Introduction
If you are going to search in Google with this word ‘SOLID’, you will get tons of articles, videos, tutorials on this topic. My aim is to make this article as simple as possible and provide small but good examples.
And I can provide all those examples in C#. But don’t worry, get the concept and keep practicing with the programming language that you are familiar with.
1. About SOLID
SOLID is basically 5 principles, which will help to create a good software architecture. You can see that all design patterns are based on these principles. SOLID is basically an acronym of the following:
- S is single responsibility principle (SRP)
- O stands for open closed principle (OCP)
- L Liskov substitution principle (LSP)
- I interface segregation principle (ISP)
- D Dependency injection principle (DIP)
I believe that with pictures, with examples, an article will be more approachable and understandable.
1.1 Single responsibility principle (SRP)
A class should take one responsibility and there should be one reason to change that class. Now what does that mean? I want to share one picture to give a clear idea about this.
Now see this tool is a combination of so many different tools like knife, nail cutter, screw driver, etc. So will you want to buy this tool? I don’t think so. Because there is a problem with this tool, if you want to add any other tool to it, then you need to change the base and that is not good. This is a bad architecture to introduce into any system. It will be better if nail cutter can only be used to cut the nail or knife can only be used to cut vegetables.
Now I want to give one C# example on this principle:
namespace SRP
{
public class Employee
{
public int Employee_Id { get; set; }
public string Employee_Name { get; set; }
public bool InsertIntoEmployeeTable(Employee em)
{
return true;
}
public void GenerateReport(Employee em)
{
}
}
}
‘Employee
’ class is taking 2 responsibilities, one is to take responsibility of employee database operation and another one is to generate employee report. Employee
class should not take the report generation responsibility because suppose some days after your customer asked you to give a facility to generate the report in Excel or any other reporting format, then this class will need to be changed and that is not good.
So according to SRP, one class should take one responsibility so we should write one different class for report generation, so that any change in report generation should not affect the ‘Employee
’ class.
public class ReportGeneration
{
public void GenerateReport(Employee em)
{
}
}
2.2 Open closed principle (OCP)
Now take the same ‘ReportGeneration
’ class as an example of this principle. Can you guess what is the problem with the below class!!
public class ReportGeneration
{
public string ReportType { get; set; }
public void GenerateReport(Employee em)
{
if (ReportType == "CRS")
{
}
if (ReportType == "PDF")
{
}
}
}
Brilliant!! Yes you are right, too much ‘If
’ clauses are there and if we want to introduce another new report type like ‘Excel
’, then you need to write another ‘if
’. This class should be open for extension but closed for modification. But how to do that!!
public class IReportGeneration
{
public virtual void GenerateReport(Employee em)
{
}
}
public class CrystalReportGeneraion : IReportGeneration
{
public override void GenerateReport(Employee em)
{
}
}
public class PDFReportGeneraion : IReportGeneration
{
public override void GenerateReport(Employee em)
{
}
}
So if you want to introduce a new report type, then just inherit from IReportGeneration
. So IReportGeneration
is open for extension but closed for modification.
2.3 Liskov substitution principle (LSP)
This principle is simple but very important to understand. Child class should not break parent class’s type definition and behavior. Now what is the meaning of this!! Ok let me take the same employee
example to make you understand this principle. Check the below picture. Employee
is a parent class and Casual
and Contractual employee
are the child classes, inhering from employee
class.
Now see the below code:
public abstract class Employee
{
public virtual string GetProjectDetails(int employeeId)
{
return "Base Project";
}
public virtual string GetEmployeeDetails(int employeeId)
{
return "Base Employee";
}
}
public class CasualEmployee : Employee
{
public override string GetProjectDetails(int employeeId)
{
return "Child Project";
}
public override string GetEmployeeDetails(int employeeId)
{
return "Child Employee";
}
}
public class ContractualEmployee : Employee
{
public override string GetProjectDetails(int employeeId)
{
return "Child Project";
}
public override string GetEmployeeDetails(int employeeId)
{
throw new NotImplementedException();
}
}
Up to this is fine right? Now, check the below code and it will violate the LSP principle.
List<Employee> employeeList = new List<Employee>();
employeeList.Add(new ContractualEmployee());
employeeList.Add(new CasualEmployee());
foreach (Employee e in employeeList)
{
e.GetEmployeeDetails(1245);
}
Now I guess you got the problem. Yes right, for contractual employee
, you will get not implemented exception and that is violating LSP. Then what is the solution? Break the whole thing in 2 different interfaces, 1. IProject
2. IEmployee
and implement according to employee
type.
public interface IEmployee
{
string GetEmployeeDetails(int employeeId);
}
public interface IProject
{
string GetProjectDetails(int employeeId);
}
Now, contractual employee
will implement IEmployee
not IProject
. This will maintain this principle.
2.4 Interface segregation principle (ISP)
This principle states that any client should not be forced to use an interface which is irrelevant to it. Now what does this mean, suppose there is one database for storing data of all types of employees (i.e. Permanent, non-permanent), now what will be the best approach for our interface?
public interface IEmployee
{
bool AddEmployeeDetails();
}
And all types of employee
class will inherit this interface for saving data. This is fine right? Now suppose that company one day told to you that they want to read only data of permanent employees. What you will do, just add one method to this interface?
public interface IEmployeeDatabase
{
bool AddEmployeeDetails();
bool ShowEmployeeDetails(int employeeId);
}
But now we are breaking something. We are forcing non-permanent employee
class to show their details from database. So, the solution is to give this responsibility to another interface.
public interface IAddOperation
{
bool AddEmployeeDetails();
}
public interface IGetOperation
{
bool ShowEmployeeDetails(int employeeId);
}
And non-permanent employee
will implement only IAddOperation
and permanent employee
will implement both the interface.
2.5 Dependency inversion principle (DIP)
This principle tells you not to write any tightly coupled code because that is a nightmare to maintain when the application is growing bigger and bigger. If a class depends on another class, then we need to change one class if something changes in that dependent class. We should always try to write loosely coupled class.
Suppose there is one notification system after saving some details into database.
public class Email
{
public void SendEmail()
{
}
}
public class Notification
{
private Email _email;
public Notification()
{
_email = new Email();
}
public void PromotionalNotification()
{
_email.SendEmail();
}
}
Now Notification
class totally depends on Email
class, because it only sends one type of notification. If we want to introduce any other like SMS then? We need to change the notification system also. And this is called tightly coupled. What can we do to make it loosely coupled? Ok, check the following implementation.
public interface IMessenger
{
void SendMessage();
}
public class Email : IMessenger
{
public void SendMessage()
{
}
}
public class SMS : IMessenger
{
public void SendMessage()
{
}
}
public class Notification
{
private IMessenger _iMessenger;
public Notification()
{
_ iMessenger = new Email();
}
public void DoNotify()
{
_ iMessenger.SendMessage();
}
}
Still Notification
class depends on Email
class. Now, we can use dependency injection so that we can make it loosely coupled. There are 3 types to DI, Constructor injection, Property injection and method injection.
Constructor Injection
public class Notification
{
private IMessenger _iMessenger;
public Notification(Imessenger pMessenger)
{
_ iMessenger = pMessenger;
}
public void DoNotify()
{
_ iMessenger.SendMessage();
}
}
Property Injection
public class Notification
{
private IMessenger _iMessenger;
public Notification()
{
}
public IMessenger MessageService
{
private get;
set
{
_ iMessenger = value;
}
}
public void DoNotify()
{
_ iMessenger.SendMessage();
}
}
Method Injection
public class Notification
{
public void DoNotify(IMessenger pMessenger)
{
pMessenger.SendMessage();
}
}
So, SOLID principle will help us to write loosely coupled code which is highly maintainable and less error prone.