Approved.

1. Actual & Expected are flipped in the random test.

   test.assert_approx_equals

   test.assert_approx_equals(my_celsius_to_romer(1.0*temp), celsius_to_romer(1.0*temp))
   #                         ^^^^^^^^^^^^^^^^^^^            ^^^^^^^^^^^^^^^^
   

2. Please use random values for the random tests.

Rejected, since issue already resolved.

Diff --> this line added to the description.

Rejected, merge conflict & author changed the description without approving this fork.

_{Have 2 consolation upvotes for the effort instead. 🙂}

Approved

Fixed in your fork.

Approved.

Nice! Two more things.

1. Actual & expected are swapped in the random test

           expect    = new Person(personGuy).greet(whoGuy)
           actual    = "Hello #{whoGuy}, my name is #{personGuy}"
   #       ^^^^^^
   

2. This is not test breaking, but you can simplify the random test by using _.sample instead (Lodash ref)

           personGuy = persons[genIdx(persons.length-1)]
           whoGuy    = who[genIdx(who.length-1)]
           #           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
           #           can be replaced with _.sample(who), for example
   

Already approved by someone

Should use mocha (describe + it) & chai (assert.strictEqual) instead of the deprecated Codewars Test framework (Test.expect).

You can look at the Authoring Examples to see the modern way to structure the tests.

Fixed in this fork

• Removed unnecessary random seeding in the test _{(See this issue)}

• Enabled Python 3.11

This warning is only present in Crystal 0.34 (link to fork), but not in the current version (Crystal 1.0).

• Crystal 0.34.0 - Exhaustive Case

• Crystal 0.35.0 - Exhaustive Case (take 2)

Resolving...