Repository Pattern with Entity Framework 4.1 and Code First

Introduction

A popular pattern for ORM data access is the Repository pattern. Repositories are currently very popular even in EF for the reasons below:

Hide EF from upper layer

Make code better testable

The big disadvantage of EF is rigid architecture which can be hardly mocked, so if you want to unit test upper layer you must wrap EF somehow to allow mocking its implementation.

The solution I'll show is a very simple implementation of this pattern using EF 4.1 and code first to generate the database.

Using the code

We will start using a classic implementation of the Repository Pattern. In this section, I'll show a very generic one, but you can of course refactor it to be better suitable for your needs.

Below the Repository Interface


Collapse | Copy Code

public interface IRepository<TEntity> : IDisposable where TEntity : IEntity
{
IQueryable<TEntity> GetAll();
void Delete(TEntity entity);
void Add(TEntity entity);
}

As you can see, we got a generic

GetAll()

method which returns an

IQuerable

that allows to retrieve and query any entity in our model. Because the Repository uses generics, we have to constrain it(as you can see, this is a

TEntity

).


Collapse | Copy Code

public interface IEntity
{
int Id { get; set; }
}

Now let's see how to implement the repository pattern.


Collapse | Copy Code

public class Repository<TEntity> : IRepository<TEntity> where TEntity : class, IEntity
{
private IDbContext_context;
public Repository(IDbContextcontext)
{
_context = context;
}
private IDbSet<TEntity> DbSet
{
get
{
return _context.Set<TEntity>();
}
}
public IQueryable<TEntity> GetAll()
{
return DbSet.AsQueryable();
}
public void Delete(TEntity entity)
{
DbSet.Remove(entity);
}
public void Add(TEntity entity)
{
DbSet.Add(entity);
}

public void Dispose()
{
Dispose(true);
GC.SuppressFinalize(this);
}
protected virtual void Dispose(bool disposing)
{
if (disposing)
{
if (_context != null)
{
_context.Dispose();
_context = null;
}
}
}
}

The repository has a constructor that takes an

IDbContext

object. That interface adds an abstraction layer on the EF

DbContext

as shown below:


Collapse | Copy Code

 public interface IDbContext{

IDbSet<TEntity> Set<TEntity>() where TEntity : class;
int SaveChanges();
void Dispose();
} 

As you can see, the

IDbContext

has an

IDbSet

property. This property represents an entity set that is used to perform create, read, update, and delete operations.

In EF, when you are using code first you need to create a Context which will contain all the

DbSet

properties that are used to access the entities in your model. The context class must inherit from

System.Data.EntityDbContext

which provides facilities for querying and working with entity data as objects.
Below is the implementation of

DbContext

which implements the above interface and Inherits from

DbContext

.


Collapse | Copy Code

public class ApplicationContext : DbContext, IDbContext
{
public DbSet<User> User { get; set; }
public new IDbSet<TEntity> Set<TEntity>() where TEntity : class
{
return base.Set<TEntity>();
}
}

The

DbContext

will contain all the properties to access the entity in our context. In this example, it will access only "User". In this example, we use code first also to generate the DataBase: (only the entity that has got a relative DbSetwill be generated)

At this point, all we need to implement the repository pattern is done. At this point, we probably need to create the DataBase using code first approach.

For this scope, I have created another interface that will contain the DataBase initialization strategy.


Collapse | Copy Code

public interface IDatabaseInitializer<in TContext> where TContext :
IDbContext
{
// Summary:
//   Executes the strategy to initialize
//   the database for the given context.
// Parameters:
//   context: The context.
void InitializeDatabase(TContext context);
}

This interface is used to create the DataBase and allow the user to specify a kind of strategy. Let's assume we want to delete and create the DataBase every time we only need to implement the interface to specify what we need to do within the

InitializeDatabase

method as shown below:


Collapse | Copy Code

public class DataBaseInitializer : IDatabaseInitializer<ApplicationContext>
{
public void InitializeDatabase(ApplicationContext context)
{
context.Database.Delete();
context.Database.Create();
}
}

Now we got all we need.
First we need to initialize the DataBase (you need to add a connection string in your configuration file) as below:


Collapse | Copy Code

<connectionStrings>
<add
name="PeluSoft.ApplicationContext"
providerName="System.Data.SqlClient"
connectionString="Server=myServer;Database=TestUser;Persist Security Info=True;"/>

</connectionStrings>

At this point, we are ready to initialize the DataBase:


Collapse | Copy Code

var context=new ApplicationContext();
var testDataBaseInitializer = new TestDataBaseInitializer();
testDataBaseInitializer.InitializeDatabase(context);

Then you can use the repository for the common operation on the entities.


Collapse | Copy Code

var context=new ApplicationContext();
var userRepo=new Repository<User>(context);
var user =new User()
{
Username = "TestName"
};
userRepo.Add(user);
context.SaveChanges();