claps

mmmmmmm ... there are many ways to resolve this task without using subqueries. I dont think this is a "best practices".

Right. But things go wrong. Or there's maintenance. And if there's something to change it ought to be encouraged^^

The process of making an update kind of sets an expecation of getting everything right the first time and never changing your mind later

A translation is essentially supposed to be just a "port" of the original - if you want to be more creative then you should author a new Kata ;-)

Approved again.

The process of making an update kind of sets an expecation of getting everything right the first time and never changing your mind later, makes me feel silly when there's something to change x).

I snuck in some other edits to the error reporting (learned some new things about how error messages can be supplied because of this bug :D)

It no longer loops infinitely on error (and shows what input was used when that happens), but I still changed the default code so that it does not crash because the error message does get a little verbose and I know some people refuse to read when they get a bit on the dense side (even though the text is both plain and relevant)

Please fork, else I will mess it up. I don't know Haskell. I will approve asap.

Oh uhm. Found a bug. I created an infinite loop in the shrinker for empty input.
(When the input is "" it suggests shrinking to "" which fails so it shrinks "" into "" and so on)
Do I fork or is that something to edit directly?

  shrink (TestZoo []) = []  -- add this line
  shrink (TestZoo zoo) =
    let items = splitOn "," zoo
     in TestZoo . intercalate "," <$> subOneLists items

Approved. Thanks!

Oh. Okay. Good catch. That's definitely me not paying attention. Gone now!

randomUnknown = elements ["unidentified flying object", "cucumber", "peanut"]
randomKnown = elements ["busker", "banana", "bicycle", "antelope", "big-fish", "bug", "bear", "chicken", "cow", "fox", "giraffe", "grass", "leaves", "lion", "little-fish", "panda", "sheep"]

There are already some unknown things at the zoo - "busker", "banana", "bicycle"... So I don't understand what is the purpose of the randomUnknown elements that was introduced?

I also guess the instruction was not clear enough

I'm exploiting a defecting in your testing to pass this. I suggest you improve the test harness.

Walking every group entirely ( length ) when all you're interested in is the second : constructor is not a Best Practice.

There is no advantage and this is not at all best practice, it just makes the query less readable.

The sql optimizer is so good that in 99.99999999999% of cases you never have to think about optimization, only readability.

x2