Posted by : Sushanth Monday 17 January 2022

 

Chain of Responsibility:

Scenario: Design the software for a system that approves purchasing requests. The approval authority depends on the dollar amount of the purchase. The approval authority for a given dollar amount could change at any time and the system should be flexible enough to handle this situation.

                  Solution: PurchaseRequest objects forward the approval request to a PurchaseApproval object. Depending on the dollar amount, the PurchaseApproval object may approve the request or forward it on to the next approving authority in the chain. The approval authority at any level in the chain can be easily modified without affecting the original PurchaseRequest object. => Chain of Responsibility pattern.

 Intent:

Avoid coupling the sender of a request to its receiver by giving more than

one object a chance to handle the request. Chain the receiving objects and

pass the request along the chain until an object handles it.

 Applicability

Use Chain of Responsibility:

Ø  When more than one object may handle a request and the actual handler is not know in advance

Ø  When requests follow a “handle or forward” model - that is, some requests can be handled where they are generated while others must be forwarded to another object to be handled

 Advantages:

Ø  Reduced coupling between the sender of a request and the receiver – the sender and receiver have no explicit knowledge of each other

Ø  The chain of handlers can be modified dynamically

DisAdvantages:

Receipt is not guaranteed - a request could fall off the end of the chain without being handled

Structure:

The derived classes know how to satisfy Client requests. If the “current” object is not available or sufficient, then it delegates to the base class, which delegates to the “next” object, and the circle of life continues.



Multiple handlers could contribute to the handling of each request. The request can be passed down the entire length of the chain, with the last link being careful not to delegate to a “null next”.

Code Flow:

  1. The base class maintains a “next” pointer.
  2. Each derived class implements its contribution for handling the request.
  3. If the request needs to be “passed on”, then the derived class “calls back” to the base class, which delegates to the “next” pointer.
  4. The client (or some third party) creates and links the chain (which may include a link from the last node to the root node).
  5. The client “launches and leaves” each request with the root of the chain.
  6. Recursive delegation produces the illusion of magic.

 

#include <iostream>

using namespace std;

 

enum ErrorStates { ANALYZE=0,FIX,VERIFY,CLOSE };

 

 

class ErrorReport {

 

private:

       ErrorStates state;

 

public:

       ErrorReport(ErrorStates state)

       {

              this->state = state;

       }

       ErrorStates GetState()

       {

              return state;

       }

       void setState(ErrorStates state)

       {

              this->state= state;

       }

};

 

class Error {

protected:

       ErrorStates state;

       Error *next;

 

public:

       //Constructor

       Error(ErrorStates astate){state = astate;}

 

       //set the successor

       void setnext(Error *n) {

              this->next = n;

       }

 

       //pure virtual function to be implemented by concrete classes

       virtual void ProcessError(ErrorReport &Err) = 0;

};

 

class FixError : public Error {

 

public:

       FixError() : Error(FIX) {}

       void ProcessError(ErrorReport &Err)

       {

              if(Err.GetState() == FIX)

                     cout << "Error to be fixed "<< endl;

              else

              {

                     cout <<"error not fixed"<< endl;

 

 

                     next->ProcessError(Err);

              }

       }

};

class AnalyzeError : public Error

{

public:

       AnalyzeError() : Error(ANALYZE){}

       void ProcessError(ErrorReport &Err)

       {

              if(Err.GetState() == ANALYZE)

                     cout << "Error to be analyzed "<< endl;

              else

              {

                     cout <<"error not handled in analysis"<< endl;

                     next->ProcessError(Err);

              }

       }

};

 

class VerifyError : public Error

{

public:

       VerifyError() : Error(VERIFY){}

       void ProcessError(ErrorReport &Err)

       {

              if(Err.GetState() == VERIFY)

                     cout << "Error to be verified "<< endl;

              else

              {

                     cout <<"error not handled in verify"<< endl;

                     next->ProcessError(Err);

              }

       }

};

class CloseError : public Error

{

public:

       CloseError() : Error(CLOSE){}

       void ProcessError(ErrorReport &Err)

       {

              if(Err.GetState() == CLOSE)

                     cout << "Error to be closed "<< endl;

              else

              {

                     cout <<"error not handled in close"<< endl;

                     next->ProcessError(Err);

              }

       }

};

 

int main() {

       AnalyzeError *AE = new AnalyzeError();

       FixError *FE = new FixError();

 

 

     CloseError *CE= new CloseError();

     VerifyError *VE = new VerifyError();

 

     AE->setnext(FE);

     FE->setnext(VE);

     VE->setnext(CE);

 

     ErrorReport *ERA = new ErrorReport(ANALYZE);

     AE->ProcessError(*ERA);

 

     ErrorReport *ERC = new ErrorReport(CLOSE);

     AE->ProcessError(*ERC);

 

     delete AE;

     delete FE;

     delete CE;

     delete VE;

     delete ERA;

     delete ERC;

 

     cout << "Stack Implementation" << endl; // prints Stack Implementation

     return 0;

}

 





Leave a Reply

Subscribe to Posts | Subscribe to Comments

- Copyright © Technical Articles - Skyblue - Powered by Blogger - Designed by Johanes Djogan -