I am trying to understand Ninject Contextual Binding. I understand the scenarios where I know my context at design time. e.g. I understand that I can use Named Attributes to Bind the DB object to a mock DB when I want to use it in a test class and to a SQL DB when I use it from my actual code.
However, I don't know how to handle contextual Binding at runtime. e.g. let's say I am writing software for a shopping center. The shopkeeper can use a keyboad for billing or a barcode scanner. I don't know which one he will use beforehand. And he might add other ways of scanning like RFID sometime in the future.
So I have the following:
interface IInputDevice
{
public void PerformInput();
}
class KeyboardInput : IInputDevice
{
public void PerformInput()
{
Console.Writeline("Keyboard");
}
}
class BarcodeInput : IInputDevice
{
public void PerformInput()
{
Console.Writeline("Barcode");
}
}
class Program
{
static void Main()
{
IKernel kernel = new StandardKernel(new TestModule());
var inputDevice = kernel.Get<IInputDevice>();
inputDevice.PerformInput();
Console.ReadLine();
}
}
public class TestModule : Ninject.Modules.NinjectModule
{
public override void Load()
{
Bind<IInputDevice>().To<....>();
}
}
So, how can I go about it in the least amount of custom code? I would like to request specific code examples and not links to articles/wikis/tutorials on contextual binding.
You will need some criteria to decide which one shall be used. E.g. App.config or device detection. Then use conditional bindings:
Bind<IInputDevice>().To<KeyboardInput>().When(KeyboardIsConfigured);
Bind<IInputDevice>().To<BarcodeInput>().When(BarcodeReaderIsConfigured);
public bool KeyboardIsConfigured(IContext ctx)
{
// Some code to decide if the keyboard shall be used
}
public bool BarcodeReaderIsConfigured(IContext ctx)
{
// Some code to decide if the barcode reader shall be used
}