C# Probably Getting New "Safe Navigation" Operator "?."

It looks as if the Visual Studio dev team may be implementing a new operator in a future .NET release. This is due in large part to community demand, which is pretty cool because it shows that the VS team is listening to their customer base; a key part of a successful product.

This new operator is likely going to take the syntax of ?. and is known as the Safe Navigation Operator.

Its purpose is rather simple, and it eliminates a load of boiler-plate code by making the compiler do it for you.

Say you have these classes:

public class A
    public B B { get; set; }

public class B
    public C C { get; set; }

public class C
    public int D { get; set; }

It’s plain to see above that it’s possible that an instance of A may have a null B property, or even that an instance of B may have a null C property. If you want to get D from A safely (so as to avoid null reference exceptions), your code often ends up looking something like this:

int? result = A.B == null ? (int?)null : (A.B.C == null ? (int?)null : A.B.C.D);

What an incredibly annoying pain to write out, and even to read. You could maybe make it clearer, at the cost of more verbosity and lines of code:

int? result = null;
if (A.B != null)
    if (A.B.C != null)
        result = A.B.C.D;

This is still a bit convoluted and boiler-plate for my liking. Really, all we want to do in English is get the integer value of D if it exists on A. It shouldn’t be so challenging!

Enter the new Safe Navigation Operator and how it might function:

int? result = A?.B?.C.D;

Note the ?. syntax that is placed at the point where the A instance references B and B instance references C. This is much more readable, and intuitive to look at. To me, it says “if A’s B is not null and B’s C is not null, return D, else return null” which is pretty streamlined. Upon simple inspection, it also clearly says that A.B is nullable and B.C is nullable, which are great side effects of the shortened syntax.

Note that the final product, if included in C#, may behave differently than I’ve demonstrated and hypothesized here. The code in this post is mostly made on assumptions of compiler rules for this operator, and could in the future prove to be wrong. However, even still, this operator will be a great addition to C# and .NET if it makes the cut!

See also

comments powered by Disqus