Add quickfix that swaps actual and expected value

[inyutin-maxim]

For example:

"video1234_1234_test".Should().Be(result);

should be replaced with:

result.Should().Be("video1234_1234_test");

[meziantou]

Good catch! What is the original assert code? Also, is the code using NUnit or xUnit?

[inyutin-maxim]

Good catch! What is the original assert code? Also, is the code using NUnit or xUnit?

This code from my very old project, tests writen used NUnit.
Original Assert:

Assert.AreEqual(result, "video1234_1234_test");
                ^^^^^^^, ^^^^^^^^
               expected, actual

but i think this problem can reproduce and for xUnit

[meziantou]

It seems you swapped the expected and actual parameters in the original call (documentation).

The analyzer uses the following semantic:

Assert.AreEqual(expected, actual); // expected is the first parameter
actual.Should().Be(expected)

Assert.AreEqual(result, "video1234_1234_test");
// Should be
Assert.AreEqual("video1234_1234_test", result);

[inyutin-maxim]

It seems you swapped the expected and actual parameters in the original call (documentation).

I did this deliberately to show what the problem was, but in those distant times it was not at all obvious to me.
I started this problem to have a convenient tool for automatic replacement. I understand that a deeper semantic analysis is needed to smartly identify a potential error, and a warning need not always be shown, but this could make life easier. If you consider the problem redundant, you can close it

[inyutin-maxim]

I see an order like this:

1. I apply the fix for the original code:

Assert.AreEqual(result, "video1234_1234_test");

2. I get the following code:

"video1234_1234_test".Should().Be(result);

3. I apply the next fix for the new code:

"video1234_1234_test".Should().Be(result);

4. I get the following code:

result.Should().Be("video1234_1234_test");

[meziantou]

I thought you were reporting a bug, but it's a feature request!

Something we can try is detecting constant values on the left side and report them because I don't think there is a valid reason to have a constant before "Should".

[inyutin-maxim]

yes you are right this new functionality, perhaps I did not exactly express myself when describing