {
            return (gestureSample.GestureType & gestureType) == gestureType;
        }

        private static List<GestureSample> GetGestures()
        {
            var gestures = new List<GestureSample>();
            while (TouchPanel.IsGestureAvailable)
                gestures.Add(TouchPanel.ReadGesture());
            return gestures;
        }
    }
}

When we read first implementation we can easily discover the flow of the program and the order of particular pieces of code look natural. In the pure XNA approach it is really hard to understand how the game works internally and how to extend them. Let’s image that we would like to display sequentially “3” “2” “1” “start” just after first tap in 1 second periods. Think how to do it in each of these two approaches ?

So now lets go to F# implementation. F# has something called asynchronous workflow which was available for F# developers long before Async CTP (and in fact it was the inspiration for Async CTP authors) the F# implementation is very similar to Async CTP one.

namespace SmileyGame.FSharp

open SmileyGame.Common
open AsyncXnaIntegration
open SmileyGame.FSharp.Extensions
open Microsoft.Xna.Framework.Input.Touch

type FSharpGame() as this =
  inherit SmileyGameBase()
  let _gameAwaiter = new GameAwaiter(this)
  do this.Components.Add(_gameAwaiter)

  // access to protected members
  member private this.set_smileyPosition p = this._smileyPosition <- p
  member private this.set_time t = this._time <- t 
  member private this.set_score s = this._score <- s
  member private this.set_message m = this._message <- m
  member private this.set_isRunning r = this._isRunning  <- r
  member private this.inc_score() = this._score <- this._score + 1  
  member private this.get_score = this._score
  member private this.get_TimerPeriod = SmileyGameBase.TimerPeriod
  member private this.get_GameDuration = SmileyGameBase.GameDuration  
  member private this.get_ChangePositionPeriod = SmileyGameBase.ChangePositionPeriod  
  member private this.IsSmileyClicked g = base.IsSmileyClicked g
  member private this.GetRandomPostion() = base.GetRandomPostion()
  
  member private this.StartTimer n =
    let n = ref n
    async {      
      while !n >= 0 do        
        this.set_time !n
        do! _gameAwaiter.Delay(this.get_TimerPeriod) |> Async.AwaitTask
        n := !n - 1
    }

  member private this.StartGame() =
    async {
    do this.set_score 0
    let timer = this.StartTimer(this.get_GameDuration) |> Async.ToTask
    let c = ref true
    while !c do
      let! m = _gameAwaiter.WhenAny(timer,_gameAwaiter.Delay(this.get_ChangePositionPeriod), _gameAwaiter.Gesture(this.IsSmileyClicked)) |> Async.AwaitTask
      match m.Index with
      | 0 -> c := false
      | 1 -> this.set_smileyPosition(this.GetRandomPostion())
      | 2 -> this.set_smileyPosition(this.GetRandomPostion()) ; this.inc_score()
      | _ -> ()
    }

  member private this.StartMenu() =
    async {
      while true do
        this.set_message "tap to start the game ... "
        let! _ = _gameAwaiter.Gesture(GestureType.Tap) |> Async.AwaitTask
        this.set_isRunning true
        do! this.StartGame()
        this.set_isRunning false
        this.set_message ("your score : " + (this.get_score).ToString() )
        do! _gameAwaiter.Delay(float 4000) |> Async.AwaitTask
    }

  override this.LoadContent() =
    base.LoadContent()
    this.StartMenu() |> Async.StartImmediate 

Fine, but how does it work? All I had to do was to implement class GameAwaiter deriving from GameComponent which means that we can add it to the components stored inside Game class and its Update method will be called automatically when Game’s Update method is being called. GameAwaiter gives us basically a few public methods such as Gesture, Delay and WhenAny. All of them return Task object which means that at some point in the future the result will come. Let’s see the details:

using System;
using System.Collections.Generic;
using System.Linq;
using System.Threading.Tasks;
using Microsoft.Xna.Framework;
using Microsoft.Xna.Framework.Input.Touch;

namespace AsyncXnaIntegration
{
    using AsyncXnaIntegration;

    public class GameAwaiter : GameComponent
    {
        private readonly List<Job> _jobs = new List<Job>();

        public GameAwaiter(Game game)
            : base(game)
        { }

        public Task Delay(TimeSpan interval)
        {
            return RegisterJob(new TimerJob { Interval = interval });
        }
        
        public Task Delay()
        {
            return RegisterJob(new TimerJob());
        }
     
        public Task While(Func<bool> condition)
        {
            return RegisterJob(new WhileJob { Condition = condition });
        }
  
        public Task<GestureSample[]> Gesture(Func<GestureSample, bool> predicate)
        {
            return RegisterJob(new GestureJob { GestureCondition = predicate });
        }

        public void TryDisposeTask(Task tt)
        {
            var job = _jobs.FirstOrDefault(j => j.TaskO == tt);
            if (job != null)
                job.IsDisposed = true;
        }

        public override void Update(GameTime gametime)
        {
            var state = new State { GameTime = gametime, TouchState = TouchPanel.GetState(), };
            while (TouchPanel.IsGestureAvailable)
                state.Gestures.Add(TouchPanel.ReadGesture());

            foreach (var job in _jobs.ToArray())
            {      
                if (job.IsDisposed || job.Update(state))                
                    _jobs.Remove(job);
            }
        }


        private Task<T> RegisterJob<T>(Job<T> job)
        {
            _jobs.Add(job);
            return job.Task;
        }

        #region inner types

        private class State
        {
            public GameTime GameTime;
            public TouchCollection TouchState;
            public readonly List<GestureSample> Gestures = new List<GestureSample>();
        }

        private abstract class Job
        {
            public bool IsDisposed { get; set; }
            public abstract Task TaskO { get; }

            public abstract bool Update(State state);
        }

        private abstract class Job<T> : Job
        {
            protected TaskCompletionSource<T> _source = new TaskCompletionSource<T>();
            public Task<T> Task { get { return _source.Task; } }
            public override Task TaskO { get { return Task; } }
        }

        private class TimerJob : Job<object>
        {
            private TimeSpan? StartTime;
            public TimeSpan? Interval;

            public override bool Update(State state)
            {
                if (!Interval.HasValue)
                {
                    _source.TrySetResult(null);
                    return true;
                }

                if (!StartTime.HasValue)
                    StartTime = state.GameTime.TotalGameTime;

                if (state.GameTime.TotalGameTime - StartTime >= Interval)
                {
                    _source.TrySetResult(null);
                    return true;
                }

                return false;
            }
        }

        private class WhileJob : Job<object>
        {
            public Func<bool> Condition;

            public override bool Update(State state)
            {
                if (Condition())
                {
                    _source.TrySetResult(null);
                    return true;
                }
                return false;
            }
        }

        private class GestureJob : Job<GestureSample[]>
        {
            public Func<GestureSample, bool> GestureCondition;

            public override bool Update(State state)
            {
                if (state.Gestures.Count > 0)
                {
                    if (GestureCondition == null)
                    {
                        _source.TrySetResult(state.Gestures.ToArray());
                        return true;
                    }

                    var gestures = state.Gestures.Where(GestureCondition).ToArray();
                    if (gestures.Length > 0)
                    {
                        _source.TrySetResult(gestures);
                        return true;
                    }
                }

                return false;
            }
        }

        #endregion
    
    }


public static class GameAwaiterExtensions
    {
        public static Task<Branch> WhenAny(this GameAwaiter gameAwaiter, params Task[] tasks)
        {
            return GameAwaiterExtensions.WhenAny(tasks)
                .ContinueWith(t =>
                {
                    try
                    {
                        foreach (var tt in tasks)
                            gameAwaiter.TryDisposeTask(tt);
                        return t.Result;
                    }
                    catch (Exception exception)
                    {
                        Debugger.Break();
                        return t.Result;
                    }
                }, TaskContinuationOptions.ExecuteSynchronously);
        }

        public static Task<Branch> WhenAny(params Task[] tasks)
        {
            return TaskEx
                .WhenAny(tasks)
                .ContinueWith(t =>
                {
                    var task = t.Result;
                    var taskType = task.GetType();
                    object value = null;

                    // we cannot read nonpublic types via reflection in silverlight 
                    if (taskType.IsGenericType && taskType.GetGenericArguments()[0].IsPublic)
                        value = task.GetType().GetProperty("Result").GetValue(task, null);

                    return new Branch { Index = Array.IndexOf(tasks, task), Value = value };                    
                }, TaskContinuationOptions.ExecuteSynchronously);
        }

        public static Task Delay(this GameAwaiter gameAwaiter, double milliseconds)
        {
            return gameAwaiter.Delay(TimeSpan.FromMilliseconds(milliseconds));
        }

        public static Task<GestureSample[]> Gesture(this GameAwaiter gameAwaiter, GestureType gestureType)
        {
            return gameAwaiter.Gesture(g => (g.GestureType & gestureType) == gestureType);
        }

        public static  Task<GestureSample[]> Gesture(this GameAwaiter gameAwaiter)
        {
            return gameAwaiter.Gesture(g => true);
        }
    }


    public class Branch
    {
        public int Index;
        public object Value;
    }
}

The last thing I would like to mention is a really tricky stuff :) What was really important for me when writing GameAwaiter, was the linear execution of the code. I didn’t want to introduce any new threads or use the SynchronizationContext object underneath. XNA framework calls Update and Draw methods one by one in the single thread many times per second and next iteration can start after the previous one has finished, so it’s really easy to debug such a single-threaded application. The problem is that, by default, Async CTP uses SynchronizationContext‘s Post method when the continuation delegate passed to TaskAwaiter object is called. Of course that happens if any context exists and in XNA case the “SynchronizationContext.Current” property returns the instance of the context. The latest version of Async CTP gives us the ability to configure that behavior but we would need to call “task.ConfigureAwait(false)” in every place where we use await keyword. That would be pretty inconvenient. So I have implemented my own extension method for Task object, that creates appropriately configured awaiter object.

namespace AsyncXnaIntegration
{    
    public static class Extensions
    {
        public static ConfiguredTaskAwaitable.ConfiguredTaskAwaiter GetAwaiter(this Task task)
        {
            return task.ConfigureAwait(false).GetAwaiter();
        }

        public static ConfiguredTaskAwaitable<T>.ConfiguredTaskAwaiter GetAwaiter<T>(this Task<T> task)
        {
            return task.ConfigureAwait(false).GetAwaiter();
        }
    }
}

Now we need to find a way to force C# compiler to use our extension method instead of that provided in AsyncCtpLibrary_Phone.dll library. We can do it for example by placing “using AsyncXnaIntegration;” statement inside the namespace declaration in all files where we are using Async CTP. Thanks to that little trick our method will be “closer” than all others defined outside the namespace declaration.

Of course GameAwaiter class can be extended (and it most likely will be) by many other useful methods simplifying use of phone’s Keyboard, Geo-Position or Accelerometer APIs, but this is just a proof of concept. As always I encourage you to download the source code and definitely play the Smiley game :) Thanks for reading this post and I hope it will open your eyes for many new very interesting scenarios where the Async CTP can change XNA game development.

downlaod

Friday, January 21, 2011

Async CTP on WP7

In this post I will show you how to use Async CTP released during last PDC conference inside your Windows Phone application. Async CTP extends two .Net platform languages C# and VB giving us a new way of writing asynchronous code. There are two main aspects of this project: new C#/VB compiler (two new keywords: async, await) and library AsyncCtpLibrary.dll. When the compiler sees asynchronous method it generates IL code (which we will see in details further) and that code is using some types from the mentioned library. The problem is that so far we have only two versions of library: .Net and Silverlight. So in this post I will show you how we can use  Async CTP on Windows Phone 7 despite these limitations.

The first thing we need to do before we write our first asynchronous method in the WP7 project is we need to add the reference to Silverlight version of the library AsyncCtpLibrary_Silverlight.dll. Just after doing it Visual Studio shows  a warning: “Adding a reference to a Silverlight assembly may lead to unexpected application behavior. Do you want to continue?”. The problem is that the assembly is compiled under full version of Silverlight and the Windows Phone does not contain the full version but some subset of it running on the top of .Net compact framework. So if we execute some code from such a assembly which is using types that are not available on the phone we will get the runtime exception. The key element of the AsyncCtpLibrary are Tasks and unfortunately they are not running on the phone throwing exception at runtime. The question is: can we use Async CTP without Tasks ? Of course we can! The whole mechanism behind asynchronous methods doesn’t force us to use the instance of type Task followed by await keyword. That type needs to contain method (instance or extension method) called GetAwaiter returning any type that contains appropriate pair of methods (instance or extension) BeginAwait and EndAwait. And all I did is I implemented such a matching type.

public enum AwaiterResultType
{
    Completed,
    Failed,
    Cancelled
}

public class AwaiterResult<T>
{
    public AwaiterResultType ResultType { get; private set; }
    public Exception Exception { get; private set; }
    public T Value { get; private set; }

    private AwaiterResult() { }
    public static AwaiterResult<T> Completed(T result)
    {
        return new AwaiterResult<T> { ResultType = AwaiterResultType.Completed, Value = result };
    }
    public static AwaiterResult<T> Failed(Exception exception)
    {
        return new AwaiterResult<T> { ResultType = AwaiterResultType.Failed, Exception = exception };
    }
    public static AwaiterResult<T> Cancelled()
    {
        return new AwaiterResult<T> { ResultType = AwaiterResultType.Cancelled };
    }
}

public class Awaiter<T>
{
    private SynchronizationContext Context { get; set; }
    private bool IsSynchronized { get; set; }
    private Action _continuation;

    public AwaiterResult<T> Result { get; private set; }

    private Awaiter() { }

    public Awaiter<T> GetAwaiter()
    {
        return this;
    }

    public bool BeginAwait(Action continuation)
    {
        if (Result != null)
            return false;

        _continuation += continuation;
        return true;
    }

    public T EndAwait()
    {
        if (Result == null)
            throw new InvalidOperationException("Awaiter does not contain the result yet.");

        if (Result.ResultType == AwaiterResultType.Completed)
            return Result.Value;

        if (Result.ResultType == AwaiterResultType.Failed)
            throw Result.Exception;

        return default(T); // if cancelled
    }

    private void OnResult(AwaiterResult<T> result)
    {
        if (result == null) throw new ArgumentNullException("result");
        if (Result != null) throw new InvalidOperationException("The result can be provided only once.");

        Result = result;
        if (IsSynchronized && Context != null)
        {
            Context.Post(o =>
                             {
                                 if (_continuation != null)
                                     _continuation();
                                 _continuation = null;
                             }, null);
            Context = null;
        }
        else
        {
            if (_continuation != null)
                _continuation();
            _continuation = null;
        }
    }

    public static Awaiter<T> Create(Action<Action<AwaiterResult<T>>> resultProvider, bool synchronizeWithCurrentContext = true)
    {
        if (resultProvider == null) throw new ArgumentNullException("resultProvider");

        var awaiter = new Awaiter<T> { IsSynchronized = synchronizeWithCurrentContext };
        if (synchronizeWithCurrentContext)
            awaiter.Context = SynchronizationContext.Current;
        resultProvider(awaiter.OnResult);
        return awaiter;
    }
}

The key class here is Awaiter class containing appropriate three methods: GetAwaiter, BeginAwait and EndAwait. This is an abstraction of asynchronous work (very similar to Task type) which at some point in time is finishing its work in one of three possible states: an exception could be thrown, maybe someone cancelled the execution or the work has been completed correctly returning some result. The constructor is private so the factory method called Create is the only way to create an instance of Awaiter type. So let’s see how we can use it:

public static class AwaiterEx
{
    public static Awaiter<T> Run<T>(Func<T> action)
    {
        return Awaiter<T>.Create(resultProvider =>
        {
            ThreadPool.QueueUserWorkItem(o =>
            {
                try
                {
                    var result = action();
                    resultProvider(AwaiterResult<T>.Completed(result));
                }
                catch (Exception exception)
                {
                    resultProvider(AwaiterResult<T>.Failed(exception));
                }
            },null);
        });         
    }

    public static Awaiter<object> Delay(TimeSpan timeSpan)
    {
        return Awaiter<object>.Create(resultProvider =>
        {
            var timer = new Timer(o => resultProvider(AwaiterResult<object>.Completed(null)));
            timer.Change((int)timeSpan.TotalMilliseconds, -1);
        });
    }
}

async private static void Sample()
{
    int intResult = await AwaiterEx.Run(() =>
    {
        Thread.Sleep(3000);
        return 123;
    });
    MessageBox.Show("Completed: " + intResult);

    await AwaiterEx.Delay(TimeSpan.FromSeconds(3));
    MessageBox.Show("After 3 seconds...");
}

Now let’s look at simplified version of what compiler is generating underneath to allow us to write synchronously looking code running asynchronously.

private static void SampleInternals()
{
    Awaiter<int> awaiter1 = null;
    Awaiter<object> awaiter2 = null;
    int state = 0;
    
    Action action = null;
    action = () =>
    {
        if (state == 1) goto JUMP_LABEL_1;
        if (state == 2) goto JUMP_LABEL_2;
        
        awaiter1 = AwaiterEx.Run(() =>
        {
            Thread.Sleep(3000);
            return 123;
        }).GetAwaiter();
        state = 1;
        
        if (awaiter1.BeginAwait(action))
            return;
        
        JUMP_LABEL_1:
        var intResult = awaiter1.EndAwait();
        MessageBox.Show("Completed: " + intResult);


        awaiter2 = AwaiterEx.Delay(TimeSpan.FromSeconds(3)).GetAwaiter();
        state = 2;

        if (awaiter2.BeginAwait(action))
            return;

        JUMP_LABEL_2:
        awaiter2.EndAwait();
        
        MessageBox.Show("After 3 seconds...");
    };

    action();         
}

The last thing worth mentioning is how to write asynchronous method returning some value. Asynchronous methods have two limitations: void, Task and Task<T> are the only valid return types and out parameters are not allowed. As we said before Tasks under WP7 don’t work so we cannot return Task object. In fact the solution is very simple:

public static class AwaiterEx
{
    public static Awaiter<object> BeginWriteAwaiter(this Stream stream, byte[] buffer, int offset, int numBytes)
    {
        return Awaiter<object>.Create(resultProvider =>
        {
            try
            {
                stream.BeginWrite(buffer, offset, numBytes, o =>
                {
                    try
                    {
                        stream.EndWrite(o);
                        resultProvider(AwaiterResult<object>.Completed(null));
                    }
                    catch (Exception exception)
                    {
                        resultProvider(AwaiterResult<object>.Failed(exception));
                    }
                }, null);
            }
            catch (Exception exception)
            {
                resultProvider(AwaiterResult<object>.Failed(exception));
            }
        });
    }
}

public Awaiter<object> WriteFileAwaiter(string path, string text)
{
    return Awaiter<object>.Create(async resultProvider =>
    {
        try
        {
            using (var stream = IsolatedStorageFile.OpenFile(path, FileMode.Create))  
                // caution: in fact the stream is running synchronously on WP7
            {
                var data = Encoding.Unicode.GetBytes(text);
                await stream.BeginWriteAwaiter(data, 0, data.Length);
                resultProvider(AwaiterResult<object>.Completed(null));
            }
        }
        catch (Exception exception)
        {
            resultProvider(AwaiterResult<object>.Failed(exception));
        }
    });
}

I encourage you to download sources with attached samples and play with Async CTP on WP7 because it changes a lot in terms of writing asynchronous code. AsyncCtpLibrary library provides TaskEx class with few very useful methods so you can find counterpart in my library called AwaiterEx with following API:

public static class AwaiterEx
{
    public static Awaiter<string> DownloadStringAwaiter(this WebClient webClient, Uri uri) { ... }
    public static Awaiter<object> BeginWriteAwaiter(this Stream stream, byte[] buffer, int offset, int numBytes) { ... }
  
    public static Awaiter<T> ToAwaiter<T>(this IObservable<T> observable) { ... }
    public static Awaiter<T[]> ToAwaiterAll<T>(this IObservable<T> observable) { ... }
    public static IObservable<T> ToObservable<T>(this Awaiter<T> awaiter) { ... }
    
    public static Awaiter<T[]> WhenAll<T>(this IEnumerable<Awaiter<T>> awaiters) { ... }
    public static Awaiter<T[]> WhenAll<T>(params Awaiter<T>[] awaiters) { ... }
    public static Awaiter<T> Run<T>(Func<T> action) { ... }
    public static Awaiter<object> Delay(TimeSpan timeSpan) { ... }
}

Have fun and let me know if you liked it or not. Download