Approved

As I know all correct input data have only one possible solution.
As for 1010 question is where to take input sample for 1010?

I can't create a generator of random solution, which can be used for random generating test.
I can't find regularity and think up an quick algorithm.
Now it takes about 50-300Ms for one new correct random clues. I think it is not enought for tests.
But I can generate in advance any amout of solutions
For example https://github.com/ekaruk/codewars/blob/master/Scrapers.txt
As I understand, these are right examples of Skyscrapers and clues with one solution.
I use little brute force for generation, so time of generation is not fixed.

You can add hundred solutions and use any random 5-10 from these fixed solutions for every test

People already hardcoded ten solutions; twenty won't be enough. A hundred maybe.

If you can code a solver that takes 54 ms, maybe you can code a similarly fast random generator? I'd love to have one.

cheaters. Reported, thx.

I posted a solution to this Kata this afternoon. Now I can see that another user by the name of GtpOtyf has posted a "variation" on my solution that's line-for-line identical. How is this possible?

Such problems are very well solved step by step.
Decision one.
Therefore on each step we can improve the last decision without brute force.

Interesting Kat series (44; 66; 77).
I am proud of myself :)
The first solution for 44 works with the minimum changes in indexes both on 66 and on 77.
For the size 7*7 calculation lasts on PC 54 milliseconds so possible will work also for the bigger size.
Plus 1 Kyu for one solution :)

One of options.
It is possible to add 20 in advance generated tests, but to check every time, for example, any random 5 of them.

Java translation done

Interesting approach. Two identical kata with different complexity
At first I solved it in easier version https://www.codewars.com/kata/52756e5ad454534f220001ef
The solution of a Performance version kata was much shorter and 50 times faster, than the decision it in easier version

Thank you.
The first time I haven't understood the instructions correctly

Brackets signifies the difference between both results.

I have get in Java version result of test
expected: <a[c]> but was:<a[]>
Result must be String. What does it mean a[c]? String "a" or "c" ?
Correct wrong result usually look like this : expected:<[ab]> but was:<[abc]>