I got the following block
container.RegisterType<IService, ServiceA>("a");
container.RegisterType<IService, ServiceB>("b");
I want to have a Dictionary of type Dictionary<string,IService>.
I will receive service name by parameter in an API rest and my idea is based on that parameter get the implementation I need from the Dictionary.
I can't figure out how to inject the Dictionary (with the resolved classes inside) into my business class.
I want to do something like this.
private readonly IDictionary<string,IService> serviceDictionary;
public ClassConstructor (IDictionary<string,IService> dictionary)
{
this.serviceDictionary = dictionary;
}
You should not inject IDictionary<string,IService> into your component, but instead an application-tailored abstraction:
public interface IServiceProvider
{
IService GetService(string key);
}
This way you can create an implementation for Unity as follows:
public class UnityServiceProvider : IServiceProvider
{
public IUnityContainer Container { get; set; }
public IService GetService(string key) => Container.Resolve<IService>(key);
}
Now you can complete your registration as follows:
container.RegisterType<IService, ServiceA>("a");
container.RegisterType<IService, ServiceB>("b");
container.RegisterInstance<IService>(new UnityServiceProvider { Container = container });
container.RegisterType<ClassConstructor>();
Related
What is the best practice for placing business logic in message based design?
Im using servicestack for building my api.
The wiki shows the example of placing the RequiredRole Attribute on the message instead of the service handling it.
In a sense this [RequiredRole]/[Authenticate] is business logic/security attached to the message.
Concrete example
Say for example i would add DeleteAddress message:
public class DeleteAddress : IReturn<bool>
{
public int AddressId { get; set; }
}
But for this to be properly secure i want to check either Admin Role, permission to ManageAllAddresses or that the AddressId is linked to this user (maybe in session, maybe through a db call).
How would i best go about this?
Proposition
Is the following code the good practice and if so how would i implement it?
[RequiredRole("Admin")]
[RequiredPermission("ManageAllAddresses ")]
[RequiredAddressLinkedToAccount]
public class DeleteAddress : IReturn<bool>
{
public int AddressId { get; set; }
}
ServiceStack's recommendation is to keep your ServiceModel free of dependencies so we'd recommend to annotate your Service implementation classes instead which you can annotate either on the Service class to apply to all Operations or on the individual methods to apply just to that operation, e.g:
[RequiredRole("Admin")]
public class AddressServices : Service
{
[RequiredPermission("ManageAllAddresses ")]
[RequiredAddressLinkedToAccount]
public object Any(DeleteAddress request)
{
}
}
Please note ServiceStack requires your Services to return reference types, which is typically a Response DTO but can also be a string, e.g:
public class DeleteAddress : IReturn<string>
{
public int AddressId { get; set; }
}
To finish of this question. I could make a request filter and add it on the service.
Either inherit from AuthenticateAttribute or Directly from RequestFilterAttribute.
public class RequiredAddressLinkedToAccount : AuthenticateAttribute
{
public RequiredRoleAttribute(ApplyTo applyTo)
{
this.ApplyTo = applyTo;
this.Priority = (int)RequestFilterPriority.RequiredRole;
}
public override void Execute(IRequest req, IResponse res, object requestDto)
{
var dto = requestDto as ILinkedToAccount;
var session = req.GetSession();
if(dto.AccountId == session.Id)
return; //we dont want anything to be blocked if the account Id is there.
//Implement like RequireRoleAttribute
if (DoHtmlRedirectIfConfigured(req, res))
return;
res.StatusCode = (int)HttpStatusCode.Forbidden;
res.StatusDescription = "Address does not belong to you";
res.EndRequest();
}
}
Given below are my different class declarations and how I am trying to setup unity container configuration to get a Interface to Concrete class implementation. The code currently throws either an stackoverflow exception or suggests that a interface cannot be constructed.
Please help me fix, either the class structure or the container configuration.
CodesController Class -
public class CodesController : ApiController
{
private readonly IUnitOfWorkAsync unitOfWork;
private readonly ICodeRepository repository;
public CodesController(IUnitOfWorkAsync unitOfWork, ICodeRepository codeRepository)
{
if (unitOfWork == null)
{
throw new ArgumentNullException("unitOfWork");
}
this.unitOfWork = unitOfWork;
this.repository = codeRepository;
}
//Other class level methods here
}
CodeRepository class -
public class CodeRepository : ICodeRepository
{
private readonly ICodeRepository codeRepository;
public CodeRepository(ICodeRepository repository)
{
this.codeRepository = repository;
}
public virtual async Task<IEnumerable<Code>> GetCodeAsync(string codeKey)
{ //Some implementation here}
}
ICodeRepository Interface -
public interface ICodeRepository : IRepositoryAsync<Code>
{
Task<IEnumerable<Code>> GetCodeAsync(string codeKey);
}
IRepositoryAsync Interface -
public interface IRepositoryAsync<TEntity> : IRepository<TEntity> where TEntity : class, IPersistenceHint
{
Task<bool> DeleteAsync(params object[] keyValues);
Task<bool> DeleteAsync(CancellationToken cancellationToken, params object[] keyValues);
Task<TEntity> FindAsync(params object[] keyValues);
Task<TEntity> FindAsync(CancellationToken cancellationToken, params object[] keyValues);
}
Unity Container Configuration-
container.RegisterType<IUnitOfWorkAsync, UnitOfWork>(
"test",
new TransientLifetimeManager(),
new InjectionConstructor(container.Resolve<IDataContextAsync>("test")));
container.RegisterType<ICodeRepository, CodeRepository>();
container.RegisterType<CodesController, CodesController>();
With this given configuration and class structure, based on my experimentation with container config, I get following exception -
JSON
exceptionMessage=An error occurred when trying to create a controller of type 'CodesController'. Make sure that the controller has a parameterless public constructor.
exceptionType=System.InvalidOperationException
innerException
exceptionMessage=Type '<Namespace>.Api.Controllers.CodesController' does not have a default constructor
stackTrace= at System.Linq.Expressions.Expression.New(Type type)
at System.Web.Http.Internal.TypeActivator.Create[TBase](Type instanceType)at System.Web.Http.Dispatcher.DefaultHttpControllerActivator.GetInstanceOrActivator(HttpRequestMessage request, Type controllerType, Func`1& activator)
at System.Web.Http.Dispatcher.DefaultHttpControllerActivator.Create(HttpRequestMessage request, HttpControllerDescriptor controllerDescriptor, Type controllerType)
Please suggest, if anything is wrong here, so that I can fix the same. Already struggling many days on this.
You're injecting ICodeRepository to CodeRepository, which probably causes to stackoverflow exception, since it will keep generating ICodeRepositories. It will generate a recursive call. Somewhat like this one:
public class BaseFoo
{
public BaseFoo(BaseFoo foo){ }
}
public class Foo : BaseFoo
{
public Foo() : base(new Foo()) { }
}
And regarding the "does not have a default constructor"-exception, have you registered a DependencyResolver for Web API? See one of these questions for more detailed information how to do it:
Using Unity with Web Api 2 gives error does not have a default constructor
Unity.WebApi | Make sure that the controller has a parameterless public constructor
ASP.Net MVC 4 Web API controller dosn't work with Unity.WebApi
As a side note, you shouldn't have to register the CodesController in your unity registration.
I am currently working on a DDD-based application using Unity IOC container and need a way to pass my custom Principal object to the repository and service layers that would allow for unit testing. How should this be done? My current thoughts are to create a property on the service and repository classes of type IPrincipal. Then use Unity on Application_Start to set and pass in the Principal.
For one, am I on the right track in my thinking?
Two, if not at application_start, which seems like that is not the right place since I need a person to login first before the injections occur, where should this occur?
Three, for Unity, what should the container.RegisterType look like for getting the Principal from Thread.CurrentPrincipal or HttpContext.Current.User?
You can have a PrincipalDto class that will contain the relevant IPrincipal properties you need to use in your Service layer and map the values from the IPrincipal to the PrincipalDto. This way you do not need to include the reference assembly of IPrincipal to the other layers.
Below is an example that uses auto mapping.
public class PrincipalDto
{
public UserId { get; set; }
public Username { get; set; }
public RoleId { get; set; }
}
public class SomeService
{
public void SomeServiceMethod(PrincipalDto principal)
{
// do work here
}
}
public class SomeConsumer()
{
public void SomeConsumerMethod()
{
// where User is the IPrincipal object instance
var principal = Mapper.Map<PrincipalDto>(User);
var service = new Service();
service.SomeServiceMethod(principal);
}
}
If I'm trying to serialize a normal CLR object, and I do not want a particular member variable to be serialized, I can tag it with the
[NonSerialized]
attribute. If I am creating a table services entity, is there an equivalent attribute I can use to tell Azure table services to ignore this property?
For Version 2.1 there is a new Microsoft.WindowsAzure.Storage.Table.IgnoreProperty attribute. See the 2.1 release notes for more information: http://blogs.msdn.com/b/windowsazurestorage/archive/2013/09/07/announcing-storage-client-library-2-1-rtm.aspx.
There's no equivalent I know of.
This post says how you can achieve the desired effect - http://blogs.msdn.com/b/phaniraj/archive/2008/12/11/customizing-serialization-of-entities-in-the-ado-net-data-services-client-library.aspx
Alternatively, if you can get away with using "internal" rather than "public" on your property then it will not get persisted with the current SDK (but this might change in the future).
For version 2.0 of the Table Storage SDK there is a new way to achieve this.
You can now override the WriteEntity method on TableEntity and remove any entity properties that have an attribute on them. I derive from a class that does this for all my entities, like:
public class CustomSerializationTableEntity : TableEntity
{
public CustomSerializationTableEntity()
{
}
public CustomSerializationTableEntity(string partitionKey, string rowKey)
: base(partitionKey, rowKey)
{
}
public override IDictionary<string, EntityProperty> WriteEntity(Microsoft.WindowsAzure.Storage.OperationContext operationContext)
{
var entityProperties = base.WriteEntity(operationContext);
var objectProperties = this.GetType().GetProperties();
foreach (PropertyInfo property in objectProperties)
{
// see if the property has the attribute to not serialization, and if it does remove it from the entities to send to write
object[] notSerializedAttributes = property.GetCustomAttributes(typeof(NotSerializedAttribute), false);
if (notSerializedAttributes.Length > 0)
{
entityProperties.Remove(property.Name);
}
}
return entityProperties;
}
}
[AttributeUsage(AttributeTargets.Property)]
public class NotSerializedAttribute : Attribute
{
}
Then you can make use of this class for your entities like
public class MyEntity : CustomSerializationTableEntity
{
public MyEntity()
{
}
public string MySerializedProperty { get; set; }
[NotSerialized]
public List<string> MyNotSerializedProperty { get; set; }
}
We have a class Event (it's actually named differently, but I'm just making abstraction):
public class Event
{
public string Name { get; set; }
public string Description { get; set; }
public EventType EventType { get; set; }
}
We need to build an instance of a Message class with this object, but depending on the EventType, we use a different builder:
switch (event.EventType)
{
case EventType.First:
message = FirstMessageBuilder.Build(event);
break;
case EventType.Second:
message = SecondMessageBuilder.Build(event);
break;
}
Do you think this is acceptable, or should we take the following approach:
Make an abstract class:
public class Event
{
public string Name { get; set; }
public string Description { get; set; }
public abstract Message BuildMessage();
}
Then derive two classes: class FirstMessage and class SecondMessage and make the domain objects responsible for building the message.
I hope it isn't too abstract. The bottom line is we need to transform one class to another. A simple mapper won't do, because there are properties with XML content and such (due to a legacy application making the events). Just accept what we're trying to do here.
The real question is: can a domain object be responsible for such a transformation, or would you not recommend it? I would avoid the ugly switch statement, but add complexity somewhere else.
Whilst I agree with Thomas, you might want to look at the following design patterns to see if they help you:
Vistor Pattern
Double-Dispatch Pattern
Builder Pattern
Strictly speaking, a domain object shouldn't be responsible for anything other than representing the domain. "Changing type" is clearly a technical issue and should be done by some kind of service class, to maintain a clear separation of concerns...
In order to gain the readability of
var message = eventInstance.AsMessage();
as well following the single responsibility principle, you could define AsMessage() as an extension method of the event type.
There are few possible solutions. To use abstract factory:
public interface IMessageFactory
{
Message Create();
}
public class FirstMessageFactory : IMessageFactory
{
public Message Create()
{
//...
}
}
public class SomeService
{
private readonly IMessageFactory _factory;
public SomeService(IMessageFactory factory)
{
_factory = factory;
}
public void DoSomething()
{
var message = _factory.Create();
//...
}
}
Now you can wire IoC container to right factory for requested service.
To use Assembler which makes the transformation:
public interface IAssembler<TSource, TDestination>
{
TDestination Transform(TSource source);
}
This is quite similar to factory pattern, but if you are dependent on EventType, its possible to do it like:
public interface IAssembler<TEventType>
{
object Transform(object source);
}
I would encapsulate the logic into a separate Factory/Builder class, and use an extension method on Event to call the builder.
This would give you the best of both worlds.