Suppose I have an interface for fetching data, and an implementation of it:
interface IResourceProvider
{
string Get( Uri uri );
}
class HttpResourceProvider : IResourceProvider
{
public string Get( Uri uri )
{
// Download via HTTP.
}
}
I can register this in Castle Windsor as follows:
container.Register
( Component.For<IResourceProvider>().ImplementedBy<HttpResourceProvider>()
);
(Which is all fine).
If I then decided I wanted a caching implementation as follows:
class CachingResourceProvider : IResourceProvider
{
public CachingResourceProvider( IResourceProvider resourceProvider )
{
_resourceProvider = resourceProvider;
}
public string Get( Uri uri )
{
// Return from cache if it exists.
// Otherwise use _resourceProvider and add to cache.
}
private readonly IResourceProvider _resourceProvider;
}
How would I register these nested dependencies? i.e., I want to say an IResourceProvider is implemented by a CachingResourceProvider, except where in the constructor, where it's a HttpResourceProvider.
Just register CachingResourceProvider before HttpResourceProvider - e.g.
container.Register(Component
.For<IResourceProvider>()
.ImplementedBy<CachingResourceProvider>());
container.Register(Component
.For<IResourceProvider>()
.ImplementedBy<HttpResourceProvider>());
BTW - this is know as the Decorator design pattern.
This is known as the Decorator Pattern...
See this link as an example: http://mikehadlow.blogspot.com/2010/01/10-advanced-windsor-tricks-4-how-to.html
Related
I have a small class which I am using to make sure the strings sent and received by a service remain URL safe without additional encoding (see below).
Ideally I would like to just apply this type to my DTOs and have Servicestack be smart enough to use the implicit operators.
public class MyDto {
Base64UrlString myString;
}
var dto = new MyDto() { myString = "hello i am url safe"; }
On the client this is received as myString: {}
Is there a more elegant way to do this? I had hoped applying a type this way would "just work"
// used only for Base64UrlEncoder
using Microsoft.IdentityModel.Tokens;
namespace MyDto.ServiceModel.Types
{
public class Base64UrlString
{
private readonly string _base64UrlString;
public Base64UrlString(string str)
{
_base64UrlString = Base64UrlEncoder.Encode(str);
}
public static implicit operator string(Base64UrlString base64UrlString) => base64UrlString.ToString();
public static implicit operator Base64UrlString(string str) => new(str);
public override string ToString() => Base64UrlEncoder.Decode(_base64UrlString);
}
}
You'll need to change your class to a struct to make use of the custom struct behavior you're trying to use in your example.
Also ServiceStack.Text serializers only serializes public properties by default so your DTO should use public properties:
public class MyDto {
public Base64UrlString MyString { get; set; }
}
Alternatively you can configure it to serialize public fields with:
JsConfig.Init(new Config {
IncludePublicFields = true
});
I am trying to inject the IApplicationConfigurationSection implementation into this MVC5 Controller, so that I can have access to some of the information (various strings) from my web.config custom section in all of my views:
public class BaseController : Controller
{
public IApplicationConfigurationSection AppConfig { get; set; }
public BaseController()
{
ViewBag.AppConfig = AppConfig; // AppConfig is always null
}
}
I want to use setter injection so I don't have to clutter up my derived Controller constructors with parameters that they don't really care about.
Note: If there is a better way to inject base class dependencies, please let me know. I admit I may not be on the right track here.
In my Global.asax I load my StructureMap configurations:
private static IContainer _container;
protected void Application_Start()
{
_container = new Container();
StructureMapConfig.Configure(_container, () => Container ?? _container);
// redacted other registrations
}
My StructureMapConfig class loads my registries:
public class StructureMapConfig
{
public static void Configure(IContainer container, Func<IContainer> func)
{
DependencyResolver.SetResolver(new StructureMapDependencyResolver(func));
container.Configure(cfg =>
{
cfg.AddRegistries(new Registry[]
{
new MvcRegistry(),
// other registries redacted
});
});
}
}
My MvcRegistry provides the mapping for StructureMap:
public class MvcRegistry : Registry
{
public MvcRegistry()
{
For<BundleCollection>().Use(BundleTable.Bundles);
For<RouteCollection>().Use(RouteTable.Routes);
For<IPrincipal>().Use(() => HttpContext.Current.User);
For<IIdentity>().Use(() => HttpContext.Current.User.Identity);
For<ICurrentUser>().Use<CurrentUser>();
For<HttpSessionStateBase>()
.Use(() => new HttpSessionStateWrapper(HttpContext.Current.Session));
For<HttpContextBase>()
.Use(() => new HttpContextWrapper(HttpContext.Current));
For<HttpServerUtilityBase>()
.Use(() => new HttpServerUtilityWrapper(HttpContext.Current.Server));
For<IApplicationConfigurationSection>()
.Use(GetConfig());
Policies.SetAllProperties(p => p.OfType<IApplicationConfigurationSection>());
}
private IApplicationConfigurationSection GetConfig()
{
var config = ConfigurationManager.GetSection("application") as ApplicationConfigurationSection;
return config; // this always returns a valid instance
}
}
I have also "thrown my hands up" and tried using the [SetterProperty] attribute on the BaseController - that technique failed as well.
Despite my best efforts to find a solution, the AppConfig property in my controller's constructor is always null. I thought that
`Policies.SetAllProperties(p => p.OfType<IApplicationConfigurationSection>());`
would do the trick, but it didn't.
I have found that if I discard setter injection and go with constructor injection, it works as advertised. I'd still like to know where I'm going wrong, but I'd like to stress that I'm not a StructureMap guru - there may be a better way to avoid having to constructor-inject my base class dependencies. If you know how I should be doing this but am not, please share.
While constructor injection in this scenario appears to be the better solution to the stated problem as it follows The Explicit Dependencies Principle
Methods and classes should explicitly require (typically through method parameters or constructor parameters) any collaborating objects they need in order to function correctly.
The mention of only needing to access the AppConfig in your views leads me to think that this is more of an XY problem and a cross cutting concern.
It appears that the controllers themselves have no need to use the dependency so stands to reason that there is no need to be injecting them into the controller explicitly just so that the dependency is available to the View.
Consider using an action filter that can resolve the dependency and make it available to the View via the same ViewBag as the request goes through the pipeline.
public class AccessesAppConfigAttribute : ActionFilterAttribute {
public override void OnActionExecuting(ActionExecutingContext filterContext) {
var resolver = DependencyResolver.Current;
var appConfig = (IApplicationConfigurationSection)resolver.GetService(typeof(IApplicationConfigurationSection));
filterContext.Controller.ViewBag.AppConfig = appConfig;
}
}
This now makes the required information available to the views with out tight coupling of the controllers that may have a use for it. Removing the need to inject the dependency into derived classes.
Either via adorning Controller/Action with the filter attribute
[AccessesAppConfig] //available to all its actions
public class HomeController : Controller {
//[AccessesAppConfig] //Use directly if want to isolate to single action/view
public ActionResult Index() {
//...
return View();
}
}
or globally for all requests.
public class FilterConfig {
public static void RegisterGlobalFilters(GlobalFilterCollection filters) {
filters.Add(new AccessesAppConfigAttribute());
}
}
At this point it really does not matter which IoC container is used. Once the dependency resolver has been configured, Views should have access to the required information in the ViewBag
Let's say we have a situation where a service would call other services in ServiceStack.
From reading around, this is how one would call another service:
public class CompanyService : Service
{
public SetupCompanyResponse Any(SetupCompany request)
{
var employeeService = base.ResolveService<EmployeeService>();
// Do something with employeeService
var response = employeeService.Any(new SetupEmployees());
return new SetupCompanyResponse { NumOfEmployeesCreated = response.Count };
}
}
Question: How do I mock EmployeeService if I'm unit-testing CompanyService?
Easiest way I could think of is to generate an IEmployeeService interface so that it's easily mockable. However I'm not sure if base.ResolveService<T> will be able to properly resolve and auto-wire a ServiceStack service, given its interface instead, like so:
var employeeService = base.ResolveService<IEmployeeService>();
Especially when we are registering services this way (which I assume is configuring the object resolution based on concrete class, and not the interface e.g IEmployeeService)
public HelloAppHost() : base("Hello Web Services", typeof(HelloService).Assembly) { }
Update:
Apparently I'm able to somewhat achieve this simply by:
1) Registering the service interface with its implementation
public static void SetupServices(Container container)
{
container.RegisterAs<EmployeeService, IEmployeeService>();
}
2) Using the same exact code to resolve, except now I pass in the interface. The service is successfully resolved.
var employeeService = base.ResolveService<IEmployeeService>();
3) All I need to do now is to override the resolver, and the service dependency should be totally mockable.
Question: Is this also a valid approach? Why or why not?
ServiceStack's Service class resolves all its dependencies from an IResolver, defined by:
public interface IResolver
{
T TryResolve<T>();
}
This can be injected in ServiceStack's Service class in an number of ways as seen by the implementation:
public class Service : IService, IServiceBase, IDisposable
{
public static IResolver GlobalResolver { get; set; }
private IResolver resolver;
public virtual IResolver GetResolver()
{
return resolver ?? GlobalResolver;
}
public virtual Service SetResolver(IResolver resolver)
{
this.resolver = resolver;
return this;
}
public virtual T TryResolve<T>()
{
return this.GetResolver() == null
? default(T)
: this.GetResolver().TryResolve<T>();
}
...
}
Which you can use to control how Services resolve dependencies.
An easy way to unit test Services is to use a BasicAppHost as seen on the Testing wiki, i.e:
appHost = new BasicAppHost().Init();
var container = appHost.Container;
container.Register<IDbConnectionFactory>(
new OrmLiteConnectionFactory(":memory:", SqliteDialect.Provider));
container.RegisterAutoWired<CompanyService>();
container.RegisterAutoWiredAs<StubEmployeeService, EmployeeService>();
Where StubEmployeeService is your stub implementation of EmployeeService, e.g:
public class StubEmployeeService : EmployeeService
{
public SetupEmployeesResponse Any(SetupEmployees request)
{
return new SetupEmployeesResponse { ... };
}
}
You can also register Services using any of the registration methods ServiceStack's IOC Supports if you prefer to use your own mocking library.
I'm struggling with implementing a factory object. Here's the context :
I've in a project a custom store. In order to read/write records, I've written this code in a POCO model/separated repository:
public class Id { /* skip for clarity*/} // My custom ID representation
public interface IId
{
Id Id { get; set; }
}
public interface IGenericRepository<T> where T : IId
{
T Get(Id objectID);
void Save(T #object);
}
public interface IContext
{
TRepository GetRepository<T, TRepository>()
where TRepository : IGenericRepository<T>
where T:IId;
IGenericRepository<T> GetRepository<T>()
where T:IId;
}
My IContext interface defines two kind of repositories.
The former is for standard objects with only get/save methods, the later allows me to define specifics methods for specific kind of objects. For example :
public interface IWebServiceLogRepository : IGenericRepository<WebServiceLog>
{
ICollection<WebServiceLog> GetOpenLogs(Id objectID);
}
And it the consuming code I can do one of this :
MyContext.GetRepository<Customer>().Get(myID); --> standard get
MyContext.GetRepository<WebServiceLog, IWebServiceLogRepository>().GetOpenLogs(myID); --> specific operation
Because most of objects repository are limited to get and save operations, I've written a generic repository :
public class BaseRepository<T> : IGenericRepository<T>
where T : IId, new()
{
public virtual T Get(Id objectID){ /* provider specific */ }
public void Save(T #object) { /* provider specific */ }
}
and, for custom ones, I simply inherits the base repository :
internal class WebServiceLogRepository: BaseRepository<WebServiceLog>, IWebServiceLogRepository
{
public ICollection<WebServiceLog> GetByOpenLogsByRecordID(Id objectID)
{
/* provider specific */
}
}
Everything above is ok (at least I think it's ok). I'm now struggling to implement the MyContext class. I'm using MEF in my project for other purposes. But because MEF doesn't support (yet) generic exports, I did not find a way to reach my goal.
My context class is looking like by now :
[Export(typeof(IContext))]
public class UpdateContext : IContext
{
private System.Collections.Generic.Dictionary<Type, object> m_Implementations;
public UpdateContext()
{
m_Implementations = new System.Collections.Generic.Dictionary<Type, object>();
}
public TRepository GetRepository<T, TRepository>()
where T : IId
where TRepository : IGenericRepository<T>
{
var tType = typeof(T);
if (!m_Implementations.ContainsKey(tType))
{
/* this code is neither working nor elegant for me */
var resultType = AppDomain.CurrentDomain.GetAssemblies().SelectMany(
(a) => a.GetTypes()
).Where((t)=>t.GetInterfaces().Contains(typeof(TRepository))).Single();
var result = (TRepository)resultType.InvokeMember("new", System.Reflection.BindingFlags.CreateInstance, null, null, new object[] { this });
m_Implementations.Add(tType, result);
}
return (TRepository)m_Implementations[tType];
}
public IGenericRepository<T> GetRepository<T>() where T : IId
{
return GetRepository<T, IGenericRepository<T>>();
}
}
I'd appreciate a bit of help to unpuzzle my mind with this quite common scenario
Not sure if I've understood you correctly, but I think you're perhaps over complicating things. To begin with, make sure you've designed your code independent of any factory or Dependency Injection framework or composition framework.
For starters lets look at what you want your calling code to look like, this is what you said:
MyContext.GetRepository<Customer>().Get(myID); --> standard get
MyContext.GetRepository<WebServiceLog, IWebServiceLogRepository>().GetOpenLogs(myID);
You don't have to agree with my naming choices below, but it indicates what I undertand from your code, you can tell me if I'm wrong. Now, I feel like the calling would be simpler like this:
RepositoryFactory.New<IRepository<Customer>>().Get(myId);
RepositoryFactory.New<IWebServiceLogRepository>().GetOpenLogs(myId);
Line 1:
Because the type here is IRepository it's clear what the return type is, and what the T type is for the base IRepository.
Line 2:
The return type here from the factory is IWebServiceLogRepository. Here you don'y need to specify the entity type, your interface logically already implements IRepository. There's no need to specify this again.
So your interface for these would look like this:
public interface IRepository<T>
{
T Get(object Id);
T Save(T object);
}
public interface IWebServiceLogRepository: IRepository<WebServiceLog>
{
List<WebServiceLog> GetOpenLogs(object Id);
}
Now I think the implementations and factory code for this would be simpler as the factory only has to know about a single type. On line 1 the type is IRepository, and in line 2, IWebServiceLogRepository.
Try that, and try rewriting your code to simply find classes that implement those types and instantiating them.
Lastly, in terms of MEF, you could carry on using that, but Castle Windsor would really make things much simpler for you, as it lets you concentrate on your architecture and code design, and its very very simple to use. You only ever reference Castle in your app startup code. The rest of your code is simply designed using the Dependency Injection pattern, which is framework agnostic.
If some of this isn't clear, let me know if you'd like me to update this answer with the implementation code of your repositories too.
UPDATE
and here's the code which resolves the implementations. You were making it a bit harder for yourself by not using the Activator class.
If you use Activator and use only one Generic parameter as I've done in the method below, you should be ok. Note the code's a bit rough but you get the idea:
public static T GetThing<T>()
{
List<Type> assemblyTypes = AppDomain.CurrentDomain.GetAssemblies()
.SelectMany(s => s.GetTypes()).ToList();
Type interfaceType = typeof(T);
if(interfaceType.IsGenericType)
{
var gens = interfaceType.GetGenericArguments();
List<Type> narrowed = assemblyTypes.Where(p => p.IsGenericType && !p.IsInterface).ToList();
var implementations = new List<Type>();
narrowed.ForEach(t=>
{
try
{
var imp = t.MakeGenericType(gens);
if(interfaceType.IsAssignableFrom(imp))
{
implementations.Add(imp);
}
}catch
{
}
});
return (T)Activator.CreateInstance(implementations.First());
}
else
{
List<Type> implementations = assemblyTypes.Where(p => interfaceType.IsAssignableFrom(p) && !p.IsInterface).ToList();
return (T)Activator.CreateInstance(implementations.First());
}
}
I'm trying to design an application following Misko Heverys insights. It's an interesting experiment and a challenge. Currently I'm struggling with my ViewHelper implementation.
The ViewHelper decouples the model from the view. In my implementation it wraps the model and provides the API for the view to use. I'm using PHP, but I hope the implementation is readable for everyone:
class PostViewHelper {
private $postModel;
public function __construct(PostModel $postModel) {
$this->postModel = $postModel;
}
public function title() {
return $this->postModel->getTitle();
}
}
In my template (view) file this could be called like this:
<h1><?php echo $this->post->title(); ?></h1>
So far so good. The problem I have is when I want to attach a filter to the ViewHelpers. I want to have plugins that filter the output of the title() call. The method would become like this:
public function title() {
return $this->filter($this->postModel->getTitle());
}
I need to get observers in there, or an EventHandler, or whatever service (in what I see as a newable, so it needs to be passed in through the stack). How can I do this following the principles of Misko Hevery? I know how I can do this without it. I'm interested in how for I can take it and currently I don't see a solution. ViewHelper could be an injectable too, but then getting the model in there is the problem.
I didn't find the blog post you referenced very interesting or insightful.
What you are describing seems more like a Decorator than anything to do with dependency injection. Dependency injection is how you construct your object graphs, not their state once constructed.
That said, I'd suggest taking your Decorator pattern and running with it.
interface PostInterface
{
public function title();
}
class PostModel implements PostInterface
{
public function title()
{
return $this->title;
}
}
class PostViewHelper implements PostInterface
{
public function __construct(PostInterface $post)
{
$this->post = $post;
}
public function title()
{
return $this->post->title();
}
}
class PostFilter implements PostInterface
{
public function __construct(PostInterface $post)
{
$this->post = $post;
}
public function title()
{
return $this->filter($this->post->title());
}
protected function filter($str)
{
return "FILTERED:$str";
}
}
You'd simply use whatever DI framework you have to build this object graph like so:
$post = new PostFilter(new PostViewHelper($model)));
I often use this approach when building complex nested objects.
One problem you might run into is defining "too many" functions in your PostInterface. It can be a pain to have to implement these in every decorator class. I take advantage of the PHP magic functions to get around this.
interface PostInterface
{
/**
* Minimal interface. This is the accessor
* for the unique ID of this Post.
*/
public function getId();
}
class SomeDecoratedPost implements PostInterface
{
public function __construct(PostInterface $post)
{
$this->_post = $post;
}
public function getId()
{
return $this->_post->getId();
}
/**
* The following magic functions proxy all
* calls back to the decorated Post
*/
public function __call($name, $arguments)
{
return call_user_func_array(array($this->_post, $name), $arguments);
}
public function __get($name)
{
return $this->_post->get($name);
}
public function __set($name, $value)
{
$this->_post->__set($name, $value);
}
public function __isset($name)
{
return $this->_post->__isset($name);
}
public function __unset($name)
{
$this->_post->__unset($name);
}
}
With this type of decorator in use, I can selectively override whatever method I need to provide the decorated functionality. Anything I don't override is passed back to the underlying object. Multiple decorations can occur all while maintaining the interface of the underlying object.