Programming Puzzles & Code Golf Stack Exchange is a question and answer site for programming puzzle enthusiasts and code golfers. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top


Write a function or program that, given an array (or list), containing only integers, returns or output an array with all sub-elements reversed. That is, reverse all elements of the deepest array, then the second deepest, etc. The dimensions need not be specified, but the function or program must work for jagged arrays in your programming language's native format.



[[1,2], [3,4]]

Would become this:

[[4,3], [2,1]]


[[[ 1, 2, 3], [ 4, 5, 6], [ 7, 8, 9]],
 [[10,11,12], [13,14,15], [16,17,18]],
 [[19,20,21], [22,23,24], [25,26,27]],
 [[28,29,30], [31,32,33], [34,35,36]],
 [[37,38,39], [40,41,42], [43,44,45]],
 [[46,47,48], [49,50,51], [52,53,54]]]

Would become this:

[[[54,53,52], [51,50,49], [48,47,46]],
 [[45,44,43], [42,41,40], [39,38,37]],
 [[36,35,34], [33,32,31], [30,29,28]],
 [[27,26,25], [24,23,22], [21,20,19]],
 [[18,17,16], [15,14,13], [12,11,10]],
 [[ 9, 8, 7], [ 6, 5, 4], [ 3, 2, 1]]]


 [[3,4], [5]],
 [[6,7,8], [9], [10,11]],
 [[[12,13], [14,15]], [16], [17,18,19,20]],

Would become this:

 [[20,19,18,17], [16], [[15,14], [13,12]]],
 [[11,10], [9], [8,7,6]],
 [[5], [4,3]],


This will hopefully encourage answers in some object-oriented programming languages...

-50% Bytecount If your program can take as input an array (or list) with its member's of various types (these can be in the form of objects) and successfully reverse all arrays.



Would become this:

share|improve this question
In your bonus example, you treat strings as atoms. I would argue that they are sub-array, and thus should be reversed too. This is in fact what my APL solution does when fed normal strings, as APL does not have a string datatype, only character datatypes. Strings are therefore one-dimensional character arrays. If you want strings to remain in normal order, you just have to make them objects with a display form. – Adám Mar 17 at 12:27
@Nᴮᶻ Do you believe that the -50% Bytecount is too generous? I may do something along the lines of -30% Bytecount for various data types, and something like -10% Bytecount for reversing Strings, -15% Bytecount for reversing an integer type (123 -> 321) and -15% Bytecount for reversing a floating type (3.14 -> 41.3). – Mr Public Mar 17 at 12:32
I generally dislike bonuses. Reversing integers and floats is ... interesting. – Adám Mar 17 at 12:36
Leave it for now, but next time, you may want to use the sandbox to figure out such things. – Adám Mar 17 at 12:38
About bonuses... – Martin Ender Mar 17 at 13:07

10 Answers 10

Pyth, 11 - 50% = 5.5 bytes


Try it online: Demonstration or Test Suite.

This defines a function y. The additional 3 bytes <newline>yQ simply call the function with the input list and therefore doesn't need to be counted towards the bytes total.


L             define a function y(b), that returns:
 ?+IbY           if b + [] == b (test if b is a list):
      _yMb           recursively call y on all elements in b, then reverse the list
          b      else: b
share|improve this answer

Dyalog APL, 14 - 50% = 7 bytes


⌽⍵ reverse argument
⍣(×|≡⍵) if the argument is not an atom (sign of the [absolute] depth)...
∇¨ ... apply the function to each element (of the reversed argument).

If ⎕ML←3 (IBM style), which is the case on systems that migrated from APL2, one byte can be saved by removing |.

Try APL online.

For curiosity, the proposed int and float reversing:


The inner function:

0::⌽⍵ if any error occurs, just return the revesed argument
⍎⌽⍕ make into string, reverse, make into number

share|improve this answer

Python 2, 40 - 50% = 20

f=lambda x:map(f,x)[::-1]if"">x>[]else x

Only some minor modifications needed from the basic way to do it are needed to get the bonus. Uses the fact that all lists are less than the empty string, and all numbers are less than the empty list.

All test cases

share|improve this answer
Just a note that the version without the bonus is f=lambda x:map(f,x)[::-1]if x>[]else x. – mbomb007 Mar 17 at 18:12

Prolog, 40 - 50% = 20 bytes


This recursively calls predicate a/2 with maplist/3, for each member of the list, until reverse/2 fails (i.e. last element wasn't a list).

share|improve this answer

Emacs Lisp, 46 bytes * 0.5 = 23

(defun g(x)(if(atom x)x(mapcar'g(reverse x))))

Usage example: (g '((1 2) 3 (four 5))) -> ((5 four) 3 (2 1))

Classic recursive approach: if the argument isn't a list, take it unchanged, if it's a list, map the function over the reverse of the list.

share|improve this answer

Clojure 43/2=21.5 bytes

(defn f[x](if(coll? x)(reverse(map f x))x))
share|improve this answer

JavaScript ES6, 42 - 50% = 21 bytes

My score is perfect in so many ways. Implements a function r which recursively applies itself to the members of its input.


If we assume no object has the property pop, then this becomes (31 - 50% = 15.5), thanks to dev-null:


Or, if we assume the object has a sane reverse property, we could do that as well (35 - 50% = 17.5):

share|improve this answer
I guess you could safely check for an array like this: a.pop?a.reverse().map(r):a. Assuming there is no need to handle void 0 and custom objects. – andlrc Mar 17 at 14:18

Mathematica, 34/2=17 bytes


Or just Reverse//@#& if you want a ton of errors and Reverses everywhere.

share|improve this answer

Lua, 111 99 * .5 = 55.5 49.5 bytes

function r(a)local t={}for i,v in next,a do t[#a+1-i]=type(v)=="table"and r(v)or v end return t end

Good bit of recursion

share|improve this answer

CJam, 20 bytes * 50% = 10


Defines the named block F which can be applied to an array on top of the stack (or anything else, in which case it's a no-op).

Test it here.

share|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.