I'm having a hard time getting my head around what seems like an obvious pattern problem/limitation when it comes to typical constructor dependency injection. For example purposes, lets say I have an ASP.NET MVC3 controller that looks like:
Public Class MyController
Inherits Controller
Private ReadOnly mServiceA As IServiceA
Private ReadOnly mServiceB As IServiceB
Private ReadOnly mServiceC As IServiceC
Public Sub New(serviceA As IServiceA, serviceB As IServiceB, serviceC As IServiceC)
Me.mServiceA = serviceA
Me.mServiceB = serviceB
Me.mServiceC = serviceC
End Sub
Public Function ActionA() As ActionResult
' Do something with Me.mServiceA and Me.mServiceB
End Function
Public Function ActionB() As ActionResult
' Do something with Me.mServiceB and Me.mServiceC
End Function
End Class
The thing I'm having a hard time getting over is the fact that the DI container was asked to instantiate all three dependencies when at any given time only a subset of the dependencies may be required by the action methods on this controller.
It's seems assumed that object construction is dirt-cheep and there are no side effects from object construction OR all dependencies are consistently utilized. What if object construction wasn't cheep or there were side effects? For example, if constructing IServiceA involved opening a connection or allocating other significant resources, then that would be completely wasted time/resources when ActionB is called.
If these action methods used a service location pattern (or other similar pattern), then there would never be the chance to unnecessarily construct an object instance that will go unused, of course using this pattern has other issues attached making it unattractive.
Does using the canonical constructor injection + interfaces pattern of DI basically lock the developer into a "limitation" of sorts that implementations of the dependency must be cheep to instantiate or the instance must be significantly utilized? I know all patterns have their pros and cons, is this just one of DI's cons? I've never seen it mentioned before, which I find curious.
If you have a lot of fields that aren't being used by every member this means that the class' cohesion is low. This is a general programming concept - Constructor Injection just makes it more visible. It's usually a pretty good indicator that the Single Responsibility Principle is being violated.
If that's the case then refactor (e.g. to Facade Services).
You don't have to worry about performance when creating object graphs.
When it comes to side effects, (DI) constructors should be simple and not have side effects.
Generally speaking, there should be no major costs or side effects of object construction. This is a general statement that I believe applies to most (not all) objects, but is especially true for services that you would inject via DI. In other words, constructing a service class automatically makes a database/service call, or changes the state of your system in a way that would have side effects is (at least) a code smell.
Regarding instances that go unused: it's hard to create a system that has perfect utilization of instances within dependent classes, regardless of whether you use DI or not. I'm not sure achieving this is very important, as long as you are adhering to the Single Responsibility Principle. If you find that your class has too many services injected, or that utilization is really uneven, it might be a sign that your class is doing too much and needs to be split into two or more smaller classes with more focused responsibilities.
No you are not tied to the limitations you have listed. As of .net 4 you do have Lazy(Of T) at your disposal, which will allow you to defer instantiation of your dependencies until required.
It is not assumed that object construction is dirt-cheap and consequently some DI containers support Lazy(Of T) out of the box. Whilst Unity 2.0 supports lazy initialization out of the box through automatic factories, there is a good article here on an extension supporting Lazy(Of T) the author has on MSDN.
Isn't your controller a singleton though? That is the normal way to do it in Java. There is only one instance created. Also you could split the controller into multiple controllers if the roles of the actions is so distinct.
Related
I am trying to understand single responsibility principle. I have following questions.
The Single Responsibility Principle (SRP) states that there should never be
more than one reason for a class to change.
Usually our Resource,Service and Repository classes have
create,read,update and delete method. We are changing each class to
modify code for any any of these operations. Is it violating SRP? Do we need
separate class for each action?
When I run sonar lint, I have seen below message.
Classes should not coupled to too many other classes.
Here I am injecting other classes using spring DI. Is there any limit on
number of dependencies?
I may be missing crux of this concept. Please suggest a good resource for understanding this concept better with examples
The SRP states that the class should only do one thing, like persisting entities in the case of repositories. I guess that you've confused "class" and "object" here: if you have several methods that could change the object's state this could be in accordance with the SRP. However the only reason for a repository class to change should have something to do with its purpose, namely persisting or retrieving entities in this case.
The Wikipedia article about the Single Responsibility Principle puts it very well.
To your second point: there is no such thing as a maximum number of dependencies a class can have, but it could be a sign for a design weakness if there are many of them.
The Single Responsibility Principle (SRP) states that there should
never be more than one reason for a class to change.
The Single Responsibility principle doesn't mean a single method or a single type of action by component/class.
It means a single responsibility in the scope of a matter.
Persistence operations makes part of the same matter.
So putting all of them in a single class doesn't violate necessary the principle.
Now if you have dozen and dozen of specific database operations, it would make sense to divide them into distinct classes having a well defined responsibility such as selecting operations, updating operations, and so for.
Usually our Resource,Service and Repository classes have
create,read,update and delete method. We are changing each class to
modify code for any any of these operations. Is it violating SRP?
These are distinct layers.
If you change the model of a layer, others are very often impacted as data passes between layers.
It is like if you add an information in your database, you necessary need to change your GUIs and your processing if you want see/manipulate them.
Now if you change implementation of layers, other layers should have no or very few consequences.
I'm starting to develop website that use the spring framework.I have three controller.There are newCustomerController,editCustomerController and deleteCustomerController.These controllers are mapped with view that use for create update and delete, but I create only customer.
So, I would like to know.Is it appropriate to declare the controllers like this.
Thank
The answer to this question is subjective and maybe more a topic for https://softwareengineering.stackexchange.com/. However, there is something very spring related about it that I would like to comment.
There are a few principles that attempt at guiding developers of how to strike a good balance when thinking about designing the classes. One of those is the Single responsibility principle.
In object-oriented programming, the single responsibility principle
states that every class should have a single responsibility, and that
responsibility should be entirely encapsulated by the class. All its
services should be narrowly aligned with that responsibility
A catchier explanation is
A class or module should have one, and only one, reason to change.
However, its still often hard to reason about it properly.
Nevertheless, Spring gives you means for it (think of this statement as a poetic freedom of interpretation). Embrace constructor based dependency injection. There are quite a few reasons why you should consider constructor based dependency injection, but the part relevent to your question is adressed in the quote from the blog
An often faced argument I get is: “Constructors just get too verbose
if I have 6 or 7 dependencies. With fields only, this is fine”.
Awesome, you’ve effectively worked around a clear indicator that the
code you write is doing way too much. An increase in the number of
dependencies a type has should hurt, as it makes you think about
whether you should split up the component into multiple ones.
In other words, if you stick to constructor based injection, and your constructor turns a bit ugly, the class is most likely doing too much and you should consider redesigning.
The same works the other way around, if your operations are a part of the logical whole (like CRUD operations), and they use the same dependencies (now "measurable" by the count and the type of the injected deps) with no clear ideas of what can cause the operations to evolve independently of each other, than no reason to split to separate classes/components.
It should be better if you define one controller for Customer class and in that class you should have all methods related to customer operations (edit,delete,create and read).
I have an MVC3 project that uses Unity for dependency injection.
There is a main MVC3 project, a “domain” class library that sits between MVC3 and the data tier, and a bunch of class libraries that make up the data tier.
(MVC3) – (domain) – (data tier)
This is an example of one of the service constructors in the domain class:
public DomainModelCacheServices(
Data.Interface.ICountryRepository countryRepository,
Data.Interface.ILanguageRepository languageRepository,
Data.Interface.ISocialNetRepository socialNetRepository
)
Every time a controller is called that has DomainModelCacheServices in its constructor, a new DomainModelCacheServices object is constructed, plus the three repository classes in the constructor of DomainModelCacheServices.
I cannot believe this is efficient!
What makes this worse is that the class DomainModelCacheServices is a cache class. It loads lists of data that never change, and holds them as statics. But it still needs to construct three repository classes for every reference!
If I give DomainModelCacheServices the lifetime of a singleton (forever), I have to ensure it is thread-safe, and if the day comes when I am getting hundreds of hits, there’s going to be a lot of locking.
I could change the constructor to this:
public DomainModelCacheServices(
IServiceLocator serviceLocator
)
I don’t know why, but this doesn’t look right. The constructor becomes meaningless to the eye, and I have to reference Unity in the domain class and somehow make the domain class aware of the ServiceLocator owned by the MVC3 application. Maybe the loose-coupling can be too loose?
Maybe constructing all these classes is not as inefficient as it looks I shouldn’t worry about it?
What would be nice is if Unity supported “Lazy” constructor parameters. But it doesn’t.
So, any ideas on how to make an MVC3 + Unity project more efficient, specifically in the domain model design?
Thanks for reading!
The cache shouldn't be definied on the domain level but on the repositories implemntation level (so in DAL). So for example ICountryRepository should have two implementations in DAL : CountryRepository and ChachedCountryRepository. These should be wired as decorators in Unity (CountryRepository is inside the ChachedCountryRepository). CachedCountryRepository would check if the data is in the cache and if not it would pass the call to the inner CountryRepository.
Creating objects is not expensive and wouldn't care too much about issues as a caching is correctly definied.
Great reasoning.
However, creating objects is cheap. I would not create a singleton since you already are caching all objects in static fields. The current approach is easy to understand.
I got another question for you:
Why are you not caching in your repository classes?
The repositories are responsible for the data and all data handling should be transparent to everything else. It also makes everything easier since they are responsible of updating the data sources. How do you keep the cache in sync with changes today? Through domain events?
I would create a cache class which I would use as a private field in the repository.
I have a class called FooJob() which runs on a WCF windows service. This class has only 2 public methods, the constructor, and a Run() method.
When clients call my service, a Dim a new instance of the Job class, pass in some parameters to the ctor, then call Run()...
Run() will take the parameters, do some logic, send a (real time) request to an outside data vendor, take the response, do some business logic, then put it in the database...
Is it wise to only write a single unit test then (if even possible) on the Run() function? Or will I wind up killing myself here? In this case then should I drill into the private functions and unit test those of the FooJob() class? But then won't this 'break' the 'only test behavior' / public interface paradigm that some argue for in TDD?
I realize this might be a vague question, but any advice / guidance or points in the right direction would be much appreciated.
Drew
do some logic, send a (real time) request to an outside data vendor, take the response, do some business logic, then put it in the database
The problem here is that you've listed your class as having multiple responsibilities... to be truly unit testable you need to follow the single responsibility principle. You need to pull those responsibilities out into separate interfaces. Then, you can test your implementations of these interfaces individually (as units). If you find that you can't easily test something your class is doing, another class should probably be doing that.
It seems like you'd need at least the following:
An interface for your business logic.
An interface defining the request to the outside vendor.
An interface for your data repository.
Then you can test that business logic, the process of communicating with the outside vendor, and the process of saving to your database separately. You can then mock out those interfaces for testing your Run() method, simply ensuring that the methods are called as you expect.
To do this properly, the class's dependencies (the interfaces defined above) should ideally be passed in to its constructor (i.e. dependency injection), but that's another story.
My advice would be to let your tests help with the design of your code. If you are struggling to execute statements or functions then your class is doing too much. Follow the single-responsibility-priciple, add some interfaces (allowing you to mock out the complicated stuff), maybe even read Fowler's 'Refactoring' or Feather's 'Working With Legacy Code', these taught me more about TDD than any other book to date.
It sounds like your run method is trying to do too much I would separate it up but if you're overall design won't allow it.
I would consider making the internal members protected then inheriting from the class in your test class to test them. Be careful though I have run into gotchas doing this because it doesn't reset the classes state so Setup and TearDown methods are essential.
Simple answer is - it depends. I've written a lot of unit tests that test the behaviour of private methods; I've done this so that I can be happy that I've covered various inputs and scenarios against the methods.
Now, many people think that testing private methods is a bad idea, since it's the public methods that matter. I get this idea, but in my case the public method for these private calls was also just a simple Run() method. The logic of the private methods included reading from config files and performing tasks on the file system, all "behind the scenes".
Had I just created a unit test that called Run() then I would have felt that my tests were incomplete. I used MSTest to create accessors for my class, so that I could call the private methods and create various scenarios (such as what happens when I run out of disk space, etc).
I guess it's each to their own with this private method testing do/or don't do argument. My advice is that, if you feel that your tests are incomplete, in other words, require more coverage, then I'd recommend testing the private methods.
Thanks everyone for the comments. I believe you are right - I need to seperate out into more seperate classes. This is one of the first projects im doing using true TDD, in that I did no class design at all and am just writing stub code... I gotta admit, I love writing code like this and the fact I can justify it to my mangagment with years of backed up successful results is purely friggin awesome =).
The only thing I'm iffy about is over-engineering and suffering from class-bloat, when I could have just written unit tests against my private methods... I guess common sense and programmers gut have to be used here... ?
Warning acronym overload approaching!!! I'm doing TDD and DDD with an MVP passive view pattern and DI. I'm finding myself adding dependency after dependency to the constructor of my presenter class as I write each new test. Most are domain objects. I'm using factories for dependency injection though I will likely be moving to an IoC container eventually.
When using constructor injection (as apposed to property injection) its easy to see where your dependencies are. A large number of dependencies is usually an indicator that a class has too much responsibility but in the case of a presenter, I fail to see how to avoid this.
I've thought of wrapping all the domain objects into a single "Domain" class which would act as a middle man but I have this gut feeling that I'd only be moving the problem instead of fixing it.
Am I missing something or is this unavoidable?
Often a large number of arguments to a method (constructor, function, etc) is a code smell. It can be hard to understand what all the arguments are. This is especially the case if you have large numbers of arguments of the same type. It is very easy for them to get confused which can introduce subtle bugs.
The refactoring is called "Introduce Parameter Object". Whether that's really a domain object or not, it is basically a data transfer object that minimizes the number of parameters passed to a method and gives them a bit more context.
I only use DI on the Constructor if I need something to be there from the start. Otherwise I use properties and have lazy loading for the other items. For TDD/DI as long as you can inject the item when you need it you don't need to add it to your constructor.
I recommend always following the Law of Demeter and not following the DI myth of everything needs to be in the constructor. Misko Hevery (Agile Coach at Google) describes it well on his blog http://misko.hevery.com/2008/10/21/dependency-injection-myth-reference-passing/
Having a Layer Supertype might not be a bad idea, but I think your code smell might be indicating something else. Geofflane mentioned the refactor pattern, Introduce Parameter Object. While it's a good pattern for this sort of problem, I'm not entirely sure it's the way to go for this situation.
Question: Why are you passing in Domain Model objects to the constructor?
There is such a thing as having too much abstraction. If there's one solid layer of code you should be able to trust, it's your Domain Model. You don't need to reference 3 IEntity objects when you're dealing with Customer, Vendor, and Product classes if those are part of your basic Domain Model and you don't necessarily need polymorphism.
My advice: Pass in application and domain services. Trust your Domain Model.
EDIT:
Re-reading the problem when it's not horribly late at night, I realize your "Domain class" is already the Introduce Parameter Object refactoring and not, in fact, a Layer Supertype, as I thought at 3AM.
I also realize that perhaps you need to reference the Model objects in the application code, outside the Presenter. Perhaps you're doing some initial setup of your Model objects before passing them in to the Presenter. If this is the case, your "Domain class" idea might be best. If there is some initial setup, when moving to an IoC, you'll want to look at something like Factory Support in Castle Windsor. (Other IoC containers have similar concepts.)