封装和抽象之间的确切区别是什么?
当前回答
让我们以堆栈为例。它可以使用数组或链表来实现。但它支持的操作是推送和弹出。
Now abstraction is exposing only the interfaces push and pop. The underlying representation is hidden (is it an array or a linked list?) and a well-defined interface is provided. Now how do you ensure that no accidental access is made to the abstracted data? That is where encapsulation comes in. For example, classes in C++ use the access specifiers which ensure that accidental access and modification is prevented. And also, by making the above-mentioned interfaces as public, it ensures that the only way to manipulate the stack is through the well-defined interface. In the process, it has coupled the data and the code that can manipulate it (let's not get the friend functions involved here). That is, the code and data are bonded together or tied or encapsulated.
其他回答
让我们以堆栈为例。它可以使用数组或链表来实现。但它支持的操作是推送和弹出。
Now abstraction is exposing only the interfaces push and pop. The underlying representation is hidden (is it an array or a linked list?) and a well-defined interface is provided. Now how do you ensure that no accidental access is made to the abstracted data? That is where encapsulation comes in. For example, classes in C++ use the access specifiers which ensure that accidental access and modification is prevented. And also, by making the above-mentioned interfaces as public, it ensures that the only way to manipulate the stack is through the well-defined interface. In the process, it has coupled the data and the code that can manipulate it (let's not get the friend functions involved here). That is, the code and data are bonded together or tied or encapsulated.
我读得越多,就越困惑。所以,我的理解是:
封装:
我们通常从外面看到手表,它的组件被封装在它的身体里。我们对不同的操作有某种控制。这种隐藏细节和公开控制(例如设置时间)的方式就是封装。
抽象:
到目前为止,我们一直在谈论手表。但我们没有具体说明是哪种手表。可以是数字的,也可以是模拟的,可以是手用的,也可以是墙用的。有很多可能性。我们所知道的是,这是一块手表,它显示时间,这是我们唯一感兴趣的东西,时间。这种隐藏细节和公开通用特性或用例的方法就是抽象。
封装是将复杂性包裹在一个胶囊中,即类&因此封装… 抽象是一个对象区别于其他对象的特征。
抽象可以通过使具有一个或多个抽象方法的类抽象来实现。它只是一个特性,应该由扩展它的类来实现。 例如,当你发明/设计一辆汽车时,你定义了一个特征,比如汽车应该有4个门、刹车、方向盘等,所以任何使用这种设计的人都应该包括这些特征。实现并不是抽象的首要部分。它只会定义应该包含的特征。
Encapsulation is achieved keeping data and the behaviour in one capsule that is class & by making use of access modifiers like public, private, protected along with inheritance, aggregation or composition. So you only show only required things, that too, only to the extent you want to show. i.e. public, protected, friendly & private ka funda…… e.g. GM decides to use the abstracted design of car above. But they have various products having the same characteristics & doing almost same functionality. So they write a class which extends the above abstract class. It says how gear box should work, how break should work, how steering wheel should work. Then all the products just use this common functionality. They need not know how the gear box works or break works or steering wheal works. Indivisual product can surely have more features like a/c or auto lock etc…..
两者都很强大;但是使用抽象需要比封装更多的技能,没有抽象,更大的应用程序/产品就无法生存。
我认为封装是实现抽象的一种方式。看看下面的链接。
抽象和封装
另一个例子:
假设我创建了一个不可变的Rectangle类,如下所示:
class Rectangle {
public:
Rectangle(int width, int height) : width_(width), height_(height) {}
int width() const { return width_; }
int height() const { return height_; }
private:
int width_;
int height_;
}
现在很明显,我已经封装了宽度和高度(访问受到某种限制),但我没有抽象任何东西(好吧,也许我忽略了矩形在坐标空间中的位置,但这是示例的缺陷)。
好的抽象通常意味着好的封装。
一个好的抽象例子是通用数据库连接类。它的公共接口与数据库无关,非常简单,但允许我对连接做我想做的事情。你看到了吗?这里还有封装,因为类内部必须有所有低级句柄和调用。