Polymorphic model binding - asp.net-mvc

This question has been asked before in earlier versions of MVC. There is also this blog entry about a way to work around the problem. I'm wondering if MVC3 has introduced anything that might help, or if there are any other options.
In a nutshell. Here's the situation. I have an abstract base model, and 2 concrete subclasses. I have a strongly typed view that renders the models with EditorForModel(). Then I have custom templates to render each concrete type.
The problem comes at post time. If I make the post action method take the base class as the parameter, then MVC can't create an abstract version of it (which i would not want anyways, i'd want it to create the actual concrete type). If I create multiple post action methods that vary only by parameter signature, then MVC complains that it's ambiguous.
So as far as I can tell, I have a few choices on how to solve this proble. I don't like any of them for various reasons, but i will list them here:
Create a custom model binder as Darin suggests in the first post I linked to.
Create a discriminator attribute as the second post I linked to suggests.
Post to different action methods based on type
???
I don't like 1, because it is basically configuration that is hidden. Some other developer working on the code may not know about it and waste a lot of time trying to figure out why things break when changes things.
I don't like 2, because it seems kind of hacky. But, i'm leaning towards this approach.
I don't like 3, because that means violating DRY.
Any other suggestions?
Edit:
I decided to go with Darin's method, but made a slight change. I added this to my abstract model:
[HiddenInput(DisplayValue = false)]
public string ConcreteModelType { get { return this.GetType().ToString(); }}
Then a hidden automatically gets generated in my DisplayForModel(). The only thing you have to remember is that if you're not using DisplayForModel(), you'll have to add it yourself.

Since I obviously opt for option 1 (:-)) let me try to elaborate it a little more so that it is less breakable and avoid hardcoding concrete instances into the model binder. The idea is to pass the concrete type into a hidden field and use reflection to instantiate the concrete type.
Suppose that you have the following view models:
public abstract class BaseViewModel
{
public int Id { get; set; }
}
public class FooViewModel : BaseViewModel
{
public string Foo { get; set; }
}
the following controller:
public class HomeController : Controller
{
public ActionResult Index()
{
var model = new FooViewModel { Id = 1, Foo = "foo" };
return View(model);
}
[HttpPost]
public ActionResult Index(BaseViewModel model)
{
return View(model);
}
}
the corresponding Index view:
#model BaseViewModel
#using (Html.BeginForm())
{
#Html.Hidden("ModelType", Model.GetType())
#Html.EditorForModel()
<input type="submit" value="OK" />
}
and the ~/Views/Home/EditorTemplates/FooViewModel.cshtml editor template:
#model FooViewModel
#Html.EditorFor(x => x.Id)
#Html.EditorFor(x => x.Foo)
Now we could have the following custom model binder:
public class BaseViewModelBinder : DefaultModelBinder
{
protected override object CreateModel(ControllerContext controllerContext, ModelBindingContext bindingContext, Type modelType)
{
var typeValue = bindingContext.ValueProvider.GetValue("ModelType");
var type = Type.GetType(
(string)typeValue.ConvertTo(typeof(string)),
true
);
if (!typeof(BaseViewModel).IsAssignableFrom(type))
{
throw new InvalidOperationException("Bad Type");
}
var model = Activator.CreateInstance(type);
bindingContext.ModelMetadata = ModelMetadataProviders.Current.GetMetadataForType(() => model, type);
return model;
}
}
The actual type is inferred from the value of the ModelType hidden field. It is not hardcoded, meaning that you could add other child types later without having to ever touch this model binder.
This same technique could be easily be applied to collections of base view models.

I have just thought of an intersting solution to this problem. Instead of using Parameter bsed model binding like this:
[HttpPost]
public ActionResult Index(MyModel model) {...}
I can instead use TryUpdateModel() to allow me to determine what kind of model to bind to in code. For example I do something like this:
[HttpPost]
public ActionResult Index() {...}
{
MyModel model;
if (ViewData.SomeData == Something) {
model = new MyDerivedModel();
} else {
model = new MyOtherDerivedModel();
}
TryUpdateModel(model);
if (Model.IsValid) {...}
return View(model);
}
This actually works a lot better anyways, because if i'm doing any processing, then I would have to cast the model to whatever it actually is anyways, or use is to to figure out the correct Map to call with AutoMapper.
I guess those of us who haven't been using MVC since day 1 forget about UpdateModel and TryUpdateModel, but it still has its uses.

It took me a good day to come up with an answer to a closely related problem - although I'm not sure it's precisely the same issue, I'm posting it here in case others are looking for a solution to the same exact problem.
In my case, I have an abstract base-type for a number of different view-model types. So in the main view-model, I have a property of an abstract base-type:
class View
{
public AbstractBaseItemView ItemView { get; set; }
}
I have a number of sub-types of AbstractBaseItemView, many of which define their own exclusive properties.
My problem is, the model-binder does not look at the type of object attached to View.ItemView, but instead looks only at the declared property-type, which is AbstractBaseItemView - and decides to bind only the properties defined in the abstract type, ignoring properties specific to the concrete type of AbstractBaseItemView that happens to be in use.
The work-around for this isn't pretty:
using System.ComponentModel;
using System.ComponentModel.DataAnnotations;
// ...
public class ModelBinder : DefaultModelBinder
{
// ...
override protected ICustomTypeDescriptor GetTypeDescriptor(ControllerContext controllerContext, ModelBindingContext bindingContext)
{
if (bindingContext.ModelType.IsAbstract && bindingContext.Model != null)
{
var concreteType = bindingContext.Model.GetType();
if (Nullable.GetUnderlyingType(concreteType) == null)
{
return new AssociatedMetadataTypeTypeDescriptionProvider(concreteType).GetTypeDescriptor(concreteType);
}
}
return base.GetTypeDescriptor(controllerContext, bindingContext);
}
// ...
}
Although this change feels hacky and is very "systemic", it seems to work - and does not, as far as I can figure, pose a considerable security-risk, since it does not tie into CreateModel() and thus does not allow you to post whatever and trick the model-binder into creating just any object.
It also works only when the declared property-type is an abstract type, e.g. an abstract class or an interface.
On a related note, it occurs to me that other implementations I've seen here that override CreateModel() probably will only work when you're posting entirely new objects - and will suffer from the same problem I ran into, when the declared property-type is of an abstract type. So you most likely won't be able to edit specific properties of concrete types on existing model objects, but only create new ones.
So in other words, you will probably need to integrate this work-around into your binder to also be able to properly edit objects that were added to the view-model prior to binding... Personally, I feel that's a safer approach, since I control what concrete type gets added - so the controller/action can, indirectly, specify the concrete type that may be bound, by simply populating the property with an empty instance.

Using Darin's method to discriminate your model types via a hidden field in your view, I would recommend that you use a custom RouteHandler to distinguish your model types, and direct each one to a uniquely named action on your controller. For example, if you have two concrete models, Foo and Bar, for your Create action in your controller, make a CreateFoo(Foo model) action and a CreateBar(Bar model) action. Then, make a custom RouteHandler, as follows:
public class MyRouteHandler : IRouteHandler
{
public IHttpHandler GetHttpHandler(RequestContext requestContext)
{
var httpContext = requestContext.HttpContext;
var modelType = httpContext.Request.Form["ModelType"];
var routeData = requestContext.RouteData;
if (!String.IsNullOrEmpty(modelType))
{
var action = routeData.Values["action"];
routeData.Values["action"] = action + modelType;
}
var handler = new MvcHandler(requestContext);
return handler;
}
}
Then, in Global.asax.cs, change RegisterRoutes() as follows:
public static void RegisterRoutes(RouteCollection routes)
{
routes.IgnoreRoute("{resource}.axd/{*pathInfo}");
AreaRegistration.RegisterAllAreas();
routes.Add("Default", new Route("{controller}/{action}/{id}",
new RouteValueDictionary(
new { controller = "Home",
action = "Index",
id = UrlParameter.Optional }),
new MyRouteHandler()));
}
Then, when a Create request comes in, if a ModelType is defined in the returned form, the RouteHandler will append the ModelType to the action name, allowing a unique action to be defined for each concrete model.

Related

Difficulty repopulating viewmodels on post

I have various viewmodels that have properties that must be populated. For example, imagine a database table full of countries. I have a create user page, and one of the properties on the viewmodel is List<string> Countries.
Originally I populated this in a parameterless constructor, ie.
public CreateUserViewModel()
{
this.Countries = new CountryManager().GetCountries();
}
But I read this is a bad practice, and I should pass them in.
public CreateUserViewModel(IEnumerable<string> countries)
{
this.Countries = countries;
}
But in my post, this data is lost, and if validation fails it redirects to the view, but then the countries property is null.
I am wondering how I should be repopulating this value. Manually putting some code into the controller post method seems bad, eg.
[HttpPost]
public ActionResult CreateUser(CreateUserViewModel vm)
{
if (Model.IsValid)
{
new UserManager().CreateUser(vm);
}
else
{
vm.Countries = new CountryManager().GetCountries();
return View(vm);
}
}
I am struggling to google what is probably a very common question. Any ideas?
The rationale behind avoiding parameterless constructors is for inversion of control. The logic in this case would be that Countries is a dependency and by externalizing that dependency (so that it is inject into the class, instead), you make your class less brittle and more open to extension.
However, I would argue that doesn't apply actually in your scenario, because Countries is not really a dependency of your class, but rather of your view. The view model is there to serve the view and is somewhat closed and unextensible anyways as a result. In other words, do follow inversion of control for things like services, repositories, utility classes, etc., but for view models, it's not really necessary or important.
Anyways, here's how I handle this type of thing:
public class FooViewModel
{
...
// Countries is not initialize by a constructor
public IEnumerable<SelectListItem> Countries { get; set; }
}
Then in your controller:
public class FooController : Controller
{
internal void PopulateCountryChoices(FooViewModel model)
{
// fetch countries
model.Countries = countries.Select(m => new SelectListItem
{
Text = m.Name,
Value = m.Id.ToString()
});
}
public ActionResult Bar()
{
var model = new FooViewModel();
PopulateCountryChoices(model);
return View(model);
}
[HttpPost]
public ActionResult Bar(FooViewModel model)
{
if (ModelState.IsValid)
{
// save and redirect
}
PopulateCountryChoices(model);
return View(model);
}
}
Instead of populating this static data in model, We can populate Dropdownlist values in the view itself -
#Html.DropDownListFor(model => model.State,
new SelectList(Utils.GetCountries()),
"value",
"text",
2)
Where Utils is a Helper class which returns all the countries.
This way of populating all the static data (I mean all options of Dropdownlist or Listbox etc) will be taken over by the View, which makes model to be free from holding this data.

ASP.NET MVC ViewModel with SelectList(s) best practice

I noticed that in the NerdDinner application that if the ModelState is invalid for a dinner, it merely returns the view for the model:
if (ModelState.IsValid) {
...
return RedirectToAction("Details", new { id=dinner.DinnerID });
}
return View(dinner);
However, in my application the model (a view model in this situation) contains multiple SelectLists. These lists are not instantiated at this point because this view model was just populated from the form submission. What is the recommended way to repopulate this SelectLists before sending them back to the user?
This is what I want my controller to do:
public ActionResult Save(MyModel model)
{
if (ModelState.IsValid)
{
businessClass.Save(model);
return RedirectToAction("Index", "Home");
}
// This won't work because model has uninstantiated SelectLists
return View("MyView", model);
}
I don't want to send the model to my business logic if the ModelState is invalid, but it doesn't seem to make sense to put SelectList population code in my controller. Should I create a public method in my business logic solely for doing this kind of stuff on my view model(s)?
Personally I like to keep it simple:-
[HttpGet]
public Edit(int id) {
EditForm form = new EditForm();
// Populate from the db or whatever...
PopulateEditPageSelectLists(form);
return View(form);
}
[HttpPost]
public Edit(EditForm form) {
if (ModelState.IsValid) {
// Do stuff and redirect...
}
PopulateEditPageSelectLists(form);
return View(form);
}
public void PopulateEditPageSelectLists(form) {
// Get lookup data from the db or whatever.
}
If the logic to populate the select lists is all kinds crazy it might be worthwhile moving to a separate class or whatever it but as a first step this is the best place to start.
You dont say how much reusability would you like. But personally, i like things "clear" (dont invading controller) and reausable as possible, and that in MVC means - filters.
Look at this :
public class SupplyLanguagesAttribute : System.Web.Mvc.ActionFilterAttribute
{
public override void OnActionExecuting(System.Web.Mvc.ActionExecutingContext filterContext)
{
filterContext.Controller.ViewData["languagesList"] =
someService.LoadLanguagesAsDictionary();
base.OnActionExecuting(filterContext);
}
}
then you just use it with every action method where you "might" need languages :
[SupplyLanguages]
public ActionResult DoSomething()
{
...
}
And then in view, you can use the data directly for DropDownList from ViewData, or you can even "wrap" this too (and avoid "magic strings" in views), with custom reusable DropDown :
public static MvcHtmlString LanguageDropDown(this HtmlHelper html, string name, object selectValue, bool defaultOption = false)
{
var languages = html.ViewData["languagesList"] as IDictionary<string,string>;
if (languages == null || languages.Count() == 0)
throw new ArgumentNullException("LanguageDropDown cannot operate without list of languages loaded in ViewData. Use SupplyLanguages filter.");
var list = new SelectList(languages, "Key", "Value", selectValue);
return SelectExtensions.DropDownList(html, name, list);
}
My controllers populate the SelectLists on my Model if the ModelState is not valid.
Following Separation of Concerns, your business classes shouldn't know anything about the view model at all. If your view needs a list of employees your controller gets a list of employees from your business layer and creates the SelectList that your view needs.
Example
public ActionResult Save(MyModel model)
{
if (ModelState.IsValid)
{
businessClass.Save(model);
return RedirectToAction("Index", "Home");
}
model.PossibleEmployees
= _employeeRepository.All().Select(e =>
new SelectListItem{Text=e.Name,
Value=e.Id});
return View("MyView", model);
}
Update
If your select list population code is determining WHICH options to present I think you probably should move that to a service in your business layer. If reusability is the big concern, rouen's answer looks like it has the most possibility for reuse.
I use to fill lists even when the model is invalid. One other possible solution is to have an action returning the json information and build the select via ajax. SOmetimes I've also resorted to static properties / cached collections. I guess it's always depending on the particular case.
PS: You can use a local Model in each action, so I can leave initialization inside the Model constructor. (often I override a base model with [NonAction] utilities as well).
For example, I have an Employee list used widely in your application.
I've added some utility method in a base controller to build up SelectListItems and the likes. Since each and every model inherits from the base, I've got them almost everywhere in the app. Of course the Collection is filled via a dedicated business objec.
What I do is I have a static function in a class that returns a SelectList. The method accepts an Enum value which defines which SelectList to return. In the View the DropDownList or DropDownListFor functions call this function to get the SelectList.
The static function looks like this:
class HelperMethods
{
enum LookupType {Users, Companies, States};
public static SelectList CommonSelectList(LookupType type, int? filterValue = null)
//filterValue can be used if the results need to be filtered in some way
var db = new WhateverEntities();
switch (type)
{
case LookupType.Users:
var list = db.Users.OrderBy(u => u.LastName).ToList()
return new SelectList(list, "ID", "FullName")
break;
case LookupType.Companies
var list = db.Companies.OrderBy(u => u.Name).ToList()
return new SelectList(list, "ID", "Name")
break;
//and so on...
}
}
}
And the view contains this:
#Html.DropDownListFor(m => m.UserID, HelperMethods.CommonSelectList(LookupType.Users))
This way the Model and Controller does not need code to configure a SelectList to send over to the View. It makes it very easy to reuse a SelectList that has already been configured. Also, if a View needs to loop through a list of objects, then this same function can be used to get a list for that. This is the simplest and most convenient way I found of doing this.

MVC UpdateModel and Sub Classes vs Base Class

I'm looking to use the UpdateModel method to a Sub Class that retrieved at runtime, would be great if someone could shed the light on whether I'm making a total hash of it and/or whether or not what I'm trying to do is possible.
I'm using a generic action to control the validation of a bunch of partial views; I'm trying to get away from having a specific action per partial view.
Each partial view has a unique Model that derives from a Base Model:
public class ModelA : ModelBase{
[Required]
public string SomeStringProperty{get;set;}
...
}
public class ModelB : ModelBase{
[Required]
public DateTime? SomeDateProperty{get;set;}
...
}
public class ModelBase{
public Guid InstanceId{get;set;}
}
I'm using the FormCollection on the Action to get the submitted form elements and their values, this includes the type of model that the View should be using to validate its request. Ignore the security implications of this for this example, I'm aware of them and this is an internal only proof of concept
[HttpPost]
public ActionResult ChangeCaseState(int id, FormCollection formCollection)
{
Guid instanceId = new Guid(formCollection["instanceId"]);
string modelType = formCollection["modelType"];
//Return a specific Model class based on the event/modelType
var args = GetStateModelClass(modelType, instanceId);
try
{
UpdateModel(args);
if(Model.IsValid){
...
}
catch (Exception)
{
return View("~/Views/Shared/StateForms/" + modelType + ".ascx", args);
}...
And here is the code I'm using to return a Sub Class based on the modelType passed to the controller.
private static ModelBase StateModelClassFactory(string stateModelTypeName, Guid instanceId)
{
switch (stateModelTypeName)
{
case "modelTypeA":
return new ModelA(workflowInstanceId);
case "modelTypeB":
return new ModelB(workflowInstanceId);
...
}
Because the return type of the StateModelClassFactory method is of the Base Class, even though I'm actually returning a Sub Class, the Model Binder used by the UpdateModel method only binds against the values within the Base Class.
Any ideas on how I can solve this problem?
UPDATE:
I created a Customer Model Binder:
public class CustomModelBinder : DefaultModelBinder
{
public override object BindModel(ControllerContext controllerContext, ModelBindingContext bindingContext)
{
And Assigned the new Model Binder to the correct Base Class to see what was going on a little more under the hood:
ModelBinders.Binders.Add(typeof(ModelBase), new CaseController.CustomModelBinder());
When I debug the model binder and inspect the bindingContext, the Model property represets the correct Sub Class, but the ModelType property is that of the Base Class. Should I be looking at changing the ModelType within the BindModel method? If so any pointers on how to do this, the setter on the ModelType seems to have been made redundant. I also noticed that the SomeDateProperty from the Sub Class is actaully in the PropertyMetadata property....Seems so close to behaving as I'd like.
I just ran into this particular issue and found that a better general approach would be just to cast your model to dynamic while passing it into UpdateModel:
[HttpPost]
public ActionResult ChangeCaseState(int id, FormCollection formCollection)
{
...try
{
UpdateModel((dynamic)args);//!!notice cast to dynamic here
if(Model.IsValid){
...
}
catch...
This appears to set all available properties of my type, regardless of whether my variable is delcared with the base type.
There's a work item filed in CodePlex for this issue: http://aspnet.codeplex.com/workitem/8277?ProjectName=aspnet
So I think I've solved my problem. Basically because of the way that I'm retrieving the Model class before calling the UpdateModel, the Model Binder is binding the BaseClass even though the Model was that of the SubClass - this is the code I used to solve my particular problem:
public class SubClassModelBinder : DefaultModelBinder
{
public override object BindModel(ControllerContext controllerContext, ModelBindingContext bindingContext)
{
var model = bindingContext.Model;
var metaDataType = ModelMetadataProviders.Current.GetMetadataForType(null, model.GetType());
bindingContext.ModelMetadata = metaDataType;
bindingContext.ModelMetadata.Model = model;
return base.BindModel(controllerContext, bindingContext);
}
}
And in the Global.asax:
ModelBinders.Binders.Add(typeof(ModelBase), new SubClassModelBinder ());
Thanks to Darin for his inital pointer.
To solve this problem you could write a custom model binder for the base type which based on the value of the string property will return the correct child instance.

Bestpractice DI with ASP.NET MVC and StructureMap - How to inject dependencies in an ActionResult

I edited my whole question, so do not wonder :)
Well, I want to have an ActionResult that takes domain model data and some additional parameters, i.e page index and page size for paging a list. It decide itself if it returns a PartialViewResult or a ViewResult depending on the kind of web request (ajax request or not).
The reffered data shall be mapped automatically by using an IMappingService, which is responsible for transforming any domain model data into a view model.
The MappingService uses AutoMapper for simplicity.
MappingActionResult:
public abstract class MappingActionResult : ActionResult
{
public static IMappingService MappingService;
}
BaseHybridViewResult:
public abstract class BaseHybridViewResult : MappingActionResult
{
public const string defaultViewName = "Grid";
public string ViewNameForAjaxRequest { get; set; }
public object ViewModel { get; set; }
public override void ExecuteResult(ControllerContext context)
{
if (context == null) throw new ArgumentNullException("context");
var usePartial = ShouldUsePartial(context);
ActionResult res = GetInnerViewResult(usePartial);
res.ExecuteResult(context);
}
private ActionResult GetInnerViewResult(bool usePartial)
{
ViewDataDictionary viewDataDictionary = new ViewDataDictionary(ViewModel);
if (String.IsNullOrEmpty(ViewNameForAjaxRequest))
{
ViewNameForAjaxRequest = defaultViewName;
}
if (usePartial)
{
return new PartialViewResult { ViewData = viewDataDictionary, ViewName = ViewNameForAjaxRequest };
}
return new ViewResult { ViewData = viewDataDictionary };
}
private static bool ShouldUsePartial(ControllerContext context)
{
return context.HttpContext.Request.IsAjaxRequest();
}
}
AutoMappedHybridViewResult:
public class AutoMappedHybridViewResult<TSourceElement, TDestinationElement> : BaseHybridViewResult
{
public AutoMappedHybridViewResult(PagedList<TSourceElement> pagedList)
{
ViewModel = MappingService.MapToViewModelPagedList<TSourceElement, TDestinationElement>(pagedList);
}
public AutoMappedHybridViewResult(PagedList<TSourceElement> pagedList, string viewNameForAjaxRequest)
{
ViewNameForAjaxRequest = viewNameForAjaxRequest;
ViewModel = MappingService.MapToViewModelPagedList<TSourceElement, TDestinationElement>(pagedList);
}
public AutoMappedHybridViewResult(TSourceElement model)
{
ViewModel = MappingService.Map<TSourceElement, TDestinationElement>(model);
}
public AutoMappedHybridViewResult(TSourceElement model, string viewNameForAjaxRequest)
{
ViewNameForAjaxRequest = viewNameForAjaxRequest;
ViewModel = MappingService.Map<TSourceElement, TDestinationElement>(model);
}
}
Usage in controller:
public ActionResult Index(int page = 1)
{
return new AutoMappedHybridViewResult<TeamEmployee, TeamEmployeeForm>(_teamEmployeeRepository.GetPagedEmployees(page, PageSize));
}
So as you can see the IMappingService is hidden. The controller should not know anything about the IMappingService interface, when AutoMappedHybridViewResult is used.
Is the MappingActionResult with the static IMappingServer appropriate or am I violating the DI principle?
I think a better design is to have a ViewResultFactory that depends on IMappingService, then you can inject that into your controller. Then you call it like so:
public class MyController : Controller
{
IViewResultFactory _viewResultFactory;
ITeamEmployeeRepository _teamEmployeeRepository;
public MyController(IViewResultFactory viewResultFactory)
{
_viewResultFactory = viewResultFactory;
}
public ActionResult MyAction(int page, int pageSize)
{
return
_viewResultFactory.GetResult<TeamEmployee, TeamEmployeeForm>(
_teamEmployeeRepository.GetPagedEmployees(page, pageSize));
}
}
The implementation would like this (you would need to create overloads for each of your HybridViewResult constructors):
public HybridViewResult<TSourceElement, TDestinationElement> GetResult<TSourceElement, TDestinationElement>(PagedList<TSourceElement> pagedList)
{
return new HybridViewResult<TSourceElement, TDestinationElement>(_mappingService, pagedList);
}
That way you hide the implementation from your controllers, and you don't have to depend on the container.
There are a few different points that you could inject IMappingService. http://codeclimber.net.nz/archive/2009/04/08/13-asp.net-mvc-extensibility-points-you-have-to-know.aspx is a good site for help in picking the appropriate extensibility points for .NET MVC.
If you want to stick with having this functionality be a derived ActionResult, then I think you could put the dependency in the ActionInvoker if you want to, but the Controller makes more sense to me. If you don't want the IMappingService in the Controller, you could always wrap it in a HybridViewResultFactory, and access that object in the Controller. In that case your shortcut methods would look like:
public HybridViewResult<TSourceElement, TDestinationElement> AutoMappedHybridView<TSourceElement,TDestinationElement>(PagedList<TSourceElement> pagedList, string viewNameForAjaxRequest)
{
HybridViewResultFactory.Create<TSourceElement, TDestinationElement>(pagedList, viewNameForAjaxRequest);
}
etc.
I'm not sure why you need to use an ActionResult, but if there is no reason that makes it explicitly necessary, you could create a HybridViewModel class and a HybridViewModelBinder class that is injected with the mapping service dependency.
I am assuming you want to use constructor injection, but if you have the StructureMap dependency in your UI assembly, you could access a static dependency resolver class (like Clowers said).
This question would be easier to give a definite answer to if I understood why you using an ActionResult.
It seems like you are using the action result to handle two functionalities that do not necessarily go together all the time, and that could be used separately. Also, there is not a clear indication that it needs to be in an ActionResult.
Presumably, you could (a) leverage the Automapper functionality for results other than html (ViewResult) output, and (b) you could leverage the functionality of auto-detecting ajax requests without needing to automap the model.
It seems to me like the automapping of the view model could be used to inject the view model into the controller action directly, thus removing the controller's dependency on the IMappingService. What you would need is a ModelBinder class to be injected with your IMappingService (the implementation of which I assume contains a repository or datastore type dependency).
Here is a good article explaining how to leverage model binders: http://odetocode.com/blogs/scott/archive/2009/04/27/6-tips-for-asp-net-mvc-model-binding.aspx.
Then you can overwrite the DefaultModelBinder in the classes that need to be Automapped as follows:
public ActionResult DoItLikeThis([AutoMap(typeof(MyDomainModelClass))]MyViewModelClass viewModel){
//controller action logic
}
Now, regarding the HybridViewResult, I would suggest that you handle this with an Action Filter instead. So, you could just use ActionResult or ViewResultBase as the Result type of your action method and decorate it with an action filter, i.e.:
[AutoSelectViewResult]
public ViewResultBase AndDoThisLikeSo(){
//controller action logic
}
I think overall this will be a much better solution than coupling these two functionalities to an ActionResult.

ASP.NET MVC - Custom Model Binder on Interface Type

I'm not sure if this behavior is expected or not, but it seems that custom model binding doesn't work when the binding is assigned to an interface type. Has anyone experimented with this?
public interface ISomeModel {}
public class SomeModel : ISomeModel {}
public class MvcApplication : HttpApplication {
protected void Application_Start(object sender, EventArgs e) {
ModelBinders.Binders[typeof(ISomeModel)] = new MyCustomModelBinder();
}
}
With the above code when I bind to a model of type SomeModel, MyCustomModelBinder is never hit; however, if I change the above code and substitute typeof(ISomeModel) for typeof(SomeModel) and post the exact same form MyCustomModelBinder is called as expected. Does that seem right?
Edit
I found myself back in this predicament over a year after I originally asked this question, and now I have a solution that works. Thank you Matt Hidinger!
http://www.matthidinger.com/archive/2011/08/16/An-inheritance-aware-ModelBinderProvider-in-MVC-3.aspx
I was experimenting with this issue and I came up with a solution of sorts. I made a class called InterfaceModelBinder:
public class InterfaceModelBinder : DefaultModelBinder
{
public override object BindModel(ControllerContext controllerContext, ModelBindingContext bindingContext)
{
ModelBindingContext context = new ModelBindingContext(bindingContext);
var item = Activator.CreateInstance(
Type.GetType(controllerContext.RequestContext.HttpContext.Request.Form["AssemblyQualifiedName"]));
Func<object> modelAccessor = () => item;
context.ModelMetadata = new ModelMetadata(new DataAnnotationsModelMetadataProvider(),
bindingContext.ModelMetadata.ContainerType, modelAccessor, item.GetType(), bindingContext.ModelName);
return base.BindModel(controllerContext, context);
}
}
Which I registered in my Application_Start as so:
ModelBinders.Binders.Add(typeof(IFormSubmission), new InterfaceModelBinder.Models.InterfaceModelBinder());
The interface and a concrete implementation look like this:
public interface IFormSubmission
{
}
public class ContactForm : IFormSubmission
{
public string Name
{
get;
set;
}
public string Email
{
get;
set;
}
public string Comments
{
get;
set;
}
}
The only downside to this whole approach (as you might have gathered already) is that I need to get the AssemblyQualifiedName from somewhere, and in this example it is being stored as a hidden field on the client side, like so:
<%=Html.HiddenFor(m => m.GetType().AssemblyQualifiedName) %>
I'm not certain though that the downsides of exposing the Type name to the client are worth losing the benefits of this approach. An Action like this can handle all my form submissions:
[HttpPost]
public ActionResult Process(IFormSubmission form)
{
if (ModelState.IsValid)
{
FormManager manager = new FormManager();
manager.Process(form);
}
//do whatever you want
}
Any thoughts on this approach?
Suddenly, an MVC3 solution appears:
http://www.matthidinger.com/archive/2011/08/16/An-inheritance-aware-ModelBinderProvider-in-MVC-3.aspx
I'm not sure if its directly related but yes there are things that you need to think about when using model binding and interfaces... I ran into similar problems with the default model binder, but it may not be directly related depending on how you are doing things...
Have a look at the following:
ASP.net MVC v2 - Debugging Model Binding Issues - BUG?
ASP.net MVC v2 - Debugging Model Binding Issues - BUG?

Resources