关于多态的实现原理以及实例见前几天更新的C++多态一文： C++多态
使用依赖颠倒原则之后，虽然代码实现比原来更加复杂了，但是实现了高层业务逻辑层与底层模块之间的解耦合，更改或者更换底层模块不在需要改动高层模块的代码就可以实现，整个系统的更加的稳定。
传统的过程式设计倾向于使用高层的模块依赖低层次的模块，抽象层依赖于具体层
没有使用依赖颠倒之前，上层业务逻辑依赖底层业务逻辑 依赖颠倒原则 使用了依赖颠倒之后，上层业务逻辑和底层模块都依赖于抽象层，代码复杂了，但是系统易于维护了
依赖颠倒的实现也是基于多态的基础之上，，具体的Cpp代码实现如下：
#include <iostream> using namespace std; // 让computer框架和具体的厂商 进行解耦合 // 1. 定义虚类，主要目的是为了规定后期子类的行为 class HardDisk { public: virtual void work(void) = 0; }; class Memory { public: virtual void work(void) = 0; }; class Cpu { public: virtual void work(void) = 0; }; class Computer { public: Computer(HardDisk *hardDisk, Memory *memory, Cpu *cpu) { m_hardDisk = hardDisk; m_memory = memory; m_cpu = cpu; } void work() { m_hardDisk->work(); m_memory->work(); m_cpu->work(); } private: HardDisk *m_hardDisk; Memory *m_memory; Cpu *m_cpu; }; class InterCpu : public Cpu { public: void work(void) { cout << "inter cpu work status ok." << endl; } }; class XSDisk : public HardDisk { public: void work(void) { cout << "xs data work status ok." << endl; } }; class KingStonMemory : public Memory { public: void work(void) { cout << "KingSton memory status is ok." << endl; } }; int main(int argc, char const *argv[]) { // 基于多态实现依赖颠倒 HardDisk *hardDisk = NULL; Memory *memory = NULL; Cpu *cpu = NULL; hardDisk = new XSDisk; memory = new KingStonMemory; cpu = new InterCpu; Computer *myComputer = new Computer(hardDisk, memory, cpu); myComputer->work(); delete myComputer; delete cpu; delete memory; delete hardDisk; cout << "hello world." << endl; return 0; }执行结果：
xs data work status ok. KingSton memory status is ok. inter cpu work status ok. hello world.