How to avoid code duplication in MVVM? - silverlight

Question

This question is similar to Where does business logic sit in MVVM?
However, I didn't want to create a comment chain on that one
Lets say for example that I have a table of invoices and I want to get this data and perform some apportionment on it ready for use in 2 totally separate reports and 3 screens.
In our current web application I would have put this in the Data Service Layer, and all of my reports and screens would have called this
In MVVM people seem to suggest that the model should not be bloated out and that logic should be put in the view models. But in this case I be duplicating the code 5 times?
In his answer to my other question Reed states "Anything that's specific to the domain or business should be reusable by other applications, using other architectures."
Can Reed or someone clarify what my approach should be? Can MVVM be combined with other architectures?
I am using Silverlight 5 with the Simple MVVM Toolkit
Paul

Answer 1

The ViewModel is not for business logic. It is for user interface logic. As its name is given, it is representing the View. If you have five different reports that presents similar data, give this data a name and make all five ViewModels understand how to work with the data objects.
Since we are talking about reports, data is understood to e only displayed. The ViewModel can just be a simple data source with minimal user interface interaction, basically a very thin layer.

Answer 2

I faced the same problem and decided to go this way:
I created classes like controllers in MVC (performing some actions with my model) and I work with them in all ViewModels.
For example: our application has a list of books. We need to add/edit/delete them.
So we have a model:
public class Book {
public int BookId { get; set; }
public string Title { get; set; }
public string Author { get; set; }
}
Then we have a controller class:
public class BookController {
string dbPath = ...;
public void AddBook(string title, string author)
{
var book = new Book() { Title = title, Author = author };
AddBook(book);
}
public void DeleteBook(int id)
{
using (var db = new SQLiteConnection(dbPath))
{
db.Delete<Book>(id);
}
}
public void DeleteBook(Book book)
{
using (var db = new SQLiteConnection(dbPath))
{
DeleteBook(book.BookId);
}
}
public List<Book> GetAllBooks()
{
using (var db = new SQLiteConnection(dbPath))
{
return db.Table<Book>().ToList();
}
}
public Book FindBook(string title, string author, int id)
{
.....
}
}
Now we can use it wherever we need, e.g.:
public class BookListViewModel : ViewModelBase {
public BookListViewModel() {
GetData();
}
void GetData()
{
BookController bc = new BookController(); // here we start using our controller.
_books = new List<Book>();
_books = bc.GetAllBooks();
}
}
Such approach helps us:
1) keep all business logic separately (in controller class)
2) avoid code duplication