Suppose the following situation:
// [Validator(typeof(AddressForSupplierValidator))] <-- Cannot put both
// [Validator(typeof(AddressForCompanyValidator))] <--
// But even if I skip both of them it works great for SERVER SIDE
public class Address
{
public string AddressName { get; set; }
public string StreetName { get; set; }
}
[Validator(typeof(SupplierValidator))]
public class Supplier
{
public string Name { get; set; }
public Address Address { get; set; }
}
public class SupplierValidator: AbstractValidator<Supplier>
{
public SupplierValidator() {
RuleFor(x => x.Name).NotEmpty();
RuleFor(x => x.Address).SetValidator(new AddressForSupplierValidator());
// AddressForSupplierValidator - Requires AddressName to be .NotEmpty()
}
}
[Validator(typeof(CompanyValidator))]
public class Company
{
public string Name { get; set; }
public Address Address { get; set; }
}
public class CompanyValidator: AbstractValidator<Company>
{
public CompanyValidator() {
RuleFor(x => x.Name).NotEmpty();
RuleFor(x => x.Address).SetValidator(new AddressForCompanyValidator());
// AddressForCompanyValidator - Requires StreetName to be .NotEmpty()
}
}
As illustrated Address entity is being used by Company and by Supplier.
Now I need to validate Address depending on its context.
The above solution works for SERVER SIDE only. But for CLIENT SIDE:
It doesn't generate data-val- attributes on HTML elements. This is
because it lacks any [Validator(..)] on top Address class
definition. Adding [Validator(..)] solves the data-val- generation
but breaks the ideea of reusing Address entity and validate it
depending on context
Questions:
Why the absence of [Validator()] ontop of Address breaks CLIENT SIDE validaton, even with .SetValidator()?
Why do we need RuleFor(x => x.Address).SetValidator(..); if anyway the validator must be specified on top of the class ?
Some thoughts:
My belief is that specifying .SetValidator(..) for a nested property should be enough. The requirement with adding [Validator()] breaks the ideea of reusing same ViewModel with many validation rules depending on context.
The author already did great job integrating the client validation. And they work if leaving the ideea of reusing the viewmodel.
Temporary solution:
public class AddressBase
{
public string AddressName { get; set;}
public string StreetName { get; set;}
}
[Validator(typeof(AddressForSupplierValidator))]
public class SupplierAddress : AddressBase
{}
[Validator(typeof(AddressForCompanyValidator))]
public class CompanyAddress : AddressBase
{}
It works, but again whats the reason for .SetValidator(new AddressForSupplierValidator()); then !? Kind of redundant.
References:
Related SO question which demonstrates this issue is place at:
Unobtrusive client validation data attributes are not rendered for nested property rules
Related
I'm getting started with the concept of mapping domain models to view models in ASP.NET MVC after watching a recommendation to do this to pass specific viewModels to the views.
I've been able to manage a basic mapping of one domain model to a simpler viewmodel with less properties but now need to produce a more complex viewmodel and can't figure it out. I have the following domain models
public class Club
{
public int ClubID { get; set; }
public string FullName { get; set; }
public string Description { get; set; }
public string Telephone { get; set; }
public string URL { get; set; }
public DateTime CreatedDate { get; set; }
public virtual ICollection<Member> Members{ get; set; }
}
public class Member
{
public int MemberID{ get; set; }
public string Name { get; set; }
public MemberType Membership{ get; set; }
public virtual Club Club { get; set; }
public virtual int ClubID { get; set; }
}
public enum MemberType
{
Standard,
Special,
Limited
}
I want to map to a view model such as this (note: I've split it like this because I think it makes sense but I'm not sure)...
public class ClubDetailsViewModel
{
public int ClubID { get; set; }
public string FullName { get; set; }
public string Description { get; set; }
public IList<ClubDetailsMemberSummaryViewModel> Members { get; set; }
}
public class ClubDetailsMemberSummaryViewModel
{
public MemberType Membership { get; set; }
public int MemberCount { get; set; }
}
What I'm trying to end up with is a page which displays some of the club details plus a summary report of the member types at the club with a count of the members. Such as:
Some Club Name
Description of the club.....
CLUB MEMBERS
Limited - 15
Standard - 100
So I think the viewmodel makes sense for this (although might be a better way to do it). Where I'm struggling is how to map the elements. I can get the Club to map the main fields to the club viewmodel but really can't work out how to map the result of the list of clubs onto their view model and then add that to the main view model as a list.
I'm getting the clubs from my repository using this
var clubs = _clubRepository.GetClubByID(ID);
Then I can transform the Courts which are returned using an include in the data access layer from entity framework using this
var grpCourts = from c in clubs.Members
group c by c.Membership into grp
select new { st = grp.Key, count = grp.Distinct().Count() };
How would I loop through the resulting records and map those to the ClubDetailsMemberSummaryViewModel and then add the list of those to the main ClubDetailsViewModel?
Your mapping from Club to ClubDetailsViewModel will be trivial with the exception of Members. For that property, you could write a quick resolver inline or write your own custom resolver. An inline resolver would look something like this:
Mapper.CreateMap<Club, ClubDetailsViewModel>()
.ForMember(dest => dest.Members, opt => opt.ResolveUsing(src =>
{
return src.Members
.GroupBy(m => m.Membership)
.Select(grp => new ClubDetailsMemberSummaryViewModel
{
Membership = grp.Key,
MemberCount = grp.Distinct().Count()
});
}));
I think it's good practice to refactor more complex resolvers like this out to their own classes:
public class MembershipClubDetailsResolver : ValueResolver<Club, IList<ClubDetailsMemberSummaryViewModel>>
{
protected override IList<ClubDetailsMemberSummaryViewModel> ResolveCore (Club source)
{
return source.Members
.GroupBy (m => m.Membership)
.Select(grp => new ClubDetailsMemberSummaryViewModel
{
Membership = grp.Key,
MemberCount = grp.Distinct().Count()
})
.ToList();
}
}
And then use that resolver in your mapping:
Mapper.CreateMap<Club, ClubDetailsViewModel>()
.ForMember(dest => dest.Members, opt => opt.ResolveUsing<MembershipClubDetailsResolver>());
Your mapping appears to be rather complex, I think I would use the .ConvertUsing method of automapper
Mapper.CreateMap<List<Club>,List<ClubDetailsViewModel>>()
.ConvertUsing<ClubToClubDetailsViewModel>();
The conversion class has the following inheritance
public class ClubToClubDetailsViewModel: TypeConverter<List<Club>,List<ClubDetailsViewModel>>
{
....
}
Alternatively you can tinker with creating two "simple" mappings
Mapper.CreateMap<Club,ClubDetailsViewModel>()
That will map everything except the property called Members
Then you need to create a mapping for the members to ClubDetailsMemberSummaryViewModel, you can do that mapping manually or you can configure this in automapper aswell.
For more specific details on automapper you can visit https://github.com/AutoMapper/AutoMapper/wiki
I have a DbDataController which delivers a List of Equipment.
public IQueryable<BettrFit.Models.Equipment> GetEquipment() {
var q= DbContext.EquipmentSet.OrderBy(e => e.Name);
return q;
}
In my scaffolded view everything looks ok.
But the Equipment contains a HashSet member of EquipmentType. I want to show this type in my view and also be able to add data to the EquipmentType collection of Equipment (via a multiselect list).
But if I try to include the "EquipmentType" in my linq query it fails during serialisation.
public IQueryable<BettrFit.Models.Equipment> GetEquipment() {
var q= DbContext.EquipmentSet.Include("EquipmentType").OrderBy(e => e.Name);
return q;
}
"Object Graph for Type EquipmentType Contains Cycles and Cannot be Serialized if Reference Tracking is Disabled"
How can I switch on the "backtracking of references"?
Maybe the problem is that the EquipmentType is back-linking through a HashSet? But I do not .include("EquipmentType.Equipment") in my query. So that should be ok.
How is Upshot generating the model? I only find the EquipmentViewModel.js file but this does not contain any model members.
Here are my model classes:
public class Equipment
{
public Equipment()
{
this.Exercise = new HashSet<Exercise>();
this.EquipmentType = new HashSet<EquipmentType>();
this.UserDetails = new HashSet<UserDetails>();
}
public int ID { get; set; }
public string Name { get; set; }
public string Description { get; set; }
public string Picture { get; set; }
public string Link { get; set; }
public string Producer { get; set; }
public string Video { get; set; }
public virtual ICollection<EquipmentType> EquipmentType { get; set; }
public virtual ICollection<UserDetails> UserDetails { get; set; }
}
public class EquipmentType
{
public EquipmentType()
{
this.Equipment = new HashSet<Equipment>();
this.UserDetails = new HashSet<UserDetails>();
}
public int ID { get; set; }
public string Name { get; set; }
public string Description { get; set; }
public virtual ICollection<Equipment> Equipment { get; set; }
public virtual ICollection<UserDetails> UserDetails { get; set; }
}
try decorating one of the navigation properties with [IgnoreDataMember]
[IgnoreDataMember]
public virtual ICollection<Equipment> Equipment { get; set; }
The model generated by upshot can be found on the page itself. In your Index view you will see the UpshotContext HTML helper being used (given that you are using the latest SPA version), in which the dataSource and model type are specified.
When the page is then rendered in the browser, this helper code is replaced with the actual model definition. To see that, view the source code of your page in the browser and search for a <script> tag that starts with upshot.dataSources = upshot.dataSources || {};
Check here for more info about how upshot generates the client side model.
As for the "backtracking of references", I don't know :)
I figured out - partially how to solve the circular reference problem.
I just iterated over my queried collection (with Include() ) and set the backreferences to the parent to NULL. That worked for the serialisation issue which otherwise already breaks on the server.
The only problem now is the update of a data entity - its failing because the arrays of the referenced entitycollection are static...
To solve the cyclic backreference, you can use the IgnoreDataMember attribute. Or you can set the back reference to NULL before returning the data from the DbDataController
I posted a working solution to your problem in a different question, but using Entity Framework Code First.
https://stackoverflow.com/a/10010695/1226140
Here I show how to generate your client-side model manually, allowing to you to map the data however you please
Here's a sample of how it works now:
[MetadataType(typeof(PersonMetadata))]
public class Person
{
public string Name { get; set; }
public int Age { get; set; }
}
public class PersonMetadata
{
[Required]
public string Name { get; set; }
[Range(0,150]
public int Age { get; set; }
}
However I don't want the MetadataAttribute to be put on the Person class, instead I want some way of telling the MVC framework "hey if you encounter Person class the metadate is in the PersonMetadata class".
It's just reversing the direction which for me feels more compositional and following the Open-Close Principle since I don't have to touch the Person class to "extend" it with metadata.
I know this creates a problem of possible multiple and conflicting metadatas being attached, also increases complexity but flexibility always comes at a price.
I want either a programmatic or declarative way or preferably both :)
So one could be something like :
MetadataTypeMappings.Add(typeof(Person), typeof(PersonMetadata));
Another could be:
[MetadataFor(typeof(Person)]
public class PersonMetadata
{
[Required]
public string Name { get; set; }
[Range(0,150]
public int Age { get; set; }
}
I'm using MvcExtensions. One of the features is a nice way to describe metadata like this:
public class ProductEditModelConfiguration :
ModelMetadataConfiguration<ProductEditModel>
{
public ProductEditModelConfiguration()
{
Configure(model => model.Id).Hide();
Configure(model => model.Name)
.DisplayName(() => LocalizedTexts.Name)
.Required(() => LocalizedTexts.NameCannotBeBlank)
.MaximumLength(64, () =>
LocalizedTexts.NameCannotBeMoreThanSixtyFourCharacters);
}
}
Neat thing is - this allows to use resources for localization in strongly typed fashion.
Drawback for using this - there's bunch of other things bundled which You might not want to use.
Perhaps I'm missing something here, but it seems that anything in the object model tree 3 or more levels down, is ignored when using TryUpdateModel.
For example (simplified):
public virtual ActionResult SomeAction(int id, FormCollection form)
{
IValueProvider vpFrom = form.ToValueProvider();
/*
At this stage, vpForm contains:
1)PropertyA
2) PropertyB.SubPropertyA
3) PropertyB.SubPropertyB.SubSubPropertyA
*/
TryUpdateModel(someObjectModel, null, null, null, vpFrom);
//The first two properties are applied, number (3) seems to be ignored
Am I missing something here? If this is just the way it is, has anyone come up with a workaround?
A quick project created with the following model.
public class TestModel {
public TestModelA A { get; set; }
public string Name { get; set; }
}
public class TestModelA {
public TestModelB B { get; set; }
public string Name { get; set; }
}
public class TestModelB {
public TestModelC C { get; set; }
public string Name { get; set; }
}
public class TestModelC {
public TestModelD D { get; set; }
public string Name { get; set; }
}
public class TestModelD {
public TestModelE E { get; set; }
public string Name { get; set; }
}
public class TestModelE {
public string Name { get; set; }
}
Here's my edit - which is essentially the same as yours
[HttpPost]
public ActionResult Edit(FormCollection form) {
IValueProvider vpFrom = form.ToValueProvider();
Models.TestModel t = new Models.TestModel();
TryUpdateModel(t, null, null, null, vpFrom);
return View(t);
}
This all works exactly as expected with all the models created properly. The only problem that I can see happening is that you possibly aren't passing the same property names back from the form. (by not using <%: Html.TextBoxFor(model => model.A.B.C.CName)%> for example)
The models require parameterless constructors. But I'm sure you would have gotten an error about that - unless you're consuming the error.
So without more information about your project it will be hard to help as a basic setup produces expected results.
I believe the problem is in one of your model classes. Check, please, if PropertyB.SubPropertyB.SubSubPropertyA is really a property but not a field. A property should have get and set accessors.
Here's my checklist:
Make sure you're getting the value back in the form request. Request["A.B.C.Name"] and etc.
All the required fields are on the form.
I had deleteOnNull issue with Linq to SQL: How to set DeleteOnNull from designer for future ref if you're using L2SQL.
i'd like to know, I have a application in asp.net mvc and nhibernate. I've read about that in the Views on asp.net mvc, shouldn't know about the Domain, and it need use a DTO object. So, I'm trying to do this, I found the AutoMapper component and I don't know the correct way to do my DTOS, for some domain objects. I have a domain class like this:
public class Entity
{
public virtual int Id { get; set; }
public virtual bool Active { get; set; }
}
public class Category : Entity
{
public virtual string Name { get; set; }
public virtual IList<Product> Products { get; set; }
public Category() { }
}
public class Product : Entity
{
public virtual string Name { get; set; }
public virtual string Details { get; set; }
public virtual decimal Prince { get; set; }
public virtual int Stock { get; set; }
public virtual Category Category { get; set; }
public virtual Supplier Supplier { get; set; }
public Product() { }
}
public class Supplier : Entity
{
public virtual string Name { get; set; }
public virtual IList<Product> Products { get; set; }
public Supplier() { }
}
I'd like to get some example of how can I do my DTOs to View ? Need I use only strings in DTO ? And my controllers, it should get a domain object or a DTO and transform it on a domain to save in repository ?
Thanks a lot!
Cheers
There is no guidelines on this matter and it depends on your personal chice. I have few advices that have proven useful in practice:
1. Use flat DTOs - this means that the properties of the DTO must be as primitive as possible. This saves you the need for null reference checking.
For example if you have a domain object like this:
public class Employee
{
prop string FirstName{get; set;}
prop string LastName{get; set;}
prop Employee Boss{get; set;}
...
}
And you need to output in a grid a list of employees and display information for their 1st level boss I prefer to create a DTO
public class EmployeeDTO
{
prop string FirstName{get; set;}
prop string LastName{get; set;}
prop bool HaveABoss{get;set}
prop string BossFirstName{get; set;}
prop string BossLastName{get; set;}
...
}
or something like this (-:
2. Do not convert everything to sting - this will bind the DTO to a concrete view because you'll apply special formatting. It's not a problem to apply simple formatting directly in the view.
3. Use DTOs in your post actions and than convert them to domain objects. Usually controller's actions are the first line of deffence against incorrect data and you cannot expect to be able to allways construct a valid domain object out of the user's input. In most cases you have to do some post-processing like validation, setting default values and so on. After that you can create your DTOs.