• Replaces reference solution with something less verbose
• Removes reduntant test groups
• Replaces negated logic in solutions

lgtm, approved

I think this might be an artefact from one of translations I submitted, where I translated tests directly and used author's reference solution (because it's author's), but used my own solution as "Complete solution" snippet, because it's more compact (even if still not ultra-compact).

I would definitely go with some more compact reference solution in tests, but I did not want to change too much, considering the author is active and I already stepped on their toes by submitting a bunch of translations and self-approving a couple of them.

sixty lines of reference solution for something that can be done in about three ?!? It doesn't even fit on the screen!

It's good practice to have different example and reference solutions, but in this case, if you have validated the example solution, use that as the reference one as well and throw out the mess that is probably the Python reference solution. And try to make it a bit more compact; less to read is better readability in itself.

.

Lua translation

looks good to me; someone else has already approved it

Hi, I have made a fork, please review.

this should not allow solutions that use global variables to keep track of the state, as the goal of the kata is to use closures. could you make a fork that fixes that ?

approved

Lua translation

Approved

Lua translation

Hi, I have wrapped all the assertions now.

why do you hardcode these repetitive assertion messages instead of wrapping the assertion into a function ?