Fixed!

Yes, for me it worked now. Thanks a lot!

Is it better now? I have re-saved files.

Same problems

Same problem, please fix it, been stuck here twice...

You might be interested in the functions words and unwords,
both in Prelude.

Funny, it looks like the issue was fixed (https://github.com/Codewars/codewars-runner-cli/commit/59050f46dee6dd943e5c78cb4a0abadf1dd1cf33) but not pushed yet (2016-12-15 11:24 AM CET) to the servers used to test my submission.

I think that n and u(n) being so "closely" related that saying "going to n" in the u sequence (... where a term of u is less than ...) or "going" to u(n) is the same, no? So I consider the issue as closed. If you don't agree please tell me. Thanks for your post.

The task for comp(n) should be rephrased:

2. Given n (integer > 2)return the number of times where a term of u is less than its predecessor up to and including u(n). Call this last function comp(n)

In my C solution (and yours, too), the loop goes to n, not u(n)

All tests pass (Time: 2457ms Passed: 9 Failed: 0) yet I can't submit my final solution. The only explanation I see is that there's output to STDERR, but these are all warnings for the content of the Test.hs file over which I have no control (?).

I had to fix the examples to get rid of similar warnings.

Can you please look at the Test.hs file and fix the warnings?

STDERR:
/tmp/haskell116119-18-1h3md2h.98bxs10pb9/RPN/Test.hs:9:5: Warning:
Tab character
... snip same warning for lines
... 11, 13, 15, 17, 19, 21, 23, 25

All tests pass (Time: 3092ms Passed: 17 Failed: 0) yet I can't submit my final solution. The only explanation I see is that there's output to STDERR, but these are all warnings for the content of the Main.hs file over which I have no control (?).

[Edit 2016-12-09:] I can run the examples and don't get these warnings.

Can you please look at the Main.hs file and fix the warnings?

STDERR:
/tmp/haskell116117-18-czqumu.62eoupu8fr/Main.hs:7:1: Warning:
Tab character
... snip same warning for lines
... 8, 9, 10, 11, 17, 18, 19, 60, 61, 62, 63, 64, 65, 66, 67, 68, 71 and 72

I had the same problem with random testing using k < 1.

In the end, I got an ugly solution working...

Here are the test cases I gathered while working on it, so you have an idea what kind of values QuickCheck generates and which answers to expect

it "reverse engineers specs implied by random tests" $ do
  let posExa xs k e = josephus xs k `shouldBe` e -- these functions were used as shorthand for negative and positive examples
  let negExa xs k e = josephus xs k `shouldNotBe` e
  josephus "" (- 1) `shouldBe` ""
  josephus "abc" (- 1) `shouldBe` "bca" -- validated?
  -- examples from quickcheck
  posExa "aba" (- 2) "aab"
  posExa "aba" (- 3) "aab"
  posExa "abc" (- 3) "cab"
  posExa "aab" (- 3) "baa"
  posExa "aba" (- 4) "baa"
  posExa "aab" (- 4) "aab"
  posExa "aab" (- 5) "aab"
  posExa "aab" (- 6) "baa"
  posExa "aaab" (- 2) "aaab"
  posExa "Z,n>ToWV" (- 52) ">V,oTZWn"
  posExa "W{VZ" (- 11) "W{VZ"
  posExa "A>Ba" (- 6) ">AaB"
  posExa "" (- 6) ""
  posExa "%0f!" (- 9) "f0!%"
  posExa "WTZBt" (- 3) "TZWBt"
  posExa "(qWQrUMT" (- 36) "Qq(MUrWT"
  posExa "Zo" (- 36) "oZ"
  posExa "V=TfnnQP" (- 17) "QTfPnnV="
  posExa "Tq9QPUA>V$" (- 17) "9QTA>qV$PU"
  posExa "(qWQrUMT" (- 36) "Qq(MUrWT"
  posExa "[sTVQ PR=OS~YZX31l" (- 37) "1YP3V TZ[XQ~OsSR=l"
  posExa "VPRBKg-sYDTQbH/gzdrSce" (- 16) "gT/dScrgQ-VYzPeDHsKRbB"
  posExa "~PJcOvl~VgRn:E)Sy_avtfb}KT{ei]C" (- 18) ":KOSTJRa}]PvlcC{vneEtbf)i_~gV~y"
  posExa "ZW)VaCXabmZyd" (- 6) "XdamZ)VWZCyba"
  posExa "abc" (- 14) "acb"
  posExa "dI{\181\133\GSf[f\SOHMTw\NUL" (- 4) "\SOH\133\NUL[Ifd\GSMTf{w\181"
  posExa "\230\\\145" (- 3) "\145\230\\"
  posExa "UC\SOH\ACKV" (- 8) "CU\SOHV\ACK"
  posExa "dI{\181\133\GSf[f\SOHMTw\NUL" (- 4) "\SOH\133\NUL[Ifd\GSMTf{w\181"
  posExa "\146z\228e\183\ACK.nW7\231M" (- 14) "7\ACK\146\183n.z\228W\231eM"
  posExa "T\143\f_" (- 17) "\f_T\143"
  posExa "\253m\160\&2V\DC3\246J\172\SYN G\219" (- 9) "2\246\172\SYNJV\253\DC3 m\219\160G"
  posExa "Y\ENQZQ" (- 4) "Q\ENQYZ"
  posExa "" (- 19) ""
  posExa "\ACK\229I\ACK.+/\131=\bF)w\a\r\253" (- 4) ")/\229w+\253=\ACK\131\aI\ACK\r\b.F"
  posExa "*9\148\NAKL:\tD\ENQV\226o\226\152j\242\148\NUL#B\172" (- 26) "\242\ENQ*o\172\t\226j\152D9:V\148L#\148B\NAK\NUL\226"
  posExa "cX\142*{#+\DLEe(\ACKY\251,f\234\&1 " (- 7) "\ACK\142\251*Yc\DLEf {#X\234e(1+,"
  posExa "\205;\195(\232cL`\171\252`\ETBt]w.\236touND\212\GS\238\237\DC2" (- 1) "\237\GSDut.]\ETB\252`c(;\DC2\212ow`L\195\238\236\171\205tN\232"
  posExa ",#\186S><3XQ\DC4\FS\162{\n\235\&2`" (- 2) "\235\162Q<\186`{XS2\DC4#\FS,>\n3"
  posExa "W\SOHM\157\140\132|./\246\182\DC3a" (- 32) "|\182\DC3.Wa\157\246\140M\132\SOH/"
  posExa "e\217+\196\&1\RSD}\187\f\223\US#*j\229\135\247\240N\140\177bc\235\\X\130Ll\240" (- 31) "\240L\\\177\135\223\196c#ej\130}\229N\140\247\187lD\US\RS\235X\f+b\217*\240\&1"
  posExa "\183f8\214r*L9L\205+]M\182\223-p\253h\r\232\187_\217\224\GS{\252" (- 25) "8r*\214\183\GS\187p+{-f\182\232\224\252_\223LL\217]9\r\253\205Mh"
  posExa "+\ETB6\131}\ETBN1i\199\f\228U\191#\136\DC2\236c\RS\204h\ESC;" (- 36) "\228h\ETBU\DC2c\236\191N\ESC}\199i\ETB\f1\204\136#6+;\RS\131"
  posExa "\rG7\163\151-<\189 0\n:\DC2g\SO\vH\173r" (- 52) "\151<\163\173\&0g\DC2-\n7 H\189\SO\rr:G\v"
  posExa "\rQ-\ETXt\238\213\200V@\aD4H`" (- 34) "\a\ETXVD\200Qt`\r\238\213\&4-H@"
  -- negative examples from quickcheck
  negExa "aba" (- 4) "aba"
  negExa "aab" (- 4) "aba"
  negExa "aab" (- 4) "baa"
  negExa "aab" (- 6) "aba"
  negExa "aab" (- 6) "aab"
  negExa "aab" (- 7) "baa"
  negExa "aba" (- 5) "baa"
  negExa "aaba" (- 1) "abaa"
  negExa "aaab" (- 3) "aaab"

Add test cases to "prevent" solutions like https://www.codewars.com/kata/reviews/5486115fd8325ec04c0001ea/groups/55988c1318b9e971f300001a or https://www.codewars.com/kata/reviews/5486115fd8325ec04c0001ea/groups/57cd9854fa9fc50f580002b7 which do not preserve the order of the input strings.

E.g. lcs "12345" "45321" == "45"

Also, a property test could be added to verify that lcs x y == lcs y x