I am getting started with Entity Framework 4, using model-first development. I am building a simple WPF demo app to learn the framework. My app has two entities, Topic and Note. A Topic is a discussion topic; it has Title, Text, and DateRevised properties. Topic also has a Notes collection property. a Note has DateCreated and Text properties.
I have used EF4 to create an EDM and data store for the app. Now I need to add just a bit of intelligence to the entities. For example, the property setter for the Topic.Text property needs to update the Topic.DateRevised property, and a Note needs to set its DateCreated property when it is instantiated--pretty simple stuff. I assume that I can't modify the generated classes directly, because my code would be lost if the entities are re-generated.
Is this the sort of thing that I can implement by modifying the T4 template that EF4 uses to generate the entities? In other words, can a T4 template be modified to add my code for performing these tasks to the entities that it generates? Can you refer me to a good tutorial or explanation of how to get started?
Most of what I have found so far talks about how to add a tt file to an EDM, so I can do that. What I am looking for is a resource that I can use to get to the next level, assuming that a T4 template can be used to customize generated entities as I have described. Thanks for your help.
You can do this without T4, using partial classes and partial methods.
Every EF property will have a partial OnPropertyNameChanged method. If you implement that in a partial class, you can add the behavior you need, and you won't lose your changes when you update.
So you would add a new file, let's say Topic.cs. There, you would write:
namespace MyNamespace
{
public partial class Topic
{
partial void OnTextChanged()
{
this.DateRevised = DateTime.Now;
}
}
}
Related
I want to hand-write my POCO's as simple classes with virtual properties so that the entity framework can generate proxies. I will enable proxy creation (which is on by default) so I can have change tracking and lazy loading.
My question is: can I have them?
More specifically, I noticed that the POCO template generator generates classes with a FixupCollection<T>. Is that necessary to have? If I make the navigational properties in my simple hand-written POCO a virtual IEnumerable<T>, will that suffice without any FixupCollection<T> stuff?
As an alternative, I could use the POCO template generator but I do not wish to, because my model is quite large and complicated. I will need to keep making changes to the model and will need to keep editing the model class definitions. If I use the POCO template generator, every time I need to regenerate the model for some reason, it will overwrite all my custom changes.
So, is it possible to just have POCO's in the real spirit of their name?
Yes, of corse it is possible.
http://msdn.microsoft.com/en-us/library/vstudio/dd456853%28v=vs.100%29.aspx
FixupCollection class can be used by the POCO classes to keep the opposite ends of a relationships in sync.
http://blogs.msdn.com/b/adonet/archive/2010/01/25/walkthrough-poco-template-for-the-entity-framework.aspx
I was wondering if there is a good explanation somewhere as to when a Property will need to be instantiated
in a model class constructor and when it is not necessary?
Docu on the MS EF site is good and covers a lot of related topics
Relationships and Navigation Properties - http://msdn.microsoft.com/en-us/data/jj713564
Configuring relationships with API http://msdn.microsoft.com/en-us/data/jj591620%20%22Configuring%20relationships%20with%20API
Entity Framework class library http://msdn.microsoft.com/en-us/library/hh289362%28v=vs.103%29.aspx
and im also part way through reading the Book Programming Entity Framework - Code First By Julia Lerman
I would like to better understand when a property needs to be instantiated in the Model constructor.
I see Samples in MS documentation and in the Book with and without newed up properties in the Model constructor.
There is most likely a good reason why some solutions use a constructor with newed Up objects and others not.
Seems like Virtual ICollection has the new List<> and complex type has new RefObj.
But simple virtual navigation property of type ModelObject are Ok without new.
Can I stick to a pattern without knowing the internals of each convention class. Or is there another topic I should be reading?
You never need to instantiate a property in the models constructor for EF to work, but it make it more convenient.
For example, If you had a class Order and it has children of Order Detail, and you wanted to add a new Order to your database.
Without initializing OrderDetails
var order = new Order();
order.OrderDetails.Add(new OrderDetail()); // this would error because OrderDetails is null
One option could be to call order.OrderDetails = new List();
before you use Add(), but then you would need to have that code repeated everywhere you wanted to use Add(). Instead you can put that code in the constructor, and then OrderDetails will be ready to use with all newly created Orders.
I have an Entity called "Product", this entity, through table mapping, merges 6 tables that have a 1 to 1..0 relationship with "Products". This all works wonderfully. There is another property I want to add to "Products", which is sBBR_rate, this value is not in a table that has a direct 1 to 1..0 relationship, it is related through a link table as below:
When I import the two tables into the EDM, I can't see a way in the "Mapping Details" of
"Product" to reference the sBBR_rate. I can reference RatesLink and link that to the "Products" primary key, however, I cannot reference the BBR table.
The methods I can think of to work "around" this is are as follows:
Create a view, reference the view in the EDM.
Create an SP and use a function import to retrieve the BBR when it is required.
Create a "Rates" entity in the EDM that can then draw down the sBBR_rate into it. Navigate to the Products BBR through Product.Rates.sBBR_rate.
Is there a better way I can do this that doesn't feel so much like a fudge? Perhaps by directly editing the XML of the Mapping or Conceptual layers of the EDM?
Thanks for your input.
Because the multiplicities on the Product -> RatesLink and RatesLink -> BBR relationships are 0 to 1, you should be able to access the sBBR_rate from a Product instance like this:
myProductInstance.RatesLink.BBR.sBBR_rate
I can see on the EDM screenshot that RatesLink has a Product and BBR property, which would indicate this should be available - is it?
On a side note, if it makes sense for the sBBR_rate property to commonly be accessed directly from Product, you might want to follow the law of demeter and create a property on Product which returns it directly.
The model we are using is to extend entities by using partial classes which we've found useful so we can get additional properties in the autogenerated classes (we are using a POCO T4 template to autogen but I believe this would work just as well with the default entity object generation).
So we would have
//.. this one is from the T4 template
public partial class Product
{
//.. all the autogenerated methods
}
and in a separate file that isn't autogened
//.. now in a separate file created by me
public partial class Product
{
//.. my custom properties and methods to make the entities more usable
public string BBRRate
{
get {return this.RatesLink.BBR.sBBR_rate; }
}
}
This means that I can just do
myProduct.BBRRte
I know there are other ways to do this by amending the edmx file but this one we found easy to implement. You just need to watch performance because you are potentially loading extra data. Also we did this with LazyLoading turned on but with more work you wouldn't have to
We also experimented with hooking into the ObjectMaterialized event in the ObjectContext class to preload some of these properties. Using a custom interface i.e. IMaterialisable we could check if the object was of that type then call a method (Materialise) to prepopulate some of the properties. This seems like a good idea but we didn't widely use it - it was easy to load up too much stuff. If you do the load on the properties in the partial classes then it becomes more efficient. Just my experience.
Anyway - as always an interesting question and good luck again with your dev.
EDIT
There is a rule that everything in the store layer must be represented some way in your conceptual layer. Therefore removing the tables from the conceptual layer but bring through some of the properties I don't think will work in it's gross form. Therefore I can think of two further options
Create a View on the database and bring that in as you have already mentioned. TBH this is what I would do.
Use the DefiningQuery element directly in your xml (the store layer) and map the query through to a custom entity of your exact design. Julie Lerman describes this as the ultimate escape hatch for Entity Framework.
Remember though - if you manual amend the XML in point 2 then you lose the ability to automatically update the module through the IDE
I ended up creating a view and then linking this view in the EDM, this worked a treat.
I've tried to use Entity Framework 4 and POCO for my MVC 3 project. May be, I don't understand the main idea of this ORM, but the problem is following:
I added ADO .NET Entity Data Model and make model according to database.
I clicked Add Code Generation Item and added ADO .NET POCO Entity Generator.
It makes classes for every database table.
I want to add some methods to work with data (Add, Update, Delete, GetAll etc) to appropriate models.
For LINQTOSQL I added partial classes and placed them to Models. But now I can't do it because:
a) Models folder has classes with the same names, which was created by POCO.
b) If I place my partial class in the another folder, it will be another namespace - so, such classes won't be partial one.
c) If I place my code in POCO classes, it can be destroyed during update POCO.
How can I use it? Where sould I place my methods for data working?
Is the best way to make for POCO and EF the other project - http://blogs.msdn.com/b/adonet/archive/2010/01/25/walkthrough-poco-template-for-the-entity-framework.aspx?
First of all you don't have to write your CRUD inside POCO,
There are many places where you can do it like in edmx.cs file or write one more layer which is called as CRUD Services which handles the Database operations using context object.
Now coming to your questions,
Create separate Models folder and place the Model classes in there.
Your Model class may like this,
EmployeeDepartmentModel
{
prop EmpList List(Emp);
prop DeptList List(Dept);
//Emp and Dept are my POCOs
}
So now I have to fill both of these list(Your CRUD question),
For that, I will Create one method in Controller class(its better to write such logic in some another library, but for time being I suggest you to create in Controller),
FillTheModel()
{
EmployeeDepartmentModel.EmpList = EDMX.GetAllEmployees;
EmployeeDepartmentModel.DeptList = EDMX.GetAllDepartments;
}
Now you can bind this model with your view.
You can place the partial classes in another folder and modify the namespace.
I agree with allisewell, but if you really want to add parts to partial classes, give files another name,
e.g. MyPoco.Part2.cs or modify t4 template to name generated files
e.g. Poco.Generated.cs
Our organization is looking to standardize on Entity Framework once v4 comes out. As a result, I am looking at what it would take to migrate our application that uses NHibernate for persistence to EF4 using POCO support. In a couple of places we use single table inheritance (also known as Table Per Hierarchy). I have been unable to get it to work using the following.
Payment (base class [should be abstract but having trouble there as well])
CreditCardPayment (concrete implementation)
ACHPayment (concrete implementation)
CheckPayment (concrete implementation)
Right now, I am mapping them with only the base class properties. All of these classes are in the same namespace. They have a discrimimator that is called PaymentTypeId in the database, so the Payment mapping has a condition of "When PaymentTypeId = 0". Each of the subclasses have the same condition with different values (i.e. CreditCardPayment = 1, etc.).
When I try to load each a list of all payments using DataContext.Payments.ToList() (DataContext inherits from ObjectContext) I am getting the following exception:
"Object mapping could not be found for Type with identity 'DataLayer.DataModel.CreditCardPayment'."
I can't figure out what this means, as the POCO CreditCardPayment class lives in the same namespace as the POCO Payment class does (in fact in the same file).
What am I missing?
This is complaining not about database mapping, but model to CLR mapping.
The EF can't for some reason find your CreditCardPayment class.
Now one possible reason is that you haven't loaded the metadata for it yet.
For example if you have this:
Assembly 1:
- Payment
Assembly 2 references Assembly 1:
- CreditCardPayment extends Payment
Then when you query the EF doesn't know where CreditCardPayment lives.
The way to get around this is with LoadAssembly i.e:
using (DataContext ctx = new DataContext())
{
ctx.MetadataWorkspace.LoadFromAssembly(typeof(CreditCardPayment).Assembly);
// now do your query.
}
You need to tell to LoadFromAssembly every assembly that isn't directly reference by your DataContext class.
Note: typeof(Payment).Assembly is directly referenced because of the IQueryable<Payment> Payments property.
Hope this helps
Alex
Microsoft.
What I didn't represent before (I didn't think it relevant, but it was). Was that CreditCardPayment inherited from an intermediary class named "CreditPayment" and ACHPayment inherited from CashPayment. CreditCardPayment and CashPayment live in the same namespace and file, but were not represented in the EF mapping. Once I added those within the mapping file, everything worked ok.
So, even thought EF does not ever map to one of those types directly, it seems to need to know about them, because it changes the basetype of the CreditCardPayment classes et al. Thank you for your help on this.