Showing posts with label Silverlight. Show all posts
Showing posts with label Silverlight. Show all posts

2013-08-09

Silverlight: ObservableCollection.CollectionChanged Event

The ObservableCollection<T>.CollectionChanged event will only fire if the collection itself changes (items are added, removed or rearranged/moved.), NOT if an item in the collection changes (meaning one of the item's in the collection has one or more properties modified.)

It's now around 2:30am, and I've been going the rounds for a couple of hours with a Silverlight project at work; fighting a seemingly simple change that ended up not being so simple because of a major misunderstanding of the CollectionChanged event on the ObservableCollection<T>.  My ObservableCollection<T>.CollectionChanged event was not firing like I thought it should.

I imagine this will be a common mistake for others, and I'm sure I myself am doomed to repeat it again in the future; so I'm documenting it for a quick reminder later.

My thought process was this... "I need to watch my ObservableCollection for one of its items to be changed.  Once the item is changed, I need to flip a boolean flag which happens to be bound to a command which will disable a button after the change has been made."

Simple, right?
I should just be able to handle the ObservableCollection<T>.CollectionChanged event, because if one of those items in the collection gets changed, that counts... 

Or not.

See, the misconception is, the collection itself hasn't changed.  It has the exact same items it had before, it's just that the item within the collection changed.  So there are two things that can happen to your collection:

  • The collection itself is modified (Items are added, removed or moved.)
  • An item in the collection is modified (its individual properties are changed.)

Let me illustrate:
For this example we will have a class called VideoGame to keep track of titles and costs.

public class VideoGame : INotifyPropertyChanged
{
 private string _title;
 public string Title 
 { 
  get { return _title; } 
  set
  {
   _title = value;
   RaisePropertyChanged("Title");
  }
 }
 
 private decimal _cost;
 public decimal Cost 
 { 
  get { return _cost; } 
  set
  {
   _cost = value;
   RaisePropertyChanged("Cost");
  }
 }
 
 public event PropertyChangedEventHandler PropertyChanged;
 public void RaisePropertyChanged(string propertyName)
 {
  if(PropertyChanged != null)
   PropertyChanged(this, new PropertyChangedEventArgs(propertyName));
 }
}

And now some code to use it...

ObservableCollection<VideoGame> videoGames = new ObservableCollection<VideoGame>
{
 new VideoGame{ Title = "TrackMania 2: Stadium", Cost = 9.99M },
 new VideoGame{ Title = "TrackMania 2: Canyon", Cost = 19.99M },
 new VideoGame{ Title = "TrackMania 2: Valley", Cost = 19.99M },
};

videoGames.CollectionChanged += (sender, eventArgs) => {
 
 Debug.WriteLine("--- The collection changed");
};

Debug.WriteLine("Changing the cost of one of the video games in the collection...");
// The CollectionChanged event will not fire because it was just one of the collection's items that changed...
// not the collection itself.
videoGames[0].Cost = 7.99M; // the game went on sale.

Debug.WriteLine("Adding a video game to the collection...");
// The CollectionChanged event WILL fire because the collection of items changed.
videoGames.Add(new VideoGame{ Title ="ShootMania: Storm", Cost = 19.99M });

And the output...
Changing the cost of one of the video games in the collection...
Adding a video game to the collection...
--- The collection changed
When it dawned on me, at first I was frustrated; but then clarity set in.  It makes sense to me that we only want the CollectionChanged event to fire when the collection itself is modified, not its individual parts.  With it, the NotifyCollectionChangedEventArgs that are passed with the event give us a lot of good information about what happened with the change.

  • Action - Tells which action effected the change in the collection (Add, Remove, Replace, Reset)
  • NewItems - The items affected by the action
  • OldItems - The items removed or replaced in the collection
  • NewStartingIndex - Tells the index where the change occurred.
  • OldStartingIndex - Tells where a Replace or Remove action took place.
While all this is cool, what do I need to have happen so I can track when an item in the collection is changed?

I need to make sure I handle PropertyChanged on the individual items of the collection.  Make note, this can be tricky because I still need to watch for the Collection to change so I can subscribe to any new items' PropertyChanged events as they are added to the collection, and clean up after myself when the items leave the collection.

I'll let you decide how you're going to approach that; it's a discussion for another day.  But I want to drive home the concept that the CollectionChanged event is just that... when the collection itself changes, not its individual items.  Know that, and you'll save yourself a lot of time.

2012-08-29

Telerik RadNumericUpDown for Silverlight does not update binding when value is blanked out

Note: This may be fixed in a later version of Telerik controls; this issue was discovered in the Telerik.Windows.Controls library; version 2011.1.315.1040

Working on a Silverlight application, we found ourselves with some dialogs that utilize the Telerik RadNumericUpDown control.  We discovered a defect when binding the Value property of the control to a ViewModel behind the scenes.

The original binding works as expected, and if the spinner controls are used to change the value, the ViewModel updates as we expect.  If a user types a new value into the control it also updated the binding on the ViewModel correctly.

The problem arises when we blank out the value by clicking into the textbox portion of the control, highlighting it and hitting the Backspace or Delete key.  Removing focus from the control or not, the ViewModel property bound to that control does not get updated.

To work around this, we created this Event Handler:
void RadNumericUpDown_ValueChanged(object sender, 
                                   RadRangeBaseValueChangedEventArgs e)
{
    if (sender is RadNumericUpDown)
    {
        RadNumericUpDown upDownControl = sender as RadNumericUpDown;
        if (!upDownControl.Value.HasValue)
            upDownControl.Value = upDownControl.Minimum;
    }
}
Now we apply this event handler to our XAML:
<telerik:RadNumericUpDown ValueChanged="RadNumericUpDown_ValueChanged" ... />
This makes the behavior act the same as if the user had used the down spinner on the control to modify the value down to the minimum.

This may not be the ideal solution for you, but understanding this will give you direction on how you can create your own workaround.  We started down the path of handling the KeyUp event, but ValueChanged seems much more appropriate.

Telerik makes fantastic controls; and I'm not mad by any means.  Being a developer myself, I realize these things happen; but I do hope they have handled this better in a later release.  And as far as I know this is the only control in the 2011.1.315.1040 version that has this issue.

2011-11-02

Silverlight Unit Testing: TestInitialize/TestCleanup methods do not fire, tests do not run, nor pass/fail, and no details given

I had the classic bang-head-against-the-keyboard for three hours scenario today while doing Silverlight Unit Testing.

I had a test class with two unit tests in them which worked fine, until I started refactoring to get common setup/teardown functionality moved into methods with TestInitialize/TestCleanup attributes applied.

Then the weirdness started.  I started seeing this:
The tests don't run; don't pass or fail; and no details as to why.
Check out the yellow highlighted areas.
As you can see, there wasn't a whole lot of information present to assist me in my predicament.  After removing the TestInitialize/TestCleanup attributes, the tests worked fine.

I began searching the internet trying to find someone who ran into the same problem.  I found many results where developers complained of the TestInitialize methods swallowing exceptions and moving on, so I started down that path in my own code, trying to discover if there was an exception I was missing that was being eaten and causing the entire class to fail.

The only problem was, when I commented all code within the bodies of the TestInitialize/TestCleanup methods, I still ended up with the same results.
???
Digging deeper, it finally dawned on me that the test class in question was inheriting from a base test class that also happened to have methods dressed with the TestInitialize/TestCleanup attributes.  Looking at the declaration of TestInitializeAttribute revealed what I suspected:

[AttributeUsage(AttributeTargets.Method, AllowMultiple = false)] public sealed class TestInitializeAttribute : Attribute

AllowMultiple = false... which makes sense. So my code had two problems:

  1. My base class' method for TestInitialize needed to be made virtual so it could be overridden in classes that inherit from it to allow for overriding (and calling back to the base so both get their jobs done.)
  2. My child class' method was named different, so I needed to go make it match and override.
Once I fixed these problems for both the init and cleanup methods, the tests went back to working accurately.

Bottom Line: When doing unit testing with base test classes that perform common setup functionality, make sure to have your Setup/Teardown or Initialize/Cleanup methods marked virtual in the base class, so they can be overridden in the sub-classes.  Then override those methods in the sub-class, making sure to call back to the base appropriately.

Example (MSTest):

public class CommonTestSetupBase { [TestInitialize] public virtual void TestSetup() { // do common setup work for all that inherit } [TestCleanup] public virtual void TestTeardown() { // do common cleanup work for all that inherit } } //================= [TestClass] public class UnitTestedClassTests : CommonTestSetupBase { [TestInitialize] public override void TestSetup() { base.TestSetup(); // do common setup for this class } [TestCleanup] public override void TestTeardown() { // do common cleanup for this class base.TestTeardown(); } } I knew I wouldn't remember this if I bump into it again, so I had to document it here... I wish I had more time to go into more detail, but I hope this will be enough to help you and I in the future when we meet this problem again.

2011-08-30

Silverlight throws InvalidOperationException with Common_MethodFailed error when calling HtmlElement.SetAttribute for the "type" attribute

I was working on a Silverlight application the other day and needed to create and submit an HTML form to another ASPX page inside the website hosting the app.

To accomplish this, I was using the HtmlDocument.CreateElement, HtmlElement.SetAttribute and HtmlElement.AppendChild methods to set the whole thing up.

After deploying the application having success with Internet Explorer 9 (IE9), Chrome, etc. we tested on two virtual machines that had IE7 and IE8 and were met with errors and no form submission.

It turns out it was because I was trying to set the "type" attribute of the <input /> elements I was creating with the Silverlight methods after I'd called the AppendChild method.

IE7 and IE8 don't like a developer setting the "type" attribute after the element has been appended to the document and it will throw a JavaScript error which bubbles up to the Silverlight application and the user is shown an obscure InvalidOperationException error whose key is "Common_MethodFailed".

It made no sense to me, but I dug deeper until I could find the offending line.  I then tried an experiment with a XHTML/JavaScript page in the offending browsers to see if it was a Silverlight problem or a problem with the JavaScript engines of said browsers.

I ran into the same problem without Silverlight in the equation.  I wrote up a detailed article explaining it and have the experiment to try in different browsers yourself on my JavaScript Jedi website.

The problem can be solved by making sure you call AppendChild AFTER you set up all the properties/attributes on the HtmlElement, not before.

I hope this helps you avoid several hours of frustration.  :)
Happy Coding!

2011-06-17

Silverlight and Unit Testing - Reference Gotcha When Testing View Code

I'm new to Silverlight and am still learning my way around various nuances of the platform.

I like it so far and have been learning a lot. When a recent check-in was made by a co-worker to our project at work, we had around 90 of over 800 unit tests broken with this error in each of them:
The attachable property 'TransitionEffect' was not found in type 'ExtendedVisualStateManager'.

Being new to XAML, I had no idea where to start, except to go find the .xaml files resource dictionaries to try to figure out what the problem was.  Searching the internet made it sound like I had x:Name attributes with the same values from different dependent XAML files.

I should mention I was working in a UserControl XAML file to begin with, and started searching up the dependency hierarchy to see if the naming problem was the real issue.

The XAML files were quite large, so finding all the x:Name attributes and comparing them was going to take a while.  I opted to approach this using LINQPad.  I replaced quotes with double-quotes in the XAML so I could bring it into LINQPad as a string constant and loaded it into a XDocument object; using LINQ to XML to find matches from the other file's XAML for me.  After discovering one, I fixed it and re-ran my unit tests.  I still got the same error.

What I didn't understand was that the application ran fine and the effect was working... why on earth wouldn't it work in our unit tests?  It should have dawned on me that both the application project and the unit test project required the appropriate DLLs to be referenced in order for that namespace to be available.

Sure enough, the application project had them but the unit test project did not.  As soon as I added the references, the tests passed with flying colors again.

The required references, for those that get the exact same error:
Microsoft.Expression.Effects
Microsoft.Expression.Interactions
(both part of the Silverlight v4.0 Toolkit)

Moral of the story: When you add references to your application project, make sure you add them to the corresponding unit test project if necessary as well.

Now the next question... Why was I required to have those when I wasn't rendering any UI, but merely using the View classes in the unit test?  It seemed as if the XAML was still being parsed.

This makes sense, being as your View classes are declared as partial and the other piece to this partial puzzle is the XAML file.  Remember that whenever the View class is constructed, the InitializeComponent method is called, which parses the XAML and sets up everything declared there (correct me if I'm not describing that accurately!)

That said, if you're going to unit test a view file, you have to make sure that anything required by the XAML files are also referenced.  It's easy to take care of that for the C# files because you get compile-time errors, but you don't necessarily with the declarative XAML files.  A drawback, but one that can be handled if we're aware.

Bottom line... remember that both the XAML and corresponding C# file make up your View class and both files' reference requirements must be met to ensure all namespaces/code are available for your unit tests.

I hope this will save you (and myself) a lot of time in the future if I ever encounter this problem again.