We Keep Coding

Turning a method call into an observable event, good idea?

I'm learning how to incorporate IObservable into my code. Below are two different approaches for a simple class that prints out the most recent word from an IObservable<string>. Which is the cleaner approach? I don't like WordPrinterWithCache because it introduces additional state (the _lastWord) variable and the interesting code is now scattered throughout the class. I prefer WordPrinterWithExtraSubject because the interesting code is localized to the constructor. But mostly it seems more consistently functional and "reactive"; I am reacting to the combination of two "events": (1) a new word being emitted, and (2) the invocation of the PrintMostRecent method.
However I have read that using Subjects is not desirable when it is not strictly necessary and I am introducing an unnecessary Subject<Unit>. Essentially what I'm doing here is generating an observable from a method call so I can use the observable combinator functions in more places. I like the idea of building up my code using the observable combinators and subscriptions rather than using "old-style" method invocation trees. I don't know if this is a good idea.
public class WordPrinterWithCache
{
string _lastWord = string.Empty;
public WordPrinterWithCache(IObservable<string> words)
{
words.Subscribe(w => _lastWord = w);
}
public void PrintMostRecent() => Console.WriteLine(_lastWord);
}
public class WordPrinterWithExtraSubject
{
Subject<Unit> _printRequest = new Subject<Unit>();
public WordPrinterWithExtraSubject(IObservable<string> words)
{
_printRequest
.WithLatestFrom(words.StartWith(string.Empty), (_, w) => w)
.Subscribe(w => Console.WriteLine(w));
}
public void PrintMostRecent() => _printRequest.OnNext(Unit.Default);
}
The actual scenario in my code is that I have an ICommand. When the user invokes the command (maybe by clicking a button), I want to take action on the most recent value of a specific observable. For example, maybe the ICommand represents Delete and I want to delete the selected item in a list that is represented by an IObservable<Guid>. It gets ugly to keep a bunch of "last emitted value" cache variables for each observable.
The approach I'm leaning toward is an ICommand implementation kind of like what you see below. This allows me to write code like deleteCommand.WithLatestFrom(selectedItems,(d,s)=>s).Subscribe(selected=>delete(selected));
public class ObservableCommand : ICommand, IObservable<object>
{
bool _mostRecentCanExecute = true;
Subject<object> _executeRequested = new Subject<object>();
public ObservableCommand(IObservable<bool> canExecute)
{
canExecute.Subscribe(c => _mostRecentCanExecute = c);
}
public event EventHandler CanExecuteChanged; // not implemented yet
public bool CanExecute(object parameter) => _mostRecentCanExecute;
public void Execute(object parameter) => _executeRequested.OnNext(parameter);
public IDisposable Subscribe(IObserver<object> observer) => _executeRequested.Subscribe(observer);
}

So as a general rule (guideline), I strongly suggest not having an IObservable<T> as a parameter to a method. The obvious caveat being if that method is a new Rx Operator e.g. Select, MySpecialBuffer, Debounce etc.
The theory here is that IObservable<T> is a callback mechanism. It allows something to callback to another thing that it otherwise knows nothing about. However in this case, you have something that knows about both the IObservable<T> (parameter) and the other thing (WordPrinterWithCache).
So why is there this extra layer of indirection? What ever it was that was pushing values to the IObservable<T> could just instead call a method of the WordPrinterWithCache instance.
In this case, just call a method on the other thing
public class WordPrinterWithCache
{
private string _lastWord = string.Empty;
public void SetLastWord(string word)
{
_lastWord = word;
}
public void PrintMostRecent() => Console.WriteLine(_lastWord);
}
Now this class starts to look rather pointless, but that might be ok. Simple is good.
Use Rx to help you with layering.
Upstream layers depend on down stream layers. They will directly call methods (issue commands) on the downstream layers.
Downstream layers wont have access to upstream layers. So to expose data to them they can either return values from methods, or can expose callbacks. The GoF Observer patter, .NET Events and Rx are ways to provide callbacks to upstream layers.

While I'm not sure I'd use the term "turning a method call into an observable event", I have been a fan of using fully declarative, functional and Rx driven code for a while now.
Your description of an ICommand implementation matches very closely with one I wrote a couple of years ago and have used many times and very successfully since. Furthermore, this has become the basis for a pattern I refer to as "Reactive Behaviors" which provides numerous benefits. From my blog:
Promotes behavior driven development and unit testing.
Promotes functional and thread safe programming practises.
Reduces the risk of (and if done well, can eliminate) side effects as specific behaviors are isolated in a single well named method.
Stops 'code rot' as all behavior is encapsulated within specifically named methods. Want new behaviour? Add a new method. Don't want a specific behavior anymore? Just removed it. Want a specific behavior to change? Change the one method and know that you haven't broken anything else.
Provides concise mechanisms for aggregating multiple inputs and promotes asynchronous processes to first-class status.
Reduces the need for utility classes as data can be passed through the pipeline as strongly typed anonymous classes.
Prevents memory leaks as all behaviors return a disposable that when disposed removes all subscriptions and disposed all managed resources.
You can read the full article on my blog.

Explicit Event add/remove, misunderstood?

I've been looking into memory management a lot recently and have been looking at how events are managed, now, I'm seeing the explicit add/remove syntax for the event subscription.
I think it's pretty simple, add/remove just allows me to perform other logic when I subscribe and unsubscribe? Am I getting it, or is there more to it?
Also, while I'm here, any advice / best practices for cleaning up my event handles.

The add/remove properties are basically of the same logic of using set/get properties with other members.
It allows you to create some extra logic while registering for an event, and encapsulates the event itself.
A good example for WHY you'd want to do it, is to stop extra computation when it's not needed (no one is listening to the event).
For example, lets say the events are triggered by a timer, and we don't want the timer to work if no-one is registered to the event:
private System.Windows.Forms.Timer timer = new System.Windows.Forms.Timer();
private EventHandler _explicitEvent;
public event EventHandler ExplicitEvent
{
add
{
if (_explicitEvent == null) timer.Start();
_explicitEvent += value;
}
remove
{
_explicitEvent -= value;
if (_explicitEvent == null) timer.Stop();
}
}
You'd probably want to lock the add/remove with an object (an afterthought)...

Yes, the add/remove syntax allows you to implement your own subscription logic. When you leave them out (the standard notation for an event) the compiler generates standard implementations. That is like the auto-properties.
In the following sample, there is no real difference between Event1 and Event2.
public class Foo
{
private EventHandler handler;
public event EventHandler Event1
{
add { handler += value; }
remove { handler -= value; }
}
public event EventHandler Event2;
}
But this is a separate topic from 'cleaning up' handlers. It is the subscribing class that should do the unsubscribe. The publishing class can not help with this very much.
Imagine a class that would 'clean' up the subscription list of its events. It can only sensibly do this when it is Disposed itself, and then it is unlikely to be productive as a Disposed class usually becomes collectable shortly after it is Disposed.

Add/remove syntax is commonly used to "forward" an event implementation to another class.
Cleaning up subscriptions (not "event handles") is best done by implementing IDisposable.
UPDATE: There is some variation on which object should implement IDisposable. The Rx team made the best decision from a design perspective: subscriptions themselves are IDisposable. Regular .NET events do not have an object that represents the subscription, so the choice is between the publisher (the class on which the event is defined) and the subscriber (usually the class that contains the member function being subscribed). While my design instincts prefer making the subscriber IDisposable, most real-world code makes the publisher IDisposable: it's an easier implementation, and there may be cases where there isn't an actual subscriber instance.
(That is, if the code actually cleans up event subscriptions at all. Most code does not.)

c# class instance communication

When you have different instances of the same class such as 'FootballTeam' for example and you want to let another instance of this FootballTeam class know that something occured, whats the best way to go about this?
Events won't really work I guess...
EG:
FootballTeam A = new FootballTeam();
FootballTeam B = new FootballTeam();
// now A needs to let B know about it's manager change
// manager is a property inside this class...

Events could work:
FootballTeam A = new FootballTeam();
FootballTeam B = new FootballTeam();
A.ManagerChanged += B.OnOtherManagerChanged;
Definition of the event -- FootballTeam calls its OnManagerChanged method when its Manager property changes value:
class FootballTeam
{
public event EventHandler ManagerChanged;
protected virtual void OnManagerChanged(EventArgs e)
{
EventHandler handler = ManagerChanged;
if (handler != null)
handler(this, e);
}
public void OnOtherManagerChanged(object sender, EventArgs e)
{
FootballTeam otherTeam = (FootballTeam) sender;
// A manager changed on a different FootballTeam instance
// ...do something here
}
}

There are a lot of responses to this question that provide examples of how to solve this type of problem using events (or the Observer pattern). And while these are appropriate patterns to employ, choosing how and when to use them can have a significant impact on the end result.
In your example, you describe FootballTeam instances that need to be informed when relevant changes in other instances take place. However, you must decide whether it should really be the responsibility of the FootballTeam class to both listen for and respond to such events. The Single Responsibility Principle states that there should be only one reason for a class to change. Adhering to this design concept generally results in clearer separation of concerns and overall better code.
There are a number of problems that can arise in such a design.
First, having each FootballTeam be responsible for listening for changes and responding to them can quickly become problematic. For one thing, the number of direct (point-to-point) communication lines between FootballTeam instances grows with the square of the number of instances: n*(n-1). With five teams you would have 20 connections, with ten teams you would have 90, with thirty teams you would have 870.
Managing the event subscriptions (and unsubscription) for all these instances is likely to result in confusing and unmaintainable code. Also, having event subscriptions between each pair of teams can have affect garbage collection - and result in potential leaks - or in the very least, instances of objects staying in memory much longer than needed.
Another potential problem with a design where all instances subscribe to events on each other, is infinite or cyclical event chains: A notifies B, which updates itself and notifies C, which updates itself and notifies A, which updates itself and notifies B ... ad infinitum. Or until you run out of stack space. This can be a difficult problem to solve in such a design, and would likely require awkward state management to prevent cycles and recursion.
An alternative approach, is to create a separate observer class - let's call it FootballTeamObserver - that subscribes to the change events of all FootballTeam instances, and would responsible for propagating changes, as necessary, across instances. So FootballTeam would still be responsible for broadcasting when a meaningful change takes place, and FootballTeamObserver would respond to the notification. Importantly, there would ever only be once instance of FootballTeamObserver - the singleton pattern - ensuring that a central handling site exists for all notification. This both reduces the number of event subscriptions (which would only be as many as there are teams) and separates the responsibility of responding to changes in a clean way. It can also be responsible for cycle detection and making sure that update chains have a finite length.

The typical way to do this in .Net is to define an event for operations that are interesting to other objects. For instance you might define the following
public class FootballTeam {
private string _manager;
public string Manager {
get { return _manager; }
set {
if ( ManagerChanged != null ) {
ManagerChanged(this,EventArgs.Empty);
}
}
}
public event EventHandler ManagerChanged;
}
Ideally you'd want a more type safe event but there's only so much space here
You can then listen to this event in other FootballTeam instances and respond to that event.
FootballTeam a = new FootballTeam();
FootballTeam b = new FootballTeam();
a.ManagerChanged += (sender, e) => {
Console.WriteLine("A's manager changed");
};

There are a few different ways. The fact that you need to do this is sometimes indicative of a design problem, though of course pragmatism must come into play.
One simple way is to have a private static event in the FootballTeam class which itself subscribes to in the ctor:
public class FootballTeam
{
private static event EventHandler SomethingHappened;
public FootballTeam()
{
SomethingHappened += this.HandleSomethingHappened;
}
public void DoSomething()
{
SomethingHappened(); //notifies all instances - including this one!
}
}
To avoid a memory leak, make sure to clean up the event handlers by implementing IDisposable:
public class FootballTeam : IDisposable
{
//...
public void Dispose()
{
SomethingHappened -= this.HandleSomethingHappened;
//release the reference to this instance so it can be GC'd
}
}

You could use the Observer pattern. It allows objects to "subscribe" to events that they care about.

When to use callbacks instead of events in c#?

When would you favour using a callback (i.e, passing in a Func or Action), as opposed to exposing and using an event?
UPDATE
What motivated this question was the following problem:
I have a ThingsHandler class, which
can be associated with a ThingEditor.
The ThingsHandler handles a list of
Things, knows their order, which one is 'current', when new
ones are added or deleted etc.
The ThingEditors can just modify a single
thing.
The ThingsHandler needs to alert
the ThingEditor when the user selects
a new Thing to edit, and the
ThingEditor needs to alert the
ThingsHandler when the user says
'done'.
What bothers me is having these two classes holding references to each other - though I guess that's inevitable - or binding to events in both directions. I wondered if using a callback in one direction was 'cleaner'.
I suspect there is a design pattern for this.

Though the other answers thus far seem reasonable, I would take a more philosophical tack.
A class is a mechanism that models a particular kind of thing in a particular domain. It is very easy when writing the internal details of a class to conflate the implementation details of the mechanism with the semantics being modeled. A brief example of what I mean:
class Giraffe : Mammal, IDisposable
{
public override void Eat(Food f) { ... }
public void Dispose() { ... }
}
Notice how we've conflated the real-world thing being modeled (a giraffe is a kind of mammal, a giraffe eats food) with the details of the implementation (an instance of Giraffe is an object which can be disposed of with the "using" statement). I guarantee that if you go to the zoo, you will never see a giraffe being disposed of with the using statement. We've mixed up the levels here, which is unfortunate.
I try to use events (and properties) as part of the semantic model and use callback methods (and fields) as part of the mechanism. I would make GaveBirth an event of Giraffe, since that is part of the model of real-world giraffe behaviour we're attempting to capture. If I had some mechanism, like, say I wanted to implement an inorder-walk tree traversal algorithm that walked the family tree of giraffes and called a method back on every one, then I'd say that this was clearly a mechanism and not part of the model, and make it a callback rather than try to shoehorn that into the event model.

I use callbacks in a few cases where I know it will only ever fire once, and the callback is specific to a single method call (rather than to an object instance) - for example, as the return part of an async method.
This is particularly true of static utility methods (since you don't have an instance, and static events are deadly when used carelessly, and to be avoided), but of course the other option is to create a class instance with an event instead.

Generally, I use a callback if it is required, whereas an event is used when it should be optional.
Don't expose an event if you're expecting there to always be something listening.
Consider the following:
public class MyClass_Event
{
public event EventHandler MakeMeDoWork;
public void DoWork()
{
if (MakeMeDoWork == null)
throw new Exception("Set the event MakeMeDoWork before calling this method.");
MakeMeDoWork(this, EventArgs.Empty);
}
}
versus:
public class MyClass_Callback
{
public void DoWork(EventHandler callback)
{
if (callback == null)
throw new ArgumentException("Set the callback.", "callback"); // better design
callback(this, EventArgs.Empty);
}
}
The code is almost the same as the callback can be passed as null, but at least the exception thrown can be more relevant.

Callbacks are good when one object wishes to receive a single notification (e.g. an Async data read runs and then calls you with the result).
Events are good for recurring notifications that can be received by an arbitrary number of listeners.

One example is when the callback should return something. E.g. (stupid example):
public int Sum(Func<int> callbackA, Func<int> callbackB) {
return callbackA() + callbackB();
}
public void UseSum() {
return sum(() => 10, () => 20);
}

In terms of OO design and class coupling there isn't great deal of difference between a callback interface and an event.
However, I prefer events where they are things the class needs to "shout about" to whoever is interested in listening (usually multiple things) and callbacks where a specific class has requested an async operation.
Whatever you use, use them consistently across the codebase!

I would use Func or Action when I am going to call the function once or use a Lambda expression.
Events can be registered more than once which sometimes is ideal. With a callback, one has to implement a registration system for the callbacks if you want multiple.

Well, I think they are same things. There're many different tech terms to name the same concepts or things in the different languages.
So, what do you mean "Callback" or "event handler"?
According to MSDN: Callback function is code within a managed application that helps an unmanaged DLL function complete a task.
And, MADN also gives us a introduction of the difference between them.click here
Callbacks are extensibility points that allow a framework to call back into user code through a delegate. These delegates are usually passed to the framework through a parameter of a method.
Events are a special case of callbacks that supports convenient and consistent syntax for supplying the delegate (an event handler). In addition, Visual Studio’s statement completion and designers provide help in using event-based APIs
Also, in some books, such as this book, the author seemed say the same thing with MSDN.
Therefore, in my opinion, you can't say use callbacks instead of events in the C#.

C#: events or an observer interface? Pros/cons?

I've got the following (simplified):
interface IFindFilesObserver
{
void OnFoundFile(FileInfo fileInfo);
void OnFoundDirectory(DirectoryInfo directoryInfo);
}
class FindFiles
{
IFindFilesObserver _observer;
// ...
}
...and I'm conflicted. This is basically what I would have written in C++, but C# has events. Should I change the code to use events, or should I leave it alone?
What are the advantages or disadvantages of events over a traditional observer interface?

Consider an event to be a callback interface where the interface has only one method.
Only hook events you need
With events, you only need to implement handlers for events you're interested in handling. In the observer interface pattern, you'd have to implement all methods in the entire interface including implementing method bodies for notification types you don't actually care about handling. In your example, you always have to implement OnFoundDirectory and OnFoundFile, even if you only care about one of these events.
Less maintenance
Another good thing about events is you can add a new one to a particular class so that it will raise it, and you don't have to change every existing observer. Whereas if you want to add a new method to an interface, you have to go around every class that already implements that interface and implement the new method in all of them. With an event though, you only need to alter existing classes that actually want to do something in response to the new event you're adding.
The pattern is built into the language so everybody knows how to use it
Events are idiomatic, in that when you see an event, you know how to use it. With an observer interface, people often implement different ways of registering to receive notifications and hook up the observer.. with events though, once you've learnt how to register and use one (with the += operator), the rest are all the same.
Pros for interfaces
I haven't got many pros for interfaces. I guess they force someone to implement all methods in the interface. But, you can't really force somebody to implement all those methods correctly, so I don't think there's a lot of value on this.
Syntax
Some people don't like the way you have to declare a delegate type for each event. Also, standard event handlers in the .NET framework have these parameters: (object sender, EventArgs args). As sender doesn't specify a particular type, you have to down-cast if you want to use it. This often is fine in practice, it feels not quite right though because you're losing the protection of the static type system. But, if you implement your own events and don't follow the .NET framework convention on this, you can use the correct type so potential down-casting isn't required.

Hmm, events can be used to implement the Observer pattern. In fact, using events can be regarded as another implementation of the observer-pattern imho.

Events are harder to propagate through chain of objects, for example if you use FACADE pattern or delegate work to other class.
You need to be very careful with unsubscribing from events to allow object to be garbage collected.
Events are 2x time slower than simple function call, 3x slower if you do null check on every raise, and copy event delegate before null check and invocation to make it thread safe.
Also read MSDN about new (in 4.0) IObserver<T> interface.
Consider this example:
using System;
namespace Example
{
//Observer
public class SomeFacade
{
public void DoSomeWork(IObserver notificationObject)
{
Worker worker = new Worker(notificationObject);
worker.DoWork();
}
}
public class Worker
{
private readonly IObserver _notificationObject;
public Worker(IObserver notificationObject)
{
_notificationObject = notificationObject;
}
public void DoWork()
{
//...
_notificationObject.Progress(100);
_notificationObject.Done();
}
}
public interface IObserver
{
void Done();
void Progress(int amount);
}
//Events
public class SomeFacadeWithEvents
{
public event Action Done;
public event Action<int> Progress;
private void RaiseDone()
{
if (Done != null) Done();
}
private void RaiseProgress(int amount)
{
if (Progress != null) Progress(amount);
}
public void DoSomeWork()
{
WorkerWithEvents worker = new WorkerWithEvents();
worker.Done += RaiseDone;
worker.Progress += RaiseProgress;
worker.DoWork();
//Also we neede to unsubscribe...
worker.Done -= RaiseDone;
worker.Progress -= RaiseProgress;
}
}
public class WorkerWithEvents
{
public event Action Done;
public event Action<int> Progress;
public void DoWork()
{
//...
Progress(100);
Done();
}
}
}

Pros of an interface-solution:
If you add methods, existing observers needs to implement those methods. This means that you have less of a chance of forgetting to wire up existing observers to new functionality. You can of course implement them as empty methods which means you have the luxury of still doing nothing in response to certain "events". But you won't so easily forget.
If you use explicit implementation, you'll also get compiler errors the other way, if you remove or change existing interfaces, then observers implementing them will stop compiling.
Cons:
More thought has to go into planning, since a change in the observer interface might enforce changes all over your solution, which might require different planning. Since a simple event is optional, little or no other code has to change unless that other code should react to the event.

Some further benefits of events.
You get proper multicast behaviour for free.
If you change the subscribers of an event in response to that event the behaviour is well defined
They can be introspected (reflected) easily and consistently
Tool chain support for events (simply because they are the idiom in .net)
You get the option to use the asynchronous apis it provides
You can achieve all of these (except the tool chain) yourself but it's surprisingly hard. For example:
If you use a member variable like a List<> to store the list of observers.
If you use foreach to iterate over it then any attempt to add or remove a subscriber within one of the OnFoo() method callbacks will trigger an exception unless you write further code to deal with it cleanly.

The best way to decide is this: which one suits the situation better. That might sound like a silly or unhelpful answer, but I don't think you should regard one or the other as the "proper" solution.
We can throw a hundred tips at you. Events are best when the observer is expected to listen for arbitrary events. An interface is best when the observer is expected to listed to all of a given set of events. Events are best when dealing with GUI apps. Interfaces consume less memory (a single pointer for multiple events). Yadda yadda yadda. A bulleted list of pros and cons is something to think about, but not a definitive answer. What you really need to do is try both of them in actual applications and get a good feel for them. Then you can choose the one that suits the situation better. Learn form doing.
If you have to use a single defining question, then ask yourself which better describes your situation: A set of loosely related events any of which may be used or ignored, or a set of closely related events which will all generally need to be handled by one observer. But then, I'm just describing the event model and interface model, so I'm back at square one: which one suits the situation better?

Pros are that events are more 'dot-netty'. If you are designing non-visual components that can be dropped onto a form, you can hook them up using the designer.
Cons are that an event only signifies a single event - you need a separate event for each 'thing' that you want to notify the observer about. This doesn't really have much practical impact except that each observed object would need to hold a reference for every observer for every event, bloating memory in the case where there are lots of observed objects (one of the reasons they made a different way of managing the observer/observable relationship in WPF).
In your case I'd argue it doesn't make much difference. If the observer would typically be interested in all those events, use an observer interface rather than separate events.

I prefer an event base solution for the following reasons
It reduces the cost of entry. It's much easier to say "+= new EventHandler" than to implement a full fledged interface.
It reduces maintenance costs. If you add a new event into your class that's all that needs to be done. If you add a new event to an interface you must update every single consumer in your code base. Or define an entirely new interface which over time gets annoying to consumers "Do I implement IRandomEvent2 or IRandomEvent5?"
Events allow for handlers to be non-class based (ie a static method somewhere). There is no functional reason to force all event handlers to be an instance member
Grouping a bunch of events into an interface is making an assumption about how the events are used (and it's just that, an assumption)
Interfaces offer no real advantage over a raw event.

Java has language support for anonymous interfaces, so callback interfaces are the thing to use in Java.
C# has support for anonymous delegates - lambdas - and so events are the thing to use in C#.

A benefit of interfaces is that they are easier to apply decorators to. The standard example:
subject.RegisterObserver(new LoggingObserver(myRealObserver));
compared to:
subject.AnEvent += (sender, args) => { LogTheEvent(); realEventHandler(sender, args); };
(I'm a big fan of the decorator pattern).

If your objects will need to be serialized in some way that retains references such as with NetDataContractSerializer or perhaps protobuf events will not be able to cross the serialization boundary. Since observer pattern relies on nothing more than just object references, it can work with this type of serialization with no problem if that is what is desired.
Ex. You have a bunch of business objects that link to each other bidirectionally that you need to pass to a web service.