tempusfugit (128) [Avatar] Offline
#1
I realize that the code isn't production strength and that it needs to be simple to get the point across. However there is one aspect about the code that strikes me as inconsistent. FuncList.Cons takes advantage of the immutability of the "tail list" and simply reuses it to construct the new list. However the most common "tail list", the empty list (instance) is always duplicated whenever you start a new list.

A couple more lines of code could make "The Empty List<T>" (that is, the empty list for a particular type parameter T) reusable without getting into a separate class or using the Special Case/Null Object pattern. Granted, there is little to be gained from saving a few instances of empty list object instances. However I think a shared empty list better communicates the idea that it is safe to share any "tail list" between lists thanks to the immutability of the lists themselves.

[pre]
public class FuncList<T> {

public FuncList(T head, FuncList<T> tail) {
IsEmpty = false;
Head = head;
Tail = tail;
}

public bool IsEmpty { get; private set; }
public T Head { get; private set; }
public FuncList<T> Tail { get; private set; }

// Enforced "Just Create One"
// http://butunclebob.com/ArticleS.UncleBob.SingletonVsJustCreateOne
//
public static FuncList<T> EmptyList { get { return emptyList; } }

private static readonly FuncList<T> emptyList = new FuncList<T>();
private FuncList() { IsEmpty = true; } // For the EmptyList only
}

public static class FuncList {

public static FuncList<T> Empty<T>() {
return FuncList<T>.EmptyList;
}

public static FuncList<T> Cons<T>(T head, FuncList<T> tail) {
return new FuncList<T>(head, tail);
}
}
[/pre]