to create the dictionary only once, you could write like in the kumite

Thank you, switching to C# 10 solved the problem.

Try switching between C# 10 and C# 12, one of these should work.

I will take a look later why the non-working version is available, it should not.

The solution is wrong, depending on n. I accidentally submitted it.

I can't factor the solution because I get this error message:
"Assert does not contain a definition for 'AreEqual'"

c#: in the base tests, “Expected” and “But what” are interchanged

yes, funny, there are also no invalidatet solutions for this kata.

I'm not sure how to answer this, other than to say they did.

(edit) It's been a few months since I solved this, I see your objection now. I believe I solved this assuming all shoe sizes (per foot) were distinct. Reviewing my solution, this is not a safe assumption and adding tests with duplicated sizes may aid in weeding out false positive solutions.

c#:
the function Ipow() called in TestRandomNumbers() in Test Cases (but not in Test or Attempt) doesn't seem to be correct, it raises an OverflowException if b is rather big. Could Test Cases get updated?:

 private static int Ipow(int b, int e)
 {
     int p = 1;
     for (; e > 0; e >>= 1)
     {
         if ((e & 1) == 1) p *= b;
         b *= b;
     }
     return p;
 }

This solution is displayed at the top when sorting by newest. I would just be curious how this solution could pass tests.

Looks like sample tests expect uint as argument, but Attempt tests expect int. You should modify line 9 static void Act(uint expected, uint n) to static void Act(uint expected, int n), and you should raise this as an issue with the kata :P

c# 10/2024
In “Test” mode, but not in “ATTEMPT” mode, this error message

appears after reset:

tests/Fixture.cs(10,55): error CS1503: Argument 1: cannot convert from 'uint' to 'int'

Should I change the type of the argument n from int to uint?

Changing the return type to long is insufficient, but required. (unless you're really lucky with tests?)
You must also cast the long to int, as some random tests expect the overflow value.

The c# solution is currently not solvable without this realization.

CS:
I think the labeling in the error message is reversed for the following tests: RandomBig, RandomFull, RandomTestsAlmostEmpty, and RandomTestsManyShare

Example:
RandomTestsAlmostEmpty

Test Failed
Expected: null

But was: < 66, 16, 4, 55, 76, 55, 36, 5, 34, 1, 35, 57, 60, 86, 6, 12, 16, 66, 60, 26, 19, 84, 40, 66, 26, 43, 26, 60, 26, 66, 43, 23, 77, 26, 50, 23, 66, 60, 74, 88, 43, 23, 12, 38, 6, 60, 85, 43, 57, 60, 6, 38, 21, 32, 20, 50, 74, 50, 50, 16, 40, 6, 2, 79, 43, 6, 66, 83, 77, 79, 12, 66, 2, 66, 32, 43, 40, 50, 26, 79, 86, 57, 6, 21, 79, 12, 43, 36 >

㊗