您还可以在接口声明中使用事件，但委托则不是这样。

An event in .net is a designated combination of an Add method and a Remove method, both of which expect some particular type of delegate. Both C# and vb.net can auto-generate code for the add and remove methods which will define a delegate to hold the event subscriptions, and add/remove the passed in delegagte to/from that subscription delegate. VB.net will also auto-generate code (with the RaiseEvent statement) to invoke the subscription list if and only if it is non-empty; for some reason, C# doesn't generate the latter.

Note that while it is common to manage event subscriptions using a multicast delegate, that is not the only means of doing so. From a public perspective, a would-be event subscriber needs to know how to let an object know it wants to receive events, but it does not need to know what mechanism the publisher will use to raise the events. Note also that while whoever defined the event data structure in .net apparently thought there should be a public means of raising them, neither C# nor vb.net makes use of that feature.

注意:如果你有c# 5.0 Unleashed，请阅读第18章“事件”中的“委托的普通使用限制”，以更好地理解两者之间的区别。

---

举一个简单而具体的例子总是对我有帮助。这是给社区的。首先，我将展示如何单独使用委托来完成事件为我们做的事情。然后，我将展示相同的解决方案如何使用EventHandler的实例。然后我解释为什么我们不想做我在第一个例子中解释的事情。这篇文章的灵感来自John Skeet的一篇文章。

例1:使用公共委托

假设我有一个只有一个下拉框的WinForms应用程序。下拉列表绑定到List<Person>。其中Person具有Id，名称，昵称，头发颜色的属性。在主窗体上有一个自定义用户控件，显示此人的属性。当某人在下拉菜单中选择某人时，用户控件中的标签将更新以显示所选人员的属性。

下面是它的工作原理。我们有三个文件可以帮助我们把这些放在一起:

cs——静态类保存委托 Form1.cs——主要形式 cs——用户控件显示所有细节

下面是每个类的相关代码:

class Mediator
{
    public delegate void PersonChangedDelegate(Person p); //delegate type definition
    public static PersonChangedDelegate PersonChangedDel; //delegate instance. Detail view will "subscribe" to this.
    public static void OnPersonChanged(Person p) //Form1 will call this when the drop-down changes.
    {
        if (PersonChangedDel != null)
        {
            PersonChangedDel(p);
        }
    }
}

这是我们的用户控件:

public partial class DetailView : UserControl
{
    public DetailView()
    {
        InitializeComponent();
        Mediator.PersonChangedDel += DetailView_PersonChanged;
    }

    void DetailView_PersonChanged(Person p)
    {
        BindData(p);
    }

    public void BindData(Person p)
    {
        lblPersonHairColor.Text = p.HairColor;
        lblPersonId.Text = p.IdPerson.ToString();
        lblPersonName.Text = p.Name;
        lblPersonNickName.Text = p.NickName;

    }
}

最后，我们在Form1.cs中有以下代码。这里我们正在调用OnPersonChanged，它调用任何订阅到委托的代码。

private void comboBox1_SelectedIndexChanged(object sender, EventArgs e)
{
    Mediator.OnPersonChanged((Person)comboBox1.SelectedItem); //Call the mediator's OnPersonChanged method. This will in turn call all the methods assigned (i.e. subscribed to) to the delegate -- in this case `DetailView_PersonChanged`.
}

好的。这就是不使用事件只使用委托就能让它工作的方法。我们只是把一个公共委托放到一个类中——你可以让它是静态的，或者是单例的，等等。太好了。

但是，但是，但是，我们不想做我上面描述的事情。因为公共字段有很多不好的原因。那么我们有什么选择呢?正如约翰·斯基特所描述的，我们有以下几种选择:

A public delegate variable (this is what we just did above. don't do this. i just told you above why it's bad) Put the delegate into a property with a get/set (problem here is that subscribers could override each other -- so we could subscribe a bunch of methods to the delegate and then we could accidentally say PersonChangedDel = null, wiping out all of the other subscriptions. The other problem that remains here is that since the users have access to the delegate, they can invoke the targets in the invocation list -- we don't want external users having access to when to raise our events. A delegate variable with AddXXXHandler and RemoveXXXHandler methods

第三个选项本质上是事件给我们的。当我们声明一个EventHandler时，它给了我们对委托的访问，不是公开的，不是作为属性，而是作为我们称之为事件的东西它有添加/删除访问器。

让我们看看同样的程序是什么样子，但是现在使用Event而不是public委托(我也将我们的Mediator更改为单例):

例2:使用EventHandler代替公共委托

中介:

class Mediator
{

    private static readonly Mediator _Instance = new Mediator();

    private Mediator() { }

    public static Mediator GetInstance()
    {
        return _Instance;
    }

    public event EventHandler<PersonChangedEventArgs> PersonChanged; //this is just a property we expose to add items to the delegate.

    public void OnPersonChanged(object sender, Person p)
    {
        var personChangedDelegate = PersonChanged as EventHandler<PersonChangedEventArgs>;
        if (personChangedDelegate != null)
        {
            personChangedDelegate(sender, new PersonChangedEventArgs() { Person = p });
        }
    }
}

注意，如果你在EventHandler上F12，它会显示你的定义只是一个泛型的委托，带有额外的"sender"对象:

public delegate void EventHandler<TEventArgs>(object sender, TEventArgs e);

用户控件:

public partial class DetailView : UserControl
{
    public DetailView()
    {
        InitializeComponent();
        Mediator.GetInstance().PersonChanged += DetailView_PersonChanged;
    }

    void DetailView_PersonChanged(object sender, PersonChangedEventArgs e)
    {
        BindData(e.Person);
    }

    public void BindData(Person p)
    {
        lblPersonHairColor.Text = p.HairColor;
        lblPersonId.Text = p.IdPerson.ToString();
        lblPersonName.Text = p.Name;
        lblPersonNickName.Text = p.NickName;

    }
}

最后，这是Form1.cs的代码:

private void comboBox1_SelectedIndexChanged(object sender, EventArgs e)
{
        Mediator.GetInstance().OnPersonChanged(this, (Person)comboBox1.SelectedItem);
}

因为EventHandler想要和EventArgs作为参数，我创建这个类只有一个属性:

class PersonChangedEventArgs
{
    public Person Person { get; set; }
}

希望这向您展示了为什么我们有事件，以及它们作为委托如何不同——但功能相同。

Delegate是一个类型安全的函数指针。事件是使用委托的发布者-订阅者设计模式的实现。

对于生活在2020年的人们来说，想要一个干净的答案……

定义:

Delegate:定义一个函数指针。 事件:定义 (1)受保护接口，以及 (2)操作(+=，-=)，和 (3)优点:你不再需要使用new关键字。

关于形容词protected:

// eventTest.SomeoneSay = null;              // Compile Error.
// eventTest.SomeoneSay = new Say(SayHello); // Compile Error.

还要注意来自微软的这一部分:https://learn.microsoft.com/en-us/dotnet/standard/events/#raising-multiple-events

代码示例:

委托:

public class DelegateTest
{
    public delegate void Say(); // Define a pointer type "void <- ()" named "Say".
    private Say say;

    public DelegateTest() {
        say  = new Say(SayHello);     // Setup the field, Say say, first.
        say += new Say(SayGoodBye);
        
        say.Invoke();
    }
    
    public void SayHello() { /* display "Hello World!" to your GUI. */ }
    public void SayGoodBye() { /* display "Good bye!" to your GUI. */ }
}

---

事件:

public class EventTest
{
    public delegate void Say();
    public event Say SomeoneSay;  // Use the type "Say" to define event, an 
                                  // auto-setup-everything-good field for you.
    public EventTest() {
         SomeoneSay += SayHello;
         SomeoneSay += SayGoodBye;

         SomeoneSay();
    }
    
    public void SayHello() { /* display "Hello World!" to your GUI. */ }
    public void SayGoodBye() { /* display "Good bye!" to your GUI. */ }
}

参考:

事件和委托——解释c#中事件和委托模式之间的重要区别，以及它们为什么有用。: https://dzone.com/articles/event-vs-delegate

这是另一个很好的参考链接。 http://csharpindepth.com/Articles/Chapter2/Events.aspx

简单地说，本文的要点是——事件是对委托的封装。

引用自文章:

Suppose events didn't exist as a concept in C#/.NET. How would another class subscribe to an event? Three options: A public delegate variable A delegate variable backed by a property A delegate variable with AddXXXHandler and RemoveXXXHandler methods Option 1 is clearly horrible, for all the normal reasons we abhor public variables. Option 2 is slightly better, but allows subscribers to effectively override each other - it would be all too easy to write someInstance.MyEvent = eventHandler; which would replace any existing event handlers rather than adding a new one. In addition, you still need to write the properties. Option 3 is basically what events give you, but with a guaranteed convention (generated by the compiler and backed by extra flags in the IL) and a "free" implementation if you're happy with the semantics that field-like events give you. Subscribing to and unsubscribing from events is encapsulated without allowing arbitrary access to the list of event handlers, and languages can make things simpler by providing syntax for both declaration and subscription.