C＃MVVM，了解何时不制作模型

Question

I Everyone.

I am a big fan of MVVM But need to clarify how I should Implement a general method that could be used across several view models.

Currently I have implemented a method in to a model to take a screen shot like so EG

Interface

public interface ICapture {
 void CaptureMethod();
}

This is the model that implements that interface

public class CaptureModel : ICapture {
    void CaptureMethod(){ //implement the code to take screen shot }
}

Now I want to execute a screen shot on one or multiple view models, so I would need to instantiate the and call the function like so.

public ViewModel(){
 void TakeScreenShotMethodOrCommandDontCare(){
   ICapture captureClass = new CaptureClass();
   captureClass.captureMethod
  }
}

Which just feels WRONG having to instantiate an instance of an object in order to call a function which will be taking a screenshot.

I suppose i could go in the view model and have capture screenshot method but that would lead to code duplication across all view models with that feature.

Maybe a static utility class but surely that would be expensive on the app.

I feel if I had an interface that was implemented by a base class then my view model could inherit from that base class and implement the base implementation on any view model I desire.

So to sum up my question, if you wanted to implement a method that captures screen shots and you like mvvm then where would you implement ??

Again many thanks to all that have taken the time to shine some light on this for me.

Answer 1

This is actually what I would consider well within the MVVM standard. When dealing with MVVM, I find it best to think about it this way. Your View should know only about your ViewModel and the ViewModel should know nothing of the View. The ViewModel should know about your model (or models) but the model should know nothing of the ViewModel or View.

So, here your Model is being instantiated in the ViewModel which is perfectly valid! In this case, your Model is the CaptureClass and the ViewModel knows that it can take a string from that and provide it to some View that may want to use it.

Model:

class CaptureClass : ICapture
{       
  public string CaptureShot()
  {
     string s = DependencyService.Get<ICapture>().CaptureShot();
     return s;
  }      
}

ViewModel:

public class CaptureViewModel(){

   private _captureText;

   public CaptureText
   {
       get{ return _captureText; }
       set{
           _captureText = value;
       }
   }

   public CaptureViewModel()
   {
       CaptureClass cc = new CaptureClass();
       _captureText = cc.CaptureShot();
   }
}

Now all you've got to do is hook it up to your View. Remember that the View knows which property to bind to so just set your DataContext either in the code behind or in XAML and then just bind the property in XAML like so:

<Label Content="{Binding CaptureText}" />

Answer 2

The answer has been found and involved injecting a service class in to the view model I wish to use the methods in.

Link to answer.

https://softwareengineering.stackexchange.com/questions/380036/c-mvvm-understanding-when-not-to-make-a-model