我想说明如何在Rails中实现这三个术语。ActiveRecord将两个模型之间的任何类型的关系称为关联。在阅读文档或文章时，人们不会经常发现术语组合和聚合与ActiveRecord相关。通过向类的主体中添加一个关联类宏来创建关联。其中一些宏是belongs_to, has_one, has_many等。

If we want to set up a composition or aggregation, we need to add belongs_to to the owned model (also called child) and has_one or has_many to the owning model (also called parent). Wether we set up composition or aggregation depends on the options we pass to the belongs_to call in the child model. Prior to Rails 5, setting up belongs_to without any options created an aggregation, the child could exist without a parent. If we wanted a composition, we needed to explicitly declare this by adding the option required: true:

class Room < ActiveRecord::Base
  belongs_to :house, required: true
end

在Rails 5中，这一点被改变了。现在，声明belongs_to关联在默认情况下创建了一个组合，子元素不能没有父元素而存在。所以上面的例子可以重写为:

class Room < ApplicationRecord
  belongs_to :house
end

如果我们想要允许子对象在没有父对象的情况下存在，我们需要通过选项optional显式地声明这一点

class Product < ApplicationRecord
  belongs_to :category, optional: true
end

我知道这个问题被标记为c#，但概念是非常一般的问题，就像这里的这个重定向。所以我将在这里提供我的观点(从java的角度来看，我更舒服)。

当我们想到面向对象的本质时，我们总是想到对象、类(对象蓝图)以及它们之间的关系。对象之间通过方法相互关联和交互。换句话说，一个类的对象可以使用另一个类的对象提供的服务/方法。这种关系被称为关联。

聚合和组合是关联的子集，这意味着它们是关联的特定情况。

In both aggregation and composition object of one class "owns" object of another class. But there is a subtle difference. In Composition the object of class that is owned by the object of it's owning class cannot live on it's own(Also called "death relationship"). It will always live as a part of it's owning object where as in Aggregation the dependent object is standalone and can exist even if the object of owning class is dead. So in composition if owning object is garbage collected the owned object will also be which is not the case in aggregation.

困惑吗?

组合示例:以一辆汽车为例，它的发动机是这辆汽车特有的(意味着它不能用于任何其他汽车)。Car和SpecificEngine类之间的这种关系称为Composition。Car类的对象如果没有SpecificEngine类的对象就不能存在，而如果没有Car类，SpecificEngine的对象就没有意义。简单地说，Car类单独“拥有”SpecificEngine类。

聚合示例:现在考虑类Car和类Wheel。Car需要一个Wheel对象来运行。这意味着Car对象拥有Wheel对象，但我们不能说没有Car对象，Wheel对象就没有意义。它可以很好地用于自行车，卡车或不同的汽车对象。

总结一下

总而言之，关联是一个非常通用的术语，用于表示一个类使用另一个类提供的功能。如果一个父类对象拥有另一个子类对象，并且没有父类对象，这个子类对象就不能有意义地存在，我们就说它是复合。如果可以，则称为聚合。

详情请点击这里。 我是http://opensourceforgeeks.blogspot.in的作者，并在上面添加了相关帖子的链接以获取更多信息。

组合(如果你删除“整体”，“部分”也会自动删除-“所有权”)

在新类中创建现有类的对象。这称为组合，因为新类是由现有类的对象组成的。 通常使用普通成员变量。 如果组合类自动处理负责创建/销毁子类的分配/回收，则可以使用指针值。

c++中的复合

#include <iostream>
using namespace std;
/********************** Engine Class ******************/
class Engine
{
    int nEngineNumber;
    public:
    Engine(int nEngineNo);
    ~Engine(void);
};
Engine::Engine(int nEngineNo)
{
    cout<<" Engine :: Constructor " <<endl;
}
Engine::~Engine(void)
{
    cout<<" Engine :: Destructor " <<endl;
}
/********************** Car Class ******************/
class Car
{
    int nCarColorNumber;
    int nCarModelNumber;
    Engine objEngine;
    public:
    Car (int, int,int);
    ~Car(void);
};
Car::Car(int nModelNo,int nColorNo, int nEngineNo):
nCarModelNumber(nModelNo),nCarColorNumber(nColorNo),objEngine(nEngineNo)
{
    cout<<" Car :: Constructor " <<endl;
}
Car::~Car(void)
{
    cout<<" Car :: Destructor " <<endl;
    Car
    Engine
    Figure 1 : Composition
}
/********************** Bus Class ******************/
class Bus
{
    int nBusColorNumber;
    int nBusModelNumber;
    Engine* ptrEngine;
    public:
    Bus(int,int,int);
    ~Bus(void);
};
Bus::Bus(int nModelNo,int nColorNo, int nEngineNo):
nBusModelNumber(nModelNo),nBusColorNumber(nColorNo)
{
    ptrEngine = new Engine(nEngineNo);
    cout<<" Bus :: Constructor " <<endl;
}
Bus::~Bus(void)
{
    cout<<" Bus :: Destructor " <<endl;
    delete ptrEngine;
}
/********************** Main Function ******************/
int main()
{
    freopen ("InstallationDump.Log", "w", stdout);
    cout<<"--------------- Start Of Program --------------------"<<endl;
    // Composition using simple Engine in a car object
    {
        cout<<"------------- Inside Car Block ------------------"<<endl;
        Car objCar (1, 2,3);
    }
    cout<<"------------- Out of Car Block ------------------"<<endl;
    // Composition using pointer of Engine in a Bus object
    {
        cout<<"------------- Inside Bus Block ------------------"<<endl;
        Bus objBus(11, 22,33);
    }
    cout<<"------------- Out of Bus Block ------------------"<<endl;
    cout<<"--------------- End Of Program --------------------"<<endl;
    fclose (stdout);
}

输出

--------------- Start Of Program --------------------
------------- Inside Car Block ------------------
Engine :: Constructor
Car :: Constructor
Car :: Destructor
Engine :: Destructor
------------- Out of Car Block ------------------
------------- Inside Bus Block ------------------
Engine :: Constructor
Bus :: Constructor
Bus :: Destructor
Engine :: Destructor
------------- Out of Bus Block ------------------
--------------- End Of Program --------------------

聚合(如果你删除“整体”，“部分”可以存在-“无所有权”)

聚合是一种特定类型的组合，其中不暗示复杂对象和子对象之间的所有权。当一个聚合被销毁时，子对象不会被销毁。 通常使用指针变量/引用变量指向在聚合类作用域之外的对象 可以使用指向在聚合类范围之外的对象的引用值吗 不负责创建/销毁子类

c++中的聚合代码

#include <iostream>
#include <string>
using namespace std;
/********************** Teacher Class ******************/
class Teacher
{
    private:
    string m_strName;
    public:
    Teacher(string strName);
    ~Teacher(void);
    string GetName();
};
Teacher::Teacher(string strName) : m_strName(strName)
{
    cout<<" Teacher :: Constructor --- Teacher Name :: "<<m_strName<<endl;
}
Teacher::~Teacher(void)
{
    cout<<" Teacher :: Destructor --- Teacher Name :: "<<m_strName<<endl;
}
string Teacher::GetName()
{
    return m_strName;
}
/********************** Department Class ******************/
class Department
{
    private:
    Teacher *m_pcTeacher;
    Teacher& m_refTeacher;
    public:
    Department(Teacher *pcTeacher, Teacher& objTeacher);
    ~Department(void);
};
Department::Department(Teacher *pcTeacher, Teacher& objTeacher)
: m_pcTeacher(pcTeacher), m_refTeacher(objTeacher)
{
    cout<<" Department :: Constructor " <<endl;
}
Department::~Department(void)
{
    cout<<" Department :: Destructor " <<endl;
}
/********************** Main Function ******************/
int main()
{
    freopen ("InstallationDump.Log", "w", stdout);
    cout<<"--------------- Start Of Program --------------------"<<endl;
    {
        // Create a teacher outside the scope of the Department
        Teacher objTeacher("Reference Teacher");
        Teacher *pTeacher = new Teacher("Pointer Teacher"); // create a teacher
        {
            cout<<"------------- Inside Block ------------------"<<endl;
            // Create a department and use the constructor parameter to pass the teacher to it.
            Department cDept(pTeacher,objTeacher);
            Department
            Teacher
            Figure 2: Aggregation
        } // cDept goes out of scope here and is destroyed
        cout<<"------------- Out of Block ------------------"<<endl;
        // pTeacher still exists here because cDept did not destroy it
        delete pTeacher;
    }
    cout<<"--------------- End Of Program --------------------"<<endl;
    fclose (stdout);
}

输出

--------------- Start Of Program --------------------
Teacher :: Constructor --- Teacher Name :: Reference Teacher
Teacher :: Constructor --- Teacher Name :: Pointer Teacher
------------- Inside Block ------------------
Department :: Constructor
Department :: Destructor
------------- Out of Block ------------------
Teacher :: Destructor --- Teacher Name :: Pointer Teacher
Teacher :: Destructor --- Teacher Name :: Reference Teacher
--------------- End Of Program --------------------

在一个非常简单的句子中: 聚合和组合是关联的子集。

A使用B ->这是一个聚合 A需要B ->是复合。

点击这里阅读更多。

这些答案的问题在于，它们只说了一半:它们解释了聚合和组合是关联的形式，但没有说一个关联是否可能不是这两种形式。

基于对SO和一些UML文档的一些简要阅读，我收集到类关联有4种主要的具体形式:

A是由A组成的;没有A, B就不存在，就像家里的房间一样 聚合:A有A B;B可以没有A而存在，就像教室里的学生一样 依赖:A使用B;A和B之间没有生命周期依赖关系，比如方法调用参数、返回值或方法调用期间创建的临时对象 泛化:A是A

当两个实体之间的关系不是其中之一时，它可以被称为一般意义上的“关联”，并以其他方式进一步描述(注意，原型等)。

我的猜测是，“通用关联”主要用于两种情况:

when the specifics of a relationship are still being worked out; such relationship in a diagram should be converted as soon as possible to what it actually is/will be (one of the other 4). when a relationship doesn't match any of those 4 predetermined by UML; the "generic" association still gives you a way of representing a relationship that is "not one of the other ones", so that you aren't stuck using an incorrect relationship with a note "this is not actually aggregation, it's just that UML doesn't have any other symbol we could use"

依赖性(引用) 这意味着两个物体之间没有概念上的联系。例如:enroll service对象引用Student & Course对象(作为方法参数或返回类型)

public class EnrollmentService {
    public void enroll(Student s, Course c){}
}

协会(有) 这意味着对象之间几乎总是有一个链接(它们是相关联的)。 Order对象有一个Customer对象

public class Order {
    private Customer customer
}

聚合(has-a + whole-part) 两个对象之间存在整体-部分关系的一种特殊联想。但他们可能没有彼此。

public class PlayList {
    private List<Song> songs;
}

OR

public class Computer {
    private Monitor monitor;
}

注意:最棘手的部分是区分聚合和普通关联。老实说，我认为这有不同的解释。

组成(has-a +整体部分+所有权) 一种特殊的聚合。公寓是由一些房间组成的。没有公寓，就没有房间。删除一个公寓时，所有关联的房间也会被删除。

public class Apartment{
    private Room bedroom;
    public Apartment() {
       bedroom = new Room();
    }
}