As soon as I started writing LightSwitch applications I noticed that many times I was repeating the same code over and over for trivial tasks. So after all this time I have collected a number of extension methods that I widely use in my apps.
For me reusing code is a
must and although the implementation of LS (IMHO) does not provide for this out of the box the underlying framework is ideal for writing extension classes and methods that are a major step towards code reusability. If you have downloaded any of my samples from msdn or have seen my
Application Logo post, you already suspect I am an “extension method fanatic”.
So I will present a small series (I don’t know how small) of posts with extension methods from my Base.LightSwitch.Client library.
The first method is one of the first (if not the first one) extension methods I wrote. As soon as you want to override the code for default commands like Edit and Delete for a collection (let’s name it MyCollection) you have to write something like that:
partial void MyCollectionDelete_CanExecute(ref bool result){
if (this.MyCollection is null || this.MyCollection.SelectedItem == null)
result = false;
else
result = true;
}
This is the minimum code (it can be written more elegantly I know but this is the concept) you need to write. I don’t take into account the permissions issue.
A similar chunk of code has to be written for Edit.
Isn’t the code listed below easier to read:
partial void MyCollectionDelete_CanExecute(ref bool result){
result = this.HasSelection("MyCollection");
}
It’s not only easier to read but is less error prone than the original. Plus you can inject any security logic in this HasSelection method.
And this is the code:
public static bool HasSelection(this IScreenObject screen, string collectionName) {
if (!screen.Details.Properties.Contains(collectionName))
return false;
IVisualCollection collection =
screen.Details.Properties[collectionName].Value as IVisualCollection;
return collection != null &&
collection.SelectedItem != null &&
(collection.SelectedItem as IEntityObject).Details.EntityState != EntityState.Deleted;
}
For the ones that dislike using string property names I can suggest this version :
public static bool HasSelection<T>(this VisualCollection<T> collection) where T : class, IEntityObject {
return collection != null &&
collection.SelectedItem != nul &&
collection.SelectedItem.Details.EntityState != EntityState.Deleted;
}
This version is more concrete is generic and also does not have to do the (out of thin air) conversion of SelectedItem to IEntityObject. If you use this version though you have to change your partial method as you cannot extend null:
partial void MyCollectionDelete_CanExecute(ref bool result){
result = MyCollection!= null && MyCollection.HasSelection();
}
The choice is yours…