Updated the description: case-insensitively-alphabetical-order-of-appearance ought to be enough of a tearse spec for anyone with decent understanding of the spec language (English in the original case). Translations of katas is a feature I think could be really useful, but of course as with all translations, the output depends not only on the original text but also on the translator :)

Thanks for the good feedback @JohanWiltink. Each time I write a kata I have already forgotten what is and is not included in the "CodewarsSpec" test class, and for some reason the documentation always seems far/further than one click away so... 😄 When I eventually checked it (because I wondered about where you were supposed to stash the reference solution that you need in case of random test input) I did not notice the Test.randomize(array) → Array, Test.randomNumber() → Integer, Test.randomToken() → String, Test.sample(array) → object since they were are just tacked onto the end without even so much as a comment.

Thanks again anyway, I'll see if I can make use of them in the future.

good point, sortBy is what we need - should be possible to do this without lodash as well

localeCompare, sensitivity base, wow I didn't know about that

I was renaming the function / the docs were lagging. Dang you guys are fast on the ball, hardly had time to hit Save :D

OK if that's how you wanna play this game - I can see some hardcoded solutions, so I'm gonna have to add radomized inputs for the final test.

EDIT Done. Should invlidate some existing solutions @JohanWiltink ;-) Also reclassified as "Puzzle", in case people now have so many issues with reading the spec.

Can you explain what this instability involves, and what it means for the kata solution?

I'm not opposed to tagging it "puzzle". But it would be better if we could figure out a spec that explains the desired output better.

Expected BbeehHilloTy do you want capitals first or lowercase?

I want it exacly as it is: the letter present in the input, in alphabetic order, stacked in the order (from left-to-right) that they appear in the input. Obviously irrelevant of casing, since letters excisted prior to computer encodings.

Avoid getting trapped only thinking in terms of ASCII codes. ;)

There are no a:s. For b:s, the uppercase one appears prior to the lowercase one, so all the b:s add up to "Bb". The first "h" is lowercase, the second one uppercase, so the h:s stack up to "hH".

It would have been more interesting if it had a negative number in the final test cases as well, but ok, perhaps better to keep it simple.

Again, any solution utilizing numbers.splice() mutates the original array, which specifically was not allowed according to the instructions. But the kata author forgot to assert that...

Or just use the spread operator Math.min(... numbers) instead.

(Ref. https://developer.mozilla.org/en/docs/Web/JavaScript/Reference/Operators/Spread_operator)

The solution is still invalid because it mutates the original array. The final kata test should have had a guard against that...

I seem to be one of the few to use .reduce() for accumulating the output! :-) I started out with .join() too, but decided to refactor it out.

Other observations: my normal linter would kill me if it saw me using && for side-effects and unary ops (++/--) inside predicates, but they're fun to use as shortcuts here! It also pleases me that I got a chance to use do/while in JS - something which doesn't seem to happen often.

If you go with just return 0 and just press RUN SUITE, the test output looks as follows:

Sampling getRandomInRange(0, 1) shows incorrect return values - Expected: '[0, 1]', instead got: '[0]'

and I can't think of any way to make that more clear: it says that it sampled calling with parameters (0, 1) and it did not return the expected (random) values [0 and 1], but instead got [only the value 0].

I suppose it looks more cryptic if the reader is less familiar with xUnit style assertion errors... Thanks for the input anyway, I will think about it - open to suggestions! :)

OMG what have I done?!? Such pain and misery! :-D