using the unit of work and repository patterns i recently came across the issue, that changes to the unit of work are not reflected to subsequent queries. Example:
var ctx = DIContainer.Current.Resolve<IB2bContext>();
var rep = DIContainer.Current.Resolve<IRepository<Word>>(
new DependencyOverride<IB2bContext>(ctx));
rep.Add(new Word () { "One" };
rep.Add(new Word () { "Two" };
rep.GetAll().ToList().ForEach(i =>
Console.Write(i.text)); // nothing seen here
So in other words, unless i call SaveChanges() to persist the objects into the Database, i dont see them. Well ofcause i can fiddle around with the ChangeTracker and/or do things like context.Entry(foo).Property(...).CurrentValue. But does that play with a ddd like decoupling of layers? I dont think so. And where is my consistent dataview that once was called a database transaction?
Please enlighten me.
Armin
Your repository exposes some GetAll method. The method itself executes database query. If you want to see local data not inserted to database you must add them to result set. For example like:
public IEnumerable<Word> GetAll()
{
DbSet<Word> set = context.Set<Word>();
return set.AsEnumerable().Concat(set.Local);
}
The query execution is only responsible for returning persisted (real) data.
Related
Setting:
One db with many tables, grouped by modules. The moduleTableGroups are independent, there are no relationships in between.
Design:
1: The modules have each there own Context, to keep them small and clear.
2: Because it is an very old Database with manual administration of PKeys and sequences it must be possible to nest SaveChanges, when handling those things in the context.
E.G. AContext has a table with an entity which has a ValueGenerator which reads an incremented value from a table. Because it should be finished before AContext.SaveChanges(), is called, it must be done in an seperate SeqContext in the same db which can handle the concurency issues which might occur.
The SeqContext depends on the DbOptionsContext of AContext.
Wishes:
1: Instead of create Options for each Context which is currently needed, becaus a context seems only to accept context specific options, I'd like to either use the same DbContextOptions for many context or copy or cast the options to a new .
DbContextOptions<AContext> aContextOptions = new DbContextOptionsBuilder<AContext>(). . .Options;
// casting:
DbContextOptions<BContext> bContextOptions = (DbContextOptions<BContext>) aContextOptions;
// copy construction:
DbContextOptions<BContext> bContextOptions = new DbContextOptionsBuilder<BContext>(aContextOption).Options;
2: Because of the dependency of SeqContext to AContext especially in testing scenarios it would be useful to do something from within a AContext like:
SeqContext = new SeqContext((DbContextOptions<SeqContext>) this.getOptions());
You might get the idea. How can I archive to share options between different TContext right now?
I have a web api project that I'm building on an N-Tier system. Without causing too many changes to the overall system, I will not be touching the data server that has access to the database. Instead, I'm using .NET remoting to create a tcp channel that will allow me to send requests to the data server, which will then query the database and send back a response object.
On my application, I would like to use entity framework to create my datacontexts (unit of work), then create a repository pattern that interfaces with those contexts, which will be called by the web api project that I created.
However, I'm having problems with entity framework as it requires me to have a connection with the database. Is there anyway I can create a full entity framework project without any sqlconnections to the database? I just need dbcontexts, which I will be mapping my response objects and I figure that EF would do what I needed (ie help with design, and team collabs, and provide a nice graphical designer); but it throws an error insisting that I need a connection string.
I've been searching high and low for tutorials where a database is not needed, nor any sql connection string (this means no localdb either).
Okay as promised, I have 3 solutions for this. I personally went with #3.
Note: Whenever there is a repository pattern present, and "datacontext" is used, this is interpreted as your UnitOfWork.
Solution 1: Create singletons to represent your datacontext.
http://www.breezejs.com/samples/nodb
I found this idea after going to BreezeJS.com's website and checked out their samples. They have a sample called NoDb, which allows them to create a singleton, which can create an item and a list of items, and a method to populate the datacontext. You create singletons that would lock a space in memory to prevent any kind of thread conflicts. Here is a tid bit of the code:
//generates singleton
public class TodoContext
{
static TodoContext{ }
private TodoContext() { }
public static TodoContext Instance
{
get
{
if (!__instance._initialized)
{
__instance.PopulateWithSampleData();
__instance._initialized = true;
}
return __instance;
}
}
public void PopulateWithSampleData()
{
var newList = new TodoItem { Title = "Before work"};
AddTodoList(newList);
var listId = newList.TodoListId;
var newItem = new TodoItem {
TodoListId = listId, Title = "Make coffee", IsDone = false };
AddTodoItem(newItem);
newItem = new TodoItem {
TodoListId = listId, Title = "Turn heater off", IsDone = false };
AddTodoItem(newItem);
}
//SaveChanges(), SaveTodoList(), AddTodoItem, etc.
{ ... }
private static readonly Object __lock = new Object();
private static readonly TodoContext __instance = new TodoContext();
private bool _initialized;
private readonly List<TodoItem> _todoLists = new List<TodoItem>();
private readonly List<KeyMapping> _keyMappings = new List<KeyMapping>();
}
There's a repository included which directs how to save the context and what needs to be done before the context is saved. It also allows the list of items to be queryable.
Problem I had with this:
I felt like there was higher maintenance when creating new datacontexts. If I have StateContext, CityContext, CountryContext, the overhead of creating them would be too great. I'd have problems trying to wrap my head around relating them to each other as well. Plus I'm not too sure how many people out there who agree with using singletons. I've read articles that we should avoid singletons at all costs. I'm more concerns about anyone who'd be reading this much code.
Solution 2: Override the Seed() for DropCreateDatabaseAlways
http://www.itorian.com/2012/10/entity-frameworks-database-seed-method.html
For this trick, you have to create a class called SampleDatastoreInitializer that inherits from System.Data.Entity.DropCreateDatabaseAlways where T is the datacontext, which has a reference to a collection of your POCO model.
public class State
{
[Key()]
public string Abbr{ get; set; }
public string Name{ get; set; }
}
public class StateContext : DbContext
{
public virtual IDbSet<State> States { get; set; }
}
public class SampleDatastoreInitializer : DropCreateDatabaseAlways<StateContext>
{
protected override void Seed (StateContext context)
{
var states = new List<State>
{
new State { Abbr = "NY", Name = "New York" },
new State { Abbr = "CA", Name = "California" },
new State { Abbr = "AL", Name = "Alabama" },
new State { Abbr = "Tx", Name = "Texas" },
};
states.ForEach(s => context.States.Add(s));
context.SaveChanges();
}
}
This will actually embed the data in a cache, the DropCreateDatabaseAlways means that it will drop the cache and recreate it no matter what. If you use some other means of IDatabaseInitializer, and your model has a unique key, you might get an exception error, where you run it the first time, it works, but run it again and again, it will fail because you're violating the constraints of primary key (since you're adding duplicate rows).
Problem I had with this:
This seems like it should only be used to provide sample data when you're testing the application, not for production level. Plus I'd have to continously create a new initializer for each context, which plays a similar problem noted in solution 1 of maintainability. There is nothing automatic happening here. But if you want a way to inject sample code without hooking up to a database, this is a great solution.
Solution 3: Entity framework with Repository (In-memory persistence)
I got this solution from this website:
http://www.roelvanlisdonk.nl/?p=2827
He first sets up an edmx file, using EF5 and the code generator templates for EF5 dbcontexts you can get from VS extension libraries.
He first uses the edmx to create the contexts and changes the tt templates to bind to the repository class he made, so that the repository will keep track of the datacontext, and provide the options of querying and accessing the data through the repository; in his website though he calls the repository as MemoryPersistenceDbSet.
The templates he modified will be used to create datacontexts that will bind to an interface (IEntity) shared by all. Doing it this way is nice because you are establishing a Dependency Injection, so that you can add any entity you want through the T4 templates, and there'd be no complaints.
Advantage of this solution:
Wrapping up the edmx in repository pattern allows you to leverage the n-tier architecture, so that any changes done to the backend won't affect the front end, and allows you to separate the interface between the front end and backend so there are no coupled dependencies. So maybe later on, I can replace my edmx with petapoco, or massive, or some other ORM, or switch from in-memory persistence to fetching data from a database.
I followed everything exactly as explained. I made one modification though:
In the t4 template for .Context.tt, where DbSetInConstructor is added, I had the code written like this:
public string DbSetInConstructor(EntitySet entitySet)
{
return string.Format(
CultureInfo.InvariantCulture,
“this.{1} = new BaseRepository();”,
_typeMapper.GetTypeName(entitySet.ElementType), entitySet);
}
Because in my case I had the entityset = Persons and entityname = Person. So there’d be discrepancy. But this should cover all bases.
Final step:
So whether you picked solution 1, 2, or 3. You have a method to automatically populate your application. In these cases, the stubs are embedded in the code. In my case, what I've done is have my web server (containing my front end app), contact my data server, have the data server query the database. The data server will receive a dataset, serialize it, and pass it back to the web server. The web server will take that dataset, deserialize it, and auto-map to an object collection (list, or enumberable, or objectcollection, etc).
I would post the solutions more fully but there's way too much detail between all 3 of these solutions. Hopefully these solutions would point anyone in the right direction.
Dependency Injection
If anyone wants some information about how to allow DI to api controllers, Peter Provost provides a very useful blog that explains how to do it. He does a very very good job.
http://www.peterprovost.org/blog/2012/06/19/adding-ninject-to-web-api/
few more helpful links of repository wrapping up edmx:
http://blogs.msdn.com/b/wriju/archive/2013/08/23/using-repository-pattern-in-entity-framework.aspx
http://www.codeproject.com/Articles/688929/Repository-Pattern-and-Unit-of
I have the following scenario:
I am using EF with Repository pattern and Unit Of Work from here
Then I extended my partial class and have created a Finder class which have GetAll and other methods.
You can this see below:
Below you can see that I am using a unit of work class with repositories of all classes to get the instance from the generic repository:
protected Repository<Category> _CategoryRepository;
public Repository<Category> CategoryRepository
{
get { return _CategoryRepository ?? (_CategoryRepository = new Repository<Category>(context)); }
}
So in this way I had different repositories and when getting an entity from db and updating it from a different context caused problems. So I used this method to use for context lifetime management. And it resolved that issue.
Now the problem I am facing is in the following code:
var cat = Instance.Category.GetSingle(c => c.CategoryID == 7);
var orignal = cat.CategoryName;
var expected = cat.CategoryName + " Test Catg Update";
cat.CategoryName = expected;
cat.Update(); //This doesn't actually update due to a validation in place (which is correct)
cat = Instance.Category.GetSingle(c => c.CategoryID == 7);
Assert.AreEqual(cat.CategoryName, expected);
When I use Update and I have some validators in it, and the Update fails (for example due to size of the string exceeds 15 characters). When I try to call GetSingle (2nd last line of above code) again, it brings me the same old record which is [cat.CategoryName + " Test Catg Update"]. Is this a normal way? If not how can this be fixed so I can reload the object from database.
Let me know if you need any other code or reference.
in your Update() method, if the validation fails, simply set the state of the entity to EntityState.Unchanged. Under the covers, changing the state of an entity from Modified to Unchanged first sets the values of all properties to the original values that were read from the database when it was queried, and then marks the entity as Unchanged. This will also reject changes to FK relationships since the original value of the FK will be restored.
You may be able to take advantage of the ObjectContext.Refresh() method to handle this as well. However, this is on the ObjectContext, so you would need a method to handle it in your UoW. The refresh method would be something like context.Refresh(RefreshMode.ServerWins, entity);
My question is this: How can you implement a default server-side "filter" for a navigation property?
In our application we seldom actually delete anything from the database. Instead, we implement "soft deletes" where each table has a Deleted bit column. If this column is true the record has been "deleted". If it is false, it has not.
This allows us to easily "undelete" records accidentally deleted by the client.
Our current ASP.NET Web API returns only "undeleted" records by default, unless a deleted argument is sent as true from the client. The idea is that the consumer of the service doesn't have to worry about specifying that they only want undeleted items.
Implementing this same functionality in Breeze is quite simple, at least for base entities. For example, here would be the implementation of the classic Todo's example, adding a "Deleted" bit field:
// Note: Will show only undeleted items by default unless you explicitly pass deleted = true.
[HttpGet]
public IQueryable<BreezeSampleTodoItem> Todos(bool deleted = false) {
return _contextProvider.Context.Todos.Where(td => td.Deleted == deleted);
}
On the client, all we need to do is...
var query = breeze.EntityQuery.from("Todos");
...to get all undeleted Todos, or...
var query = breeze.EntityQuery.from("Todos").withParameters({deleted: true})
...to get all deleted Todos.
But let's say that a BreezeSampleTodoItem has a child collection for the tools that are needed to complete that Todo. We'll call this "Tools". Tools also implements soft deletes. When we perform a query that uses expand to get a Todo with its Tools, it will return all Tools - "deleted" or not.
But how can I filter out these records by default when Todo.Tools is expanded?
It has occurred to me to have separate Web API methods for each item that may need expanded, for example:
[HttpGet]
public IQueryable<Todo> TodoAndTools(bool deletedTodos = false, bool deletedTools = false)
{
return // ...Code to get filtered Todos with filtered Tools
}
I found some example code of how to do this in another SO post, but it requires hand-coding each property of Todo. The code from the above-mentioned post also returns a List, not an IQueryable. Furthermore this requires methods to be added for every possible expansion which isn't cool.
Essentially what I'm looking for is some way to define a piece of code that gets called whenever Todos is queried, and another for whenever Tools is queried - preferably being able to pass an argument that defines if it should return Deleted items. This could be anywhere on the server-side stack - be it in the Web API method, itself, or maybe part of Entity Framework (note that filtering Include extensions is not supported in EF.)
Breeze cannot do exactly what you are asking for right now, although we have discussed the idea of allowing the filtering of "expands", but we really need more feedback as to whether the community would find this useful. Please add this to the breeze User Voice and vote for it. We take these suggestions very seriously.
Moreover, as you point out, EF does not support this.
But... what you can do is use a projection instead of an expand to do something very similar:
public IQueryable<Object> TodoAndTools(bool deleted = false
,bool deletedTools = false) {
var baseQuery = _contextProvider.Context.Todos.Where(td => td.Deleted == deleted);
return baseQuery.Select(t => new {
Todo: t,
Tools: t.Tools.Where( tool => tool.Deleted = deletedTools);
});
}
Several things to note here:
1) We are returning an IQueryable of Object instead of IQueryable of ToDo
2) Breeze will inspect the returned payload and automatically create breeze entities for any 'entityTypes' returned (even within a projection). So the result of this query will be an array of javascript objects each with two properties; 'ToDo' and 'Tools' where Tools is an array of 'Tool' entities. The nice thing is that both ToDo and Tool entities returned within the projection will be 'full' breeze entities.
3) You can still pass client side filters based on the projected property names. i.e.
var query = EntityQuery.from("TodoAndTools")
.where("Todo.Description", "startsWith", "A")
.using(em);
4) EF does support this.
I have read so much (dozens of posts) about one thing:
How to unit test business logic code that has Entity Framework code in it.
I have a WCF service with 3 layers :
Service Layer
Business Logic Layer
Data Access Layer
My business logic uses the DbContext for all the database operations.
All my entities are now POCOs (used to be ObjectContext, but I changed that).
I have read Ladislav Mrnka's answer here and here on the reasons why we should not mock \ fake the DbContext.
He said:
"That is the reason why I believe that code dealing with context / Linq-to-entities should be covered with integration tests and work against the real database."
and:
"Sure, your approach works in some cases but unit testing strategy must work in all cases - to make it work you must move EF and IQueryable completely from your tested method."
My question is - how do you achieve this ???
public class TaskManager
{
public void UpdateTaskStatus(
Guid loggedInUserId,
Guid clientId,
Guid taskId,
Guid chosenOptionId,
Boolean isTaskCompleted,
String notes,
Byte[] rowVersion
)
{
using (TransactionScope ts = new TransactionScope())
{
using (CloseDBEntities entities = new CloseDBEntities())
{
User currentUser = entities.Users.SingleOrDefault(us => us.Id == loggedInUserId);
if (currentUser == null)
throw new Exception("Logged user does not exist in the system.");
// Locate the task that is attached to this client
ClientTaskStatus taskStatus = entities.ClientTaskStatuses.SingleOrDefault(p => p.TaskId == taskId && p.Visit.ClientId == clientId);
if (taskStatus == null)
throw new Exception("Could not find this task for the client in the database.");
if (taskStatus.Visit.CustomerRepId.HasValue == false)
throw new Exception("No customer rep is assigned to the client yet.");
TaskOption option = entities.TaskOptions.SingleOrDefault(op => op.Id == optionId);
if (option == null)
throw new Exception("The chosen option was not found in the database.");
if (taskStatus.RowVersion != rowVersion)
throw new Exception("The task was updated by someone else. Please refresh the information and try again.");
taskStatus.ChosenOptionId = optionId;
taskStatus.IsCompleted = isTaskCompleted;
taskStatus.Notes = notes;
// Save changes to database
entities.SaveChanges();
}
// Complete the transaction scope
ts.Complete();
}
}
}
In the code attached there is a demonstration of a function from my business logic.
The function has several 'trips' to the database.
I don't understand how exactly I can strip the EF code from this function out to a separate assembly, so that I am able to unit test this function (by injecting some fake data instead of the EF data), and integrate test the assembly that contains the 'EF functions'.
Can Ladislav or anyone else help out?
[Edit]
Here is another example of code from my business logic, I don't understand how I can 'move the EF and IQueryable code' out from my tested method :
public List<UserDto> GetUsersByFilters(
String ssn,
List<Guid> orderIds,
List<MaritalStatusEnum> maritalStatuses,
String name,
int age
)
{
using (MyProjEntities entities = new MyProjEntities())
{
IQueryable<User> users = entities.Users;
// Filter By SSN (check if the user's ssn matches)
if (String.IsNullOrEmusy(ssn) == false)
users = users.Where(us => us.SSN == ssn);
// Filter By Orders (check fi the user has all the orders in the list)
if (orderIds != null)
users = users.Where(us => UserContainsAllOrders(us, orderIds));
// Filter By Marital Status (check if the user has a marital status that is in the filter list)
if (maritalStatuses != null)
users = users.Where(pt => maritalStatuses.Contains((MaritalStatusEnum)us.MaritalStatus));
// Filter By Name (check if the user's name matches)
if (String.IsNullOrEmusy(name) == false)
users = users.Where(us => us.name == name);
// Filter By Age (check if the user's age matches)
if (age > 0)
users = users.Where(us => us.Age == age);
return users.ToList();
}
}
private Boolean UserContainsAllOrders(User user, List<Guid> orderIds)
{
return orderIds.All(orderId => user.Orders.Any(order => order.Id == orderId));
}
If you want to unit test your TaskManager class, you should employ the Repository dessign pattern and inject repositories such as UserRepository or ClientTaskStatusRepository into this class. Then instead of constructing CloseDBEntities object you will use these repositories and call their methods, for example:
User currentUser = userRepository.GetUser(loggedInUserId);
ClientTaskStatus taskStatus =
clientTaskStatusRepository.GetTaskStatus(taskId, clientId);
If yout wanto to integration test your TaskManager class, the solution is much more simple. You just need to initialize CloseDBEntities object with a connection string pointing to the test database and that's it. One way how to achieve this is injecting the CloseDBEntities object into the TaskManager class.
You will also need to re-create the test database before each integration test run and populate it with some test data. This can be achieved using Database Initializer.
There are several misunderstandings here.
First: The Repository Pattern. It's not just a facade over DbSet for unit testing! The repository is a pattenr strongly related to Aggregate and Aggreate Root concepts of Domain Driven Design. An aggregate is a set of related entities that should stay consistent to each other. I mean a business consistency, not just only a foreign keys validity. For example: a customer who have made 2 orders should get a 5% discount. So we should somehow manage the consistency between the number of order entities related to a customer entity and a discount property of the customer entity. A node responsible for this is an aggregate root. It is also the only node that should be accessible directly from outside of the aggregate. And the repository is an utility to obtain an aggregate root from some (maybe persistent) storage.
A typical use case is to create a UoW/Transaction/DbContext/WhateverYouNameIt, obtain one aggregate root entity from the repository, call some methods on it or access some other entities by traversing from the root, Commit/SaveChanges/Whatever. Look, how far it differs from yur samples.
Second: The Business Logic. I've already showed you one example: a customer who have made 2 orders should get a 5% discount. In contrary: your second code sample is not a business logic. It's just a query. The responsibility of this code is to obtain some data from the storage. In such a case, the storage technology behind it does matter. So I would recomend integration tests here rather than pretending the storage doesn't matter when interacting with the storage is the sole purpose of this function.
I would also encapsulate that in a Query Object that was already suggested. Then - such a query object could be mocked. Not just DbContext behind it. The whole QO.
The first code sample is a bit better because it probably ivolves some business logic, but that's dificult to identify. Wich leads us to the third problem.
Third: Anemic Domain Model. Your domain doesnt' look very object oriented. You have some dumb entities and transaction scripts over them. With 7 parameters! Thats pure procedural programming.
Moreover, in your UpdateTaskStatus use case - what is the aggregate root? Befere you answer that, the most important question first: what exactly do you want to do? Is that... hmm... marking a current task of a user done when he was visited? Than, maybe there should be a method Visit() inside a Customer Entity? And this method should have something like this.CurrentTaskStatus.IsCompleted = true?
That was just a random guess. If I missed, that would clearly show another issue. The domain model should use the ubiquitous language - something common for the programmer and a business. Your code doesn't have that expressive power that a common language gives. I just don't know what is going on there in UpdateTaskStatus with 7 parameters.
If you place proper expressive methods for performing business operations in your entities that will also enforce you to not use DbContext there at all, as you need your entities to stay persistence ignorant. Then the problem with mocking disappears. You can test the pure business logic without persistence concerns.
So the final word: Reconsider your model first. Make your API expressive by using ubiquitous language first.
PS: Please don't treat me as an authority. I may be completely wrong as I'm just starting to learn DDD.