I apologize if this has been answered in a previous post; I just couldn't find the answer specific to my question.
I was watching a MVC tutorial video and in one of their controllers, they kept instantiating a repository in every method that they needed it for. I'm used to this from previous ASP.NET DAL, but they made this comment:
... I shouldn't be instantiating the repository in every method, I
should be making use of MVC's Services...
I'm curious to know what this means exactly and if instantiating a repository in my controller methods where I need it will be bad practice.
I'm not an expert, but I think you should instantiate your repository only once in the Controller's constructor like this:
public interface ISampleRepository { }
public class SampleRepository : ISampleRepository { }
public class HomeController : Controller
{
private ISampleRepository myRepository;
public HomeController()
{
myRepository = new SampleRepository();
}
}
You don't have to instance your repository in every Action Method. I'm not sure but I think every time you instantiate your repository a new Database Context is created.
So, instantiating the same repository multiple times seems overkilling to me.
Related
I'm creating a Web API and I'm using dependency inject wit Ninject.
I have:
IRTWRepository
IModelFactory
I'm injecting those 2 into my controllers like this.
private static void RegisterServices(IKernel kernel)
{
kernel.Bind<IRTWRepository>().To<RTWRepository>();
kernel.Bind<RTWAPIContext>().To<RTWAPIContext>();
kernel.Bind<IModelFactory>().To<ModelFactory>();
}
My RTWRepository constructor looks like this
public class RTWRepository :IRTWRepository
{
private RTWAPIContext _context;
public RTWRepository(RTWAPIContext context)
{
_context = context;
}}
My ModelFactory constructor looks like this
public class ModelFactory : IModelFactory
{
private IRTWRepository _repo;
public ModelFactory(IRTWRepository repo)
{
_repo = repo;
}
}
I have a Controller that looks like this:
public MaterialsController(IRTWRepository repository,IModelFactory modelFactory)
: base(repository,modelFactory)
{
}
Now, my question is : Is Ninject creating 2 separate contexts when creating an instance of my RTWRepository and also when creating an instance of ModelFactory?.
The reason that I'm asking that is because I'm having a problem when I try to save an entity that has a dependency to another object which was previously retrieve from the db.
I'm saving the entity in my controller but I'm creating it in my model factory along with is dependency.
public class RecycleCenter
{
public RecycleCenter()
{
}
public int MyProperty { get; set; }
[Required]
public virtual Address Address { get; set; }
}
The code above is for the entity Recycle Center which has an Address, this recycle center entity is created in my model factory and then in my controller I try to save it but when my repository execute this line
_context.RecycleCenters.Add(entity);
I'm getting this error
An entity object cannot be referenced by multiple instances of IEntityChangeTracker
So, somewhere in my code I'm using 2 context instead of 1 and I think is when creating the ModelFactory and RTWRepository, is this assumption correct?, if so how do I fix it?
TL;DR;
You probably need to change this line:
kernel.Bind<RTWAPIContext>().To<RTWAPIContext>();
to
kernel.Bind<RTWAPIContext>().To<RTWAPIContext>().InRequestContext();
Explanation:
When you define a binding in Ninject, you also specify how that object's lifecycle should be handled.
If you don't explicitly define it, Ninject's default lifecycle is Transient. Transient means that each time an instance is required, it will create a new one. In your case, you need to two instances: one for the RTWRepository of the ModelFactory and one for the RTWRepository of the MaterialsController.
You can modify the lifestyle to one of these options:
Singleton ==> kernel.Bind<RTWAPIContext>().To<RTWAPIContext>().InSingleTonScope();
Request ==> kernel.Bind<RTWAPIContext>().To<RTWAPIContext>().InRequestScope();
Thread ==> kernel.Bind<RTWAPIContext>().To<RTWAPIContext>().InThreadScope();
Named, Call, Parent, Custom
In your case, I think you need InRequestScope, but you have to check the necessary lifecycle as it depends on the application.
For further information please check out the documentation here: https://github.com/ninject/ninject/wiki/Object-Scopes
Most probably, it is. There's no annotation that is telling to Ninject "Hey, stop, when you have created the instance once, reuse it". You should agree that in most cases, you would want multiple instances of an object and that it is a rare case, where you want it only once.
If you want to reuse the instance, use the singleton pattern. Ninject is familiar with it, so you can bind the object mapping to a method
kernel.Bind<RTWAPIContext>().ToMethod(c => RTWAPIContext.GetInstance());
There is also a ToSingleton binding, but I bet you cannot make your context constructor private and implement C# specific singleton due to other ASP.NET problems (e.g. ASP.NET Identity will try to invoke the context's method for object creation).
We have a Disposable property in a controller is needed in other controllers too. For example, suppose the following property in a controller:
public ClassName<Template> Name { get; set; }
Which will be instantiated in the same controller's constructor like this:
public SomeController()
: this(new ClassName<Template>())
{
}
public SomeController(ClassName<Template> name)
{
Name = name;
}
Can I use the property by instantiating the container controller in other controllers?
If your question is: can I create an instance of the controller and use it in another one, the answer is yes, controllers are just classes which you can instantiate and use. Just like any other.
The thing that concerns me though, is that it seems you are using a property that should be a proper singleton as something that isn't, by re-instantiating another class that seems slightly related. If you can, move the property to an other class. Make it singleton, and use it accordingly.
I just have a quick question. Im trying to use Unity with my asp.net MVC project. Im coming across a problem when using the Unit of Work pattern with an EF context.
Say i inject the uow in the constructor, but have 4 or 5 actions in the controller that need to use the UnitOfWork in a using statement. This isnt going to work! Because Id have to do a
new UnitOfWork() in each action method.
should i be injecting a UnitOfWork into each action method? or into just the constructor? or should I even be injecting this at all!! The problem im facing is that i want to be able to unit test my controller with Mock data, and i can only do this if I inject the UnitOfWork or the DBContext.
Inject factory instead. This way you still achieve separation of concerns and loose coupling, yet you won't face any issues with using statements:
private IUnitOfWorkFactory factory;
public MyController(IUnitOfWorkFactory factory)
{
this.factory = factory;
}
public ActionResult MyAction()
{
using (var uow = factory.CreateUnitOfWork())
{
// ...
}
}
Edit:
Natural advantage of such approach is its configurability - you can register whichever factory you like to serve different controllers and wire it up at composition root:
// Note: this isn't unity syntax, but I hope my point is clear
container.Register<ISessionFactory, ReusableSessionFactory>("Reusable");
container.Register<ISessionFactory, FreshSessionFactory>("Fresh");
container.Register<IController, LoginController>().With("Fresh");
container.Register<IController, HomeController>().With("Reusable");
Now,
LoginController will use factory that under the hood serves new session upon each request
HomeController on the other hand will reuse the same session for all its lifespan
It's worth noting that from the controller point of view, it's irrelevant which factory serves the session as it's a mere implementation detail. That's why we hide session factory dependency behind abstraction (interface in this example) and perform all the object-to-dependency binding at application's root.
If I understand correctly you simply want to be able to test the UOW with something like Moq?
In that case for good design principles and proper separation of concerns you should create a base context for your database that each repository class uses.
Then you should create a repository interface for each domain model entity. Then you can implement the interface in a seperate repository library (this way you can implement a POCO model)
Finally you either create a service layer between your domain objects and your action methods or just use the required repository interfaces within the action methods.
I answer it like this because it depends on your application infrastructure. If you have no service layer then the best practice is to do the following:
public class AccountController : Controller
{
private readonly IAccountRepository _accountrepository;
public AccountController(IAccountRepository repository)
{
_accountrepository = repository;
}
}
I hope this helps.
I have a base class that has a repository injected into the constructor, now any class that I derive off of it now also needs those parameters, but the repository is only used by the base class, is there a way to bind Ninject to just the base class and not go through the constructor? And most importantly, is it a good idea?
public class HtmlPageModel
{
private readonly IHtmlPageRepository _repository;
public HtmlPageModel (IHtmlPageRepository repository)
{
_repository = repository;
}
}
public class VideoPageViewModel : HtmlPageModel
{
public VideoPageViewModel(IHtmlPageRepository repository) : base(repository)
{
}
}
View models shouldn't have dependencies at all. Inject the repository into the controller and assign the values from there.
Also if many pages are using the same base view model this indicates that some part of the page is shown in many situations or even all the time. In this case it is better to have a custom controller and view for this area and use Html.RenderAction to render this part.
I still stand to my previous comment: You have to rethink your class hierarchy, this is not something you should work around with Ninject.
If your base class accepts a repository, so should all of your derived classes. As an alternative you could however inject a specially NullRepository into your VideoPageViewModel which basically does nothing (see Null Object pattern)
Binding by target type can be achieved using WhenInjectedInto() in Ninject:
kernel.Bind<IHtmlPageRepository>()
.To<HtmlPageRepository>()
.WhenInjectedInto<HtmlPageModel>();
kernel.Bind<IHtmlPageRepository>()
.To<NullRepository>()
.WhenInjectedInto<VideoPageViewModel>();
This is basically the same answer as brokenGlass. Why not just create another base class, and move the attributes/functionality from the your current base class into that one, but don't move the constructor/functionality that relates to the repository.
I'm trying to work out a way of passing the web current http context to a service class (or initialising the class with a reference to it). I am doing this to abstract the rest of the app away from needing to know anything about the http context.
I also want the service to be testable using TDD, probably using one of the Mockable frameworks. Hence it would be preferable to use an interface rather than an actual class.
An example of what I'd like to achieve:
class WebInstanceService
{
private IHttpContext _Context;
public WebInstanceService( ... , IHttpContext HttpContext )
{
....
_Context = HttpContext;
}
// Methods...
public string GetInstanceVariable(string VariableName)
{
return _Context.Current.Session[VariableName];
}
}
One of the main issues I have is that there is no IHttpContext, the .net http context is a subclass of an abstract class which can't be mocked (easily?).
Another issue is that I can't initialise global instances of the class as then the context won't be relevant for most requests.
I could make the class static, and require the Context to be passed to each function as it is called i.e.
public static string GetInstanceVariable(string VariableName, HttpContext Context)
{ ... }
but this doesn't make the class any easier to test, I still need to create an HttpContext and additionally any non-web-aware services which want to use this class suddenly need to be able to retrieve the Context requiring them to be closely coupled to the web server - the whole reason for wanting to create this class in the first place.
I'm open to ALL suggestions - particularly those which people know facilitate easy tdd testing. How would people suggest I tackle this problem?
Cheers
This is why HttpContextBase and HttpContextWrapper were introduced. You probably want to use HttpContextBase and when passing the real context in, use new HttpContextWrapper( httpContext ), although, I think that what is available to you in the controller is already of type HttpContextBase. I would create one of these in my controller each time rather than trying to reference the current context from the static, global HttpContext.Current instance. If you need it in your view, pass a reference to your strongly typed context in ViewData.
I mock up HttpContextBase frequently in my tests.
class WebInstanceService
{
private HttpContextBase _Context;
public WebInstanceService( ... , HttpContextBase HttpContext )
{
....
_Context = HttpContext;
}
// Methods...
public string GetInstanceVariable(string VariableName)
{
return _Context.Session[VariableName];
}
}
What we do is spin one of these up http://haacked.com/archive/2007/06/19/unit-tests-web-code-without-a-web-server-using-httpsimulator.aspx
Easy as pie, just instanciate an HttpSimulator and fill in the values, and HttpContext.Current gets filled up with whatever you specify.
IHttpContext is something that is in MVC, and aparently one day will be in webforms. Hopefully that day will be .net 4
ASP.NET comes with System.Web.Abstractions that include HttpContextBase that you can use for dealing with the HttpContext in a testing situation.
I would personally abstract away the direct dependency on the HttpContext.