I have tried to implement Castle Windsor into my ASP.NET Web API project by following the guide by Mark Seemann. But when I try to run the code it gives me a ComponentNotFoundException exception. I mean that I should have registered the dependency right.
I really hope that someone has a solution to my problem. I have tried to search for a solution but with out any luck.
Global.asax
public class WebApiApplication : System.Web.HttpApplication
{
private readonly IWindsorContainer _container;
public WebApiApplication()
{
_container = new WindsorContainer().Install(new ControllerInstaller());
}
protected void Application_Start()
{
GlobalConfiguration.Configure(WebApiConfig.Register);
GlobalConfiguration.Configuration.Services.Replace(typeof(IHttpControllerActivator), new WindsorCompositionRoot(_container));
}
public override void Dispose()
{
_container.Dispose();
base.Dispose();
}
}
IHttpControllerActivator implementation
public class WindsorCompositionRoot : IHttpControllerActivator
{
private readonly IWindsorContainer _container;
public WindsorCompositionRoot(IWindsorContainer container)
{
_container = container;
}
public IHttpController Create(
HttpRequestMessage request,
HttpControllerDescriptor controllerDescriptor,
Type controllerType)
{
var controller =
(IHttpController)_container.Resolve(controllerType);
request.RegisterForDispose(
new Release(
() => _container.Release(controller)));
return controller;
}
private class Release : IDisposable
{
private readonly Action release;
public Release(Action release)
{
this.release = release;
}
public void Dispose()
{
release();
}
}
}
ControllerInstaller
public class ControllerInstaller : IWindsorInstaller
{
public void Install(IWindsorContainer container, IConfigurationStore store)
{
container.Register(Component.For<IGymnastDataAccess>().ImplementedBy<GymnastDataAccess>());
}
}
Controller
public class GymnastController : ApiController
{
private readonly IGymnastDataAccess _gymnastDataAccess;
public GymnastController(IGymnastDataAccess gymnastDataAccess)
{
_gymnastDataAccess = gymnastDataAccess;
}
public IEnumerable<string> Get()
{
_gymnastDataAccess.Load();
return new string[] { "value1", "value2" };
}
}
Castle Windsor do not support automatic resolve of concrete classes out of the box, so you should register your controller class in container:
container.Register(Component.For<GymnastController>());
or implement ILazyComponentLoader like here to get automatic resolve of concrete classes.
Related
I'm using MVC 5, Core 3.1
I have 'AddDbContext' added to my service in Startup.cs.
I then have a Class library core 3.1 project which is my ADO Dal layer.
This is added as a service as well in The ConfigureServices of Startup.cs.
I want to inject the Connection String into the DAL application.
I have:
public partial class ContainerContext : DbContext
{
public ContainerContext()
{
}
public ContainerContext(DbContextOptions<ContainerContext> options) : base(options)
{
}
}
In Startup.cs
public void ConfigureServices(IServiceCollection services)
{
var connection = Configuration.GetConnectionString("DefaultConnection");
services.AddDbContext<ContainerContext>(options => options.UseSqlServer(connection));
services.AddDAL();
}
In the Dal project:
public static class ServiceCollectionExtensions
{
// Add parameters if required, e.g. for configuration
public static IServiceCollection AddDAL(this IServiceCollection services)
{
// Register all services as required
services.AddScoped<ILeaseBll, LeaseDal>();
return services;
}
}
The Dal class.
public class LeaseDal : ILeaseBll
{
private string conString;
public LeaseDal(???????)
{
// Some validation for the Context maybe (isNull etc?) throw new ArgumentNullException("conString");
//this.connectionString = conString;
}
How would / should it be done?
Thanks
There is a philosophy change with Dot-Net-Core and Dot-Net-Framework....
public class LeaseDal : ILeaseBll
{
private string conString;
This is not best practice in dot-net-CORE.
You do NOT inject your "connection string" in your concrete DataAccessLayer object.
You inject the db-context.
(and the db-context already has been wired to the Ioc...with its correct connection string)
Something like this:
public interface IDepartmentQueryDomainData()
{
Task<int> GetCountAsync(CancellationToken token);
}
..
public class DepartmentQueryEntityFrameworkDomainDataLayer : IDepartmentQueryDomainData
{
public const string ErrorMessageILoggerFactoryIsNull = "ILoggerFactory is null";
public const string ErrorMessageMyCoolDbContextIsNull =
"MyCoolDbContext is null";
private readonly ILogger<DepartmentQueryEntityFrameworkDomainDataLayer> logger;
private readonly MyCoolDbContext entityDbContext;
public DepartmentQueryEntityFrameworkDomainDataLayer(
ILoggerFactory loggerFactory,
MyCoolDbContext context
{
if (null == loggerFactory)
{
throw new ArgumentNullException(ErrorMessageILoggerFactoryIsNull, (Exception)null);
}
this.logger = loggerFactory.CreateLogger<DepartmentQueryEntityFrameworkDomainDataLayer>();
this.entityDbContext = context ?? throw new ArgumentNullException(
ErrorMessageMyCoolDbContextIsNull,
(Exception)null);
}
public async Task<int> GetCountAsync(CancellationToken token)
{
int returnValue = await this.entityDbContext.Departments.AsNoTracking().CountAsync(token);
this.logger.Log(
new LogEntry(
LoggingEventTypeEnum.Trace,
string.Format(
LogMessages.Count,
returnValue)));
return returnValue;
}
}
You can also "see" this here:
https://learn.microsoft.com/en-us/ef/core/dbcontext-configuration/
public class MyController
{
private readonly ApplicationDbContext _context;
public MyController(ApplicationDbContext context)
{
_context = context;
}
}
I would never inject the dbContext into a "controller"...(I agree with you that the Dal should be a separate layer)...
but besides that "miscue" on the microsoft example, you do see that you inject the dbContext.
Also see:
https://hovermind.com/aspnet-core/using-dbcontext-with-dependency-injection.html
I'm running into an issue while unit testing where if I run multiple tests at once, the DbContext will lose track of records I've added during unit tests and I think this may have to do with how services are registered in my ServiceCollection.
I have the following setup:
IUnitOfWork:
public interface IUnitOfWork : IDisposable
{
IUserRepository Users { get; }
int Complete();
}
UnitOfWork
public class UnitOfWork : IUnitOfWork
{
private readonly MyDbContext _context;
public IUserRepository Users { get; }
public UnitOfWork(MyDbContext context,
IUserRepository userRepository)
{
_context = context;
Users = userRepository;
}
public void Dispose() => _context.Dispose();
public int Complete() => _context.SaveChanges();
}
UserRepository
public class UserRepository : Repository<User>, IUserRepository
{
public UserRepository(MyDbContext context) : base(context) { }
public MyDbContext MyDbContext => Context as MyDbContext;
public Task<User?> GetUserDetailsAsync(int userID)
{
var user = MyDbContext.Users.Where(user => user.Id == userID)
.Include(user => user.Emails)
.Include(user => user.PhoneNumbers).FirstOrDefault();
return Task.FromResult(user);
}
}
Here is my base test:
public abstract class BaseTest : IDisposable
{
protected ServiceProvider ServiceProvider { get; }
private MyDbContext MyDbContext { get; }
protected IUnitOfWork UnitOfWork { get; }
public BaseTest()
{
var serviceCollection = new ServiceCollection();
serviceCollection.AddScoped<IUserService, UserService>()
.AddScoped<IUnitOfWork, UnitOfWork>()
.AddScoped(typeof(IRepository<>), typeof(Repository<>))
.AddScoped<IOrganizationRepository, OrganizationRepository>()
.AddScoped<IExercisePostRepository, ExercisePostRepository>()
.AddScoped<IUserRepository, UserRepository>()
.AddTransient<IRestClient, RestClient>()
.AddAutoMapper(typeof(Startup).Assembly)
.AddDbContext<MyDbContext>(options =>
options.UseInMemoryDatabase("Core")
.EnableSensitiveDataLogging());
ServiceProvider = serviceCollection.BuildServiceProvider();
SpotcheckrCoreContext = ServiceProvider.GetRequiredService<MyDbContext>();
MyDbContext.Database.EnsureCreated();
UnitOfWork = ServiceProvider.GetRequiredService<IUnitOfWork>();
}
public void Dispose()
{
MyDbContext.Database.EnsureDeleted();
UnitOfWork.Dispose();
}
}
Sample test:
public class UserServiceTests : BaseTest
{
private readonly IUnitOfWork UnitOfWork;
private readonly IUserService Service;
public UserServiceTests()
{
UnitOfWork = ServiceProvider.GetRequiredService<IUnitOfWork>();
Service = ServiceProvider.GetRequiredService<IUserService>();
}
[Fact]
public async void GetUserAsync_WithValidUser_ReturnsUser()
{
var user = new User
{
FirstName = "John",
LastName = "Doe"
};
UnitOfWork.Users.Add(user);
UnitOfWork.Complete();
var result = await Service.GetUserAsync(user.Id);
Assert.Equal(user.Id, result.Id);
}
}
If I run this test by itself, then it will correctly pass and I can see the user in the repository. However if I run it with other tests and debug, then that user is lost once I inspect UnitOfWork.Users in the repository but I do see it in the UnitOfWork.Users in the test.
What is the correct approach here?
Edit 1:
Tried some other changes but no luck yet. Adjusted UnitOfWork to take in the interfaces of each repository and registering them in BaseTest as scoped services. Also tried marking BaseTest as implementing IDisposable and then executing:
public void Dispose()
{
MyDbContext.Database.EnsureDeleted();
UnitOfWork.Dispose();
}
In the service layer I'll see the Users just fine but as soon as I step into the repository layer I'll lose the Users :/ I have a suspicion it is related to dependency injection AddScoped vs AddTransient and how all of that works with running multiple unit tests.
Edit 2:
Tried some more things...Used IClassFixture<BaseTest> on each test class and then ensured that each test class implemented IDisposable and in there I would ensure the Context database was deleted; also ensured in the test class constructor that it was created. With this I ended up with the following error:
The instance of entity type cannot be tracked because another instance with the same key value for {'Id'} is already being tracked
And so I added .UseQueryTrackingBehavior(QueryTrackingBehavior.NoTracking) but the problem still persisted.
This is very annoying to setup.
This is what has resolved it for me for now.
Summary: Created a new ServiceFixture. This ServiceFixture is applied to a BaseTest class as IClassFixture<ServiceFixture>. The ServiceFixture is responsible for initializing the service collection and allowing for it to be reused across different test classes. The purpose of the BaseTest is to allow for disposal of the database and other clean up that is necessary after each test. The Dispose method of this class will detach entity state and also delete the database.
ServiceFixture.cs
public class ServiceFixture
{
public ServiceProvider ServiceProvider { get; }
public ServiceFixture()
{
var serviceCollection = new ServiceCollection();
serviceCollection.AddScoped<IUserService, UserService>()
.AddScoped<ICertificationService, CertificationService>()
.AddScoped<IOrganizationService, OrganizationService>()
.AddScoped<ICertificateService, CertificateService>()
.AddScoped<IUnitOfWork, UnitOfWork>()
.AddScoped<IUserRepository, UserRepository>()
.AddScoped<IExercisePostRepository, ExercisePostRepository>()
.AddScoped<IEmailRepository, EmailRepository>()
.AddScoped<IPhoneNumberRepository, PhoneNumberRepository>()
.AddScoped<ICertificationRepository, CertificationRepository>()
.AddScoped<ICertificateRepository, CertificateRepository>()
.AddScoped<IOrganizationRepository, OrganizationRepository>()
.AddTransient<IRestClient, RestClient>()
.AddSingleton<NASMCertificationValidator>()
.AddAutoMapper(typeof(Startup).Assembly)
.AddDbContext<SpotcheckrCoreContext>(options =>
options.UseInMemoryDatabase("Spotcheckr-Core")
.UseQueryTrackingBehavior(QueryTrackingBehavior.NoTracking)
.EnableSensitiveDataLogging());
ServiceProvider = serviceCollection.BuildServiceProvider();
}
}
BaseTest.cs
public abstract class BaseTest : IClassFixture<ServiceFixture>, IDisposable
{
protected readonly ServiceProvider ServiceProvider;
protected readonly IUnitOfWork UnitOfWork;
private readonly SpotcheckrCoreContext Context;
public BaseTest(ServiceFixture serviceFixture)
{
ServiceProvider = serviceFixture.ServiceProvider;
Context = serviceFixture.ServiceProvider.GetRequiredService<SpotcheckrCoreContext>();
UnitOfWork = serviceFixture.ServiceProvider.GetRequiredService<IUnitOfWork>();
Context.Database.EnsureCreated();
}
public void Dispose()
{
Context.ChangeTracker.Entries().ToList().ForEach(entry => entry.State = EntityState.Detached);
Context.Database.EnsureDeleted();
}
}
UserServiceTests.cs
public class UserServiceTests : BaseTest
{
private readonly IUserService Service;
public UserServiceTests(ServiceFixture serviceFixture) : base(serviceFixture)
{
Service = serviceFixture.ServiceProvider.GetRequiredService<IUserService>();
}
[Fact]
public async void GetUserAsync_WithInvalidUser_ThrowsException()
{
Assert.ThrowsAsync<InvalidOperationException>(() => Service.GetUserAsync(-1));
}
[Fact]
public void CreateUser_UserTypeAthlete_CreatesAthleteUser()
{
var result = Service.CreateUser(Models.UserType.Athlete);
Assert.IsType<Athlete>(result);
}
[Fact]
public void CreateUser_UserTypePersonalTrainer_CreatesPersonalTrainerUser()
{
var result = Service.CreateUser(Models.UserType.PersonalTrainer);
Assert.IsType<PersonalTrainer>(result);
}
}
I am trying to use Dependency Injection for DB context. I am not sure what i am doing wrong but even after following all the steps i still get the error
Below are the steps that i follow ,suggest me where its going wrong. I am using multi tier project hence my repositories are in my DB access layer and controller in a mvc api application
My DB Context class
public partial class TestDbContext: DbContext
{
public TestDbContext(DbContextOptions<TestDbContext> options)
: base(options)
{
}
public virtual DbSet<Table1> Table1{ get; set; }
}
public interface IRepository<T> where T : class
{
IQueryable<T> GetDbSet();
}
public class Repository<T> : IRepository<T> where T : class
{
protected DbContext _entities;
protected readonly DbSet<T> _dbset;
public Repository(DbContext context)
{
_entities = context;
_dbset = context.Set<T>();
}
public virtual IQueryable<T> GetDbSet()
{
return _dbset;
}
}
pulbic interface IUserRepository
{
List<UsersInfo> GetUsers();
}
public class UserRepository:IUserRepository
{
private readonly IRepository<Table1> table1repo;
public UserRepository(IRepository<Table1> _table1Repo)
{
table1repo = _table1Repo;
}
public List<UsersInfo> GetUsers()
{
return table1repo.GetDbSet().ToList();
}
}
public class MyController : : ControllerBase
{
private readonly IUserRepository _UserRepo;
public MyController (IUserRepository UserRepo)
{
_UserRepo= clientInfo;
}
[HttpGet]
public async Task<IActionResult> Get()
{
try
{
var result = _UserRepo.GetUsers();
return new JsonResult(result) { SerializerSettings = new JsonSerializerSettings() { Formatting = Formatting.Indented } };
}
catch(Exception e)
{
throw e;
}
}
}
Startup.cs
public void ConfigureServices(IServiceCollection services)
{
services.AddSingleton<IConfiguration>(Configuration);
services.Configure<IISOptions>(options =>
{
options.AutomaticAuthentication = false;
});
services.AddDbContext<TestDbContext>(options => options.UseSqlServer(Configuration.GetConnectionString("ConnectionString")));
services.AddScoped<IUserRepository, UserRepository>();
services.AddScoped(typeof(IRepository<>), typeof(Repository<>));
services.AddMvc().SetCompatibilityVersion(CompatibilityVersion.Version_2_2);
}
Your context type in your repository class should be TestDbContext instead of DbContext.
public class Repository<T> : IRepository<T> where T : class
{
protected TestDbContext _entities;
protected readonly DbSet<T> _dbset;
public Repository(TestDbContext context)
{
_entities = context;
_dbset = context.Set<T>();
}
public virtual IQueryable<T> GetDbSet()
{
return _dbset;
}
}
I am about to start a small/medium sized project. I am by no means a software architect. But i tend to question every move i make at times. Since i want to do things correct.
I found a way to implement a simple repository, and i wanted to know if this is a "correct" way of doing it. I came to this solution, since i know what is going on, and not taking in something to complex before i have the knowledge :)
Here it goes.
Unit of work
Where i make sure i to keep all my repositories under the same dbcontext. In my uof i can access all repo's when calling it from the controller.
public class UnitOfWork : IDisposable
{
private ContactRepository _contactRepo;
private ApplicationDbContext _entities;
public UnitOfWork(ApplicationDbContext entities)
{
_entities = entities;
}
public ContactRepository ContactRepo
{
get
{
if (_contactRepo == null)
{
_contactRepo = new ContactRepository(_entities);
}
return _contactRepo;
}
}
public void Save()
{
_entities.SaveChanges();
}
private bool disposed = false;
protected virtual void Dispose(bool disposing)
{
if (!this.disposed)
{
if (disposing)
{
_entities.Dispose();
}
}
this.disposed = true;
}
public void Dispose()
{
Dispose(true);
GC.SuppressFinalize(this);
}
}
ContactRepository
This is a short example of a repository, where it recieves the dbcontext and uses it to grab whatever data i want
public class ContactRepository
{
private ApplicationDbContext _entities;
public ContactRepository(ApplicationDbContext entities)
{
_entities = entities;
}
public IEnumerable<Contact> GetAll()
{
return _entities.Contacts;
}
}
Controller
Short example of controller
public class ContactController : Controller
{
UnitOfWork uow = new UnitOfWork(new ApplicationDbContext());
public ActionResult Index()
{
var contacts = uow.ContactRepo.GetAll();
return View(contacts);
}
protected override void Dispose(bool disposing)
{
uow.Dispose();
base.Dispose(disposing);
}
}
In this way i will have access to all my repositories under the same dbcontext which i was aiming for.
I know things can be done smarter/different. With for example a extendable generic repo. But in this case i am aiming for something simple and understandable. But still dont want to make a huge mistake, if there is a major flaw.
Do you see any major flaws with this way of handling data trough entity framework?
If you're aiming for something simple then just use Entity Framework but if you're going to use the repository pattern I would encourage you to do it properly.
Two of the biggest motivators for using repository are:
You want to simplify CRUD applications to your database.This is
done through the use of interfaces and generics
You want to the ability to test the business logic in isolation
from external dependencies.Again, this is done through the use of
interfaces
Below will take you two minutes to implement but then at least you know you're doing it right, because at the moment you're trying to implement a great pattern in an ineffective way.
Generic interface:
public interface IRepository<T>
{
T GetById(int id);
IEnumerable<T> List();
IEnumerable<T> List(Expression<Func<T, bool>> predicate);
void Add(T entity);
void Delete(T entity);
void Update(T entity);
}
Generic repository:
public abstract class EntityBase
{
}
public class DBRepository<T> : IRepository<T> where T : EntityBase
{
private readonly DbContext _dbContext;
public DBRepository(DbContext dbContext)
{
_dbContext = dbContext;
}
public virtual T GetById(int id)
{
return _dbContext.Set<T>().Find(id);
}
public virtual IEnumerable<T> List()
{
return _dbContext.Set<T>().AsEnumerable();
}
public virtual IEnumerable<T> List(System.Linq.Expressions.Expression<Func<T, bool>> predicate)
{
return _dbContext.Set<T>()
.Where(predicate)
.AsEnumerable();
}
public void Add(T entity)
{
_dbContext.Set<T>().Add(entity);
}
public void Update(T entity)
{
_dbContext.Entry(entity).State = EntityState.Modified;
}
public void Delete(T entity)
{
_dbContext.Set<T>().Remove(entity);
}
Unit of work:
public class UnitOfWork : IDisposable
{
private bool disposed = false;
private ApplicationDbContext context = new ApplicationDbContext();
private IRepository<Contact> _contactRepository;
public IRepository<Contact> ContactRepository
{
get
{
if (this._contactRepository == null)
this._contactRepository = new DBRepository<Contact>(context);
return _contactRepository;
}
}
public void Save()
{
context.SaveChanges();
}
protected virtual void Dispose(bool disposing)
{
if (!this.disposed)
{
if (disposing)
context.Dispose();
}
this.disposed = true;
}
public void Dispose()
{
Dispose(true);
GC.SuppressFinalize(this);
}
}
I created an EntitySetController that looks like this:
public class OrdersController : EntitySetController<Order,Guid>
{
private readonly PizzaCompanyEntities _context = Factories.DataFactory.GetPizzaContext();
protected override void Dispose(bool disposing)
{
if (disposing)
{
_context.Dispose();
}
base.Dispose(disposing);
}
public override IQueryable<Order> Get()
{
return _context.Orders;
}
protected override Order GetEntityByKey(Guid key)
{
var result = _context.Orders.FirstOrDefault(o => o.Id == key);
if (result == null)
{
throw new HttpResponseException(HttpStatusCode.NotFound);
}
return result;
}
}
In an existing MVC 4 web application.
I configure the route as follows:
public static class WebApiConfig
{
public static void Register(HttpConfiguration config)
{
config.Routes.MapODataRoute("PizzaApi", "odata", GetImplicitEdm());
}
private static IEdmModel GetImplicitEdm()
{
var builder = new ODataConventionModelBuilder();
builder.EntitySet<Order>("Orders");
builder.EntitySet<Pizza>("Pizzas");
builder.EntitySet<Pizzas_To_Orders>("PizzasToOrders");
builder.EntitySet<Size>("Sizes");
builder.EntitySet<Status>("Statuses");
builder.EntitySet<Pizzas_To_Toppings>("PizzasToToppings");
return builder.GetEdmModel();
}
}
And execute the configuration as follows:
public class MvcApplication : System.Web.HttpApplication
{
protected void Application_Start()
{
WebApiConfig.Register(GlobalConfiguration.Configuration);
FilterConfig.RegisterGlobalFilters(GlobalFilters.Filters);
RouteConfig.RegisterRoutes(RouteTable.Routes);
BundleConfig.RegisterBundles(BundleTable.Bundles);
}
}
But when I execute my route at http://localhost:29064/odata/Orders I am getting a 404 and a message "The controller for path /odata/Orders was not found or does not implement IController.
I cannot figure out what I am missing to get the route registered and the controller running. I have done a similar application from scratch and have not had this trouble.
How do I get my OData route working?