IntelliSense uses wrong types for lambdas passed using named parameters

This issue has been moved from a ticket on Developer Community.

---

If you have a method which accepts multiple lambdas as arguments and use named parameters to pass the lambdas in a different order, IntelliSense within a passed lambda body will use the wrong types for completion against the lambda's parameters in the lambda body. That was a big sentence.

void MethodWithMultipleLambdaArguments(
    Action<(int x, int y)> action,
    Action<IEnumerable<string>> checkThings = null)
{
    checkThings?.Invoke(new[]
    {
            "a",
            "b",
        });
    action((1, 2));
}

public void DoSomething()
{
    MethodWithMultipleLambdaArguments(checkThings: things =>
    {
        things.«here»
    }, action: pair =>
    {

    });
}

If you take the above code in a C# class in Visual Studio, remove the “«here»”, and then try completion from where “«here»” was, you will see something like this:

I think that there is something like incomplete type information being calculated because the lambda has a syntax error, preventing it from being fully resolved by the C# compiler. However, in this case, with named parameters, and a limited number of possible matching method overrides (in this case, exactly one), the compiler should have enough information to calculate the correct types for the lambda’s parameter.

---

Original Comments

Visual Studio Feedback System on 9/16/2019, 03:48 AM:

Nathan Phillip Brink (binki) on 9/16/2019, 06:43 AM:

---

Original Solutions

(no solutions)